Gamma-vinyl GABA inhibits cocaine-triggered reinstatement of drug-seeking behavior in rats by a non-dopaminergic mechanism

A

i(dwrsrP

K

1

afdna2carie

0d

Available online at www.sciencedirect.com

Drug and Alcohol Dependence 97 (2008) 216–225

Gamma-vinyl GABA inhibits cocaine-triggered reinstatement ofdrug-seeking behavior in rats by a non-dopaminergic mechanism

Xiao-Qing Peng a, Xia Li a, Jeremy G. Gilbert a,Arlene C. Pak a, Charles R. Ashby Jr. b, Jonathan D. Brodie c,Stephen L. Dewey d, Eliot L. Gardner a, Zheng-Xiong Xi a,∗

a Neuropsychopharmacology Section, Intramural Research Program, National Institute on Drug Abuse,National Institutes of Health, DHHS, Baltimore, MD 21224, United States

b Department of Pharmaceutical Sciences, Saint John’s University, Jamaica, NY 11439, United Statesc Department of Psychiatry, New York University School of Medicine, New York, NY 10016, United States

d Medical Department, Brookhaven National Laboratory, Upton, NY 11973, United States

Received 22 March 2007; received in revised form 25 September 2007; accepted 6 October 2007Available online 11 December 2007

bstract

Relapse to drug use is a core feature of addiction. Previous studies demonstrate that �-vinyl GABA (GVG), an irreversible GABA transaminasenhibitor, attenuates the acute rewarding effects of cocaine and other addictive drugs. We here report that systemic administration of GVG25–300 mg/kg) dose-dependently inhibits cocaine- or sucrose-induced reinstatement of reward-seeking behavior in rats. In vivo microdialysisata indicated that the same doses of GVG dose-dependently elevate extracellular GABA levels in the nucleus accumbens (NAc). However, GVG,hen administered systemically or locally into the NAc, failed to inhibit either basal or cocaine-priming enhanced NAc dopamine in either naıve

ats or cocaine extinction rats. These data suggest that: (1) GVG significantly inhibits cocaine- or sucrose-triggered reinstatement of reward-eeking behavior; and (2) a GABAergic-, but not dopaminergic-, dependent mechanism may underlie the antagonism by GVG of cocaine-triggeredeinstatement of drug-seeking behavior, at least with respect to GVG’s action on the NAc.ublished by Elsevier Ireland Ltd.

Relap

srceaapbane

eywords: Cocaine; Dopamine; GABA; Gamma-vinyl GABA; Reinstatement;

. Introduction

Relapse to drug use is one of the core features of addiction,nd to date, no broadly effective medications exist. One strategyor developing effective pharmacotherapies for drug abuse andrug dependence is to understand the neurobiological mecha-isms underlying drug craving and relapse, and to pharmacother-peutically target those mechanisms (O’Brien and Gardner,005; Kalivas and Volkow, 2005). Substantial evidence indi-ates that increased dopamine (DA) transmission in the nucleusccumbens (NAc) appears to be critically involved in drug
eward and relapse (Wise, 1998). For example, cocaine prim-ng significantly elevates extracellular NAc DA (Neisewandert al., 1996; McFarland et al., 2003; Xi et al., 2006). Also,
∗ Corresponding author. Tel.: +1 410 550 1749; fax: +1 410 550 5172.E-mail address: [email protected] (Z.-X. Xi).

gad

rD

376-8716/$ – see front matter. Published by Elsevier Ireland Ltd.oi:10.1016/j.drugalcdep.2007.10.004

se

ystemic or intra-NAc administration of direct or indirect DAeceptor agonists elicit, while DA receptor antagonists block,ocaine-induced reinstatement of drug-seeking behavior (Selft al., 1996; De Vries et al., 2002; Vorel et al., 2002). Inddition, repeated cocaine administration produces enduringlterations in basal DA transmission, DA response to cocaineriming and/or expression of DA receptors in reward-relatedrain regions (Kuhar and Pilotte, 1996; Lu et al., 2003; Staleynd Mash, 1996). These data suggest that DA-related mecha-isms play an important role in drug craving and relapse (Shalevt al., 2002; Anderson and Pierce, 2005). In addition to DA,rowing evidence indicates that glutamate-related mechanismsre also critically involved in cocaine-triggered reinstatement of
rug-seeking behavior (see review by Kalivas, 2004).
�-Aminobutyric acid (GABA) is the primary inhibitory neu-otransmitter in the brain and exerts inhibitory control overA and glutamate release in the NAc (Xi et al., 2003). �-

mailto:[email protected]

dx.doi.org/10.1016/j.drugalcdep.2007.10.004

ohol Dependence 97 (2008) 216–225 217

VGeeb(eecMtswrrbD

2

2

wvcwt(dAwo

2

sA2td

2d

og

2smttspn2spsp

Fig. 1. General experimental protocol and effects of GVG on cocaine- orsucrose-triggered reinstatement of reward-seeking behavior. Panel A: generalexperiment protocol showing sequential experimental phases. Panel B: meannumbers of active lever presses during last session of cocaine self-administration,last session of extinction, and cocaine-primed reinstatement in presence of vehi-cle or various GVG doses. Panel C: mean numbers of active lever presses duringli*

iaC

2dA6da

X.-Q. Peng et al. / Drug and Alc

inyl GABA (GVG) is an anticonvulsant that elevates brainABA levels by irreversibly inhibiting GABA transaminase, the

nzyme responsible for GABA’s metabolic degradation (Cubellst al., 1986). Systemic administration of GVG in rats haseen reported to significantly inhibit cocaine self-administrationKushner et al., 1999), cocaine-induced conditioned place pref-rence (Dewey et al., 1998), cocaine-induced enhancement oflectrical brain stimulation reward (Kushner et al., 1997), andocaine-induced increases in NAc DA (Dewey et al., 1997;organ and Dewey, 1998; Schiffer et al., 2000). However,

he effect of GVG on cocaine-triggered reinstatement of drug-eeking behavior has remained unexplored. In the present study,e examined the effects of GVG on cocaine- or sucrose-induced

einstatement of reward-seeking behavior in male Long-Evansats. Further, we examined the effects of GVG pretreatment onasal and cocaine-induced changes in extracellular GABA andA in the NAc using in vivo brain microdialysis.

. Materials and methods

.1. Animals

Male Long-Evans rats (Charles River Laboratories, Raleigh, NC, n = 130)eighing 250–300 g were used for all experiments. They were housed indi-idually in a climate-controlled animal colony room on a reversed light-darkycle (lights on at 7:00 p.m., lights off at 7:00 a.m.) with free access to food andater. The animals were maintained in a facility fully accredited by the Associa-

ion for Assessment and Accreditation of Laboratory Animal Care InternationalAAALAC International). All experimental procedures were conducted in accor-ance with the Guide for the Care and Use of Laboratory Animals (Nationalcademy of Sciences, Washington DC: National Academy Press, 1996) andere approved by the Animal Care and Use Committee of the National Instituten Drug Abuse of the U.S. National Institutes of Health.

.2. Drugs and chemicals

Cocaine HCl (Sigma/RBI, Saint Louis, MO) was dissolved in physiologicalaline. GVG (vigabatrin, Sabril®) in racemic crystalline form was obtained fromventis Pharma Inc. (Laval, QC, Canada). GVG was dissolved in 25% (w/v)-hydroxypropyl-�-cyclodextrin (Sigma/RBI, Saint Louis, MO) for i.p. injec-ions. For local intra-NAc perfusion, GVG was dissolved in the dialysis bufferescribed below.

.3. Cocaine self-administration and reinstatement ofrug-seeking behavior

The cocaine self-administration and reinstatement experiments were carriedut as we have previously described (e.g., Xi et al., 2006). Fig. 1A depicts theeneral experimental protocol followed.

.3.1. Surgery. Intravenous (i.v.) jugular catheterization was carried out usingtandard aseptic surgical technique. The i.v. catheters were constructed oficrorenathane (Braintree Scientific Inc., Braintree, MA). Rats were anes-

hetized with sodium pentobarbital (65 mg/kg i.p.) and then the skin overhe right external jugular was incised, the jugular vein exposed by blunt dis-ection and incised, and the catheter inserted into the vein and sutured intolace. The portion of catheter outside the jugular was then passed subcuta-eously to the top of the skull, where it exited into a connector (a modified
2 gauge cannula; Plastics One, Roanoke, VA) mounted to the skull withtainless steel skull screws (Small Parts Inc., Miami Lakes, FL) and cranio-lastic acrylic (Codman and Shurtleff, Raynham, MA). During experimentalessions, the catheter was connected to the infusion pump via tubing encased in arotective metal spring from the head-mounted connector to the top of the exper-
oiwSM

ast session of sucrose self-administration, last session of extinction, and sucrose-nduced reinstatement in presence of vehicle or various GVG doses. *p < 0.05,*p < 0.01, ***p < 0.001, compared to vehicle treatment group.

mental chamber. To help prevent clogging, catheters were flushed daily withgentamicin–heparin–saline solution (30 IU/ml heparin; ICN Biochemicals,leveland, OH).

.3.2. Apparatus. The i.v. cocaine self-administration experiments were con-ucted in standard operant test chambers (32 cm × 25 cm × 33 cm) (Medssociates Inc., Saint Albans, VT). Each test chamber had two levers, located.5 cm above the floor. Depression of one lever activated an infusion pump;epression of another lever was counted but had no consequence. A cue lightnd speaker were located 12 cm above the active lever. A house light was turned
n at the start of each 3 h test session. When an animal pressed a lever that resultedn a drug infusion, 2 drug-paired conditioned cues (cue light and cue sound/tone)ere automatically activated and remained on for the duration of the infusion.cheduling of experimental events and data collection was accomplished usinged Associates software (Med-PC IV).

2 ohol D

2fad4due(caaltoics

2adlucgddAacpet

2cwrtowip

2

2iiewiRdAg−acIhemAee

2Ndd1spsdt2cNc1f

2spEuuwaflloduG

2pdrmsuematI

2t(1aNawe

2

ciep

18 X.-Q. Peng et al. / Drug and Alc

.3.3. Cocaine self-administration under a fixed-ratio 2 (FR2) schedule of rein-orcement. After 5–7 days of recovery from surgery, each rat was placed into

test chamber and allowed to lever-press for i.v. cocaine (1 mg/kg/injection)elivered in 0.08 ml over 4.6 s, on a FR1 schedule of reinforcement. During the.6 s necessary for infusion, responses on the active lever were recorded butid not lead to additional infusions. Each session lasted 3 h. Subjects respondednder the FR1 schedule for 3–5 days until stable cocaine self-administration wasstablished. Then, the subjects were transferred to cocaine self-administration0.5 mg/kg/injection) under a FR2 reinforcement schedule until the followingriteria for stable cocaine-maintained responding were met: less than 10% vari-bility in inter-response interval and less than 10% variability in number ofctive lever-presses for at least 3 consecutive days. The unit cocaine dose wasowered from 1.0 mg/kg/infusion to 0.5 mg/kg/infusion in order to increasehe work demand (i.e., lever presses) on the animals for the same amountf drug intake, which in our experience significantly increases the sensitiv-ty of measuring changes in drug-taking or drug-seeking behavior. To avoidocaine-induced seizures, each animal was limited to 50 cocaine infusions peression.

.3.4. Extinction and testing for reinstatement. After stable cocaine self-dministration was established, animals were exposed to extinction conditions,uring which cocaine was replaced by saline, and the cocaine-associated cueight and tone were turned off. Daily 3 h extinction sessions for each rat contin-ed until that rat lever-pressed less than 10 times per 3 h session for at least 3onsecutive days. After meeting extinction criteria, animals were divided into 5roups (8–10 rats per group) for reinstatement testing. On the reinstatement testay, each rat received either vehicle (25% 2-hydroxypropyl-�-cyclodextrin) or 1ose of GVG (25, 50, 100, and 300 mg/kg, i.p.) 1 h before the reinstatement test.ll rats were then given a priming injection of cocaine (10 mg/kg i.p.) immedi-

tely before the initiation of reinstatement testing. During reinstatement testing,onditions were identical to those in extinction sessions. Cocaine-induced leverresses (reinstatement) were recorded, although these responses did not lead toither cocaine infusions or presentation of the conditioned cues. Reinstatementest sessions lasted 3 h.

.3.5. Sucrose-triggered reinstatement of sucrose-seeking behavior. The pro-edures for oral sucrose self-administration, extinction, and reinstatement testingere identical to the procedures for cocaine self-administration, extinction, and

einstatement testing except for the following: (1) no surgery was performed onhe animals in the sucrose experiment; (2) active lever presses led to deliveryf 0.1 ml of 5% sucrose solution into a liquid food tray on the operant chamberall; and (3) reinstatement was triggered initially by 2 “free” sucrose deliver-

es, and subsequent lever presses did not lead to either sucrose infusion or theresentation of the conditioned cue-light and tone.

.4. In vivo microdialysis

.4.1. Surgery and microdialysis probes. There were two groups of animals,.e. drug naıve and cocaine-treated animals, used for in vivo microdialysis stud-es. The cocaine-treated animals had the same cocaine self-administration andxtinction training as described above, and then the microdialysis experimentas performed during the reinstatement test. For those rats, both i.v. catheter-

zation and intracranial guide cannula (20 gauge, 14 mm length; Plastics One,oanoke, VA) implantation were performed at time of surgery, while for therug naıve rats, only intracranial guide cannula implantation was performed.nesthesia was as described above. The target implant coordinates for the NAcuide cannulae were +1.6 mm anterior to Bregma, ±1.6 mm mediolateral, and4.7 mm ventral to the skull surface, according to the rat brain atlas of Paxinos

nd Watson (1998), using a surgical approach of 6◦ from vertical. The guideannulae were fixed to the skull with 4 stainless steel skull screws (Small Partsnc., Miami Lakes, FL) and cranioplastic acrylic (Codman and Shurtleff, Rayn-am, MA). Microdialysis probes were constructed as previously described (Xi
t al., 2003). The active operational length of the semipermeable microdialysisembrane was 1.0–1.5 mm, and probe diameter was approximately 100 �m.fter animals had recovered from surgery for 5–7 days, we then examined the
ffect of GVG pretreatment on basal levels of, and cocaine-induced changes in,xtracellular NAc DA and GABA.

tomad

ependence 97 (2008) 216–225

.4.2. Microdialysis procedure. Microdialysis probes were inserted into theAc at least 12 h before onset of experimentation to minimize the effects ofamage-induced neurochemical release during the experiment. Then, micro-ialysis buffer (5 mM glucose, 2.5 mM KCl, 140 mM NaCl, 1.4 mM CaCl2,.2 mM MgCl2, 0.15% phosphate buffered saline, pH 7.4) was perfused viayringe pump (Bioanalytical Systems, Inc., West Lafayette, IN) through therobes at 2.0 �l/min for at least 2 h prior to sample collection. Microdialysisamples were collected every 20 min into 10 �l 0.5 M perchloric acid to preventegradation of the collected chemicals. After 1 h of baseline sample collec-ion, 1 of 3 doses of GVG (25, 100, and 300 mg/kg, i.p.) or vehicle (1 ml 25%-hydroxypropyl-�-cyclodextrin) was administered systemically, or differentoncentrations of GVG (1, 10, 100, and 1000 �M) were locally infused into theAc by reverse microdialysis. To evaluate the effects of GVG pretreatment onocaine-induced changes in NAc DA, cocaine (10 mg/kg, i.p.) was administered, 3 or 6 h after systemic GVG administration. After collection, all samples wererozen at −80 ◦C until analyzed.

.4.3. Quantification of GABA. Concentrations of GABA in the microdialysisamples were determined using HPLC with flourometric detection. The mobilehase consisted of 18% acetylnitrile (v/v), 100 mM Na2HPO4, and 0.1 mMDTA, pH 6.04. A VeloSep RP-18, 10 cm × 3 �m ODS reversed phase col-mn (PerkinElmer Life and Analytical Sciences, Inc., Wellesley, MA) wassed to separate the amino acids, and precolumn derivatization of amino acidsith o-phthalaldehyde was performed using an ESA Biosciences model 542

utosampler. GABA was detected by an ESA Biosciences Linear Fluor LC 305uorescence spectrophotometer. Excitation (Ex�) and emission (Em�) wave-

engths were 336 nm and 420 nm, respectively. The area under the curve (AUC)f the GABA peak was measured using the EZChrom EliteTM chromatographyata analysis system (ESA Biosciences, Inc., Chelmsford, MA). GABA val-es were quantified with an external standard curve. The limit of detection forABA was 0.1–1 pM.

.4.4. Quantification of DA. Microdialysate DA was measured using high-erformance liquid chromatography (HPLC) coupled with electrochemicaletection (ESA Biosciences Inc.). The DA mobile phase contained 4.76 mM cit-ic acid, 150 mM Na2HPO4, 3 mM sodium dodecyl sulfate, 50 mM EDTA, 10%ethanol, and 15% acetylnitrile, pH 5.6. DA was separated using an ESA Bio-

ciences MD-150 × 3.2 mm reversed phase column, and was oxidized/reducedsing an ESA Biosciences Coulochem® III electrochemical detector. Threelectrodes were used: a preinjection port guard cell (+0.25 V) to oxidize theobile phase, an oxidation analytical electrode (E1, −0.1 V), and a reduction

nalytical electrode (E2, 0.2 V). The AUC of the DA peak was measured usinghe EZChrom EliteTM chromatography data analysis system (ESA Biosciences,nc.). DA values were quantified with an external standard curve (1–1000 fM).

.4.5. Neuroanatomical verification of microdialysis probe sites. Followinghe microdialysis experiments, rats were given an overdose of pentobarbital>100 mg/kg i.p.) and transcardially perfused with 0.9% saline followed by0% formalin solution. Brains were removed and placed in 10% formalin fort least 1 week to ensure thorough fixation. The tissue was blocked around theAc and coronal sections (100 �m thick) were made by vibratome through therea of microdialysis probe implantation. The brain sections were then stainedith cresyl violet. Anatomical placement was verified by visual microscopic

xamination.

.5. Data analyses

All data are presented as means ± S.E.M. AUC measurement was used foromparison of overall effects of GVG treatment on basal levels of, or cocaine-nduced changes in, extracellular DA or GABA. The AUC% was calculated forach animal by subtracting 100 from the percent of baseline value for each dataoint, and subsequently summing all data points collected after drug adminis-
ration. One-way analysis of variance (ANOVA) was used to analyze the effectsf GVG on cocaine-induced reinstatement (Fig. 1B and C). Two-way (treat-ent × time) ANOVA with repeated measures on 1 factor (time) was used to
nalyze the data reflecting the time courses of neurochemical changes afterifferent doses of GVG and/or cocaine (Figs. 2–4). Post-ANOVA individual

X.-Q. Peng et al. / Drug and Alcohol Dependence 97 (2008) 216–225 219

Fig. 2. Effects of systemic administration of GVG on extracellular GABA, DAand cocaine-enhanced DA in the NAc in cocaine extinction rats during reinstate-ment testing. Panel A: effects of GVG (25–300 mg/kg, i.p.) on NAc extracellularGABA. Panel B: effects of GVG on NAc extracellular DA. Panel C: effects ofGVG pretreatment (25–300 mg/kg, i.p., 1 h prior to cocaine priming) on cocaine-induced increases in NAc DA. *p < 0.05, **p < 0.01, ***p < 0.001, compared tobaseline before GVG administration in each GVG dose group.

Fig. 3. Effects of systemic administration of GVG on extracellular GABA, DAand cocaine-enhanced DA in the NAc of drug naıve rats. Panel A: effects of GVG(25–300 mg/kg, i.p.) on NAc extracellular GABA. Panel B: effects of GVG onNAc extracellular DA. Panel C: effects of GVG pretreatment (300 mg/kg, i.p.,1, 3 or 6 h prior to cocaine priming) on cocaine-induced increases in NAc DA.*p < 0.05, **p < 0.01, ***p < 0.001, compared to baseline before GVG adminis-tration in each GVG dose group.

220 X.-Q. Peng et al. / Drug and Alcohol D

Fig. 4. Effects of local perfusion of GVG into the NAc on NAc extracellu-lar GABA, DA and cocaine-enhanced DA in drug naıve rats. Panel A: effectsof GVG (1–1000 �M) on extracellular GABA. Panel B: effects of the sameconcentrations of GVG on NAc extracellular DA. Panel C: effects of localperfusion of GVG (1000 �M) into the NAc on cocaine-induced increases inNAc DA. *p < 0.05, **p < 0.01, ***p < 0.001, compared to baseline before GVGadministration in each group.

pp

3

3r

3nsmdtwindstnds

3aerlnt(

3tptadptti(cn(aG

tpso(to

ependence 97 (2008) 216–225

re-planned group comparisons were carried out using the Bonferroni statisticalrocedure.

. Results

.1. Effects of GVG on cocaine- or sucrose-triggeredeinstatement of reward seeking

.1.1. Cocaine self-administration. Fig. 1B shows the meanumbers of active lever presses during the last session of cocaineelf-administration, last session of extinction, and reinstate-ent testing in the presence of vehicle or 1 of the 4 GVG

oses. Rats in all 5 groups exhibited stable responding onhe active lever during the last 5–7 self-administration daysith a within-subject variability of <10% in daily cocaine

nfusions. There were no significant differences in the meanumbers of cocaine infusions or active lever presses amongifferent groups of rats during the last 3 sessions of cocaineelf-administration (data not shown). Responding on the inac-ive lever was minimal in all groups of rats. There wereo significant differences in responding on the inactive leveruring self-administration among different groups (data nothown).

.1.2. Extinction. As noted above, cocaine and cocaine-ssociated cues were unavailable during the extinction phase. Asxtinction progressed (2–3 weeks), the number of drug-seekingesponses gradually decreased until the extinction criterion (<10ever presses per 3 h) was met. There were no differences in theumber of extinction responses among the different groups onhe last extinction session, 24 h prior to reinstatement testingF4,39 = 0.44, p > 0.05).

.1.3. Reinstatement. Reinstatement testing began 24 h afterhe last extinction session, under extinction conditions (i.e., leverresses did not produce either cocaine infusions or the presen-ation of the cocaine-associated cues). As indicated in Fig. 1B,

single noncontingent cocaine injection (10 mg/kg i.p.) pro-uced robust reinstatement of extinguished operant behaviorreviously reinforced by i.v. cocaine self-administration. Pre-reatment with GVG (25, 50, 100 and 300 mg/kg i.p., 1 h prioro reinstatement testing) dose-dependently inhibited cocaine-nduced reinstatement by 26, 44, 57, and 64%, respectivelyFig. 1B, right panel: F4,39 = 4.35, p < 0.01). Individual groupomparisons using the Bonferroni test revealed a statistically sig-ificant decrease in cocaine-induced responding after 100 mg/kgt = 3.39, p < 0.05) and 300 mg/kg (t = 3.55, p < 0.01), but notfter 25 mg/kg (t = 1.43, p > 0.05) or 50 mg/kg (t = 2.46, p > 0.05)VG.Fig. 1C shows that GVG, at 300 mg/kg, but not at other doses

ested, significantly inhibited the reinstatement of respondingreviously reinforced by sucrose. One-way ANOVA for the datahown in Fig. 1C showed a statistically significant effect of GVG
n sucrose-triggered reinstatement of sucrose-seeking behaviorF3,27 = 10.38, p < 0.001). Individual group comparisons usinghe Bonferroni test revealed a statistically significant inhibitionf sucrose seeking only after 300 mg/kg (t = 4.6, p < 0.001), but

ohol D

no

3Gr

3asGaesloctiiip2

3Gptmnti

3trttimemt

3G

erMw1i

3t

iioctiii3G

3aid(lttciobp

3timcimemt

3en

cabwiFiefcea


ot 100 mg/kg (t = 1.6, p > 0.05) or 25 mg/kg (t = 0.43, p > 0.05)f GVG administration.

.2. Effects of systemic administration of GVG on NAcABA, DA and cocaine-enhanced DA in cocaine extinction

ats during reinstatement testing

.2.1. Effects of GVG on GABA. To determine whether thentagonism by GVG of cocaine-triggered reinstatement of drugeeking is correlated with GVG-induced alterations in NAcABA or DA levels, we further observed the effects of GVG

lone or pretreatment on extracellular GABA, DA and cocaine-nhanced DA in rats during the reinstatement test. Fig. 2Ahows that GVG dose-dependently elevated NAc extracellu-ar GABA levels. A two-way ANOVA for repeated measuresver time for the data shown in Fig. 2A revealed a statisti-ally significant treatment main effect (F3,31 = 10.09, p < 0.001),ime main effect (F17,527 = 8.46, p < 0.001), and treatment × timenteraction (F51,527 = 5.71, p < 0.001). Individual group compar-sons using the Bonferroni test revealed a statistically significantncrease in extracellular GABA levels after 100 mg/kg (t = 2.65,< 0.05) and 300 mg/kg GVG (t = 5.06, p < 0.001), but not after5 mg/kg GVG (t = 1.18, p > 0.05).

.2.2. Effects of GVG on NAc DA. Fig. 2B shows the effects ofVG alone on NAc DA in rats during reinstatement testing. Sur-risingly, the same doses of GVG that altered NAc GABA failedo alter NAc extracellular DA. Two-way ANOVA for repeated

easures over time for the data shown in Fig. 2B revealed aon-significant treatment main effect (F3,36 = 1.17, p > 0.05),ime main effect (F17,388 = 0.65, p > 0.05), and treatment × timenteraction (F36,422 = 1.19, p > 0.05).

.2.3. Effects of GVG on cocaine-enhanced DA. Fig. 2C showshe effects of GVG on cocaine-enhanced DA in rats duringeinstatement test. Pretreatment with the same doses of GVGhat altered NAc GABA (25, 100, 300 mg/kg, i.p., 1 h prioro cocaine priming) did not inhibit cocaine priming-inducedncreases in DA (Fig. 2C). A two-way ANOVA for repeated

easures over time revealed a statistically significant time mainffect (F14,518 = 106.35, p < 0.001), but no significant treatmentain effect (F3,37 = 1.68, p > 0.05) nor treatment × time interac-

ion (F42,518 = 1.10, p > 0.05).

.3. Effects of systemic administration of GVG on NAcABA, DA and cocaine-enhanced DA in drug naıve rats

The above lack of significant effect of GVG on cocaine-nhanced NAc DA levels appears to conflict with previouseports of GVG’s effects in drug naıve rats (Dewey et al., 1997;

organ and Dewey, 1998; Schiffer et al., 2000). Therefore,e further observed the effects of GVG (25–300 mg/kg, i.p.,, 3 or 6 h prior to cocaine administration) on cocaine-induced
ncreases in NAc DA in drug-naıve rats.
.3.1. Effects of GVG on NAc GABA. Fig. 3A shows that sys-emic administration of GVG alone (25, 100, and 300 mg/kg

rslF

ependence 97 (2008) 216–225 221

.p.) dose-dependently elevated extracellular NAc GABA levelsn drug naıve rats. A two-way ANOVA for repeated measuresver time for the data shown in Fig. 3A revealed a statisti-ally significant treatment main effect (F3,28 = 18.22, p < 0.001),ime main effect (F20,560 = 9.24, p < 0.001), and treatment × timenteraction (F60,560 = 4.81, p < 0.001). Individual group compar-sons using the Bonferroni test revealed a statistically significantncrease in NAc GABA after 100 mg/kg (t = 3.17, p < 0.05) and00 mg/kg GVG (t = 5.59, p < 0.001), but not after 25 mg/kgVG (t = 0.49, p > 0.05).

.3.2. Effects of GVG on NAc DA. To determine whether suchn increase in NAc GABA by GVG inhibits NAc DA releasen naıve rats, we measured DA levels in the same micro-ialysis samples. As in the cocaine-experienced rats, GVG25–300 mg/kg) alone did not alter basal levels of extracellu-ar DA in the NAc (Fig. 3B). Although two-way ANOVA forhe data shown in Fig. 2B revealed a statistically significantreatment main effect (F2,24 = 3.29, p < 0.05), individual groupomparisons did not reveal a statistically significant differencen extracellular DA levels between the vehicle and any dosef GVG tested, except for a statistically significant differenceetween the 25 and 300 mg/kg GVG treatment groups (t = 3.37,< 0.05).

.3.3. Effects of GVG on cocaine-enhanced DA. Fig. 3C showshat cocaine priming (10 mg/kg i.p.) produced a significantncrease (∼400% of baseline) in NAc extracellular DA. Pretreat-

ent with GVG (300 mg/kg, i.p. administered 1, 3 or 6 h prior toocaine priming) had no effect on such cocaine priming-inducedncreases in DA (Fig. 3C). A two-way ANOVA for repeated

easures over time revealed a statistically significant time mainffect (F11,363 = 49.63, p < 0.001), but no significant treatmentain effect (F3,33 = 0.08, p > 0.05) nor treatment × time interac-

ion (F33,363 = 0.51, p > 0.05).

.4. Effects of local perfusion of GVG into the NAc onxtracellular GABA, DA and cocaine-enhanced DA in drugaıve rats

To determine whether such lack of effect of GVG onocaine-enhanced NAc DA was due to an interaction of bothdirect action in the NAc and an indirect action in other

rain regions, or difficulty of GVG penetrating to the NAc,e further observed the effects of local perfusion of GVG

nto the NAc on NAc GABA, DA and cocaine-enhanced DA.ig. 4A shows that local perfusion of GVG (1 �M–1 mM)

nto the NAc significantly elevated extracellular GABA lev-ls in a concentration-dependent manner. A two-way ANOVAor repeated measures over time revealed a statistically signifi-ant treatment main effect (F1,17 = 15.20, p < 0.001), time mainffect (F14,238 = 12.97, p < 0.001), and treatment × time inter-ction (F14,238 = 12.65, p < 0.001). Consistent with the findings
eported above after systemic GVG administration, local infu-ion of GVG (1 �M to 1 mM) into the NAc failed to alter basalevels of extracellular NAc DA (F1,18 = 0.05, p > 0.05) (Fig. 4B).urther, local continuous perfusion of GVG (1 mM) into the NAc

222 X.-Q. Peng et al. / Drug and Alcohol D

Fa

aN

3

smamo

4

ottmfiitaDsdedocDtld

4r

i

aerTlaeaIdcdo(pdodadbdca

4i

GDngGtAelgHa(1ectdctc(2t

ig. 5. Microdialysis probe placements within the NAc. The extent of eachctive semipermeable microdialysis membrane is depicted.

lso failed to inhibit cocaine-induced increases in extracellularAc DA (F1,12 = 0.10, p > 0.05) (Fig. 4C).

.5. Histology

The locations of the microdialysis probes in the NAc arehown in Fig. 5. The active semi-permeable membranes of theicrodialysis probes were located within both the NAc core

nd shell. There were no obvious differences in placement oficrodialysis probes across the different experimental groups

f rats.

. Discussion

The present study demonstrates that systemic administrationf GVG (25–300 mg/kg) dose-dependently inhibited cocaine-riggered reinstatement of drug-seeking behavior. In addition,he highest dose (300 mg/kg) of GVG also inhibited reinstate-

ent of responding previously reinforced by sucrose. Thisnding is consistent with but extends previous findings that GVG

nhibits cocaine’s acute rewarding effects as assessed by elec-rical brain-stimulation reward, conditioned place preference,nd drug self-administration (e.g., Kushner et al., 1997, 1999;ewey et al., 1998). Further, the present in vivo microdialysis

tudies demonstrate that systemic administration of GVG dose-ependently elevated extracellular GABA levels, and had noffect on either basal or cocaine-enhanced DA in the NAc inrug naıve or cocaine-experienced rats. Finally, local perfusionf GVG into the NAc also dose-dependently elevated NAc extra-ellular GABA levels, but had no effect on either basal levels ofA or cocaine-induced increases in NAc DA. These data suggest

hat a GABA-, but not a DA-, dependent mechanism may under-ie the antagonism by GVG of cocaine-primed reinstatement ofrug-seeking behavior.

.1. GVG inhibits cocaine- and sucrose-triggered
einstatement
The present experiments found that GVG significantly inhib-ted cocaine-triggered reinstatement at 100–300 mg/kg GVG,

tdse

ependence 97 (2008) 216–225

nd also inhibited sucrose-triggered reinstatement at the high-st dose of GVG (300 mg/kg), suggesting that GVG has aelatively selective action on cocaine’s action at lower doses.his is consistent with previous studies demonstrating that the

owest effective doses (300–320 mg/kg) for inhibiting oper-nt responding reinforced by food are higher than the lowestffective doses (180–200 mg/kg) for inhibiting cocaine self-dministration (Kushner et al., 1999; Barrett et al., 2005).t was reported that GVG alone (180–320 mg/kg) appears toose-dependently lower operant responding rates in the drug dis-rimination paradigm (Barrett et al., 2005). However, the sameoses of GVG did not alter the discriminative stimulus effectsf cocaine (Barrett et al., 2005), and an even higher GVG dose400 mg/kg) failed to alter operant responding for food underrogressive-ratio reinforcement (Kushner et al., 1999). Theseata suggest that a reduction in responding for cocaine/foodr in reinstatement of cocaine/food-seeking behaviors is likelyue to a specific action on reward and/or reinstatement mech-nisms, rather than a nonspecific inhibition of locomotion or aeficit in motoric ability. This conclusion is further supportedy evidence that GVG has no effect on locomotor activity at theoses (75–150 mg/kg) that produced a significant reduction inocaine-triggered reinstatement in the present study (Gardner etl., 2002).

.2. Non-DA mechanisms underlying GVG-inducednhibition of drug-seeking behavior

The present findings also suggest that the antagonism byVG of cocaine-triggered reinstatement appears to be NAcA-independent. It is well documented that NAc GABAergiceurons receive both DA input from the VTA and glutamater-ic input from the frontal cortex (Sesack and Pickel, 1992).iven the important role of NAc DA (and glutamate) in cocaine-

riggered relapse to drug-seeking behavior (Shalev et al., 2002;nderson and Pierce, 2005; Kalivas, 2004), we initially hypoth-

sized that GVG-induced increases in extracellular GABAevels might inhibit cocaine-induced increases in NAc DA (andlutamate), thereby inhibiting cocaine-induced reinstatement.owever, the present data do not support that hypothesis: (1)wide dose range of GVG, whether administered systemically

25–300 mg/kg, 1, 3 or 6 h prior to cocaine) or locally (1, 10,00 and 1000 �M) into the NAc, altered neither basal levels ofxtracellular DA nor cocaine-induced increases in NAc extra-ellular DA in either cocaine-treated rats (during reinstatementesting) or in drug naıve rats; and (2) the same doses of GVGose-dependently increased, rather than decreased, NAc extra-ellular glutamate levels (Xi et al., unpublished data). Clearly,hese findings conflict with previous reports that GVG inhibitsocaine- and other addictive drug-induced increases in NAc DADewey et al., 1997; Morgan and Dewey, 1998; Schiffer et al.,000). The precise reasons for this discrepancy are unclear. First,he present ineffectiveness of GVG on NAc DA is unlikely
o have been due to the doses being too low, because suchoses (25–300 mg/kg) dose-dependently inhibited cocaine- orucrose-triggered reinstatement (present study) and increasextracellular glutamate levels (Xi et al., unpublished data); and

ohol D

tdm(etp3ocaefebdvpatpsist(GgseciamdoatpndoGiDG

4d

poovGl

ai1dliscdtrpniStG(1XbcoiatGgttpGtmojaSimBcibGwicdaVSu


hese are the effective doses used in other experiments withrug self-administration, conditioned place preference, loco-otor sensitization and intracranial brain stimulation reward

Kushner et al., 1997, 1999; Dewey et al., 1997, 1998; Barrettt al., 2005; Gardner et al., 2002). Second, the present ineffec-iveness is also unlikely to have been due to inappropriate GVGretreatment time, because GVG, when administered either 1,or 6 h prior to cocaine priming, still failed to alter either basalr cocaine-enhanced NAc DA (the present study), but inhibitedocaine self-administration, cocaine-induced place preferencend brain stimulation reward (Kushner et al., 1997, 1999; Deweyt al., 1997, 1998; Barrett et al., 2005). Third, the present inef-ectiveness would appear to be unrelated to animals’ cocainexperience, because the same ineffectiveness was observed inoth cocaine-treated rats and drug naıve rats. We, therefore, arerawn to speculate that differences in rat strains (Long-Evanss. Sprague–Dawley) and/or brain regions of microdialysis sam-ling (medial NAc core/shell in the present study vs. unspecifiednatomic domains or sub-domains in other studies) likely con-ributed to the different results observed in the present study andrevious reports. Whatever the reasons, the present study, byimultaneously measuring extracellular GABA and DA levelsn the same microdialysis samples, appears to clearly demon-trate that GVG, when administered systemically (1–6 h prioro cocaine), or locally into the NAc, dose-dependently elevates100–600%) extracellular GABA, but fails to inhibit NAc DA.iven that GVG-induced increases in extracellular GABA andlutamate are derived predominantly from non-neuronal (glial)ources (Xi et al., unpublished data), we suggest that enhancedxtrasynaptic GABA may not diffuse sufficiently into synapticlefts to inhibit DA release. This is consistent with previous find-ngs that local NAc perfusion of GABAA or GABAB receptorgonists, but not GABA itself, inhibits NAc DA and gluta-ate (Xi and Stein, 1998, 1999; Xi et al., 2003), and similarly

rug (cocaine or heroin) self-administration and reinstatementf drug-seeking behaviors (Xi and Stein, 1999; McFarland etl., 2003; Roberts and Brebner, 2000). In other words, synap-ic GABA may modulate DA or glutamate release as shownreviously (Xi et al., 2003), while extra-synaptic GABA mayot necessarily modulate presynaptic DA or glutamate release,epending upon extra-synaptic GABA levels and the densityf extra-synaptic GABA transporters that may protect synapticABA transmission by removing extracellular GABA before

t reaches the synaptic cleft. GVG’s ineffectiveness on NAcA suggests that DA independent mechanisms may underlieVG-induced inhibition of drug-seeking behavior.

.3. GABAergic mediation of GVG’s inhibition ofrug-seeking behavior

The present findings also indicate that GVG (25–300 mg/kg)roduces a slow-onset (1 h) and long-acting (at least 6 h) dose-rderly increase in NAc GABA levels. Such a long duration
f GVG on brain GABA levels is consistent with GVG’s irre-ersible action (Cubells et al., 1986) and the low turnover rate ofABA transaminase (Kang et al., 2001). Although it has been
ong believed that GVG’s antagonism of drug-taking behavior in

sfd

ependence 97 (2008) 216–225 223

nimals is mediated by elevation of brain GABA levels, whichn turn inhibits drug-induced increases in NAc DA (Dewey et al.,997, 1998; Kushner et al., 1997, 1999), no study has previouslyirectly measured GVG-induced changes in extracellular GABAevels or GABA and DA levels in the brain reward system dur-ng drug self-administration. In the present study, we found thatystemic or intra-NAc GVG dose-dependently elevated extra-ellular GABA, which lasted for at least 5–6 h. Given that GVGid not inhibit cocaine-induced increases in NAc DA, we believehat GVG-induced increases in GABA may play a more directole in inhibiting cocaine-triggered reinstatement than we hadreviously surmised. It is well documented that GABAergiceurons in the brain’s reward circuitry play an important rolen opiate or psychostimulant reward (Koob and Bloom, 1988;elf and Nestler, 1995; Wise, 1998; Xi and Stein, 2002), and

hat cocaine, opiates and DA produce an inhibitory effect onABAergic neurons or GABA release in their projection areas

Uchimura and North, 1990; Qiao et al., 1990; White et al.,993; Cameron and Williams, 1994; Nicola and Malenka, 1997;i and Stein, 2002; Centonze et al., 2002). Based on this, weelieve that GVG-elevated GABA levels in brain reward cir-uitry might directly counteract the actions of cocaine, opiatesr DA on GABAergic neurons, thereby antagonizing cocaine-nduced reinstatement of drug-seeking behavior. The role ofn increase in extracellular GABA levels in the NAc remainso be determined. One possibility is that an increase in NAcABA levels may hyperpolarize NAc medium spiny (GABAer-ic) neurons, and therefore desensitize NAc neuronal responseso cocaine-induced increases in NAc DA and inhibit cocaine-riggered reinstatement. However, this is unlikely because ourrevious studies have shown that selective elevation of NAcABA levels by microinjection of GVG into the NAc failed

o inhibit heroin self-administration (Xi and Stein, 2000), whileicroinjections of GVG into the ventral tegmental area (VTA)

r the ventral palladium (VP), the primary NAc GABAergic pro-ection area (Smith and Bolam, 1990; Bennett and Bolam, 1994),lmost completely inhibit heroin self-administration (Xi andtein, 2000). These data suggest that GVG-induced increases

n extracellular GABA in the VTA or VP, but not in the NAc,ay play a critical role in attenuating heroin’s rewarding effects.ased on this, we believe that GVG-induced increases in extra-ellular GABA in the VTA and the VP may play a similarlymportant role in attenuating cocaine-triggered reinstatementy counteracting cocaine-induced reductions in extracellularABA levels as shown previously (Tang et al., 2005). Althoughe did not measure extracellular GABA levels in the VTA or VP

n the present study, we believe that a similar increase in extra-ellular GABA after GVG can be expected based upon the wideistribution of GABA transaminase in rat brain (Reijnierse etl., 1975; Chan-Palay et al., 1979) and the antagonism by intra-TA or intra-VP GVG of heroin self-administration (Xi andtein, 2000). Further experiments will be required to shed lightpon this hypothesis.

In conclusion, the present data – and their preliminary pre-entations in abstract form (Peng et al., 2004, 2006) – show,or the first time, that systemic administration of GVG dose-ependently inhibits cocaine-induced relapse to drug-seeking

2 ohol D

bNtDGsopep

C

faBBrastLtncAiaufifl

A

bI(HtASfN

napsrmGlarP

fia

R

A

B

B

C

C

C

C

D

D

D

G

K

K

K

K

K

K

K

24 X.-Q. Peng et al. / Drug and Alc

ehavior, an effect correlated with an increase in NAc GABA.Ac DA does not appear to be involved in GVG’s inhibi-

ion of cocaine-primed relapse. Given that acute cocaine orA produces an inhibitory effect on GABAergic transmission,VG-induced increases in GABA may functionally antagonize

uch a reduction in GABAergic transmission, thereby antag-nizing cocaine-primed relapse. By whatever mechanism, theresent data add further support to the preclinical animal modelvidence suggesting a potential anti-addiction pharmacothera-eutic utility for GVG.

onflict of interest

Authors Brodie, Dewey, and Ashby are the holders of a patentor the use of GVG as an antiaddiction pharmacotherapeuticgent for humans. All rights to said patent have been assigned torookhaven National Laboratory, under terms whereby authorsrodie, Dewey, and Ashby do not now and will not in the future

eceive any royalties that may be earned from the use of GVG asn anti-addiction agent in humans. Authors Brodie and Deweyerve on the Scientific Advisory Board of Catalyst Pharmaceu-ical Partners, which holds a license from Brookhaven Nationalaboratory for the development of GVG as an anti-addiction

herapeutic agent in humans. Authors Brodie and Dewey haveo fiduciary responsibilities with respect to Catalyst Pharma-eutical Partners, and are not significant stockholders therein.uthors Brodie, Dewey, and Gardner have received de min-

mus honoraria for scholarly presentations (e.g., at universitiesnd medical schools) regarding GVG’s possible anti-addictiontility. The remaining authors declare no financial activities ornancial holdings that could be perceived as constituting a con-ict of interest.

cknowledgements

Role of funding source: Funding for this study was providedy funds from the Intramural Research Program of the Nationalnstitute on Drug Abuse (NIDA), National Institutes of HealthNIH), U.S. Public Health Service (PHS), U.S. Department ofealth and Human Services (DHHS). Preliminary planning of

hese experiments was supported by research grants from thearon Diamond Foundation of New York City and from the Oldtones Foundation of Greenwich, Connecticut, and by fundsrom the Medical and Chemistry Departments of the Brookhavenational Laboratory, Upton, New York.Contributors: Authors Peng, Ashby, Brodie, Dewey, Gard-

er, and Xi designed the study. Authors Peng, Ashby, Gardner,nd Xi carried out the literature searches and summaries ofrevious work. Authors Peng, Li, Gilbert, and Pak carried outurgery on the experimental animals. Authors Peng, Li, and Xian animals in the behavioral paradigms, carried out the in vivoicrodialysis experiments, and collected the data. Authors Peng,ardner, and Xi carried out the statistical analyses of the col-

ected data. Author Peng wrote the first draft of the manuscript,nd authors Peng, Gardner, and Xi contributed to correcting,evising, re-writing, and proof-reading the manuscript. Authorseng, Gardner, and Xi contributed to the drafting, revising, and

L

ependence 97 (2008) 216–225

nalizing of the graphs and figures. All authors contributed tond have approved the final manuscript.

eferences

nderson, S.M., Pierce, R.C., 2005. Cocaine-induced alterations in dopaminereceptor signaling: implications for reinforcement and reinstatement. Phar-macol. Ther. 106, 389–403.

arrett, A.C., Negus, S.S., Mello, N.K., Caine, S.B., 2005. Effect of GABAagonists and GABA-A receptor modulators on cocaine- and food-maintainedresponding and cocaine discrimination in rats. J. Pharmacol. Exp. Ther. 315,858–871.

ennett, B.D., Bolam, J.P., 1994. Synaptic input and output of parvalbumin-immunoreactive neurons in the neostriatum of the rat. Neuroscience 62,707–719.

ameron, D.L., Williams, J.T., 1994. Cocaine inhibits GABA release in the VTAthrough endogenous 5-HT. J. Neurosci. 14, 6763–6767.

entonze, D., Picconi, B., Baunez, C., Borrelli, E., Pisani, A., Bernardi, G.,Calabresi, P., 2002. Cocaine and amphetamine depress striatal GABAergicsynaptic transmission through D2 dopamine receptors. Neuropsychophar-macology 26, 164–175.

han-Palay, V., Wu, J.-Y., Palay, S.L., 1979. Immunocytochemical localiza-tion of gamma-aminobutyric acid transaminase at cellular and ultrastructurallevels. Proc. Natl. Acad. Sci. U.S.A. 76, 2067–2071.

ubells, J.F., Blanchard, J.S., Smith, D.M., Makman, M.H., 1986. In vivo actionof enzyme-activated irreversible inhibitors of glutamic acid decarboxylaseand �-aminobutyric acid transaminase in retina vs. brain. J. Pharmacol. Exp.Ther. 238, 508–514.

e Vries, T.J., Schoffelmeer, A.N.M., Binnekade, R., Raasø, H., Vanderschuren,L.J.M.J., 2002. Relapse to cocaine- and heroin-seeking behavior mediatedby dopamine D2 receptors is time-dependent and associated with behavioralsensitization. Neuropsychopharmacology 26, 18–26.

ewey, S.L., Chaurasia, C.S., Chen, C.-E., Volkow, N.D., Clarkson, F.A.,Porter, S.P., Straughter-Moore, R.M., Alexoff, D.L., Tedeschi, D., Russo,N.B., Fowler, J.S., Brodie, J.D., 1997. GABAergic attenuation ofcocaine-induced dopamine release and locomotor activity. Synapse 25,393–398.

ewey, 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., Brodie, J.D., 1998. Anovel strategy for the treatment of cocaine addiction. Synapse 30, 119–129.

ardner, E.L., Schiffer, W.K., Horan, B.A., Highfield, D., Dewey, S.L., Brodie,J.D., Ashby Jr., C.R., 2002. Gamma-vinyl GABA, an irreversible inhibitorof GABA transaminase, alters the acquisition and expression of cocaine-induced sensitization in male rats. Synapse 46, 240–250.

alivas, P.W., 2004. Glutamate systems in cocaine addiction. Curr. Opin. Phar-macol. 4, 23–29.

alivas, P.W., Volkow, N.D., 2005. The neural basis of addiction: a pathologyof motivation and choice. Am. J. Psychiatry 162, 1403–1413.

ang, T.-C., Park, S.-K., Bahn, J.H., Chang, J.S., Cho, S.-W., Choi, S.Y., Won,M.H., 2001. Comparative studies on the GABA-transaminase immunoreac-tivity in rat and gerbil brains. Mol. Cells 11, 321–325.

oob, G.F., Bloom, F.E., 1988. Cellular and molecular mechanisms of drugdependence. Science 242, 715–723.

uhar, M.J., Pilotte, N.S., 1996. Neurochemical changes in cocaine withdrawal.Trends Pharmacol. Sci. 17, 260–264.

ushner, S.A., Dewey, S.L., Kornestsky, C., 1999. The irreversible �-aminobutyric acid (GABA) transaminase inhibitor �-vinyl-GABA blockscocaine self-administration in rats. J. Pharmacol. Exp. Ther. 290, 797–802.

ushner, S.A., Dewey, S.L., Kornetsky, C., 1997. Gamma-vinyl GABA atten-uates cocaine-induced lowering of brain stimulation reward thresholds.
Psychopharmacology 133, 383–388.
u, L., Grimm, J.W., Shaham, Y., Hope, B.T., 2003. Molecular neuroadaptationsin the accumbens and ventral tegmental area during the first 90 days offorced abstinence from cocaine self-administration in rats. J. Neurochem.85, 1604–1613.

ohol D

M

M

N

N

O

P

P

P

Q

R

R

S

S

S

S

S

S

S

T

U

V

W

W

X

X

X

X


cFarland, K., Lapish, C.C., Kalivas, P.W., 2003. Prefrontal glutamaterelease into the core of the nucleus accumbens mediates cocaine-induced reinstatement of drug-seeking behavior. J. Neurosci. 23,3531–3537.

organ, A.E., Dewey, S.L., 1998. Effects of pharmacologic increases inbrain GABA levels on cocaine-induced changes in extracellular dopamine.Synapse 28, 60–65.

eisewander, J.L., O’Dell, L.E., Tran-Nguyen, L.T.L., Castaneda, E., Fuchs,R.A., 1996. Dopamine overflow in the nucleus accumbens during extinctionand reinstatement of cocaine self-administration behavior. Neuropsy-chopharmacology 15, 506–514.

icola, S.M., Malenka, R.C., 1997. Dopamine depresses excitatory andinhibitory synaptic transmission by distinct mechanisms in the nucleusaccumbens. J. Neurosci. 17, 5697–5710.

’Brien, C.P., Gardner, E.L., 2005. Critical assessment of how to study addictionand its treatment: human and non-human animal models. Pharmacol. Ther.108, 18–58.

axinos, G., Watson, C., 1998. The Rat Brain in Stereotaxic Coordinates, fourthed. Academic Press, San Diego.

eng, X.-Q., Li, X., Gilbert, J., Pak, A., Ashby, C., Xi, Z.-X., Gardner, E.L., 2006.Gamma-vinyl GABA inhibits cocaine-primed relapse by a DA-independentmechanism. In: Abstracts of the 68th Annual Scientific Meeting of the Col-lege on Problems of Drug Dependence, 2006, Phoenix, Arizona (Abstract #616).

eng, X.-Q., Xi, Z.-X., Gilbert, J., Campos, A., Dewey, S.L., Schiffer, W.K.,Brodie, J.D., Ashby, C.R., Gardner, E.L., 2004. Gamma-vinyl GABA, but notgabapentin, inhibits cocaine-triggered reinstatement of drug-seeking behav-ior in the rat. In: Abstracts of the 34th Annual Meeting of the Society forNeuroscience, 2004, San Diego, CA (Abstract # 691.8).

iao, J.-T., Dougherty, P.M., Wiggins, R.C., Dafny, N., 1990. Effects ofmicroiontophoretic application of cocaine, alone and with receptor antago-nists, upon the neurons of the medial prefrontal cortex, nucleus accumbensand caudate nucleus of rats. Neuropharmacology 29, 379–385.

eijnierse, G.L., Veldstra, H., Van den Berg, C.J., 1975. Subcellular localizationof �-aminobutyrate transaminase and glutamate dehydrogenase in adult ratbrain. Evidence for at least two small glutamate compartments in brain.Biochem. J. 152, 469–475.

oberts, D.C.S., Brebner, K., 2000. GABA modulation of cocaine self-administration. Ann. N. Y. Acad. Sci. 909, 145–158.

chiffer, W.K., Gerasimov, M.R., Bermel, R.A., Brodie, J.D., Dewey, S.L., 2000.Stereoselective inhibition of dopaminergic activity by gamma vinyl-GABA
following a nicotine or cocaine challenge: a PET/microdialysis study. LifeSci. [Pharmacol. Lett.] 66, PL169–PL173.
elf, D.W., Barnhart, W.J., Lehman, D.A., Nestler, E.J., 1996. Opposite mod-ulation of cocaine-seeking behavior by D1- and D2-like dopamine receptoragonists. Science 271, 1586–1589.

X

X

ependence 97 (2008) 216–225 225

elf, D.W., Nestler, E.J., 1995. Molecular mechanisms of drug reinforcementand addiction. Annu. Rev. Neurosci. 18, 463–495.

esack, S.R., Pickel, V.M., 1992. Prefrontal cortical efferents in the rat synapseon unlabeled neuronal targets of catecholamine terminals in the nucleusaccumbens septi and on dopamine neurons in the ventral tegmental area. J.Comp. Neurol. 320, 145–160.

halev, U., Grimm, J.W., Shaham, Y., 2002. Neurobiology of relapse to heroinand cocaine seeking: a review. Pharmacol. Rev. 54, 1–42.

mith, A.D., Bolam, J.P., 1990. The neural network of the basal ganglia asrevealed by the study of synaptic connections of identified neurones. TrendsNeurosci. 13, 259–265.

taley, J.K., Mash, D.C., 1996. Adaptive increase in D3 dopamine receptorsin the brain reward circuits of human cocaine fatalities. J. Neurosci. 16,6100–6106.

ang, X.-C., McFarland, K., Cagle, S., Kalivas, P.W., 2005. Cocaine-inducedreinstatement requires endogenous stimulation of �-opioid receptors in theventral pallidum. J. Neurosci. 25, 4512–4520.

chimura, N., North, R.A., 1990. Actions of cocaine on rat nucleus accumbensneurones in vitro. Br. J. Pharmacol. 99, 736–740.

orel, S.R., Ashby Jr., C.R., Paul, M., Liu, X., Hayes, R., Hagan, J.J., Mid-dlemiss, D.N., Stemp, G., Gardner, E.L., 2002. Dopamine D3 receptorantagonism inhibits cocaine-seeking and cocaine-enhanced brain reward inrats. J. Neurosci. 22, 9595–9603.

hite, F.J., Hu, X.-T., Henry, D.J., 1993. Electrophysiological effects of cocainein the rat nucleus accumbens: microiontophoretic studies. J. Pharmacol. Exp.Ther. 266, 1075–1084.

ise, R.A., 1998. Drug-activation of brain reward pathways. Drug AlcoholDepend. 51, 13–22.

i, Z.-X., Gilbert, J.G., Peng, X.-Q., Pak, A.C., Li, X., Gardner, E.L., 2006.Cannabinoid CB1 receptor antagonist AM251 inhibits cocaine-primedrelapse in rats: role of glutamate in the nucleus accumbens. J. Neurosci.26, 8531–8636.

i, Z.-X., Ramamoorthy, S., Shen, H., Lake, R., Samuvel, D.J., Kali-vas, P.W., 2003. GABA transmission in the nucleus accumbens isaltered after withdrawal from repeated cocaine. J. Neurosci. 23, 3498–3505.

i, Z.-X., Stein, E.A., 1998. Nucleus accumbens dopamine release modulationby mesolimbic GABAA receptors—an in vivo electrochemical study. BrainRes. 798, 156–165.

i, Z.-X., Stein, E.A., 1999. Baclofen inhibits heroin self-administration behav-ior and mesolimbic dopamine release. J. Pharmacol. Exp. Ther. 290,
1369–1374.
i, Z.-X., Stein, E.A., 2000. Increased mesolimbic GABA concentration blocksheroin self-administration in the rat. J. Pharmacol. Exp. Ther. 294, 613–619.

i, Z.-X., Stein, E.A., 2002. GABAergic mechanisms of opiate reinforcement.Alcohol Alcohol 37, 485–494.

Gamma-vinyl GABA inhibits cocaine-triggered reinstatement of drug-seeking behavior in rats by a non-dopaminergic mechanism

Documents