at SciVerse ScienceDirect

Neuropharmacology 63 (2012) 945e957

Contents lists available

Neuropharmacology

journal homepage: www.elsevier .com/locate/neuropharm

Lesions of the lateral habenula dissociate the reward-enhancingand locomotor-stimulant effects of amphetamine

Anthony J. Gifuni, Solmaz Jozaghi, Anne-Catherine Gauthier-Lamer, Sandra M. Boye*

Department of Psychiatry, Université de Montréal, 90 Vincent-d’Indy, F-429-3, Outremont, Montreal, Quebec, Canada H2V 2S9

a r t i c l e i n f o

Article history:Received 14 September 2011Received in revised form11 July 2012Accepted 16 July 2012

Keywords:Lateral habenulaMedial habenulaRewardLocomotionAmphetaminePosterior mesencephalonMedial forebrain bundle

* Corresponding author. Tel.: þ1 514 343 6111x172E-mail address: sandra.boye@umontreal.ca (S.M. B

0028-3908/$ e see front matter � 2012 Elsevier Ltd.http://dx.doi.org/10.1016/j.neuropharm.2012.07.032

a b s t r a c t

Midbrain dopamine neurons play a key role in goal-directed behaviour as well as in some psychiatricdisorders. Recent studies have provided electrophysiological, anatomical and biochemical evidence thatthe lateral habenula (LHb) exerts strong inhibitory control over midbrain dopamine neurons. However,the behavioural relevance of this inhibitory input is poorly understood. Our aim was to examine thecontribution of the LHb to dopamine-sensitive behaviour. Here, we characterized the locomotor-stimulant and reward-enhancing properties of amphetamine in rats with and without neurotoxiclesions of the LHb. Amphetamine-induced forward locomotion and reward were respectively measuredin automated activity cages and with intracranial self-stimulation. Adult, male Sprague-Dawley rats werebilaterally infused with ibotenic acid in the LHb and allowed 7e10 days post-operative recovery. Thelocomotor-stimulant and reward-enhancing properties of amphetamine (0, 0.5 and 1.0 mg/kg, ip) werethen tested in different groups of lesioned and sham-lesioned rats. Neurotoxic lesions of the LHb causeda significant enhancement of the locomotor-stimulant effect of amphetamine, an effect not seenfollowing lesions of the medial habenula. Conversely, the reward-enhancing properties of amphetaminedid not differ between lesioned and sham-lesioned rats responding for rewarding electrical stimulationof the posterior mesencephalon or the medial forebrain bundle. The dissociation between the locomotor-stimulant and reward-enhancing effects of amphetamine following LHb lesions suggests the contributionof two distinct substrates that are functionally dissociable and differentially sensitive to LHb modulation.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

The habenula, located centrally along the dorsal diencephalicpathway, serves as a relay between forebrain afferents of limbic andbasal ganglia origins and monoamimergic midbrain nuclei intri-cately involved in the control of emotions and behaviour (Geislerand Trimble, 2008; Hikosaka et al., 2008; Li et al., 1993; Wangand Aghajanian, 1977). Most notably, habenular involvement hasbeen cited in maternal behaviour (Matthews-Felton et al., 1995),responses to stress and anxiety (Shumake and Gonzalez-Lima,2003; Wirtshafter et al., 1994), reward (Gomita and Gallistel,1982) and reward error processing (Matsumoto and Hikosaka,2007; Ullsperger and von Cramon, 2003), as well as in psychiatricdisorders such as schizophrenia (Sandyk, 1992; Shepard et al.,2006) and depression (Roiser et al., 2009; Sartorius et al., 2010).

The habenula comprises a medial and a lateral aspect, each withdistinct afferents and efferents (Sutherland, 1982). Recently, the

80.oye).

All rights reserved.

lateral habenula (LHb) has received renewed research attentionduemainly to its increasingly evident role in controlling the activityof mesolimbic and nigrostriatal dopamine (DA) neurons. Thus,earlier descriptions of habenular innervation of ventral midbrain(Bunney and Aghajanian, 1976; Herkenham and Nauta, 1979;Phillipson, 1979) have been confirmed and extended to show thatthis innervation is predominantly glutamatergic and preferentiallysynapses on GABAergic neurons of the ventral tegmental area(VTA), substantia nigra and rostromedial tegmental nucleus(Brinschwitz et al., 2010; Jhou et al., 2009); all three GABAergicgroups in turn provide inhibitory control over DAergic activity.Thus, stimulation of the LHb results in phasic inhibition of DAneurons, an effect that is mediated via GABAa receptors (Ji andShepard, 2007) and is blocked following electrolytic lesions of thefasciculus retroflexus or chemical lesions of the LHb (Christophet al., 1986; Ji and Shepard, 2007). In addition, a modest butdirect glutamatergic innervation of midbrain DA neurons has alsobeen described (Brinschwitz et al., 2010; Omelchenko et al., 2009).Despite this, the predominant consequence of habenular stimula-tion is near-absolute inhibition of DAergic firing activity in anes-thetized (Christoph et al., 1986; Ji and Shepard, 2007) as well as in

A.J. Gifuni et al. / Neuropharmacology 63 (2012) 945e957946

awake animals (Matsumoto and Hikosaka, 2007). Others havesuggested that the LHb exerts near-maximal tonic inhibition of DAneurons under basal conditions, based on the observation thatacute inhibition of LHb input to VTA results in a significant increasein forebrain DA neurotransmission (Lecourtier et al., 2008;Nishikawa et al., 1986).

Comparatively less is known about the behavioural relevance ofLHb signals transmitted to VTA DA neurons, but this knowledge isof potential relevance given the key role of this neuronal populationin goal-directed behaviour as well as in some serious psychiatricdisorders. Recent studies in behaving monkeys have illustrated ingreat detail how habenular neurons are excited by signals associ-ated with the absence of expected reward or the presence ofpunishment, an event followed closely in time by silencing ofmidbrain DA neurons. Conversely, predictable reward or the pres-ence of reward-related conditioned stimuli inhibit the habenulabut enhance midbrain DA activity (Matsumoto and Hikosaka, 2007,2009a). These findings have led to the hypothesis that intrinsic LHbneurons direct approach behaviour towards potential rewards viainhibition/disinhibition of VTA DA activity (see Hikosaka, 2010).

Midbrain DA neurons are critically important for the rewardingand motivating properties of drugs of abuse as well as of electricalbrain stimulation (Wise and Rompré, 1989). However, the fewstudies that have addressed the contribution of the habenula tothese functions have provided inconsistent results. For instance,operant responding for sucrose and cocaine on a fixed intervalschedule is increased by low-frequency habenular stimulation, isunaltered by high frequency stimulation and is reduced by mixedpatterns of stimulation (Friedman et al., 2010, 2011). Others haveshown that rats will self-administer trains of electrical pulsesdirectly to the habenula (Vachon and Miliaressis, 1992), suggestingthat the stimulation is rewarding. Lastly, whereas neurotoxiclesions of the habenula cause resistance to extinction of bothsucrose and cocaine self-administration (Friedman et al., 2010,2011), electrolytic lesions lead to long-lasting reductions in brainstimulation (Morissette and Boye, 2008) but not heroin reward(Wang et al., 2009).

Our objective was to understand the contribution of the LHb tobehaviour that is sensitive to changes in mesolimbic DA neuro-transmission. To do this, we studied the locomotor-activating aswell as the reward-enhancing properties of amphetamine, in ratswith and without neurotoxic lesions of the LHb. Forward locomo-tion following amphetamine was used to assess LHb contributionsto DA-mediated psychostimulant activity. Intracranial self-stimulation was used in combination with curve-shift scaling(Miliaressis et al., 1986), allowing us to selectively assess thereward-enhancing properties of amphetamine independent of itseffect on the rate of operant responding.

2. Material and methods

2.1. Subjects

Subjects were 250 male Sprague-Dawley rats (Charles River, St Constant,Quebec) weighing between 300 and 350 g at the time of surgery. Rats were kept ina temperature (21 �C) and humidity (50%) controlled room with a 12 h light/darkcycle (lights on at 6:30 a.m.), had unrestricted access to food and water and wereallowed to habituate to the animal colony for at least five days prior to surgery. Allprocedures followed Canadian Council on Animal Care guidelines and wereapproved by two separate Institutional Animal Care Committees.

2.2. Surgery

2.2.1. LesionsSubjects were anesthetized with a mixture of isoflurane (5%) and oxygen (0.6 L/

min) and mounted onto a stereotaxic apparatus. During surgery, the level of anes-thesia was reduced to 2e3%. The skull was drilled bilaterally over the habenula and,in those rats included in the intracranial self-stimulation experiment, in the area

overlying the posterior mesencephalon (PM) or the medial forebrain bundle (MFB).Habenular infusions were made sequentially, with a cannula (30 ga) that was con-nected to a 5 ml Hamilton syringe via PE10 tubing. Flat skull coordinates for the LHbwere: 3.3 mm posterior to bregma, �0.6 mm lateral to the midpoint of the sagittalsinus and 4.7 mm ventral to the surface of the sinus (Paxinos and Watson, 1997).Ibotenic acid (0.25 mg/0.25 ml/side; Tocris Bioscience, MO, USA) and vehicle (sterile0.9% saline, 0.25 ml/side) infusions were controlled by an infusion pump (HarvardApparatus, MA, USA) and occurred over one minute. A separate group of ratsreceived ibotenic acid or vehicle infusions into the medial habenula (MHb) at thefollowing coordinates: 3.6 mm posterior to bregma, �1.1 mm lateral to the midlineof the sagittal sinus (with a 10� angle towards midline) and 4.6 mm ventral to thesurface of the sinus. Because of the relatively smaller size of the MHb, the infusionvolume was reduced (0.25 mg/0.20 ml/side) on the basis of pilot surgeries. At the endof each infusion, an additional minute was allowed prior to cannula retraction inorder to maximize drug diffusion. The scalp was then sutured with Chromic catgut.Prior to the end of surgery, all rats received an injection of the non-steroidal anti-inflammatory analgesic ketoprofen (5 mg/kg, sc).

2.2.2. Electrode implantationRats used in intracranial self-stimulation studies were first infused with ibotenic

acid into the LHb, as described above. Once the infusions were completed, rats wereimplanted with a stimulation electrode aimed at the PM at the following coordi-nates: 7.8e8.0 mm posterior to bregma, 0.0 mm lateral to the midline and6.8e7.8 mm ventral to the surface of the skull. These coordinates correspond to theborder of the ventral portion of the dorsal raphe nucleus and the decussation of thesuperior cerebellar peduncle, a region that supports stable low-threshold self-stimulation (Rompré and Boye, 1989; Rompré and Miliaressis, 1985). For compar-ison, a separate group of rats was implantedwith a stimulation electrode in theMFB,at the level of the lateral hypothalamus, using the following coordinates: 3.0 mmposterior to bregma, 1.7 mm lateral to the midline of the sagittal sinus and 7.6 mmventral to the surface of the sinus. Stimulation electrodes were made from stainlesssteel wire (0.27 mm dia.) and were insulated with Epoxylite except at the roundedtip. A bare stainless steel wire, connected at one end to a male Amphenol pin andwrapped around six miniature screws that were threaded into the cranium, servedas the anode. Acrylic dental cement was used to chronically secure the electrodeassembly to the skull. Prior to the end of surgery, all rats received an injection ofketoprofen (5 mg/kg, sc).

2.3. Locomotor activity

Locomotor activity was measured in rectangular Plexiglas chambers (45 cmlong � 32.5 cm wide � 37.5 cm high), each equipped with two parallel infraredphotobeams situated 23 cm apart, 11 cm from the short ends of the chambers, and3 cm above the grid floor. A mesh lid allowed ventilation. Each locomotor activitychamber was encased within a sound-attenuating box that was equipped with a fanand a 7-W lightbulb. In-house software registered alternating photobeam inter-ruptions; the total locomotor activity score thus represents the number of times therat traversed the cage during the test session.

Ten days after surgery, rats were placed in the activity chambers and baseline(spontaneous) locomotor activity was counted during one hour. Rats were theninjected with amphetamine (0.5 or 1.0 mg/kg, ip) or vehicle (0.9% saline) andlocomotor activity was measured for an additional two hours.

2.4. Intracranial self-stimulation

Operant conditioning chambers (28 cm wide � 29.4 cm deep � 68.6 cm high)were constructed from PVC (back and side walls) and Plexiglas (front wall). Eachchamber was equippedwith a non-retractable lever (ENV-116M,Med Associates Inc,St Albans VT, USA) located on the left wall, 3.4 cm above the metal rod floor. Operantconditioning chambers were encased in sound-attenuating boxes (48.6 cmwide� 50.7 cm deep� 95.4 cm high)made frommelaminewith a Plexiglas windowallowing constant viewing of the rat. Each depression of the lever triggereda constant-current stimulator (PHM-152/2, Med Associates Inc, St Albans, VT, USA)to deliver a single 400-ms train of rectangular cathodal pulses of 0.1 ms in duration,delivered on a FI-1sec schedule. Current intensity was monitored on an oscilloscopeby reading the voltage drop across a 1 kU resistor in series with the electrode.

One week after surgery, rats were trained to self-administer trains of cathodalrectangular pulses by the method of successive approximations. Once the responsewas learnt, rats were allowed to self-administer the stimulation for one hour, atparameters set to support consistent responding. On the following day, rats wereallowed to self-administer the same stimulation parameters, but only during 45 strials that were separated by 30 s inter-trial intervals and preceded by five trains ofnon-contingent stimulation. All parameters of the non-contingent stimulation wereidentical to those available during the 45 s trial. Beginning on the third day, the pulsefrequency was systematically reduced by approximately 0.1 log10 units across trials,starting with a frequency that supported maximal responding and ending with onesufficiently low to produce extinction. The plot of the rate of responding as a func-tion of pulse frequency comprised a single responseefrequency curve; several ofthese curves were determined daily. Reward thresholds were defined as the pulse

Fig. 1. Ibotenic acid lesions of the lateral and medial habenula. Photomicrographs ofthionin stained sections from a control (a), a sham-lesioned (b) and two ibotenic acidlesioned rats (c,d) were taken at coronal levels corresponding to 3.3 mm (aec) and3.8 mm (d) posterior to bregma (Paxinos and Watson, 1997). Photomicrographs ofcresyl violet stained sections from a sham-lesioned (e) and a lesioned rat (f) weretaken at the level corresponding to 3.8 mm posterior to bregma. Arrowheads point tomechanical damage caused by insertion of the cannula. High magnification photomi-crographs were taken from sites immediately below cannula tips (insets). Note thereduction in neuronal density and increase in microglia in lesioned rats (i, j, l)compared to control (g) or sham-lesioned rats (h, k). Scale bar in g (50 mm) applies togel. 3V, third ventricle; MHb, medial habenula; LHb, lateral habenula.

A.J. Gifuni et al. / Neuropharmacology 63 (2012) 945e957 947

frequency required for half-maximal responding and were interpolated froma regression line fit to the rising portion of individual responseefrequency curves. Inorder to standardize the rewarding effectiveness of the stimulation across all ratsprior to the start of drug testing, current intensities were adjusted to produce rewardthresholds of approximately 50 Hz. Asymptotic response rates were defined as thehighest number of bar presses in each curve.

Once reward thresholds were stable across a minimum of three days (less than0.1 log10 unit variations), we tested the rewarding effectiveness of amphetamine. Onthe test day, we first determined four baseline reward thresholds; the first thresholdestimate was excluded from data analysis. Rats were then injected with amphet-amine (0.5 or 1.0 mg/kg, ip) or vehicle (0.9% saline) and immediately returned to theoperant conditioning chamber for six additional threshold determinations(approximately 90 min).

2.5. Histology

At the end of all testing, rats in the locomotor activity experiment were anes-thetized with urethane (1.2 g/kg, ip), decapitated and their brains were immediatelyharvested and flash frozen (2-methyl-butane, �50 �C for 10 s). For rats in theintracranial self-stimulation experiment, stimulation sites were first marked bypassing direct anodal current (100 mA, 15 s) via the stimulation electrode. Rats werethen sacrificed similarly to those in the locomotor activity experiment. Coronal brainslices (20 mm) were obtained with a cryostat and were mounted onto gelatin-coatedslides. Sections were stained for Nissl substance using thionin or cresyl violet andexamined with a light microscope.

2.6. Data analysis

Statistical analyses were carried out using Statistica 6.0 (Stat-Soft, Tulsa, OK,USA). All time course data were analysed with two-way mixed analyses of variance(ANOVAs) with lesion as the between factor and time as the repeated measure.Mean total activity scores as well as mean threshold values and mean asymptoticresponse rates were analyzed with two-way ANOVAs (lesion � treatment) followedby post hoc Tukey honestly significant difference (HSD) tests where appropriate.Mean stimulation current intensities were analyzed with t-tests for independentsamples.

3. Results

3.1. LHb lesions enhance the locomotor-stimulant effect ofamphetamine

In total, 80 rats received bilateral neurotoxin or vehicle infusionsinto the LHb. Of these, 28 were excluded from data analysis becausethey had at least one cannula track outside the LHb, resulting in 26lesioned and 26 sham-lesioned rats. Ibotenic acid lesions werecharacterized by the presence of gliosis and reduced neuronaldensities (Fig. 1i, j). Lesion boundaries were in general restricted tothe LHb and collectively, covered an area of the LHb extending fromapproximately 2.8e4.16 mm posterior to bregma (Paxinos andWatson, 1997) (Fig. 2). In one case, we were only able to locatethe lesionwithin the left LHbwith certainty, so for this rat, the rightlesion is not included in Fig. 2a. Sites of sham infusions wereidentified by locating the tip of the cannula track; theanterioreposterior distribution of these sites was similar to that ofthe lesions (Fig. 2b). No significant evidence of neuronal loss orgliosis was observed in sham-infused brains (Fig. 1h).

Fig. 3 shows the time course of the locomotor response underbaseline conditions and after injection of vehicle and amphet-amine. Lesions of the LHb did not alter spontaneous locomotion(F(1, 50) ¼ 0.43, p ¼ 0.51) or the response to vehicle (F(1, 18) ¼ 0.47,p ¼ 0.50) or 0.5 mg/kg amphetamine (F(1, 12) ¼ 0.30, p ¼ 0.59), butdid produce a hypersensitive response to the locomotor-stimulantactions of 1.0 mg/kg amphetamine (F(1, 16) ¼ 7.13, p ¼ <0.05)(Fig. 3aed). The bargraphs in Fig. 3e and f show mean total activityscores during the first and second hours post-injection, respec-tively. A two-way ANOVA on first hour scores revealed a significanteffect of treatment (F(2, 46) ¼ 46.34, p < 0.01), of lesion (F(1,46) ¼ 7.24, p < 0.01) and an interaction (F(2, 46) ¼ 3.82, p < 0.05).Tukey HSD post hoc analysis indicated that amphetamine increasedlocomotor activity in both groups (p < 0.01), but the response to

Fig. 2. Distribution of LHb lesions and sham lesions in rats included in the locomotor activity experiment. (a) Schematic diagrams of the rat brain at the level of the habenulashowing the distribution of all lesions (n ¼ 26 rats). Lesions were reconstructed from light microscopic examination of thionin stained tissue; lesion boundaries were determined bythe presence of neuronal loss and glial proliferation, viewed at a magnification of 40�. (b) Schematic representation of the sites of sham infusions (n ¼ 26 rats). Sites weredetermined by locating the tip of the cannula track under the same conditions as in (a). Numbers above each level indicate distance (mm) posterior to bregma.

A.J. Gifuni et al. / Neuropharmacology 63 (2012) 945e957948

1.0 mg/kg amphetamine was 60% greater in lesioned than in sham-lesioned rats (182.9 � 17.21 vs 115.1 � 11.26, p < 0.01) (Fig. 3e). Asimilar but nonsignificant trend was observed with the 0.5 mg/kgdose. The two-way ANOVA on activity scores of the second hourrevealed a significant effect of treatment (F(2, 46)¼ 28.77, p< 0.01)but no effect of lesion (F(1, 46) ¼ 1.93, p ¼ 0.17) (Fig. 3f).

In a few rats, we observed some lesion encroachment on theMHb (Fig. 2a). In order to ensure that the observed locomotor-enhancing effect of amphetamine was indeed due to the loss ofintrinsic LHb neurons, a second group of rats received lesions of theMHb and were tested with the 1.0 mg/kg dose or vehicle. In all, 52rats were infused and 22 were excluded due to incorrect cannulaplacements. Of the remaining 30, 16 were lesioned and 14 served assham-lesioned controls. Intra-MHb lesions did not alter sponta-neous (F(1, 28) ¼ 0.02, p ¼ 0.89), vehicle or amphetamine (F(1,12) ¼ 0.00, p ¼ 0.97) locomotion (Fig. 4aec). In the vehicle-treatedgroup, several lesioned and sham-lesioned rats did not display anyhorizontal displacements, particularly during the second hourpost-injection. The lack of variance in locomotor scores at thesetime points precluded ANOVA.Mean total activity scores during thefirst and second hours post-injection are shown in Fig. 4d and e,respectively. Two-way ANOVAs revealed significant effects oftreatment (1st hour: (F(1, 26) ¼ 33.34, p < 0.01); 2nd hour: (F(1,26) ¼ 21.17, p < 0.01)) but no effect of lesion (1st hour: (F(1,

26) ¼ 0.04, p ¼ 0.84); 2nd hour: (F(1, 26) ¼ 0.00, p ¼ 0.99)). Due tothe reduced infusion volume, lesions of the MHbwere smaller thanthose produced in the LHb (Fig. 5a). MHb lesions were locatedthroughout the dorso-ventral extent of the structure and theirdistribution ranged from approximately 2.8e4.16 mm posterior tobregma. Sham lesions of the MHb showed a similar distribution(Fig. 5b).

3.2. LHb lesions do not alter the reward-enhancing effect ofamphetamine

In total, 118 rats received bilateral neurotoxin or vehicle infu-sions into the LHb. Of these, 38 rats were excluded from dataanalysis because they had at least one cannula track outside theLHb or did not learn the operant response within the allotted time.Of the remaining rats, 43 had a stimulation electrode implanted inthe PM (Fig. 6aec) and 37 in theMFB (Fig. 6d). In those rats with PMelectrodes, the anterioreposterior distribution of LHb lesion andsham-lesion sites ranged from approximately 3.14e4.16 and3.14e3.8 mm posterior to bregma, respectively (Fig. 7a, b). Stimu-lation sites were located in and around the border of the ventraldorsal raphe nucleus and the decussation of the superior cerebellarpeduncle, approximately 7.3e8.3 mm posterior to bregma(Fig. 6aec, 7c). In those rats with MFB electrodes, the distribution

Fig. 3. Time course of the effect of amphetamine on locomotor activity in LHb lesioned and sham-lesioned rats. Alternate beam interruptions were counted during one hour underbaseline (spontaneous) conditions (a) and then during two hours following intraperitoneal injection of vehicle (0.9% saline) (b), 0.5 mg/kg amphetamine (c) or 1.0 mg/kgamphetamine (d). Each beam interruption represents one complete crossing of the locomotor activity chamber. Symbols represent mean � s.e.m locomotor activity counts inlesioned (filled symbols) and sham-lesioned (open symbols) rats during consecutive 10 min periods [spontaneous: sham (n ¼ 26), lesion (n ¼ 26); vehicle: sham (n ¼ 11), lesion(n ¼ 9); 0.5 mg/kg: sham (n ¼ 6), lesion (n ¼ 8); 1.0 mg/kg: sham (n ¼ 9), lesion (n ¼ 9)]. Mean total locomotor activity as a function of amphetamine dose is shown for the first (e)and second (f) hour post-injection in lesioned (filled bars) and sham-lesioned (open bars) rats. Bars represent the mean � s.e.m. *p < 0.01 vs respective vehicle control; yp < 0.01 vssham control.

A.J. Gifuni et al. / Neuropharmacology 63 (2012) 945e957 949

of LHb lesions and sham lesions ranged from approximately3.14e4.16 mm posterior to bregma (Fig. 8a, b). Stimulation siteswere located within the MFB, at the level of the lateral hypothal-amus, approximately 2.56e3.8 mm posterior to bregma (Fig. 8c).

Figs. 9 and 10 show the time course of changes in rewardthreshold as a function of time after injection in rats with elec-trodes in the PM and MFB, respectively. Lesions of the LHb did not

Fig. 4. Time course of the effect of amphetamine on locomotor activity in MHb lesioned anfollows: [spontaneous: sham (n ¼ 14), lesion (n ¼ 16); vehicle: sham (n ¼ 8), lesion (n ¼ 8

alter the reward-enhancing effect of amphetamine at any dose orat either site. In rats with PM electrodes, amphetamine reducedreward thresholds by approximately 25%, an effect that was stablethroughout the 90 min test. Lesions of the LHb did not alterreward thresholds in response to vehicle (F(1, 10) ¼ 0.02,p ¼ 0.89), 0.5 mg/kg (F(1, 9) ¼ 0.12, p ¼ 0.73) or 1.0 mg/kgamphetamine (F(1, 12) ¼ 2.66, p ¼ 0.13) (Fig. 9aec). The bargraphs

d sham-lesioned rats. All details are the same as in Fig. 3. Distribution of rats was as); 1.0 mg/kg: sham (n ¼ 6), lesion (n ¼ 8)].

Fig. 5. Distribution of MHb lesions and sham lesions. (a) Lesions (n ¼ 16 rats) were reconstructed from light microscopic examination of cresyl violet stained tissue, viewed ata magnification of 40�. (b) Schematic representation of the sites of sham lesions (n ¼ 14 rats). Sites were determined by locating the tip of the cannula track under the sameconditions as in (a). Numbers above each level indicate distance (mm) posterior to bregma.

A.J. Gifuni et al. / Neuropharmacology 63 (2012) 945e957950

in Fig. 9d and e show mean threshold values during the first hourand during the last half-hour of the test, respectively. A two-wayANOVA on first hour threshold values revealed only a significanteffect of treatment (F(2, 37) ¼ 17.48, p < 0.01). A similar findingwas obtained for the last half-hour (F(2, 37) ¼ 9.09, p < 0.01).Mean stimulation current intensities in lesioned (337 � 24.64 mA)and sham-lesioned rats (317.8 � 23.66 mA) did not differ(t(41) ¼ 0.56, p ¼ 0.58).

The reduction in reward threshold following amphetamine wasgreater in rats with MFB than with PM electrodes (30e50%).However, two-way ANOVAs on threshold values did not revealany effect of lesion (vehicle: F(1, 10) ¼ 0.10, p¼ 0.76; 0.5 mg/kg: F(1,10) ¼ 0.01, p ¼ 0.92; 1.0 mg/kg: F(1, 10) ¼ 0.98, p ¼ 0.34)

Fig. 6. Locations of electrode tips. Photomicrographs show sample locations of the electrodbundle (d) (Paxinos and Watson, 1997). Scale bar in (a) (100 mm) applies to aec; scale badecussation of the superior cerebellar peduncle.

(Fig. 10aec). Two-way ANOVAs revealed only a significant effect oftreatment during the first hour (F(2, 31) ¼ 46.33, p < 0.1) (Fig. 10d)and second half-hour (F(2, 31) ¼ 39.16, p < 0.01) (Fig. 10e). Meancurrent intensities in lesioned (355.5 � 22.27 mA) and sham-lesioned (362.8 � 32.3 mA) rats responding for MFB stimulationdid not differ (t(35) ¼ 0.19, p ¼ 0.85). There was also no differencebetween mean MFB (359.1 � 19.14 mA) and PM (326.7 � 16.93 mA)current intensities (t(78) ¼ 1.27, p ¼ 0.21).

Figs. 11 and 12 show the time course of changes in asymptoticresponse rates as a function of time after injection in rats withelectrodes in the PM and MFB, respectively. Overall, amphetamineincreased asymptotic response rates by approximately 15e25% inboth PM and MFB rats. LHb lesions did not alter asymptotic

e tips (arrows) within the posterior mesencephalon (aec) and in the medial forebrainr in (d) ¼ 100 mm. Aq, aqueduct of Sylvius; mlf, medial longitudinal fasciculus; xscp,

Fig. 7. Location of ibotenic acid lesions (a), sham lesions (b) and electrode tips (c) of rats with PM electrodes. Lesion and sham-lesion sites were determined as in Fig. 2. Electrode tip(stimulation) sites were determined by locating the lesion (100 mA, 15 s) produced at the tip of the electrode.

A.J. Gifuni et al. / Neuropharmacology 63 (2012) 945e957 951

response rates. In rats with PM electrodes, two-way ANOVAs onresponse rates following vehicle or 1.0 mg/kg amphetamine did notreveal any effect of lesion (vehicle: F(1, 10) ¼ 0.31, p ¼ 0.59; 1.0 mg/kg: F(1, 12) ¼ 0.98, p ¼ 0.34); at the 0.5 mg/kg dose, we observedboth an effect of lesion (F(1, 10) ¼ 8.07, p < 0.05) and time (F(5,

50) ¼ 2.75, p < 0.05) but no interaction (Fig. 11aec). Two-wayANOVAs on first hour and second half-hour mean asymptoticresponse rates revealed only significant effects of treatment (1sthour: F(2, 37) ¼ 5.81, p < 0.01; 2nd half-hour: (F(2, 37) ¼ 4.96,p < 0.05)) (Fig. 11d, e).

Fig. 8. Location of ibotenic acid lesions (a), sham infusions (b) and electrode tips (c) of rats with MFB electrodes. Lesion and sham-lesion sites were determined as in Fig. 2. Electrodetip (stimulation) sites were located as in Fig. 7.

A.J. Gifuni et al. / Neuropharmacology 63 (2012) 945e957952

Similar results were obtained in rats with MFB electrodes.Analysis of time course data did not reveal any effect of lesion(vehicle: F(1, 10) ¼ 0.27, p ¼ 0.61; 0.5 mg/kg: F(1, 10) ¼ 1.31,p ¼ 0.28; 1.0 mg/kg: F(1, 11) ¼ 1.67, p ¼ 0.22) (Fig. 12aec). Analysisof mean asymptotic response rates revealed only an effect oftreatment (F(2, 31) ¼ 4.18, p < 0.05) during the first hour but notthe last half-hour (F(2, 31) ¼ 2.07, p ¼ 0.14) (Fig. 12d, e).

4. Discussion

A fast growing body of research suggests that the habenula, itslateral aspect in particular, exerts strong inhibitory control overmidbrain DA neurons (Christoph et al., 1986; Ji and Shepard, 2007;Matsumoto and Hikosaka, 2007). Here, we examined the contri-bution of this regulation to two behaviours that are highly sensitive

Fig. 9. Time course of the effect of amphetamine on reward threshold in rats with PM electrodes. Thresholds were determined during 90 min following intraperitoneal injection ofvehicle (0.9% saline) (a), 0.5 mg/kg amphetamine (b) or 1.0 mg/kg amphetamine (c). Symbols represent mean � s.e.m threshold values in lesioned (filled symbols) and sham-lesioned (open symbols) rats, and are expressed as a percentage of baseline reward threshold [vehicle: sham (n ¼ 7), lesion (n ¼ 6); 0.5 mg/kg: sham (n ¼ 8), lesion (n ¼ 6);1.0 mg/kg: sham (n ¼ 8), lesion (n ¼ 8)]. Mean reward threshold as a function of amphetamine dose is shown for the first hour (d) and second half-hour (e) post-injection, inlesioned (filled bars) and sham-lesioned (open bars) rats. Bars represent mean � s.e.m threshold values and are expressed as a percentage of baseline.

A.J. Gifuni et al. / Neuropharmacology 63 (2012) 945e957 953

to mesolimbic DA neurotransmission, amphetamine locomotionand enhancement of brain stimulation reward. First, we show thatdestruction of intrinsic LHb neurons results in a pronouncedenhancement of the locomotor-stimulant effect of amphetamine.Lesions of the neighbouring MHb did not replicate this effect on

Fig. 10. Time course of the effect of amphetamine on reward threshold in rats with MFB eleckg: sham (n ¼ 6), lesion (n ¼ 6); 1.0 mg/kg: sham (n ¼ 6), lesion (n ¼ 7).

locomotion, a finding consistent with the distinctly differentneuroanatomical connections of this habenular subregion (Andreset al., 1999). Although MHb function is behaviourally-relevant(Fowler et al., 2011; Glick et al., 2006), it does not appear to bepertinent for psychostimulant locomotion. Second, we show that

trodes. All details are the same as in Fig. 9. Vehicle: sham (n ¼ 6), lesion (n ¼ 6); 0.5 mg/

Fig. 11. Time course of the effect of amphetamine on asymptotic response rates in rats with PM electrodes. Response rates were measured during 90 min following intraperitonealinjection of vehicle (0.9% saline) (a), 0.5 mg/kg amphetamine (b) and 1.0 mg/kg amphetamine (c). Symbols represent mean � s.e.m response rates in lesioned (filled symbols) andsham-lesioned (open symbols) rats, and are expressed as a percentage of baseline rate. Mean asymptotic response rates as a function of amphetamine dose are shown for the firsthour (d) and the second half-hour (e) post-injection in lesioned (filled bars) and sham-lesioned (open bars) rats. Bars represent mean � s.e.m response rates and are expressed asa percentage of baseline.

A.J. Gifuni et al. / Neuropharmacology 63 (2012) 945e957954

LHb lesions do not alter the reward-enhancing effect of amphet-amine on electrical stimulation of the PM or MFB. Thus, althoughour results are consistent with a role of the LHb in modulatingpsychostimulant locomotion, our findings with intracranial self-stimulation suggest that the LHb does not contribute significantly

Fig. 12. Time course of the effect of amphetamine on asymptotic response

to the acute rewarding effect of such drugs. These findingsdemonstrate, for the first time, a clear dissociation between thecontribution of the LHb to these two DA-sensitive behaviours.

Psychostimulant locomotion has not previously been studied inrats with neurotoxic LHb lesions. However, electrolytic lesions have

rates in rats with MFB electrodes. All details are the same as in Fig. 11.

A.J. Gifuni et al. / Neuropharmacology 63 (2012) 945e957 955

been shown to enhance the locomotor response to morphine (Funkand Stewart, 1992) and to apomorphine (Heldt and Ressler, 2006).Although the underlying mechanism for the heightened locomotorresponse is not yet known, it is unlikely to be due solely to disin-hibition of basal mesolimbic DA activity (Lecourtier et al., 2008),since neither we nor others (Funk and Stewart, 1992; Thorntonet al., 1994) observed any post-lesion change in spontaneouslocomotion. We also did not observe any change after administra-tion of vehicle. Similarly, single unit recording studies have shownthat neither neurotoxic lesions of the habenula (Christoph et al.,1986) nor electrolytic lesions of the fasiculus retroflexus (Ji andShepard, 2007) alter baseline DA cell firing. In the present study,a differential behavioural response between lesioned and sham-lesioned groups was only revealed following challenge withamphetamine. This finding suggests hypersensitivity within thesubstrate for amphetamine-induced locomotion in lesioned rats.One possible site for such hypersensitivity is the nucleus accum-bens, a DA terminal site critical for psychostimulant locomotion.Indeed, it has been shown that lesions of the habenula altersynaptic plasticity within the fimbria-accumbens pathway, leadingto increased long-term depression in nucleus accumbens(Lecourtier et al., 2006). Depression of synaptic strength in thisnucleus is a common action of psychostimulants (Nicola et al.,1996) and is associated with a hypersensitive (sensitized) loco-motor response following repeated exposure to these drugs(Beurrier and Malenka, 2002; Thomas et al., 2001).

In contrast, we did not observe any effect of LHb lesions onreward induced by amphetamine, at either of the two stimulationsites. It is perhaps not surprising that we obtained similar resultswith PM and MFB electrodes, since available data suggests that PMand MFB electrodes stimulate two points along the same reward-relevant axons (Boye and Rompré, 1996). We did, however,observe a greater reward-enhancing effect of amphetamine in MFBthan in PM rats. We do not have an explanation for this, but it maybe noteworthy that due to uncontrollable circumstances, the twogroups of rats were tested in different laboratories, housed indifferent animal facilities and tested more than a year apart.Regardless, histological inspection of our lesions did not revealsystematic differences in lesion spread beyond the habenula or inlesion distribution; the highest density of LHb lesions occurred oncoronal planes corresponding to 3.3e3.6 mm posterior to bregma,with generally similar overall anterioreposterior distributions.Lastly, it is unlikely that the differential effects of the lesions on thelocomotor-stimulant and reward-enhancing effects of amphet-amine are due to greater time-dependent neuronal degeneration inthe former, since intracranial self-stimulation tests were carried outlater in time, with reference to the date of surgery, than locomotiontests.

The absence of lesion-induced change in amphetamine rewardfollowing LHb lesions appears at odds with recent self-administration studies showing that similar lesions actuallyincrease sucrose and cocaine seeking, as inferred froma pronounced resistance to extinction of operant responding(Friedman et al., 2010, 2011). The reason for this discrepancy is notclear, but the use of different behavioural measures may be rele-vant. We chose to measure reward threshold with the curve-shiftparadigm because this method allows us to dissociate changes inreward function from changes in operant responding (Edmondsand Gallistel, 1974; Miliaressis et al., 1986). For instance, our datashow that despite significantly higher response rates in lesionedPM rats treated with 0.5 mg/kg amphetamine, the actual reward-enhancing capacity of this dose was not altered. Similarly, usingintravenous self-administration, Wang et al. (2009) showed thatalthough electrolytic lesions of the habenula increase fixed-ratioresponding for heroin, they do not alter heroin reward, as

measured by self-administration breaking points. It appears, then,that despite enhanced behavioural performance following LHblesions, rate-independent measures do not support a role for theLHb in the regulation of positive reward or motivation. Thesefindings further suggest that increases in the rate of operantresponding seen after habenular lesions may more closely reflectchanges in impulsivity (Lecourtier and Kelly, 2005) or behaviouralactivation (e.g. forward locomotion), than in reward function.Behavioural evidence suggesting that different neural substratesunderlie changes in reward and in operant responding, and thatthese show differential sensitivity to drugs acting on DA systems,has previously been reported (Boye and Rompré, 2000).

Earlier studies have shown that rewarding electrical stimula-tion, as well as DA agonists like amphetamine and apomorphine,reliably abate habenular metabolic activity as measuredwith 14C-2-deoxyglucose (2-DG) (Gomita and Gallistel, 1982; McCulloch et al.,1980; Wechsler et al., 1979). In contrast, reward-attenuating dosesof pimozide and haloperidol actually increase 2-DG utilization inLHb (Gomita and Gallistel, 1982; McCulloch et al., 1980; Pizzolatoet al., 1984). These findings with intracranial self-stimulation areconsistent with those of Matsumoto and Hikosaka (2007, 2009a),positing the habenula as a source of negative reward signals. Thefindings with 2-DG also suggest that our neurotoxic lesions andamphetamine treatments had functionally similar metabolic effectson the LHb. That is, reduced LHb function. Together with ourpresent results, these earlier metabolic studies further suggest thatprevious observations of intra-habenular self-stimulation (Vachonand Miliaressis, 1992) as well as attenuation of reward followingelectrolytic lesions of the LHb (Morissette and Boye, 2008) mayhave been due to the respective stimulation and destruction offibres of passage and not to direct actions on intrinsic habenularneurons.

It remains, however, that the actions of amphetamine on LHbactivity and/or accumbal DA neurotransmission appears to havedifferent consequences depending onwhat behavioural read-out ismeasured, despite the fact that both psychostimulant locomotionand reward are thought to rely on mesolimbic DA activity. Wepropose that the dissociation between lesion-induced changes inamphetamine locomotion and reward may reflect the activity ofdifferent subgroups of DA neurons. Such a proposal is tentative butfeasible, given recent findings. Electrophysiological work suggeststhat although a large population of midbrain DA neurons respondsto LHb stimulation, not all DA neurons respond in the samemanner(Christoph et al., 1986; Ji and Shepard, 2007). For instance, it hasbeen shown that different DA neurons respond preferentially topositive or to negative motivational signals (Matsumoto andHikosaka, 2009b). The existence of such heterogeneity amongmidbrain DA responses has led to the hypothesis (Bromberg-Martinet al., 2010) that different sub-populations encode distinct stimulusproperties so as to guide behaviour in a way that is congruent withcontextual cues. Since it appears that the locomotor-stimulant andrewarding effects of amphetamine are respectively mediated viaDA neurotransmission within core and shell subregions of thenucleus accumbens (Sellings and Clarke, 2003), it is plausible thatour results reflect differential control by the LHb of DA neuronsinnervating these terminal sites.

5. Conclusion

In summary, we investigated the contribution of the LHb to DA-sensitive behaviour. To this end, we measured the locomotor-stimulant and reward-enhancing effects of amphetamine in LHblesioned and sham-lesioned rats. We showed that neurotoxiclesions enhanced the locomotor response to amphetamine, whilenot altering amphetamine reward. We propose that the

A.J. Gifuni et al. / Neuropharmacology 63 (2012) 945e957956

dissociation between the locomotor-stimulant and reward-enhancing effects of amphetamine reflects the activity of twodifferent substrates that are functionally dissociable and differen-tially sensitive to LHb regulation.

Acknowledgements

This work was supported by a research grant from the NaturalSciences and Engineering Research Council of Canada (NSERC) toSMB and by a fellowship (SMB) and graduate scholarships (AJG, SJ)from the Fonds de la Recherche en Santé du Québec (FRSQ). Theauthors are grateful to Dr Baptiste Lacoste for help withphotography.

References

Andres, K.H., von, D.M., Veh, R.W., 1999. Subnuclear organization of the rat habe-nular complexes. J. Comp. Neurol. 407, 130e150.

Beurrier, C., Malenka, R.C., 2002. Enhanced inhibition of synaptic transmission bydopamine in the nucleus accumbens during behavioral sensitization to cocaine.J. Neurosci. 22, 5817e5822.

Boye, S.M., Rompré, P.P., 1996. Mesencephalic substrate of reward: axonal connec-tions. J. Neurosci. 16, 3511e3520.

Boye, S.M., Rompré, P.P., 2000. Behavioral evidence of depolarization block ofdopamine neurons after chronic treatment with haloperidol and clozapine.J. Neurosci. 20, 1229e1239.

Brinschwitz, K., Dittgen, A., Madai, V.I., Lommel, R., Geisler, S., Veh, R.W., 2010.Glutamatergic axons from the lateral habenula mainly terminate on GABAergicneurons of the ventral midbrain. Neuroscience 168, 463e476.

Bromberg-Martin, E.S., Matsumoto, M., Hikosaka, O., 2010. Dopamine in motiva-tional control: rewarding, aversive, and alerting. Neuron 68, 815e834.

Bunney, B.S., Aghajanian, G.K., 1976. The precise localization of nigral afferents inthe rat as determined by a retrograde tracing technique. Brain Res. 117,423e435.

Christoph, G.R., Leonzio, R.J., Wilcox, K.S., 1986. Stimulation of the lateral habenulainhibits dopamine-containing neurons in the substantia nigra and ventraltegmental area of the rat. J. Neurosci. 6, 613e619.

Edmonds, D.E., Gallistel, C.R., 1974. Parametric analysis of brain stimulation rewardin the rat: III. Effect of performance variables on the reward summation func-tion. J. Comp. Physiol. Psychol. 87, 876e883.

Fowler, C.D., Lu, Q., Johnson, P.M., Marks, M.J., Kenny, P.J., 2011. Habenular alpha5nicotinic receptor subunit signalling controls nicotine intake. Nature 471,597e601.

Friedman, A., Lax, E., Dikshtein, Y., Abraham, L., Flaumenhaft, Y., Sudai, E., BenTzion, M., Ami-Ad, L., Yaka, R., Yadid, G., 2010. Electrical stimulation of thelateral habenula produces enduring inhibitory effect on cocaine seekingbehavior. Neuropharmacology 59, 452e459.

Friedman, A., Lax, E., Dikshtein, Y., Abraham, L., Flaumenhaft, Y., Sudai, E., BenTzion, M., Yadid, G., 2011. Electrical stimulation of the lateral habenula producesan inhibitory effect on sucrose self-administration. Neuropharmacology 60,381e387.

Funk, D., Stewart, J., 1992. The effects of lesions of the habenular nuclei on thedevelopment of sensitization to the behavioral activational effects of repeatedlyadministered morphine in the rat. Brain Res. 583, 127e136.

Geisler, S., Trimble, M., 2008. The lateral habenula: no longer neglected. CNS. Spectr.13, 484e489.

Glick, S.D., Ramirez, R.L., Livi, J.M., Maisonneuve, I.M., 2006. 18-Methoxycoronaridine acts in the medial habenula and/or interpeduncularnucleus to decrease morphine self-administration in rats. Eur. J. Pharmacol. 537,94e98.

Gomita, Y., Gallistel, C.R., 1982. Effects of reinforcement-blocking doses of pimozideon neural systems driven by rewarding stimulation of the MFB: a 14C-2-deoxyglucose analysis. Pharmacol. Biochem. Behav. 17, 841e845.

Heldt, S.A., Ressler, K.J., 2006. Lesions of the habenula produce stress- anddopamine-dependent alterations in prepulse inhibition and locomotion. BrainRes. 1073e1074, 229e239.

Herkenham, M., Nauta, W.J., 1979. Efferent connections of the habenular nuclei inthe rat. J. Comp. Neurol. 187, 19e47.

Hikosaka, O., 2010. The habenula: from stress evasion to value-based decision-making. Nat. Rev. Neurosci. 11, 503e513.

Hikosaka, O., Sesack, S.R., Lecourtier, L., Shepard, P.D., 2008. Habenula: crossroadbetween the basal ganglia and the limbic system. J. Neurosci. 28, 11825e11829.

Jhou, T.C., Geisler, S., Marinelli, M., Degarmo, B.A., Zahm, D.S., 2009. The meso-pontine rostromedial tegmental nucleus: a structure targeted by the lateralhabenula that projects to the ventral tegmental area of Tsai and substantia nigracompacta. J. Comp. Neurol. 513, 566e596.

Ji, H., Shepard, P.D., 2007. Lateral habenula stimulation inhibits rat midbraindopamine neurons through a GABA(A) receptor-mediated mechanism. J. Neu-rosci. 27, 6923e6930.

Lecourtier, L., Defrancesco, A., Moghaddam, B., 2008. Differential tonic influence oflateral habenula on prefrontal cortex and nucleus accumbens dopamine release.Eur. J. Neurosci. 27, 1755e1762.

Lecourtier, L., Deschaux, O., Arnaud, C., Chessel, A., Kelly, P.H., Garcia, R., 2006.Habenula lesions alter synaptic plasticity within the fimbria-accumbenspathway in the rat. Neuroscience 141, 1025e1032.

Lecourtier, L., Kelly, P.H., 2005. Bilateral lesions of the habenula induce attentionaldisturbances in rats. Neuropsychopharmacology 30, 484e496.

Li, Y.Q., Takada, M., Mizuno, N., 1993. Demonstration of habenular neurons whichreceive afferent fibers from the nucleus accumbens and send their axons to themidbrain periaqueductal gray. Neurosci. Lett. 158, 55e58.

Matsumoto, M., Hikosaka, O., 2007. Lateral habenula as a source of negative rewardsignals in dopamine neurons. Nature 447, 1111e1115.

Matsumoto, M., Hikosaka, O., 2009a. Representation of negative motivational valuein the primate lateral habenula. Nat. Neurosci. 12, 77e84.

Matsumoto, M., Hikosaka, O., 2009b. Two types of dopamine neuron distinctlyconvey positive and negative motivational signals. Nature 459, 837e841.

Matthews-Felton, T., Corodimas, K.P., Rosenblatt, J.S., Morrell, J.I., 1995. Lateralhabenula neurons are necessary for the hormonal onset of maternal behaviorand for the display of postpartum estrus in naturally parturient female rats.Behav. Neurosci. 109, 1172e1188.

McCulloch, J., Savaki, H.E., Sokoloff, L., 1980. Influence of dopaminergic systems onthe lateral habenular nucleus of the rat. Brain Res. 194, 117e124.

Miliaressis, E., Rompré, P.P., Laviolette, P., Philippe, L., Coulombe, D., 1986. The curve-shift paradigm in self-stimulation. Physiol. Behav. 37, 85e91.

Morissette, M.C., Boye, S.M., 2008. Electrolytic lesions of the habenula attenuatebrain stimulation reward. Behav. Brain Res. 187, 17e26.

Nicola, S.M., Kombian, S.B., Malenka, R.C., 1996. Psychostimulants depress excitatorysynaptic transmission in the nucleus accumbens via presynaptic D1-likedopamine receptors. J. Neurosci. 16, 1591e1604.

Nishikawa, T., Fage, D., Scatton, B., 1986. Evidence for, and nature of, the tonicinhibitory influence of habenulointerpeduncular pathways upon cerebraldopaminergic transmission in the rat. Brain Res. 373, 324e336.

Omelchenko, N., Bell, R., Sesack, S.R., 2009. Lateral habenula projections to dopa-mine and GABA neurons in the rat ventral tegmental area. Eur. J. Neurosci. 30,1239e1250.

Paxinos, G., Watson, C., 1997. The Rat Brain in Stereotaxic Coordinates. AcademicPress, San Diego, CA.

Phillipson, O.T., 1979. Afferent projections to the ventral tegmental area of Tsai andinterfascicular nucleus: a horseradish peroxidase study in the rat. J. Comp.Neurol. 187, 117e143.

Pizzolato, G., Soncrant, T.T., Rapoport, S.I., 1984. Haloperidol and cerebral metabo-lism in the conscious rat: relation to pharmacokinetics. J. Neurochem. 43,724e732.

Roiser, J.P., Levy, J., Fromm, S.J., Nugent, A.C., Talagala, S.L., Hasler, G., Henn, F.A.,Sahakian, B.J., Drevets, W.C., 2009. The effects of tryptophan depletion onneural responses to emotional words in remitted depression. Biol. Psychiatry66, 441e450.

Rompré, P.P., Boye, S., 1989. Localization of reward-relevant neurons in the pontinetegmentum: a moveable electrode mapping study. Brain Res. 496, 295e302.

Rompré, P.P., Miliaressis, E., 1985. Pontine and mesencephalic substrates of self-stimulation. Brain Res. 359, 246e259.

Sandyk, R., 1992. Pineal and habenula calcification in schizophrenia. Int. J. Neurosci.67, 19e30.

Sartorius, A., Kiening, K.L., Kirsch, P., von Gall, C.C., Haberkorn, U., Unterberg, A.W.,Henn, F.A., Meyer-Lindenberg, A., 2010. Remission of major depression underdeep brain stimulation of the lateral habenula in a therapy-refractory patient.Biol. Psychiatry 67, e9ee11.

Sellings, L.H., Clarke, P.B., 2003. Segregation of amphetamine reward and locomotorstimulation between nucleus accumbens medial shell and core. J. Neurosci. 23,6295e6303.

Shepard, P.D., Holcomb, H.H., Gold, J.M., 2006. Schizophrenia in translation: thepresence of absence: habenular regulation of dopamine neurons and theencoding of negative outcomes. Schizophr. Bull. 32, 417e421.

Shumake, J., Gonzalez-Lima, F., 2003. Brain systems underlying suscepti-bility to helplessness and depression. Behav. Cogn. Neurosci. Rev. 2,198e221.

Sutherland, R.J., 1982. The dorsal diencephalic conduction system: a review of theanatomy and functions of the habenular complex. Neurosci. Biobehav. Rev. 6,1e13.

Thomas, M.J., Beurrier, C., Bonci, A., Malenka, R.C., 2001. Long-term depression inthe nucleus accumbens: a neural correlate of behavioral sensitization tococaine. Nat. Neurosci. 4, 1217e1223.

Thornton, E.W., Murray, M., Connors-Eckenrode, T., Haun, F., 1994. Dissociation ofbehavioral changes in rats resulting from lesions of the habenula versusfasciculus retroflexus and their possible anatomical substrates. Behav. Neurosci.108, 1150e1162.

Ullsperger, M., von Cramon, D.Y., 2003. Error monitoring using external feedback:specific roles of the habenular complex, the reward system, and the cingulatemotor area revealed by functional magnetic resonance imaging. J. Neurosci. 23,4308e4314.

Vachon, M.P., Miliaressis, E., 1992. Dorsal diencephalic self-stimulation: a movableelectrode mapping study. Behav. Neurosci. 106, 981e991.

Wang, R.Y., Aghajanian, G.K., 1977. Physiological evidence for habenula as major linkbetween forebrain and midbrain raphe. Science 197, 89e91.

A.J. Gifuni et al. / Neuropharmacology 63 (2012) 945e957 957

Wang, Y., Zhang, F., Tang, S., Lai, M., Hao, W., Zhang, Y., Yang, J., Zhou, W., 2009. Lackof effect of habenula lesion on heroin self-administration in rats. Neurosci. Lett.461, 167e171.

Wechsler, L.R., Savaki, H.E., Sokoloff, L., 1979. Effects of d- and l-amphetamine onlocal cerebral glucose utilization in the conscious rat. J. Neurochem. 32, 15e22.

Wirtshafter, D., Asin, K.E., Pitzer, M.R., 1994. Dopamine agonists and stress producedifferent patterns of Fos-like immunoreactivity in the lateral habenula. BrainRes. 633, 21e26.

Wise, R.A., Rompré, P.P., 1989. Brain dopamine and reward. Annu. Rev. Psychol. 40,191e225.