Cocaine-induced endocannabinoid release modulates behavioral and neurochemical sensitization in mice

Cocaine-induced endocannabinoid release modulatesbehavioral and neurochemical sensitization in mice

Maddalena Mereu*, Valeria Tronci*, Lauren E. Chun, Alexandra M. Thomas, Jennifer L. Green,Jonathan L. Katz & Gianluigi TandaPsychobiology Section, Molecular Targets & Medications Discovery Branch, Department of Health and Human Services, National Institute on Drug Abuse,National Institutes of Health, Baltimore, MD, USA

ABSTRACT

The endocannabinoid system has been implicated in the development of synaptic plasticity induced by several drugsabused by humans, including cocaine. However, there remains some debate about the involvement of cannabinoidreceptors/ligands in cocaine-induced plasticity and corresponding behavioral actions. Here, we show that a singlecocaine injection in Swiss-Webster mice produces behavioral and neurochemical alterations that are under thecontrol of the endocannabinoid system. This plasticity may be the initial basis for changes in brain processes leadingfrom recreational use of cocaine to its abuse and ultimately to dependence. Locomotor activity was monitored withphotobeam cell detectors, and accumbens shell/core microdialysate dopamine levels were monitored by high-performance liquid chromatography with electrochemical detection. Development of single-trial cocaine-inducedbehavioral sensitization, measured as increased distance traveled in sensitized mice compared to control mice, wasparalleled by a larger stimulation of extracellular dopamine levels in the core but not the shell of the nucleus accumbens.Both the behavioral and neurochemical effects were reversed by CB1 receptor blockade produced by rimonabantpre-treatments. Further, both behavioral and neurochemical cocaine sensitization were facilitated by pharmacologicalblockade of endocannabinoid metabolism, achieved by inhibiting the fatty acid amide hydrolase enzyme. In conclusion,our results suggest that a single unconditioned exposure to cocaine produces sensitization through neuronal alterationsthat require regionally specific release of endocannabinoids. Further, the present results suggest that endocannabinoidsplay a primary role from the earliest stage of cocaine use, mediating the inception of long-term brain-adaptive responses,shaping central pathways and likely increasing vulnerability to stimulant abuse disorders.

Keywords Behavioral sensitization, cannabinoid CB1 receptors, cocaine addiction, endocannabinoids, in vivodopamine microdialysis, nucleus accumbens.

Correspondence to: Gianluigi Tanda, Psychobiology Section, Molecular Targets & Medications Discovery Branch, Department of Health andHuman Services, National Institute on Drug Abuse, National Institutes of Health, 251 Bayview Blvd, Baltimore, MD 21224, USA. E-mail:[email protected]

INTRODUCTION

Repeated psychostimulant administration (Kalivas &Duffy 1993; Cadoni, Solinas & Di Chiara 2000), as wellas even a single exposure to a psychostimulant, canproduce long-lasting sensitization to acute behavioralstimulant effects in rodents (Ungless et al. 2001; Saalet al. 2003). Long-term neurobiological adaptations such

as increasing strengths of excitatory synapses in mid-brain dopaminergic regions are thought to play animportant role in various sensitized behavioral effects,including those related to the development of drugcraving and dependence (Robinson & Berridge 1993;Kauer & Malenka 2007; Bowers, Chen & Bonci 2010;Pierce & Vanderschuren 2010). Among the many neuro-transmitter systems implicated in drug abuse and

*These authors contributed equally to the present manuscript.Present address, LEC: Department of Psychology & Neuroscience, University of Colorado at Boulder, Boulder, CO; AMT: YaleM.D.-Ph.D. Program, Yale School of Medicine, New Haven, CT; JLG: Department of Psychology, University of North Carolina, ChapelHill, NC.

PRECLINICAL STUDY

bs_bs_bannerAddiction Biologydoi:10.1111/adb.12080

Published 2013. This article is a U.S. Government work and is in the public domain in the USA. Addiction Biology, 20, 91–103

dependence, recent evidence has suggested the involve-ment of the endocannabinoid system in stimulantsensitization, even though there is still a debate about itsprecise role as well as for its potential effects related tostimulant abuse and dependence (as reviewed by Arnold2005; Tanda 2007; Wiskerke et al. 2008). For instance,genetic deletion of cannabinoid CB1 receptors in mice(Soria et al. 2005) or pharmacological blockade of CB1receptors decreases the reinforcing effects of cocaine(Soria et al. 2005; but see also Cossu et al. 2001),cocaine-induced stimulation of nucleus accumbens(NAC), dopamine (DA) concentrations measured bymicrodialysis or voltammetry (Cheer et al. 2007; Li et al.2009; but see Xi et al. 2006) and the development ofcocaine-induced neurobiological adaptations (Filip et al.2006; Corbille et al. 2007; Gerdeman, Schechter &French 2008; but see Lesscher et al. 2005).

Previous studies have reported that psychostimulants,such as cocaine or amphetamine, as well as dopamine D1or D2-receptor activation may induce DA-dependentrelease of endocannabinoids in selected brain regions(Giuffrida et al. 1999; Ferrer et al. 2003; Centonze et al.2004; Thiemann et al. 2008). Thus, endocannabinoids,acting as retrograde synaptic messengers (Wilson &Nicoll 2002), may be involved in functional effects ofcocaine, including those related to synaptic plasticity.Moreover, an increase in endocannabinoid levels mightfacilitate immediate, subsecond, acute effects of cocaine(Cheer et al. 2007). Others, however, have not foundincreased levels of the endocannabinoid anandamide inseveral mesocorticolimbic regions after chronic cocaineadministration in rats (Gonzalez et al. 2002; Caille et al.2007).

In the present study, we tested the hypothesis that CB1receptors, activated by cocaine-induced endocanna-binoid release in mice (Centonze et al. 2004), might beinvolved in the very early stages of neuronal adaptationsunderlying development of behavioral sensitization. Tofocus on early stages of the process, the effects of a singlecocaine injection was assessed. Under the same experi-mental conditions, we also investigated how enhance-ment of endocannabinoid release, by blockade ofendocannabinoid metabolic enzymatic pathways, mightinfluence the development of behavioral sensitizationwhen cocaine is administered at suboptimal doses.Finally, the NAC shell and core are main terminal areas ofmesolimbic DA neurons and have documented criticalroles in cocaine reinforcement (Di Chiara 2002; Koob &Volkow 2010) that are involved from the initial exposureto drug (Pontieri, Tanda & Di Chiara 1995; Di Chiaraet al. 1998, 1999). Therefore, we also investigated theinvolvement of these areas in cocaine sensitization andthe involvement of the endocannabinoid system inmodulating DA activity in these regions.

MATERIALS AND METHODS

Subjects

Male Swiss-Webster mice (Taconic, Germantown, NY,USA), weighing 30–40 g, were housed in groups of fourper cage in a large temperature and humidity controlledcolony room, under a 12-hour light/dark cycle (lights onat 7:00 a.m.). Mice had free access to food and water,except during the behavioral and neurochemical tests.

Drugs and injections

The CB1 receptor antagonist, rimonabant (SR 141716A,abbreviated as RIM in the figures, N-(piperidin-1-yl)-5-(a-chlorophenyl)-1-(2,4-dichloro-phenyl)-4-methyl1-1H-pyrazole-3-carboxyamide HCl, NIDA Drug SupplyProgram, Bethesda, MD, USA) was dissolved in a vehiclecomprised of 2% Tween 80, 2% alcohol and saline.Cocaine HCl (Sigma Chemicals, St. Louis, MO, USA orNIDA Drug Supply Program) was dissolved in saline (0.9%NaCl). URB597 was provided by Dr. Daniele Piomelli,Department of Pharmacology, University of California,Irvine, CA, USA. It was dissolved in a vehicle comprised of2% Tween 80, 2% alcohol and saline. All injections weredelivered intraperitoneally (i.p.) (0.1 ml/10 g), and therespective vehicles served as control injections. Drug dosesand pre-treatment times are from a preliminary studypresented in abstract form (Tronci & Tanda 2007).

Study design

The main dependent measure was the amount of loco-motor activity or the change in dopamine concentrationinduced by cocaine on day 3 after various sensitizingtreatments on day 2. On day 1, all mice in all groups(behavioral and microdialysis) were moved from theanimal vivarium to the testing rooms, and after about 30minutes were anesthetized and placed in their testingcages, which housed them for the remainder of the study(see Fig. 1). Some control groups were administeredsaline on day 1 as a control for the injection of anestheticin other groups. No significant differences in cocaine-induced sensitization in locomotor activity measureswere found between the anaesthetized and saline controlgroups, and for that reason the data from groups receiv-ing saline on day 1 have been included with data from thecorresponding groups that received anesthesia on day 1.

On day 2, vehicle or cocaine was injected followingeither rimonabant or URB597 treatment. Injections onday 2 were administered 24 hours before test doses ofcocaine on day 3. All injections and testing were con-ducted during the light period (lights on at 7 am andoff at 7 pm) of the light-dark cycle. All behavioral andmicrodialysis tests were performed on day 3.

92 Maddalena Mereu et al.


Behavioral experiments

Horizontal locomotor activity (expressed as distancetraveled, cm) was measured using monitors (OmnitechElectronics Inc., Columbus, OH, USA) with a horizontalmatrix of 16 photobeams, fixed at a height of 2 cm andspaced at 5-cm intervals along the interior perimeter.Eight beams per side crossed the test chamber (transpar-ent acrylic box, 40 × 40 × 30 cm high), which waslocated within the monitor.

Effects of cannabinoid CB1 receptor blockade ondevelopment of single-trial behavioralsensitization to cocaine

On day 2, between 20 and 22 hours after their transfer tothe testing room and placement in the testing cage, micewere divided into groups that were injected with eithervehicle or rimonabant (0.3, 1.0 or 3.0 mg/kg, i.p.). Eachof these four groups was further divided into two groupsand injected 40 minutes later with saline or cocaine(20 mg/kg). On day 3 (24 hours after the cocaine orsaline injections), each group was further divided intogroups (n = 6) that were injected with cocaine (10 or20 mg/kg). Immediately after this last injection, locomo-tor activity measured by distance traveled was monitoredfor 60 minutes. Thus, this experiment assessed the devel-opment of behavioral sensitization 24 hours after a singlecocaine dose, and the potential of increasing doses ofrimonabant to antagonize the sensitization.

Effects of fatty acid amide hydrolase (FAAH) blockade ondevelopment of cocaine-induced behavioral sensitization

On day 2, between 20 and 22 hours after their transfer tothe testing room and placement in the testing cage, micewere divided into groups that were injected with eithervehicle or URB597 (1.0 or 3.0 mg/kg, i.p.). Each of thesethree groups was further divided into two groups andinjected 60 minutes later with saline or cocaine (10 mg/kg). On day 3 (24 hours after the cocaine or saline injec-tions), subjects in each group (n = 6) were injected withcocaine (20 mg/kg). Immediately after this last injection,locomotor activity was monitored for 60 minutes. Thus,

this experiment assessed the development of behavioralsensitization 24 hours after a combination of URB597and a cocaine dose that was insufficient to produce sig-nificant behavioral sensitization in order to assess thepotential of URB597 to potentiate the development ofcocaine sensitization.

Statistical analysis

Locomotor activity data were expressed as the averagedistance traveled, cm ± SEM, obtained for each experi-mental group of animals during the first 30 minutes afterinjection. Results are expressed as group means (±SEM).Data were analyzed by one- or two-way analysis ofvariance (ANOVA) with treatment and pre-treatmentas factors. Significant changes were subjected toBonferroni’s or Tukey’s post hoc tests. Significant differ-ences from control groups, P < 0.05, are represented inthe figures with a * or # mark. All results of statisticalanalyses are shown in the figure legends in order to makethe text clear and easy to follow.

Microdialysis experiments

Concentric dialysis probes were prepared with AN69fibers (Hospal Dasco, Bologna, Italy) as described previ-ously (Tanda et al. 2009; Loland et al. 2012). Briefly, two4-cm pieces of silica-fused capillary tubes (the inlet andoutlet tubing of the probes) were inserted into a 4-mmcapillary dialyzing fiber (closed by a drop of glue on theother side), with the inlet tubing set at about 0.1 mm fromthe closed, glued end of the fiber and the outlet set at 1 mmfrom the inlet tip. Note that the glue tip was cone-shaped sothat implanting the probe with a continuous infusionwould produce minimal damage to the surroundingtissue. The open end of the dialysis membrane was thenglued and the protruding two silica-fused pieces of tubingwere inserted and glued into a 22-gauge stainless steelneedle (1.7-mm length). The needle was then clipped to aCMA/10 clip (CMA Microdialysis AB, Solna, Sweden) andmounted in a stereotaxic holder. The exposed dialyzingsurface of the fiber (i.e. the portion not covered by glue)was limited to the lowest 1.0 mm portion of the probe.

day 1day 1

Mice are moved from thecolony to the locomotoractivity testing room andplaced in their testing cage.

RIM Cocaine or saline

CB1-R antagonism blocks cocaine-induced sensitization

Blockade of FAAH facilitates cocaine-induced sensitization

Cocaine

Cocaine

LMA ordialysis test

LMA ordialysis test

40 minutes

60 minutesURB597 Cocaine or saline

Mice received a dose ofanesthetic whetherimplanted with dialysisprobes or not.

day 2 day 3

Figure 1 Timeline of treatments forbehavioral and neurochemical tests

CB1R and cocaine sensitization 93


Surgery

On day 1, mice were anesthetized under a mixture ofxylazine (10 mg/kg, i.p.) and ketamine (60 mg/kg, i.p.)and a concentric dialysis probe was implanted (seeFig. 1). Mice were placed in a stereotaxic apparatus, theskull was exposed, and a small hole was drilled to exposethe dura. Mice were then randomly implanted in the rightor left side with a concentric dialysis probe aimed at theNAC shell or core (shell: A = +1.5, L = ± 0.6, V = −5.1;core: A = +1.3, L = ± 1.3, V = −4.9; anterior, A, lateral, L,vertical, V, in millimeters from bregma), according to themouse brain atlas by Paxinos & Franklin (2001). Figure 2shows drawings of forebrain sections with the bounda-ries (superimposed rectangles) within which micro-dialysis probe tracks were considered correct probeplacements. After surgery, mice were allowed to recoverin square Plexiglas cages with bedding on the floor, thatwere equipped with overhead quartz-lined fluid swivels(Instech Laboratories Inc., Plymouth Meeting, PA, USA)for connections to the dialysis probes.

Effects of cannabinoid CB1 receptor blockade oncocaine-induced stimulation of DA extracellular levels insensitized and non-sensitized mice

On day 2, treatments were administered starting about18–20 hours after probe implant (see Fig. 1). Mice weredivided into different groups that were injected witheither vehicle or rimonabant (1.0 or 3.0 mg/kg, i.p.).Each of these groups was further divided into two groupsand injected 40 minutes later with saline or cocaine(20 mg/kg). Microdialysis sampling was initiated on day3 (approximately 22–22.5 hours after day 2 treatments).Microdialysis probes were perfused with Ringer’s solution(147.0 mM NaCl, 2.2 mM CaCl2, and 4.0 mM KCl),delivered by a 1.0-ml syringe and operated by a BAS BeeSyringe Pump Controller (BAS Bioanalytical Systems,West Lafayette, IN, USA) at a constant flow rate of 1 μl/min. Collection of dialysate samples (10 μl) started after30 minutes, and samples were taken every 10 minutesand immediately analyzed, as detailed below. Cocaine(20 mg/kg) was injected on day 3 when stable DAvalues (less than 10% variability) were obtained for atleast three consecutive samples (typically after about 1hour). Dialysis sampling continued once every 10minutes for 2 hours.

Effects of inhibition of the FAAH enzyme oncocaine-induced stimulation of DA extracellular levels in theNAC core in sensitized and non-sensitized mice

As above, on day 2, treatments were administered start-ing about 18–20 hours after probe implant. Mice weredivided into different groups that were injected witheither vehicle or URB597 (1 or 3.0 mg/kg, i.p.), and

1 hour later with saline or cocaine (10 mg/kg, i.p.). TheURB597 experiment was carried out only from NACcore dialysates, as no instances of cocaine-inducedsensitization to changes in DA levels were previouslyfound in the NAC shell. Microdialysis sampling was initi-ated on day 3 as described above.

Analytical procedure

Dialysate samples (10 μl) were injected without purifica-tion into a high-performance liquid chromatographyapparatus equipped with an MD 150 × 3.2-mm column,particle size 3.0 μm (ESA Inc., Chelmsford, MA, USA)and a coulometric detector (5200a Coulochem II orcoulochem III; ESA Inc.) to quantify DA. The oxidationand reduction electrodes of the analytical cell (5014B;ESA Inc.) were set at +125 and −175 mV, respectively.The mobile phase, containing 100 mM NaH2PO4,0.1 mM Na2 ethylenediaminetetraacetic acid, 0.5 mMNa-octyl sulfate and 18% (v/v) methanol (pH adjusted to5.5 with Na2HPO4) was pumped by an ESA 582 (ESAInc.) solvent delivery module at 0.50 ml/min. Assaysensitivity for DA was 2 fmol/sample.

Histology

Histology was performed in accordance with previousstudies (Tanda et al. 2009). Brains were cut on avibratome in serial coronal slices oriented according tothe atlas by Paxinos and Franklin (2001) to identify thelocation of the probes. Only the experiments in which theprobes were appropriately located inside the NAC shell orcore boundaries have been considered and used for theDA microdialysis, results shown in the present study(see Fig. 2).

Statistical analysis

Results were expressed as a percentage of basal DAvalues, calculated as the mean of the last three to fourconsecutive samples (differing no more than 15%) imme-diately preceding the first drug or vehicle injection. Allresults are presented as group means (±SEM). Differencesin basal levels of DA between different experimentalgroups within the same brain area, or between differentbrain areas, were analyzed by one-way ANOVA. Statisti-cal analyses of experimental data were carried out withStatistica-6 software (StatSoft, Tulsa, OK, USA) using atwo-way ANOVA (drug dose and time as factors) or three-way ANOVA (drug dose for pre-treatment, drug dose fortreatment, and time) for repeated measures over time,with results from treatments showing overall changessubjected to post hoc Tukey’s test. Results were considered



significant at P < 0.05. All results of statistical analysesare shown in the figure legends in order to make the textclear and easy to follow.

RESULTS

Mice previously treated 24 hours prior with cocaine(20 mg/kg) displayed a robust and significant increase indistance traveled after the second injection of cocaine (10or 20 mg/kg) compared to those treated 24 hours priorwith saline (P < 0.05; Fig. 3a1 and 3b1; compare whiteto black bars).

Increasing doses of the CB1 receptor antagonist,rimonabant (0.3–3.0 mg/kg), administered 40 minutesbefore the otherwise sensitizing dose of cocaine (20 mg/kg, i.p.) on day 2, dose dependently (P < 0.05) attenuated(Fig. 3a2 and 3b2) the sensitized response to cocaine (10or 20 mg/kg, i.p.). In mice that did not receive a sensitiz-ing dose of cocaine (Fig. 3a3 and 3b3), rimonabantadministration, 40 minutes before saline pre-treatmentson day 2, did not significantly modify the stimulation oflocomotor activity produced by either 10 (P > 0.05), or20 (P > 0.05) mg/kg of cocaine administered 24 hourslater. As shown in the Supporting Information Fig. S1,rimonabant pre-treatment failed to significantly modifythe non-sensitized stimulation of locomotor activityinduced by a single acute i.p. injection of 10 or 20 mg/kgof cocaine (P > 0.05). Thus, it appears that under thepresent experimental conditions, rimonabant affects thesensitized response to cocaine but, across the same rangeof doses, not its acute stimulatory effects. In general, theeffects of rimonabant were statistically significant,although there was an insignificant trend for rimonabant0.3 mg/kg to attenuate the ambulatory activity inducedby 20 mg/kg of cocaine more than that induced by10 mg/kg group (see Fig. 3, a2 and b2).

In vivo microdialysis studies of DA levels in theNAC core indicated that pre-exposure to cocaine (onday 2) increased the response to cocaine when adminis-tered 24 hours later (Fig. 4a and b, compare filled andopen circles in both panels). The stimulated levels ofNAC core DA produced by cocaine (at either 10 or20 mg/kg, i.p.) were greater in mice pre-exposed tococaine (20 mg/kg, i.p., 24 hours before) than thosepre-exposed to saline under the same experimental con-ditions (P < 0.05).

The sensitized DA stimulation obtained in the NACcore was reduced to saline pre-treatment control levelswhen mice received rimonabant (1 or 3 mg/kg, i.p.) onday 2 40 minutes before the sensitizing dose of cocaine(Fig. 4a and b, compare open squares or triangles to opencircles in both panels). As with locomotor activity,rimonabant did not significantly alter the effects of acute

Figure 2 Brain microdialysis probe placements in the NAC shell orcore. Forebrain sections, redrawn from Paxinos & Franklin (2001),showing the limits of the positions of the dialyzing portions of themicrodialysis probes (superimposed rectangles). On each section, theanterior coordinate (measured from bregma) is indicated



administration of cocaine on stimulation of DA levels inthe NAC core (Supporting Information Fig. S2) in micethat were not sensitized by prior cocaine treatment(P > 0.05).

A sensitizing dose of cocaine (20 mg/kg, i.p.) on day 2did not significantly alter the stimulation of DA levels inthe NAC shell on day 3 induced by either 10 or 20 mg/kgcocaine (P > 0.05) (Fig. 4, compare filled and open circlesin panels c and d). Further, no substantial change(P > 0.05) in the dopaminergic response to cocaine in theNAC shell was produced by rimonabant pre-treatment(1.0 mg/kg, i.p.) 40 minutes before cocaine injection onday 2 (Fig. 4; compare open squares to other points inpanels c and d). That dose of rimonabant decreased thecocaine-induced sensitized locomotor response anddopaminergic response in NAC core. Also, as shown inthe Supporting Information, rimonabant administered40 minutes before cocaine (10 or 20 mg/kg, i.p.) failed tosignificantly change the stimulation of DA levels in theNAC shell after the acute injection of cocaine (P > 0.05)(Supporting Information Fig. S2).

The blockade of behavioral and neurochemicalcocaine sensitization by rimonabant suggests that thesensitization requires activation of CB1 receptors, whichcan be the result of cocaine-induced on-demand releaseof endocannabinoids (Centonze et al. 2004; Cheer et al.2007). That hypothesis was tested by pre-treating agroup of mice with a lower dose of cocaine (10 mg/kg, i.p.), which by itself did not induce behavioralsensitization (Fig. 5) possibly because it did not releaseendocannabinoids at levels sufficient to activate CB1receptors. The hypothesis predicts that the effects of a lessthan fully efficacious dose of cocaine will be enhancedby blocking endocannabinoid metabolism (Solinas et al.2006, 2010).

To test this hypothesis, a group of mice receivedon day 2 the lower (10 mg/kg, i.p.) dose of cocaine 60minutes after pre-treatment with a FAAH inhibitor,URB597 (1 and 3 mg/kg, i.p.), which blocks the metabo-lism of the endocannabinoid, anandamide (Kathuriaet al. 2003). In these mice, 10 mg/kg cocaine asexpected, did not sensitize subjects to a 20 mg/kg dose ofcocaine administered on day 3 (Fig. 5a, compare whiteand black bars). URB597 when administered beforevehicle on day 2 at either 1.0 or 3.0 mg/kg did notchange the response to cocaine on day 3 (Fig. 5a, greybars). However, URB597 at both 1.0 and 3.0 mg/kgenhanced the effect of the 10 mg/kg dose of cocaine onday 2 such that sensitization was evident on day 3 (Fig. 5,stippled bars). The effect of the challenge dose of cocaine(20 mg/kg) in mice that received URB597 with 10 mg/kg of cocaine was substantially and significantly greaterthan that produced when that dose of cocaine waspreceded by vehicle injection (Fig. 5, top panel).

Figure 3 Cocaine-induced single trial behavioral sensitization isblunted by antagonism at CB1 receptors. Locomotor activityresponse to an acute challenge with cocaine (panel a, 10 mg/kg, orpanel b, 20 mg/kg) in mice pre-treated 24 hours and 40 minutesbefore the cocaine challenge with vehicle [A two-way analysis ofvariance (ANOVA) yielded significant main effects of cocaine pre-exposure, F(1,67) = 11.235, P < 0.002, and cocaine treatment,F(2,67) = 7.527, P < 0.002, and a non-significant effect of their inter-action: F(2,67) = 2.682, P = 0.075] or increasing doses of rimonabant(RIM), 0.3 (hatched bars), 1.0 (light grey filled bars), and 3 mg/kg, i.p.,(dark grey filled bars), followed 40 minutes later with either saline(unfilled bars) or cocaine 20 mg/kg, i.p. (solid black bars). [one-wayANOVA of the effects of RIM 0.3–3.0 mg/kg pre-treatments wassignificant at both the cocaine challenges, 10 mg/kg F(3,25) = 3.689,P < 0.005), and 20 mg/kg F(3,32) = 3.042, P < 0.005]. The results areexpressed as the average distance traveled (cm/30 minutes) ±SEM obtained for each experimental group of animals during thefirst 30 minutes after injection. Significant differences from controlgroups,* = P < 0.05 versus saline, or # = P < 0.05 versus cocaine



Figure 4 Cocaine-induced single trial behavioral sensitization is accompanied by increased stimulation of NAC core, but not NAC shell, DAtransmission, which is blocked by antagonism at CB1 receptors.The figure shows the time course of the effects of systemic administration ofcocaine 10 mg/kg (top panels, a, c), or 20 mg/kg (lower panels, b,d) on extracellular levels of DA in dialysates from the NAC core (left panels)or shell (right panels) in control animals (•, number of mice and basal levels of extracellular DA expressed as fmoles/10 μl sample ± SEM were:n = 5, 41.1 ± 3.9, and n = 5, 42.9 ± 4.9 fmoles/sample, for cocaine 10 and 20 mg/kg, respectively, in the core; n = 6, 57.9 ± 6.7 fmoles/sample, andn = 9, 48.4 ± 4.7 fmoles/sample, for cocaine 10 and 20 mg/kg, respectively in the shell), or in animals sensitized to cocaine (○, n = 7, 37.4 ± 2.0fmoles/sample, and n = 6, 30 ± 3.4 fmoles/sample, for cocaine 10 and 20 mg/kg, respectively, in the core; n = 9, 48.5 ± 3.0 fmoles/sample, andn = 8, 51.3 ± 4.2 fmoles/sample, for cocaine 10 and 20 mg/kg, respectively in the shell), in animals pre-treated with RIM 1 mg/kg and cocaine20 mg/kg (□, n = 6, 52.4 ± 3.6 fmoles/sample, and n = 6, 51.8 ± 4.4 fmoles/sample, for cocaine 10 and 20 mg/kg, respectively, in the core; n = 9,44.2 ± 3.7 fmoles/sample, and n = 6, 29.1 ± 3.1 fmoles/sample, for cocaine 10 and 20 mg/kg, respectively in the shell) and in animals pre-treatedwith RIM 3 mg/kg and cocaine (△, n = 6, 59.1 ± 5.7 fmoles/sample, and n = 6, 52.1 ± 9.1 fmoles/sample, for cocaine 10 and 20 mg/kg, respectively,in the core) or pre-treated with RIM 3 mg/kg and saline (Ë, n = 4, 35.0 ± 7.7 fmoles/sample, for cocaine 20 mg/kg in the core) 24 hours beforemicrodialysis tests. [three-way analysis of variance (ANOVA) (RIM pre-treatment × cocaine treatment × time) showed in the core significantmain effects of cocaine dose F(1,42) = 15.296, P < 0.001, RIM pre-treatment F(3, 42) = 3.168, P < 0.05, and time F(6,252) = 91.066, P < 0.001;also, there were significant interactions of cocaine dose × time F(18,252) = 19.735, P < 0.001, RIM pre-treatment × time F(18,252) = 2.297,P < 0.05; non-significant interactions of RIM pre-treatment × cocaine dose F(3.42) = 0.518, NS; and pre-treatment × cocaine dose × timeinteraction F = (18,252) = 0.486, NS. Three-way ANOVA in the shell showed significant main effects of cocaine dose F(1,36) = 24.057,P < 0.001, and time F(12,432) = 101.088, P < 0.001, and a non-significant main effect of RIM pre-treatment F(2, 36) = 0.139; also, there weresignificant interactions of cocaine dose × time F(12,432) = 19.735, P < 0.001, non-significant interactions of pre-treatment × cocaine doseF(2,36) = 0.872; pre-treatment × time F(24,432) = 1.007; and pre-treatment × cocaine dose × time interaction F = (24,432) = 0.594]. Results aremeans with vertical bars representing SEM of the amount of DA in 10 minutes dialysate samples, expressed as percentage of basal values



URB597 administered 60 minutes before cocaine(10 mg/kg, i.p.) on day 2 also enhanced the effects on DAlevels of 20 mg/kg of cocaine administered on day 3(Fig. 5b, compare open squares to filled circles). Asexpected, 10 mg/kg of cocaine 60 minutes after a vehicleinjection on day 2, did not sensitize subjects to a day 3injection of 20 mg/kg of cocaine (Fig. 5b, compare openand filled circles). Similarly, URB597 administered on day2 with a vehicle injection 60 minutes later had no effecton the response to 20 mg/kg of cocaine on day 3 (Fig. 5b,compare filled squares and circles).

DISCUSSION

The results in the present study show that a context-independent behavioral sensitization that occurs in miceafter a single exposure to cocaine (Ungless et al. 2001)can be abolished by the CB1 receptor antagonist,rimonabant. Our results also indicate that a similarneurochemical sensitization, likely related to the

behavioral sensitization, occurs in the core but not in theshell of the NAC. Such sensitization in DA response in theNAC core is also attenuated by blockade of CB1 receptorswith rimonabant. Finally, the cocaine-sensitizingbehavioral and neurochemical effects were replicatedwith a small, non-sensitizing dose of cocaine under con-ditions in which the main metabolizing route for theendocannabinoid anandamide, FAAH, has been blockedby administering the FAAH-inhibitor URB597 (Solinaset al. 2006). It is interesting to notice that in agreementwith previous studies in rats (Pontieri et al. 1995; Cadoniet al. 2000; Tanda et al. 2005), this study in mice shows asignificant, larger stimulation of DA levels in the shell ascompared to the core after acute cocaine administration.Also in agreement with Cadoni et al. (2000), such signifi-cance is lost after cocaine-induced sensitization. The dif-ferences in DA response in the NAC shell and core withacute cocaine appear at variance with those by Zocchiet al. (2003). In that report, subjects were CD1 mice,implanted with a large guide cannulae less than 24 hours

Figure 5 Blockade of the FAAH enzyme by URB597 facilitates theoccurrence of cocaine-induced single trial behavioral and NAC coreneurochemical sensitization in animals pre-treated with a non-sensitizing dose of cocaine.This figure shows on panel (a) that micepre-treated with a 10 mg/kg dose of cocaine 24 hours in advance, andthen tested with a dose of cocaine, 20 mg/kg, i.p., did not displaybehavioral sensitization, measured as increased distance traveled, ascompared with mice that were pre-treated with saline 24 hoursbefore [one-way analysis of variance (ANOVA) shows NS effectof sensitization treatment, F(1,14) = 0.029, P = 0.866]. However,increased locomotor activity, indicating behavioral sensitization, wasobserved in mice tested with cocaine 20 mg/kg, i.p. when theyreceived a 24 hours pre-treatment with cocaine 10 mg/kg that waspreceded by administration 1 hour pre-treatment with either 1 or3 mg/kg i.p. doses of URB597 (which are known to block the FAAH)[two-way ANOVA shows significant main effects of URB pre-treatment, F(2,52) = 3.326, P < 0.05, cocaine treatment, F(1,52) =7.586, P < 0.01, and their interaction, F(2,52) = 3.163, P < 0.05]. Panelb shows dopamine microdialysis data from mice implanted withprobes in the NAC core that, in agreement with locomotor activitydata, display a larger DAergic response in this brain area, indicatingsensitization, obtained in animals pre-treated with the FAAH blocker,URB597 (URB) 3 mg/kg i.p. administered 1 hour before a non-sensitizing dose of cocaine, 24 hours before the testing dose ofcocaine, 20 mg/kg i.p., compared to animals that received URB vehiclepre-treatments (pre-treatment groups’ number of animals,n, and basalDA values, fmoles/sample ± SEM, were: Veh + Sal, •, n = 4, 63.18 ±8.6; Veh + cocaine 10 mg/kg, ○, n = 8, 53.2 + 5.0; URB + Sal,Ë, n = 5,64.7 ± 3.9; URB + cocaine 10 mg/kg, □, n = 6, 52.4 ± 7.9) [two-way

ANOVA for repeated measures shows significant main effects of URBpre-treatment, F(1,27) = 4.556, P < 0.05, and time, F(3,81) = 76.781,P < 0.001, and significant URB pre-treatment by time interaction,F(3,81) = 2.847, P < 0.05]. In microdialysis experiments, results areexpressed as means with vertical bars representing SEM of theamount of DA in 10 minutes dialysate samples, expressed as percent-age of basal values. * = P < 0.05, Tukey’s post-hoc test◀



before the tests, compared to our Swiss-Webster miceimplanted about 42–46 hours before test with a thin,likely more biocompatible probe (no guide cannulae, nometal and only the dialyzing membrane with silica fusedtubing in contact with brain tissue). Zocchi et al. (2003)reported that cocaine significantly increased the lowerbasal levels of DA in the NAC shell more than in the core.However, when basal variability was used as a covariateof cocaine effects, the effects in core and shell were notstatistically different. In our study, there were no signifi-cant differences in DA levels in the two NAC subregionsand the acute effects of cocaine were greater in theshell than core. The several methodological differencesbetween the present study and that of Zocchi et al. (2003)make the resolution of the differences impossible withoutfurther studies.

Conflicting results have been reported by severalresearch groups about the involvement of cannabinoidreceptors in the development of cocaine-induced sen-sitization. Compared to previous reports, the presentstudy focused on the earliest stages of cocaine-inducedsensitization, using a 1-day sensitization protocol (Unglesset al. 2001). Moreover, using a single exposure minimizedcomplications due to conditioning/contingency effects ofcocaine administration (Cossu et al. 2001; Lesscher et al.2005; Gerdeman et al. 2008).

As shown in rats (Cadoni et al. 2000), we also foundthat cocaine behavioral sensitization was paralleled by aselective neurochemical sensitization of the DA responsein the NAC core as compared to the NAC shell in mice. Inthis regard, it has been shown that cocaine-inducedbehavioral sensitization in rats is accompanied by struc-tural plasticity in the core, but not in the shell, of theNAC (Li, Acerbo & Robinson 2004). In agreement withthese reports, our data clearly show that behavioralsensitization in mice was accompanied by a largerneurochemical stimulation in the NAC core as comparedto the shell. It is interesting to notice that interactionsbetween endocannabinoids, cannabinoid CB1 receptorsand DA D1-D2 receptors have been recently demon-strated to enhance firing of NAC core neurons (Seif et al.2011), even though 2-arachidonoyl-glycerol (2-AG)more than anandamide has been involved in such inter-action. Under conditions of FAAH inhibition, an involve-ment of 2-AG seems unlikely (Caprioli et al. 2012;Wiskerke et al. 2012) in the present experiments, but itsinvolvement cannot be ruled out completely. Interactionsbetween activation of DA D2 receptors, endocannabinoidrelease and cannabinoid CB1 receptor activation havebeen described in vitro also in the ventral tegmental area(VTA) (Pan, Hillard & Liu 2008a,b), even though in thosestudies there were no distinctions between midbrain DAneurons projecting to different (i.e. shell/core) DA termi-nal areas. Dopamiergic terminal areas have been the

target for a study showing D2-induced retrogradeanandamide release being instrumental to plasticitychanges in the striatum (Gerdeman, Ronesi & Lovinger2002). The authors suggest possible roles for dysfunc-tions of the endocannabinoid system in developmentaldisorders of motor control or complex habits (Gerdemanet al. 2002). In apparent contrast with our result,Eisenstein, Holmes & Hohmann (2009) report thatanimals receiving amphetamine and URB597 showreduced behavioral sensitization compared to thosereceiving amphetamine and vehicle. It is interesting tonote that such effects occurred only from the fifth day ofURB597 treatment. In the same report, rimonabant didnot affect amphetamine-induced sensitization, during the8 days of administration. The different experimental con-ditions (drugs, animal species, number of treatments,etc.) between our study and that of Eisenstein, and thelack of significant sensitization on day 2 compared to day1 in their report make direct comparisons difficult. Inanother recent report (Luque-Rojas et al. 2013), block-ade of either FAAH or monoacylglyceryl lipase (MGL)enzymes (which selectively block the degradation of2-AG) neither facilitated nor blocked the developmentof cocaine-induced sensitization. Again, differences inexperimental conditions between these studies couldinfluence the outcomes. In the study by Luque-Rojas et al.(2013), cocaine was administered alone or in combina-tion with URB597 or URB602 (an MGL-blocker thatenhances circulating levels of 2-AG). Their model,administering cocaine repeatedly at doses already fullycapable of inducing sensitization favors the detection ofantagonism rather than sensitization (Luque-Rojas et al.2013), instead of favoring the development of cocaine-induced sensitization, with cocaine administered at non-sensitizing doses. In our model, the effects of URB597were tested acutely, conditions under which anandamidelevels released by cocaine at suboptimal doses (i.e. notable to elicit sensitization) could be enhanced by FAAHblockade. Thus, only with that enhancement wouldsufficient concentrations of anandamide be reached tosignificantly activate cannabinoid CB1 receptors whichwe hypothesized to be involved in the earliest stages ofcocaine-induced sensitization. In a different study byFourgeaud et al. (2004), using a model of cocainesensitization similar to ours, it was found that 24 hoursafter the first cocaine injection, the endocannabinoidlong-term depression was abolished in the NAC. Suchspecific alteration in the endocannabinoid system seemsdirectly related to the occurrence of cocaine-inducedsensitization, as also confirmed by Grueter, Brasnjo &Malenka (2010). These reports do not contrast with ourfindings. In our conditions, some of the delayed effects ofcocaine, including those changes related to behavioraland neurochemical sensitization in the core, were



reduced by CB1 blockade or enhanced by increasing thelevel of cannabinoids right before the first cocaine admin-istration. All of the cannabinoid/endocannabinoid inter-actions with cocaine sensitization that we have testedcould counteract or facilitate its development, 24 hourslater. Thus, compared to Fourgeaud et al. (2004), we didnot test the status or the effects of the endocannabinoidsystem 24 hours after the administration of cocaine.Indeed, we focused our attention on the endocannabinoidsystem interaction with the development and not to theexpression of cocaine-induced sensitization, althoughexpression of sensitization could be an interesting focusof future research. We want to point out that URB597and rimonabant facilitation or blockade of sensitizationwas not influenced by any long-lasting effects of thesedrugs, as we have shown that when administered alonethe drugs produced no significant change in the effects ofcocaine administered 24 hours later.

Because oleoylethanolamide (OEA) and palmitoyle-thanolamide (PEA) along with anandamide are substratesof the FAAH enzyme, URB597 would be expected toincrease circulating levels of each. As anandamide, OEAand PEA are also endogenous ligands for the nuclearreceptor PPARα, an effect mediated by that mechanismmight be an alternative explanation for the observedeffects of URB597. However, URB597 administered beforesaline did not produce a significant effect on sensitizationin the present study. A recent report showed effects ofcocaine on medium-spiny neurons in the NAC shell coun-teracted by administration of URB597 (Luchicchi et al.2010), an effect not blocked by CB1 antagonists, and likelydependent from a PPARα activation. PPARα receptors andURB597 however, did not appear to play a role in cocaineinduced firing of VTA DA neurons (Luchicchi et al. 2010)and in cocaine-induced sensitization (Fernandez-Espejo,Ramiro-Fuentes & Rodriguez de Fonseca 2009). In ourexperiments, the effects of URB597 on cocaine-inducedsensitization in the NAC core rather than the shell wastested, as significant sensitization of the effects of cocainewas only found in the former brain area.

Anandamide has been suggested to be endogenousligand of TRPV1 receptors (See for review: Toth,Blumberg & Boczan 2009). Based on Grueter et al.(2010), increased levels of anandamide after URB597might activate both CB1 and TRPV1 receptors. AlthoughTRPV1 receptors seem not to affect the rewarding effectsof cocaine, they have been shown to reduce cocaine-induced but not cue-induced reinstatement of cocaineself-administration (Adamczyk et al. 2012). We cannotexclude that, in our experimental conditions, ananda-mide released by cocaine with its levels magnified byURB597, had effects on TRPV1. While we cannot addressin this report the involvement of TRPV1 receptors oncocaine-induced sensitization, it is important to note that

in the report by Grueter et al. (2010), the effects ofanandamide in the NAC were qualitatively similar onboth presynaptic-CB1 and postsynaptic-TRPV1 recep-tors. Further, anandamide has an approximate 10-foldpreference for CB1 compared to TRPV1 receptors (Rosset al. 2001; Toth et al. 2009) and the present results showa dose-dependent blockade of cocaine behavioral andneurochemical sensitization with selective occupancy ofCB1Rs by rimonabant. Thus, although rimonabant wasnot tested against URB597, the results strongly suggestan involvement of TRPV1 receptors in cocaine sensi-tization would not interfere with that presently found forCB1Rs. Thus, we believe that the interaction of cannabi-noid CB1 and DA D1/D2 receptor activation and theenhancement of NAC core neurons firing (Gerdemanet al. 2002; Seif et al. 2011) are in perfect agreementwith our findings that endocannabinoid release and CB1activation might be required at the early stages ofcocaine-induced sensitization in specific brain regions.

There is active debate about the importance and roleof sensitization in the transition from simple drug use todrug abuse and dependence (Robinson & Berridge 1993;Di Chiara 2002; Kauer & Malenka 2007). Cocainesensitization is paralleled by the development of neuronaladaptations within the NAC and the ventral tegmentalareas that might facilitate drug seeking (e.g. Robinson &Berridge 1993, 2003; Vanderschuren & Kalivas 2000; Liet al. 2004; Bowers et al. 2010). The differences in coreand shell neurochemical sensitization (Pierce et al. 1996;Cadoni et al. 2000, and present results) might underlieimportant behavioral and neurobiological aspects of thetransition from simple drug use to subsequent abuse,drug-seeking and dependence (Di Chiara et al. 1999;Di Chiara 2002). The occurrence of cocaine-inducedsensitization of DA transmission in the NAC core might beof relevance to the habit-forming effects of drugs (Everitt& Robbins 2005) which have been suggested to be medi-ated by extrapyramidal striatal areas more related to corethan shell nuclei (Di Chiara 2002; Belin & Everitt 2008).These functions may include, for example, acquisition ormaintenance of behaviors that allow the organism tocome in repeated contact with the drug and associatedstimuli (drug-seeking behavior) (Di Chiara et al. 1993;Di Chiara 2002; Everitt & Robbins 2005).

Cocaine has been suggested to release endocan-nabinoids (Centonze et al. 2004; Cheer et al. 2007, butsee Caille et al. 2007). Blockade of cannabinoid receptorsbefore the sensitizing injection of cocaine in this reportsignificantly attenuated the occurrence of behavioral/neurochemical sensitization. Thus, our results supportthe hypothesis that neuroadaptations induced by a singleinjection of cocaine (Ungless et al. 2001) require therelease of endocannabinoids. The present results are inagreement with our working hypothesis that (1) cocaine



induces region-specific on-demand release of endocan-nabinoids, (2) antagonism of the endocannabinoidssignificantly reduces neurochemical and consequentlybehavioral sensitization induced by cocaine, and (3)blockade of endocannabinoid degradation facilitates thesensitizing effects of cocaine administered at suboptimaldoses. Taken together, the present results suggest that theendocannabinoid system plays a primary role in mediat-ing the inception of long-term brain-adaptive responsesto acute stimulant effects, which in turn might alterneuronal pathways that increase a subject’s vulnerabilityto substance use disorders (Di Chiara 2002; Koob &Volkow 2010).

Acknowledgements

We thank Patty Ballerstadt for administrative assistance,Drs. Cesar Quiroz-Molina, Takato Hiranita, JonathanSlezak and Derek Wilkinson for their helpful commentsand suggestions on an earlier version of the manuscript,and Dr. Paul Soto for his help and expertise on software-related settings of the LMA equipment, and his valuableadvice during the conduct of these studies.

This work was supported by the Intramural ResearchProgram of the National Institutes of Health, NationalInstitute on Drug Abuse, Department of Health andHuman Services.

All experiments were conducted in accordance withthe guidelines of the Institutional Animal Care and UseCommittee of the National Institute on Drug Abuse,Intramural Research Program, National Institutes ofHealth, and the Guide for Care and Use of LaboratoryAnimals of the Institute of Laboratory Animal Resources,National Research Council, National Academy Press,1997. The animal facility was accredited by AAALACInternational.

Authors Contribution

VT, MM, JLK and GT conceived and designed the experi-ments. VT, MM, AMT, JLG, LEC and GT carried out theexperiments. GT supervised all the experiments. JLK con-tributed materials and supervised the behavioral experi-ments. VT, MM, AMT, LEC, JLK and GT analyzed the data,and VT, MM, JLK and GT contributed to interpretation offindings and drafted the manuscript. All authors criticallyreviewed the first draft, edited, gave comments and sug-gestions that resulted in the final submitted version ofthe paper.

References

Adamczyk P, Miszkiel J, McCreary AC, Filip M, Papp M,Przegalinski E (2012) The effects of cannabinoid CB1, CB2and vanilloid TRPV1 receptor antagonists on cocaine addic-tive behavior in rats. Brain Res 1444:45–54.

Arnold JC (2005) The role of endocannabinoid transmission incocaine addiction. Pharmacol Biochem Behav 81:396–406.

Belin D, Everitt BJ (2008) Cocaine seeking habits depend upondopamine-dependent serial connectivity linking the ventralwith the dorsal striatum. Neuron 57:432–441.

Bowers MS, Chen BT, Bonci A (2010) AMPA receptor synapticplasticity induced by psychostimulants: the past, present, andtherapeutic future. Neuron 67:11–24.

Cadoni C, Solinas M, Di Chiara G (2000) Psychostimulantsensitization: differential changes in accumbal shell and coredopamine. Eur J Pharmacol 388:69–76.

Caille S, Alvarez-Jaimes L, Polis I, Stouffer DG, Parsons LH(2007) Specific alterations of extracellular endocannabinoidlevels in the nucleus accumbens by ethanol, heroin, andcocaine self-administration. J Neurosci 27:3695–3702.

Caprioli A, Coccurello R, Rapino C, Di Serio S, Di Tommaso M,Vertechy M, Vacca V, Battista N, Pavone F, Maccarrone M,Borsini F (2012) The novel reversible fatty acid amidehydrolase inhibitor ST4070 increases endocannabinoid brainlevels and counteracts neuropathic pain in different animalmodels. J Pharmacol Exp Ther 342:188–195.

Centonze D, Battista N, Rossi S, Mercuri NB, Finazzi-Agro A,Bernardi G, Calabresi P, Maccarrone M (2004) A criticalinteraction between dopamine D2 receptors and endocanna-binoids mediates the effects of cocaine on striatal gabaergictransmission. Neuropsychopharmacology 29:1488–1497.

Cheer JF, Wassum KM, Sombers LA, Heien ML, Ariansen JL,Aragona BJ, Phillips PE, Wightman RM (2007) Phasic dopa-mine release evoked by abused substances requires cannabi-noid receptor activation. J Neurosci 27:791–795.

Corbille AG, Valjent E, Marsicano G, Ledent C, Lutz B, Herve D,Girault JA (2007) Role of cannabinoid type 1 receptors inlocomotor activity and striatal signaling in response topsychostimulants. J Neurosci 27:6937–6947.

Cossu G, Ledent C, Fattore L, Imperato A, Bohme GA, ParmentierM, Fratta W (2001) Cannabinoid CB1 receptor knockout micefail to self-administer morphine but not other drugs of abuse.Behav Brain Res 118:61–65.

Di Chiara G (2002) Nucleus accumbens shell and core dopa-mine: differential role in behavior and addiction. Behav BrainRes 137:75–114.

Di Chiara G, Acquas E, Tanda G, Cadoni C (1993) Drugs ofabuse: biochemical surrogates of specific aspects of naturalreward? Biochem Soc Symp 59:65–81.

Di Chiara G, Tanda G, Bassareo V, Pontieri F, Acquas E, FenuS, Cadoni C, Carboni E (1999) Drug addiction as a disorderof associative learning. Role of nucleus accumbens shell/extended amygdala dopamine. Ann N Y Acad Sci 877:461–485.

Di Chiara G, Tanda G, Cadoni C, Acquas E, Bassareo V, CarboniE (1998) Homologies and differences in the action of drugs ofabuse and a conventional reinforcer (food) on dopamine trans-mission: an interpretative framework of the mechanism ofdrug dependence. Adv Pharmacol 42:983–987.

Eisenstein SA, Holmes PV, Hohmann AG (2009)Endocannabinoid modulation of amphetamine sensitizationis disrupted in a rodent model of lesion-induced dopaminedysregulation. Synapse 63:941–950.

Everitt BJ, Robbins TW (2005) Neural systems of reinforcementfor drug addiction: from actions to habits to compulsion. NatNeurosci 8:1481–1489.

Fernandez-Espejo E, Ramiro-Fuentes S, Rodriguez de Fonseca F(2009) The absence of a functional peroxisome o proliferator-activated receptor-alpha gene in mice enhances motor



sensitizing effects of morphine, but not cocaine. Neuroscience164:667–675.

Ferrer B, Asbrock N, Kathuria S, Piomelli D, Giuffrida A (2003)Effects of levodopa on endocannabinoid levels in rat basalganglia: implications for the treatment of levodopa-induceddyskinesias. Eur J Neurosci 18:1607–1614.

Filip M, Golda A, Zaniewska M, McCreary AC, Nowak E,Kolasiewicz W, Przegalinski E (2006) Involvement of can-nabinoid CB1 receptors in drug addiction: effects of rimona-bant on behavioral responses induced by cocaine. PharmacolRep 58:806–819.

Fourgeaud L, Mato S, Bouchet D, Hemar A, Worley PF, ManzoniOJ (2004) A single in vivo exposure to cocaine abolishesendocannabinoid-mediated long-term depression in thenucleus accumbens. J Neurosci 24:6939–6945.

Gerdeman GL, Ronesi J, Lovinger DM (2002) Postsynapticendocannabinoid release is critical to long-term depression inthe striatum. Nat Neurosci 5:446–451.

Gerdeman GL, Schechter JB, French ED (2008) Context-specificreversal of cocaine sensitization by the CB1 cannabinoidreceptor antagonist rimonabant. Neuropsychopharmacology33:2747–2759.

Giuffrida A, Parsons LH, Kerr TM, Rodriguez de Fonseca F,Navarro M, Piomelli D (1999) Dopamine activation of endog-enous cannabinoid signaling in dorsal striatum. Nat Neurosci2:358–363.

Gonzalez S, Cascio MG, Fernandez-Ruiz J, Fezza F, Di Marzo V,Ramos JA (2002) Changes in endocannabinoid contents inthe brain of rats chronically exposed to nicotine, ethanol orcocaine. Brain Res 954:73–81.

Grueter BA, Brasnjo G, Malenka RC (2010) Postsynaptic TRPV1triggers cell type-specific long-term depression in the nucleusaccumbens. Nat Neurosci 13:1519–1525.

Kalivas PW, Duffy P (1993) Time course of extracellular dopa-mine and behavioral sensitization to cocaine. I. Dopamineaxon terminals. J Neurosci 13:266–275.

Kathuria S, Gaetani S, Fegley D, Valino F, Duranti A, Tontini A,Mor M, Tarzia G, La Rana G, Calignano A, Giustino A, TattoliM, Palmery M, Cuomo V, Piomelli D (2003) Modulation ofanxiety through blockade of anandamide hydrolysis. Nat Med9:76–81.

Kauer JA, Malenka RC (2007) Synaptic plasticity and addiction.Nat Rev Neurosci 8:844–858.

Koob GF, Volkow ND (2010) Neurocircuitry of addiction.Neuropsychopharmacology 35:217–238.

Lesscher HM, Hoogveld E, Burbach JP, van Ree JM, Gerrits MA(2005) Endogenous cannabinoids are not involved in cocainereinforcement and development of cocaine-induced behav-ioural sensitization. Eur Neuropsychopharmacol 15:31–37.

Li X, Hoffman AF, Peng XQ, Lupica CR, Gardner EL, Xi ZX (2009)Attenuation of basal and cocaine-enhanced locomotion andnucleus accumbens dopamine in cannabinoid CB1-receptor-knockout mice. Psychopharmacology (Berl) 204:1–11.

Li Y, Acerbo MJ, Robinson TE (2004) The induction of behav-ioural sensitization is associated with cocaine-induced struc-tural plasticity in the core (but not shell) of the nucleusaccumbens. Eur J Neurosci 20:1647–1654.

Loland CJ, Mereu M, Okunola OM, Cao J, Prisinzano TE, Mazier S,Kopajtic T, Shi L, Katz JL, Tanda G, Newman AH (2012)R-modafinil (armodafinil): a unique dopamine uptake inhibi-tor and potential medication for psychostimulant abuse. BiolPsychiatry 72:405–413.

Luchicchi A, Lecca S, Carta S, Pillolla G, Muntoni AL, Yasar S,Goldberg SR, Pistis M (2010) Effects of fatty acid amide

hydrolase inhibition on neuronal responses to nicotine,cocaine and morphine in the nucleus accumbens shell andventral tegmental area: involvement of PPAR-alpha nuclearreceptors. Addict Biol 15:277–288.

Luque-Rojas MJ, Galeano P, Suarez J, Araos P, Santin LJ, deFonseca FR, Calvo EB (2013) Hyperactivity induced by thedopamine D2/D3 receptor agonist quinpirole is attenuated byinhibitors of endocannabinoid degradation in mice. Int JNeuropsychopharmacol 16:661–676.

Pan B, Hillard CJ, Liu QS (2008a) D2 dopamine receptor activa-tion facilitates endocannabinoid-mediated long-term synapticdepression of GABAergic synaptic transmission in midbraindopamine neurons via cAMP-protein kinase A signaling.J Neurosci 28:14018–14030.

Pan B, Hillard CJ, Liu QS (2008b) Endocannabinoid signal-ing mediates cocaine-induced inhibitory synaptic plasticityin midbrain dopamine neurons. J Neurosci 28:1385–1397.

Paxinos G, Franklin KBJ (2001) The Mouse Brain in StereotaxicCoordinates, 2nd edn. New York: Academic Press.

Pierce RC, Bell K, Duffy P, Kalivas PW (1996) Repeated cocaineaugments excitatory amino acid transmission in the nucleusaccumbens only in rats having developed behavioralsensitization. J Neurosci 16:1550–1560.

Pierce RC, Vanderschuren LJ (2010) Kicking the habit: theneural basis of ingrained behaviors in cocaine addiction.Neurosci Biobehav Rev 35:212–219.

Pontieri FE, Tanda G, Di Chiara G (1995) Intravenous cocaine,morphine, and amphetamine preferentially increase extra-cellular dopamine in the ‘shell’ as compared with the ‘core’ ofthe rat nucleus accumbens. Proc Natl Acad Sci U S A92:12304–12308.

Robinson TE, Berridge KC (1993) The neural basis of drugcraving: an incentive-sensitization theory of addiction. BrainRes Brain Res Rev 18:247–291.

Robinson TE, Berridge KC (2003) Addiction. Annu Rev Psychol54:25–53.

Ross RA, Gibson TM, Brockie HC, Leslie M, Pashmi G, CraibSJ, Di Marzo V, Pertwee RG (2001) Structure-activityrelationship for the endogenous cannabinoid, anandamide,and certain of its analogues at vanilloid receptors intransfected cells and vas deferens. Br J Pharmacol 132:631–640.

Saal D, Dong Y, Bonci A, Malenka RC (2003) Drugs of abuse andstress trigger a common synaptic adaptation in dopamineneurons. Neuron 37:577–582.

Seif T, Makriyannis A, Kunos G, Bonci A, Hopf FW (2011) Theendocannabinoid 2-arachidonoylglycerol mediates D1 and D2receptor cooperative enhancement of rat nucleus accumbenscore neuron firing. Neuroscience 193:21–33.

Solinas M, Justinova Z, Goldberg SR, Tanda G (2006)Anandamide administration alone and after inhibition offatty acid amide hydrolase (FAAH) increases dopamine levelsin the nucleus accumbens shell in rats. J Neurochem 98:408–419.

Solinas M, Tanda G, Wertheim CE, Goldberg SR (2010)Dopaminergic augmentation of delta-9-tetrahydrocannabinol(THC) discrimination: possible involvement of D(2)-inducedformation of anandamide. Psychopharmacology (Berl) 209:191–202.

Soria G, Mendizabal V, Tourino C, Robledo P, Ledent C,Parmentier M, Maldonado R, Valverde O (2005) Lack of CB1cannabinoid receptor impairs cocaine self-administration.Neuropsychopharmacology 30:1670–1680.



Tanda G (2007) Modulation of the endocannabinoid system:therapeutic potential against cocaine dependence. PharmacolRes 56:406–417.

Tanda G, Ebbs A, Newman AH, Katz JL (2005) Effects of4’-chloro-3 alpha-(diphenylmethoxy)-tropane on mesostria-tal, mesocortical, and mesolimbic dopamine transmission:comparison with effects of cocaine. J Pharmacol Exp Ther313:613–620.

Tanda G, Newman AH, Ebbs AL, Tronci V, Green JL, Tallarida RJ,Katz JL (2009) Combinations of cocaine with other dopamineuptake inhibitors: assessment of additivity. J Pharmacol ExpTher 330:802–809.

Thiemann G, van der Stelt M, Petrosino S, Molleman A, Di MarzoV, Hasenohrl RU (2008) The role of the CB1 cannabinoidreceptor and its endogenous ligands, anandamide and 2-arachidonoylglycerol, in amphetamine-induced behaviouralsensitization. Behav Brain Res 187:289–296.

Toth A, Blumberg PM, Boczan J (2009) Anandamide and thevanilloid receptor (TRPV1). Vitam Horm 81:389–419.

Tronci V, Tanda G (2007) Involvement of CB1 cannabinoidreceptors in cocaine-induced locomotor sensitization aftersingle pre-exposure in mice. FASEB J 21:A410–A410.

Ungless MA, Whistler JL, Malenka RC, Bonci A (2001) Singlecocaine exposure in vivo induces long-term potentiation indopamine neurons. Nature 411:583–587.

Vanderschuren LJ, Kalivas PW (2000) Alterations indopaminergic and glutamatergic transmission in the induc-tion and expression of behavioral sensitization: a criticalreview of preclinical studies. Psychopharmacology (Berl)151:99–120.

Wilson RI, Nicoll RA (2002) Endocannabinoid signaling in thebrain. Science 296:678–682.

Wiskerke J, Irimia C, Cravatt BF, De Vries TJ, Schoffelmeer AN,Pattij T, Parsons LH (2012) Characterization of the effectsof reuptake and hydrolysis inhibition on interstitialendocannabinoid levels in the brain: an in vivo microdialysisstudy. ACS Chem Neurosci 3:407–417.

Wiskerke J, Pattij T, Schoffelmeer AN, De Vries TJ (2008) Therole of CB1 receptors in psychostimulant addiction. AddictBiol 13:225–238.

Xi ZX, Gilbert JG, Peng XQ, Pak AC, Li X, Gardner EL (2006)Cannabinoid CB1 receptor antagonist AM251 inhibitscocaine-primed relapse in rats: role of glutamate in thenucleus accumbens. J Neurosci 26:8531–8536.

Zocchi A, Girlanda E, Varnier G, Sartori I, Zanetti L, Wildish GA,Lennon M, Mugnaini M, Heidbreder CA (2003) Dopamineresponsiveness to drugs of abuse: a shell-core investigation inthe nucleus accumbens of the mouse. Synapse 50:293–302.

SUPPORTING INFORMATION

Additional Supporting Information may be found in theonline version of this article at the publisher’s web-site:

Figure S1 Locomotor activity response to an acute salineor cocaine injection (10 mg/kg or 20 mg/kg) in naïvemice, pre-treated 40 minutes before with rimonabant(RIM) vehicle (unfilled bars), RIM 1 (gray bars) or3 mg/kg (black filled bars). The results are expressed asthe average distance traveled ± SEM obtained for eachexperimental group of animals during the first 30minutes after injection. Significant differences, P < 0.05,are represented in the figure with a * (Veh or RIM versusrespective Saline) or # (RIM versus Veh)Figure S2 Time course of the effects of systemic adminis-tration of RIM 40 minutes before an acute injection ofcocaine, 10 or 20 mg/kg, panels A and B, respectively, onextracellular levels of DA in dialysates from the NAC shell,or the NAC core. The figure shows that an acute injectionof RIM failed to significantly modify cocaine-inducedstimulation of extracellular DA levels in both of the NACsubregions (number of animals, n, and basal DA values,fmoles/10-μl sample ± SEM were: n = 7, 36.3 ± 5.0, andn = 4, 44.1 ± 11.6, for cocaine 10 and 20 mg/kg, respec-tively in the shell; n = 4, 35.9 ± 2.0, and n = 4, 32.7 ± 2.1,for RIM 1 and 3 mg/kg + cocaine 10 mg/kg, respectivelyin the shell; n = 5, 29.1 ± 3.6, for RIM 1 mg/kg + cocaine10 mg/kg, in the shell; n = 7, 51.2 ± 7.1 and n = 4,40.0 ± 4.1, for RIM 3 mg/kg + cocaine 10 and 20 mg/kg,respectively, in the core). Results are means with verticalbars representing SEM of the amount of DA in 10 ul ofdialysate samples, expressed as percentage of basal values



Cocaine-induced endocannabinoid release modulates behavioral and neurochemical sensitization in mice

Documents

Transcript of Cocaine-induced endocannabinoid release modulates behavioral and neurochemical sensitization in mice