BSR

SECEa

aeb

i1c

oM

Acamtctjatieindadwtmittctcrtfr

KC

*EAthac

Neuroscience 171 (2010) 1180–1186

0d

LOCKADE OF CENTRAL NICOTINE ACETYLCHOLINE RECEPTORIGNALING ATTENUATE GHRELIN-INDUCED FOOD INTAKE IN

ODENTS

Imb2vJrjccateVmttnt(

aa2stw(dbtr2sp2pdsapictaiiw

. L. DICKSON,a E. HRABOVSZKY,b C. HANSSON,a

. JERLHAG,c M. ALVAREZ-CRESPO,a K. P. SKIBICKA,a

. S. MOLNAR,b Z. LIPOSITS,b J. A. ENGELc AND

. EGECIOGLUa*

Department of Physiology/Endocrinology, Institute of Neurosciencend Physiology, The Sahlgrenska Academy at the University of Goth-nburg, Medicinaregatan 11, SE-405 30 Gothenburg, Sweden

Laboratory of Endocrine Neurobiology, Institute of Experimental Med-cine, Hungarian Academy of Sciences, 43 Szigony Street, Budapest083, Hungary

Department of Pharmacology, Institute of Neuroscience and Physiol-gy, The Sahlgrenska Academy at the University of Gothenburg,edicinaregatan 13, SE-405 30 Gothenburg, Sweden

bstract—Here we sought to determine whether ghrelin’sentral effects on food intake can be interrupted by nicotinecetylcholine receptor (nAChR) blockade. Ghrelin regulatesesolimbic dopamine neurons projecting from the ventral

egmental area (VTA) to the nucleus accumbens, partly viaholinergic VTA afferents originating in the laterodorsalegmental area (LDTg). Given that these cholinergic pro-ections to the VTA have been implicated in natural as wells drug-induced reinforcement, we sought to investigatehe role of cholinergic signaling in ghrelin-induced foodntake as well as fasting-induced food intake, for whichndogenous ghrelin has been implicated. We found that.p. treatment with the non-selective centrally activeAChR antagonist, mecamylamine decreased fasting-in-uced food intake in both mice and rats. Moreover, centraldministration of mecamylamine decreased fasting-in-uced food intake in rats. I.c.v. ghrelin-induced food intakeas suppressed by mecamylamine i.p. but not by hexame-

honium i.p., a peripheral nAChR antagonist. Furthermore,ecamylamine i.p. blocked food intake following ghrelin

njection into the VTA. Expression of the ghrelin receptor,he growth hormone secretagogue receptor 1A, was foundo co-localize with choline acetyltransferase, a marker ofholinergic neurons, in the LDTg. Finally, mecamylaminereatment i.p. decreased the ability of palatable food toondition a place preference. These data suggest that gh-elin-induced food intake is partly mediated via nAChRs andhat nicotinic blockade decreases the rewarding properties ofood. © 2010 IBRO. Published by Elsevier Ltd. All rightseserved.

ey words: GHS-R1A, dopamine, appetite, reward-seeking,PP, palatable food.

Corresponding author. Tel: �46-31-786-3876.-mail address: emil.egecioglu@neuro.gu.se (E. Egecioglu).bbreviations: ANOVA, analysis of variance; ChAT, choline acetyl-

ransferase; CPP, conditioned place preference; GHS-R1A, growthormone secretagogue receptor 1A; LDTg, laterodorsal tegmental

area; NAcc, nucleus accumbens; nAChR, nicotine acetylcholine re-eptor; VTA, ventral tegmental area.

306-4522/10 $ - see front matter © 2010 IBRO. Published by Elsevier Ltd. All rightoi:10.1016/j.neuroscience.2010.10.005

1180

t now seems clear that the orexigenic gastric-derived hor-one ghrelin targets CNS circuits involved in both energyalance (Tschop et al., 2000) and reward (Jerlhag et al.,006; Malik et al., 2008). Indeed, ghrelin appears to acti-ate the midbrain dopamine neurons (Abizaid et al., 2006;erlhag et al., 2006) that form a key component of theeward system and include a prominent dopaminergic pro-ection from the ventral tegmental area (VTA) to the nu-leus accumbens (NAcc). I.c.v. injections of ghrelin in-rease accumbal dopamine levels, measured by microdi-lysis (Jerlhag et al., 2006) and also dopamine turnover inhis area (Abizaid et al., 2006). These effects appear to bexerted, at least in part, at the level of the VTA as directTA injection of ghrelin also increases accumbal dopa-ine release (Jerlhag et al., 2007). Consistent with this,

he ghrelin receptor, growth hormone secretagogue recep-or 1A (GHS-R1A) has been identified on VTA dopamineeurons and ghrelin has been shown to increase the elec-rical activity of VTA dopamine neurons recorded in vitroAbizaid et al., 2006).

Ghrelin’s effects on the VTA dopamine system mightlso include indirect effects. Within the VTA, GHS-R1A islso present on pre-synaptic afferents (Abizaid et al.,006). Pharmacological studies investigating the identity ofuch afferents using selective blockers highlight especiallyhe importance of cholinergic and glutamatergic signalingithin the VTA for ghrelin’s dopamine-releasing effects

Jerlhag et al., 2006, 2008, 2010b). Indeed, ghrelin-in-uced accumbal dopamine release was completelylocked by mecamylamine, a non-selective nicotine ace-ylcholine receptor (nAChR) antagonist administered pe-ipherally or directly into the VTA (Jerlhag et al., 2006,008). Taken together with a further study using a moreelective nAChR antagonist, alpha conotoxin MII, to sup-ress ghrelin-induced dopamine release (Jerlhag et al.,008), it seems clear that nicotinic cholinergic signalinglays an important role in ghrelin’s effects on the midbrainopamine system. The precise origin of the ghrelin-re-ponsive cholinergic afferents remains to be elucidatedlthough it seems likely that these include the cholinergicrojection from the laterodorsal tegmental area (LDTg) as

ntra-LDTg injection of ghrelin to mice also increases ac-umbal dopamine release (Jerlhag et al., 2007). Moreover,he ability of ghrelin administered into the LDTg to increaseccumbal dopamine is attenuated by blockade of nAChRs

n the VTA (Jerlhag et al., 2008). Collectively these datandicate that the so-called “cholinergic-dopaminergic re-ard link,” a circuit conferring the incentive value of natural

nd artificial rewards (Yeomans et al., 1993; Lanca et al.,

s reserved.

2p

sdapiaGTogfra

emgtg(saso

A

AwmRiaSam2ilSeeCH

D

AsKm�eTf(Ms1o

srmheamapaKuS

S

MvoatdA

So

Awavfahsam

So

Ssvm

Sor

Fare0pmFjfr

Ca

Tfc

S. L. Dickson et al. / Neuroscience 171 (2010) 1180–1186 1181

000; Rada et al., 2000; Larsson and Engel, 2004), is aotential target for ghrelin’s effects on the reward system.

Recently, we provided evidence that the central ghrelinignaling system is important for reward from both artificialrugs such as alcohol (Jerlhag et al., 2009), cocaine andmphetamine (Jerlhag et al., 2010a) and from rewarding/alatable food (Egecioglu et al., 2010). In the latter study,

ntra-VTA ghrelin injection increased the intake of palat-ble/rewarding food whereas chronic treatment with aHS-R1A antagonist reduced preference for such foods.aken together with the aforementioned effects of ghrelinn the midbrain dopamine system, it seems likely thathrelin increases the incentive value of rewards such asood by activating the VTA dopamine system, however theole of cholinergic afferents in the ghrelin-dopamine inter-ction remains to be elucidated.

In the present study we sought evidence that cholin-rgic signaling, likely involving cholinergic afferents to theesolimbic VTA dopamine system, is important both forhrelin’s effects on food intake and for food reward. Weherefore tested the effects of nAChR blockade on (1)hrelin-induced food intake, (2) fasting-induced food intakethat appears to involve increased endogenous ghrelinignaling) and (3) the ability of palatable food to conditionplace preference, as measure of food reward. Finally, we

ought morphological evidence that cholinergic neuronsriginating in the LDTg express GHS-R1A.

EXPERIMENTAL PROCEDURES

nimals

dult male NMRI mice (25–30 g BWt, B&K, Sollentuna, Sweden)ere used for fasting and i.c.v. ghrelin-induced feeding experi-ents. Adult male Sprague–Dawley rats (�200 g BWt, Charlesiver, Germany or Taconic, Denmark) were used for fasting-

nduced, i.c.v. and intra-VTA ghrelin-induced feeding experimentsnd for food-induced conditioned place preference (CPP) studies.tandard chow (Harlan Teklad, Norfolk, England) and water werevailable ad libitum unless otherwise stated. The animal room wasaintained on a 12/12-h light/dark cycle (with lights on at 0600), at0 °C and 50% humidity. All animals were singly housed following

.c.v. surgery and during the food intake and CPP testing. Theocal Ethics Committee for Animal Experiments (Gothenburg,weden) approved all procedures involving food intake and CPPxperiments. Histological procedures to study GHS-R1A mRNAxpression were reviewed and approved by the Animal Welfareommittee of the Institute of Experimental Medicine (Budapest,ungary).

rugs and administration

cetylated rat ghrelin (Bionuclear, Bromma, Sweden) was dis-olved in vehicle solution (Ringer, NaCl 140 mM, CaCl2 1.2 mM,Cl 3.0 mM and MgCl2 1.0 mM; Merck KGaA, Darmstadt, Ger-any) and administered into the lateral ventricle at a dose of 1g/animal in 1 �l. Injections into the VTA were performed unilat-rally or bilaterally at a dose of 1 �g/animal in a volume of 500 nl.he doses of ghrelin chosen have previously been shown to affect

eeding in rodents and activate the mesolimbic dopamine systemEgecioglu et al., 2010; Naleid et al., 2005; Jerlhag et al., 2006).ecamylamine (Sigma-Aldrich, Schnelldorf, Germany) was dis-

olved in normal saline and administered by i.p. injection at 0.5,.0, 2.0 and 4.0 mg/kg in a volume of 1 ml/kg. The dose interval

f mecamylamine used was selected to cover a range previously R

hown to affect ghrelin-induced locomotor activity and dopamineelease in the NAcc (Jerlhag et al., 2006). For i.c.v. injections ofecamylamine a comparatively low dose of 3 �g was used sinceigher doses have been reported to have unspecific effects, forxample on the N-methyl-D-aspartate (NMDA) receptor (O’Dellnd Christensen, 1988; Papke et al., 2001). Hexamethonium (Sig-a-Aldrich, Munich, Germany) was dissolved in normal saline anddministered at 2 mg/kg i.p., in a volume of 1 ml/kg, a dose usedreviously in comparative studies with mecamylamine of centralnd peripheral effects of nAChR signaling (Tani et al., 1997;amens and Phillips, 2008). All central injections were performedsing a microinfusion pump (CMA 400 syringe pump, Solna,weden).

urgery

ice were implanted with guide cannulae into the dorsal thirdentricle and rats into the lateral ventricle using stereotaxic co-rdinates as described previously (Jerlhag et al., 2007; Salomé etl., 2009). In rats, the coordinates used for cannulae directed athe VTA were 6.0 mm posterior, �0.6 mm lateral and 8.4 mmeep relative to bregma according to (Paxinos and Watson, 1986).ll surgery was preformed at least 4 days before the experiments.

tudies investigating the effects of nAChR blockaden i.c.v. ghrelin-induced food intake in mice and rats

ll feeding experiments commenced in the morning. Satiated miceere i.p. injected with mecamylamine (2 mg/kg) or saline vehiclet 10 min before i.c.v. injection of ghrelin (1 �g) or an equalolume (1 �l) of vehicle. Food intake was measured at 4 hollowing central ghrelin or vehicle administration. Likewise, sati-ted rats were i.p.-injected with either mecamylamine (2 mg/kg),examethonium (2 mg/kg, a peripheral nicotinic antagonist) oraline vehicle at 10 min before i.c.v. injection of ghrelin (1 �g) orn equal volume (1 �l) of vehicle solution. Food intake waseasured at 4 h following central injections.

tudies investigating the effects of nAChR blockaden intra-VTA ghrelin-induced food intake in rats

atiated rats were i.p. injected with mecamylamine (2 mg/kg) oraline at 10 min before injections of ghrelin (1 �g) or an equalolume (0.5 �l) of vehicle solution into the VTA. Food intake waseasured at 4 h following the intra-VTA treatment.

tudies investigating the effects of nAChR blockaden re-feeding following an overnight fast in mice andats

irst, we investigated the effects of peripheral mecamylamine (0.5nd 4 mg/kg) or vehicle solution administered by i.p. injection one-feeding in overnight fasted mice (16 h) and rats (20 h). In allxperiments the injections of mecamylamine took place between900 and 0930 and food intake was followed during the first 4 host-injection in mice and first 3 h post-injection in rats. In theouse study, body weight regain was also measured at 4 h.inally, we investigated the effects of central mecamylamine in-

ection (3 �g) versus an equal volume (3 �l) of vehicle solution onood intake and body weight following an overnight (20 h) fast inats.

o-expression of GHS-R1A and ChAT (a marker ofcetylcholine containing neurons)

o generate a cRNA probe against GHS-R1A, a 837-bp cDNAragment was amplified with PCR from mouse hypothalamicDNA. The amplicon (corresponding to bases 237–1073 of GHS-

1A mRNA; NM_177330) was inserted into plasmid vector using

tkpsFC(caeacbd(clt

NhfdwawU(ewaatGc1swr

Cf

Cstfwcoc(fUcds(atiafswf(oA

hti

S

Dw(plm

nf

Fitai�t(rhoigs1e

ng

Gaommgiba

ni

Wid(dtrwmmo

S. L. Dickson et al. / Neuroscience 171 (2010) 1180–11861182

he PGEM T cloning kit from Promega (Madison, WI, USA). A 2.3b rat choline acetyltransferase (ChAT) cDNA, inserted into theSPT18 plasmid vector, was kindly made available for thesetudies by Dr. S. Berrard (Hôpital de la Pitié-Salpêtrière, Paris,rance). Both plasmids were grown in DH5� cells (Invitrogen,arlsbad, CA, USA), isolated with the QIAfilter Plasmid Maxi kit

Qiagen, Valencia, CA, USA), linearized, purified with phenol/hloroform/isoamyl alcohol (PCI), followed by chloroform/isoamyllcohol (CI) extractions, and then, precipitation with NaCl andthanol. The plasmid templates were transcribed in vitro withppropriate RNA polymerases to yield antisense probes, and forontrol purposes, sense transcripts. The ChAT template was la-eled with digoxigenin-11-UTP (Roche Diagnostics Company, In-ianapolis, IN, USA), and the GHS-R1A template with 35S-UTPNEN Life Science Products, Boston, MA, USA), using reactiononditions detailed elsewhere (Hrabovszky et al., 2004). In dual-abeling studies, the two probes were combined in the hybridiza-ion buffer.

Twelve �m thick coronal sections were prepared through theAcc and the LDTg of four adult male rats. Prehybridizationybridization and posthybridization procedures have been per-

ormed on slide-mounted sections (Hrabovszky et al., 2004). Inual-labeling experiments, the ChAT probe was visualized firstith alkaline phosphatase conjugated anti-digoxigenin antibodiesnd the BCIP/NBT chromogen system (Roche). Then the slidesere coated with 1% Parlodion (SPI-Chem, West Chester, PA,SA), air-dried, dipped into Kodak NTB nuclear track emulsion

Kodak, Rochester, NY, USA; diluted 1:1 with MQ water) andxposed for 4 weeks. The autoradiographic images were developedith Kodak processing chemicals. Dark-field images of emulsionutoradiographs from single-labeling experiments, bright-field im-ges of the Cresyl Violet-stained neighboring sections and sec-

ions dual-labeled for the simultaneous detection of ChAT andHS-R1A mRNAs were scanned with an AxioCam MRc 5 digitalamera mounted on a Zeiss AxioImager M1 microscope, using a0� objective lens and the Mozaics Modul of the AxioVision 4.6oftware (Carl Zeiss, Göttingen, Germany). The digital imagesere processed with the Adobe Photoshop 7.1 software at 300 dpi

esolution.

onditioned place preference (CPP) for palatableood

PP testing for palatable food was preformed as previously de-cribed (Egecioglu et al., 2010). In this test, the rats spend moreime in a rewarding food paired environment. Briefly, rats wereree to explore the entire two compartment apparatus for 20 minithout food present and initial preference was scored. During theonditioning phase (day 2–10) the rats were confined for 20 min tone of the two compartments in the morning and to the otherompartment in the afternoon. The least preferred compartmentdetermined from the pretest) was always paired with rewardingood (five pellets of Non-stop chocolate, Marabou, Kraft Foods,pplands Väsby, Sweden), the other side was paired to standardhow and the conditioning phase was balanced so that the con-itioned stimulus was alternated between morning and afternoonessions. On day 12 rats were i.p. injected with either mecamylamine1 or 2 mg/kg) or saline 10 min before being placed in the CPPpparatus for 20 min. On day 14 the rats received the oppositereatment in a counterbalanced design at 10 min before being placedn the CPP apparatus for 20 min. In this paradigm in which thentagonist was administered before the acute CPP test, we testor expression of a CPP response, rather than acquisition, con-olidation or extinction of the CPP response, parameters thatould require further testing. The animals had ad libitum access to

ood between sessions. During the tests on the two testing daysday 12 and 14) the animals were video recorded and the amountf time spent in each compartment was scored by visual analysis.

ll rats were habituated to the palatable food for 5 days in their (

ome cages before the CPP testing to avoid neophobia during theesting. During the conditioning phase, rats were habituated to i.p.njections of saline five times.

tatistics

ata were analyzed using Student’s t-tests, sign-rank test, one-ay, two-way or repeated measures analysis of variance

ANOVA) followed by Bonferroni post hoc test or the Fisher’srotected least significant difference (PLSD) test. The significance

evel was P�0.05 for all experiments. All data are presented aseans�SEM.

RESULTS

AChR blockade: impact on i.c.v. ghrelin-inducedeeding in rats and mice

irst we tested whether nAChRs are involved in ghrelin-nduced food intake, in studies involving peripheral pre-reatment with the centrally active non-selective nAChRntagonist, mecamylamine. Mecamylamine pretreatment

.p. attenuated 4 h i.c.v. ghrelin-induced food intake by50% in both ad lib fed mice and rats (treatment interac-

ion, two way ANOVA in mice (F(1,30)�5.1, P�0.05), in ratF(1,42)�10.4, P�0.01), Fig. 1A, B). In contrast to theesults obtained for mecamylamine, rats pretreated withexamethonium, a non-selective peripheral nAChR antag-nist, did not display an attenuated 4 h feeding response to

.c.v. ghrelin (sal/veh: 0.73�0.28 g, sal/ghrelin: 4.34�0.35, hex/veh: 0.59�0.22 g, hex/ghrelin: 3.44�0.42 g, noignificant treatment interaction, two way ANOVA (F(1,42)�.32, P�0.25), n�11–12 per group). Neither of the cholin-rgic blockers affected food intake per se.

AChR blockade: impact on intra-VTAhrelin-induced feeding in rats

iven that intra-VTA ghrelin injection increases food intakend dopamine release in the NAcc and that the latter effectn dopamine is completely blocked by pretreatment withecamylamine, we sought to determine whether meca-ylamine pretreatment suppresses the effects of intra-VTAhrelin on food intake. Injections of ghrelin into the VTA

ncreased 4 h food intake in rats; this effect was completelylocked by mecamylamine pretreatment (treatment inter-ction, two way ANOVA (F(1,69)�4.731, P�0.05), Fig. 2).

AChR blockade: impact on fasting-induced foodntake in rats and mice

e also explored the effects of nAChR blockade on foodntake that follows an overnight fast and appears to beependent on increased endogenous ghrelin signalingSalomé et al., 2009). Both doses of mecamylamine testedecreased food intake during the period immediately afterhe fast (one way ANOVA, mice (F(1,22)�11.4, P�0.001),at (F(1,14)�4.7, P�0.05) Fig. 3A, B). Furthermore, bodyeight regain in mice was decreased by both doses ofecamylamine at the 4 h time point (saline: 0.77�0.15 g,ec 0.5 mg/kg: 0.31�0.13 g, mec 4 mg/kg: 0.23�0.11 g,ne way ANOVA followed by Bonferroni post hoc test

F(2,22)�5.60, P�0.05), n�6–9). Body weights were not

rCiw1vpm7

Ct

IwdR

nplmro

nf

Gtmattmdtcbtmp

TwwngbGednss

FTitP eatment

Feknf4

S. L. Dickson et al. / Neuroscience 171 (2010) 1180–1186 1183

ecorded in the fasting-induced feeding experiment in rats.entral i.c.v. administration of mecamylamine (3 �g/rat)

nhibited fasting-induced food intake over time, comparedith vehicle treatment (veh 1 h: 17.28�0.91, mec 1 h:4.43�1.87, veh 4 h: 21.70�1.45, mec 4 h: 16.21�1.15,eh 24 h: 201.31�6.21, mec 24 h: 167.43�12.78, valuesresented are food (gram)/body weight (kg), repeatedeasures two way ANOVA (F(2,26)�3.76, P�0.05), n�–8).

olocalization of GHS-R1A on cholinergic neurons inhe LDTg

n order to identify possible neuronal pathways throughhich ghrelin might affect food intake in an acetylcholine-ependent manner, we examined the expression of GHS-1A mRNA using in situ hybridization techniques. While

ig. 1. I.c.v. Ghrelin-induced feeding is attenuated by peripheral adhe effects of i.c.v. injection of ghrelin (versus saline vehicle soluti

n (A) mice and (B) rats. In both experiments mecamylamine (2 mgreatments. (A) ** P�0.01, *** P�0.001, two-way ANOVA followed b�0.001, two-way ANOVA followed by Bonferroni post hoc test (tr

ig. 2. Intra-VTA ghrelin-induced food intake is attenuated by periph-ral administration of the nAChR antagonist, mecamylamine (2 mg/g), in rats. Food intake was measured at 4 h post ghrelin injection. ns,ot significant; * P�0.05, ** P�0.01, *** P�0.001 two-way ANOVA

eollowed by Bonferroni post hoc test (treatment interaction (F(1,69)�.731, P�0.05), n�12–22).

o GHS-R1A signal was found in the NAcc, it was ex-ressed abundantly in the LDTg of rats (Fig. 4B). Dual-

abel in situ hybridization experiments identified GHS-R1ARNA signal in ChAT mRNA expressing cholinergic neu-

ons of the LDTg (Fig. 4C). Nissl-stained reference imagesf neighboring sections are shown for orientation (Fig. 4A).

ACHR blockade: condition place preference testingor palatable food

iven that acetylcholine signaling is implicated in motiva-ional and rewarding aspects of food intake we tested ifecamylamine treatment could affect the ability of palat-ble food to condition a place preference. Expression ofhe CPP response to chocolate pellets was decreased byreatment with mecamylamine (2 mg/kg) administered im-ediately before CPP testing (Fig. 5A). There was noifference between vehicle and mecamylamine (2 mg/kg)reatment in the number of crossings between the twoompartments (Fig. 5B). However, a decrease in the num-er of crossings was noted between the pretest and thewo test days. The lower dose of mecamylamine used (1g/kg) did not effect the time spent in the rewarding food-aired compartment.

DISCUSSION

he present series of experiments are built around ourorking hypothesis that the cholinergic-dopaminergic re-ard link, a circuit that increases the incentive value ofatural and artificial rewards, mediates, at least in part,hrelin’s orexigenic effects. This hypothesis was inspiredoth by neuroanatomical studies of the distribution ofHS-R1A, that include both the VTA and the LDTg (Guant al., 1997) and by our previous finding that ghrelin’sopamine-releasing effects can be blocked by suppressingAChR signaling (Jerlhag et al., 2006, 2007, 2008). Con-istent with our hypothesis, the main finding in the presenttudy is that nAChR signaling is required for the orexigenic

on of the nAChR antagonist, mecamylamine, in both mice and rats.od intake were measured during the 4 h period following injectionhicle was administered by i.p. injection 10 min prior to the centraltest (treatment interaction (F(1,30)�5.1, P�0.05), n�7– 8). (B) ***interaction (F(1,42)�10.4, P�0.01), n�12).

ministration) on fo/kg) or vey PLSD

ffects of ghrelin exerted at the level of the VTA.

pme(2opibobtaegdgmdd

ottdaeV

drgZcoaprCtia2c2ae(ats

nwesHcnTsceergmptIcoipuleb

Fpmp3f6(

S. L. Dickson et al. / Neuroscience 171 (2010) 1180–11861184

Ghrelin’s orexigenic effects appear to be mediated byathways involved in both non-homeostatic feeding (e.g.esolimbic reward pathways) as well as those involved innergy balance (e.g. the hypothalamus and brainstem)Wren et al., 2001; Faulconbridge et al., 2003; Naleid et al.,005). For this reason, our studies investigating the effectsf nAChR blockade on ghrelin-induced food intake incor-orate both i.c.v. delivery of ghrelin as well as ghrelin

njection directly into the VTA, a key node in the mesolim-ic reward circuits. We found that peripheral administrationf a centrally active nAChR antagonist (mecamylamine)ut not a peripherally active nAChR antagonist (hexame-honium) suppresses i.c.v. ghrelin-induced food intake bys much as 50%. As mecamylamine did not fully block theffect of i.c.v. ghrelin we might infer that there also existhrelin-sensitive orexigenic pathways that are indepen-ent of nAChR signaling. When we focused specifically onhrelin’s actions at the level of the VTA, we found that i.p.ecamylamine completely blocked intra-VTA ghrelin-in-uced feeding. Thus both the orexigenic and accumbal

ig. 3. The re-feeding following an overnight fast is attenuated byeripheral administration of the nAChR antagonist, mecamylamine, inice (A) and rats (B). Food intake was measured immediately aftereripheral mecamylamine injection, during a 4 h period in mice and ah period in rat following the fast (A) ** P�0.01 one-way ANOVA

ollowed by Bonferroni post hoc test (F(1,22)�11.4, P�0.001), n�–9. (B) * P�0.05 one-way ANOVA followed by PLSD post hoc test

F(1,14)�4.7, P�0.05), n�5–6.

opamine-releasing effects of ghrelin exerted at the level a

f the VTA (Naleid et al., 2005; Jerlhag et al., 2007) appearo be dependent on nAChR signaling, implicating a role forhe cholinergic-dopaminergic reward link in ghrelin-in-uced food intake. Our findings suggest that nAChR block-de provides a novel therapeutic route to interrupt ghrelin’sffects on food intake that are exerted at the level of theTA.

Our hypothesis that ghrelin activates the cholinergic-opaminergic reward link is also supported by novel neu-oanatomical data. It is well established that GHS-R1A, thehrelin receptor, is present in the VTA (Guan et al., 1997;igman et al., 2006) where it has been shown to be colo-alized on a sub-population of dopamine cells and also onther cells in this region (Abizaid et al., 2006). GHS-R1A islso present in the LDTg (Guan et al., 1997) and, in theresent study, we show its presence on cholinergic neu-ons in this area (by colocalization between GHS-R1A andhAT mRNAs). Again, these data follow the aforemen-

ioned dopamine measurements in which ghrelin injectionnto the LDTg stimulates accumbal dopamine release in acetylcholine-dependent manner (Jerlhag et al., 2007,008). The LDTg appears to provide the major afferentholinergic projection to the VTA (Butcher and Woolf,003). The possibility remains that additional cholinergicfferents play a role in ghrelin’s orexigenic effects; cholin-rgic neurons with in the pedunculopontine tegmental areaPPTg) are activated by ghrelin in vitro (Kim et al., 2009)nd constitute an additional source of cholinergic input tohe VTA, although these neurons mainly project to theubstantia nigra (Beninato and Spencer, 1987).

It might seem rather paradoxical to propose thatAChR antagonism to suppress food intake and food re-ard, when nicotine itself has powerful anorexigenic prop-rties, with weight gain being a common occurrence aftermoking cessation (Klesges et al., 1989; Jo et al., 2002).owever, there are indications that nicotine addictionauses desensitization and a subsequent up-regulation ofAChRs (Mansvelder et al., 2002; Wooltorton et al., 2003;apper et al., 2006) and that the hyperphagia followingmoking cessation could very likely be because of in-reased expression of and hypersensitivity of nAChRs (Jot al., 2002), thereby increasing the output of the endog-nous cholinergic pathways involved in reinforcement andeward seeking. In line with a role for nAChRs in weightain following smoking cessation long term treatment withecamylamine decreases body weight in genetic and hy-othalamic lesion models of obesity, effects attributed en-irely to a decrease in food intake (Dulloo and Miller, 1986).nterestingly, feeding is associated with increased extra-ellular acetylcholine in the VTA (Rada et al., 2000). Therexigenic overlap between ghrelin and cholinergic signal-

ng via nAChRs might converge within the VTA. Howeverossible effects in other brain areas implicated in the reg-lation of food intake, such as the arcuate nucleus and the

ateral hypothalamus or the NAcc which contain cholin-rgic neurons or receive cholinergic innervations, can note ruled out.

In the present study, we tested the effects of nAChR

ntagonism on two different physiological models in which

en22nwIot(fRBch

mwgerstp

APvKsIF(FCtT

A

B

B

F((d detectionG oint to du

Fnimcdtt

S. L. Dickson et al. / Neuroscience 171 (2010) 1180–1186 1185

ndogenous central ghrelin signaling has been implicated,amely in the CPP test of food reward (Egecioglu et al.,010) and fasting-induced re-feeding (Salomé et al.,009). Thus, as shown previously for GHS-R1A antago-ists, expression of the CPP response for palatable foodas suppressed by the nAChR antagonist mecamylamine.

n accord with our findings on palatable food-induced CPPther studies have shown that peripheral mecamylamine

reatment also decreases drug-induced expression of CPPSershen et al., 2010; Zarrindast et al., 2005). The re-eeding after an overnight fast is suppressed both by GHS-1A (Salomé et al., 2009) and by nAChR antagonists.oth peripheral and central delivery of mecamylamine atomparatively low doses block fasting-induced feeding;owever, neither central nor peripheral injections of

ig. 4. Acetylcholine positive neurons in the LDTg express GHS-R1Cresyl) reference images of neighboring sections (A). Arrows point to tclusters of silver grains) formed a conspicuous cell group in the LDetection of ChAT mRNA (purple BCIP/NBT stain) with the isotopicHS-R1A expression in cholinergic neurons of the LDTg (C). Arrows p

ig. 5. Effects of peripheral pretreatment (10 min prior to test) with theAChR antagonist (mecamylamine, 2 mg/kg) on the ability of reward-

ng food to condition a place preference. (A) Pretreatment withecamylamine decreased the time spent in the palatable food paired

ompartment. (B) The activity (crossings between compartments) wasecrease by both vehicle and mecamylamine treatment compared

o the pretest. n�18, * P�0.001, ** P�0.001, *** P�0.001, Sign rankest.

ecamylamine had any effects on food intake or bodyeight in non-fasted animals. Collectively these data sug-est that nAChR signaling might be of importance for theffects of endogenous ghrelin on food intake and foodeward, although further experiments would be required tohow interaction between nAChR signaling and ghrelin inhese models and/or to exclude the role of ghrelin-inde-endent mechanisms.

cknowledgments—We thank Dr. S. Berrard at the Hôpital de laitié-Salpêtrière, Paris, France for providing the pSPT18 plasmidector containing the choline acetyltransferase (ChAT) cDNA,osrat Latife for food intake measurements. The research wasupported by Konrad and Helgfrid Johanssons Fond, Fredrik andngrid Thurings Stiftelse, the EU (FP7-HEALTH- 2009–241592,P7-KBBE-2009-3-245009), Swedish Medical Research CouncilK2007-54X-20328–013), ALF Göteborg (SU7601), the Swedishoundation for Strategic Research to Sahlgrenska Center forardiovascular and Metabolic Research (A305-188) and the Na-

ional Science Foundation of Hungary (OTKA K69127 and73002).

REFERENCES

bizaid A, Liu ZW, Andrews ZB, Shanabrough M, Borok E, ElsworthJD, Roth RH, Sleeman MW, Picciotto MR, Tschop MH, Gao XB,Horvath TL (2006) Ghrelin modulates the activity and synapticinput organization of midbrain dopamine neurons while promotingappetite. J Clin Invest 116:3229–3239.

eninato M, Spencer RF (1987) A cholinergic projection to the ratsubstantia nigra from the pedunculopontine tegmental nucleus.Brain Res 412:169–174.

utcher LL, Woolf NJ (2003) Cholinergic neurons and networks revis-ited. In: The rat nervous system (Paxinos G, ed) Amsterdam:

. The distribution of GHS-R1A mRNA was related to Nissl-stained. In single-labeling experiments, neurons expressing GHS-R1A mRNAual-labeling studies, combining the non-isotopic in situ hybridizationof GHS-R1A mRNA (autoradiographic silver grains), demonstratedal-labeled neurons. Scale bars�500 �m in (A) and 100 �m in (B, C).

A mRNAhe LDTgTg (B). D

Elsevier.

D

E

F

G

H

J

J

J

J

J

J

J

K

K

K

L

L

M

M

N

O

P

P

R

S

S

T

T

T

W

W

Y

Z

Z

S. L. Dickson et al. / Neuroscience 171 (2010) 1180–11861186

ulloo AG, Miller DS (1986) The effect of parasympathetic drugs onenergy expenditure: relevance to the autonomic hypothesis. CanJ Physiol Pharmacol 64:586–591.

gecioglu E, Jerlhag E, Salome N, Skibicka KP, Haage D, BohloolyYM, Andersson D, Bjursell M, Perrissoud D, Engel JA, Dickson SL(2010) Ghrelin increases intake of rewarding food in rodents. Ad-dict Biol 15:304–311.

aulconbridge LF, Cummings DE, Kaplan JM, Grill HJ (2003) Hy-perphagic effects of brainstem ghrelin administration. Diabetes52:2260–2265.

uan XM, Yu H, Palyha OC, McKee KK, Feighner SD, SirinathsinghjiDJ, Smith RG, Van der Ploeg LH, Howard AD (1997) Distributionof mRNA encoding the growth hormone secretagogue receptor inbrain and peripheral tissues. Brain Res Mol Brain Res 48:23–29.

rabovszky E, Kallo I, Steinhauser A, Merchenthaler I, Coen CW,Petersen SL, Liposits Z (2004) Estrogen receptor-beta in oxytocinand vasopressin neurons of the rat and human hypothalamus:immunocytochemical and in situ hybridization studies. J CompNeurol 473:315–333.

erlhag E, Egecioglu E, Dickson SL, Andersson M, Svensson L, EngelJA (2006) Ghrelin stimulates locomotor activity and accumbal do-pamine-overflow via central cholinergic systems in mice: implica-tions for its involvement in brain reward. Addict Biol 11:45–54.

erlhag E, Egecioglu E, Dickson SL, Douhan A, Svensson L, Engel JA(2007) Ghrelin administration into tegmental areas stimulates lo-comotor activity and increases extracellular concentration of dopa-mine in the nucleus accumbens. Addict Biol 12:6–16.

erlhag E, Egecioglu E, Dickson SL, Engel JA (2010a) Ghrelin recep-tor antagonism attenuates cocaine- and amphetamine-induced lo-comotor stimulation, accumbal dopamine release, and conditionedplace preference. Psychopharmacology (Berl) 211:415–422.

erlhag E, Egecioglu E, Dickson SL, Engel JA (2010b) Glutamatergicregulation of ghrelin-induced activation of the mesolimbic dopa-mine system. Addict Biol; doi: 10.1111/j.1369-1600.2010.00231.x.

erlhag E, Egecioglu E, Dickson SL, Svensson L, Engel JA (2008)Alpha-conotoxin MII-sensitive nicotinic acetylcholine receptors areinvolved in mediating the ghrelin-induced locomotor stimulationand dopamine overflow in nucleus accumbens. Eur Neuropsycho-pharmacol 18:508–518.

erlhag E, Egecioglu E, Landgren S, Salome N, Heilig M, Moechars D,Datta R, Perrissoud D, Dickson SL, Engel JA (2009) Requirementof central ghrelin signaling for alcohol reward. Proc Natl Acad SciU S A 106:11318–11323.

o YH, Talmage DA, Role LW (2002) Nicotinic receptor-mediatedeffects on appetite and food intake. J Neurobiol 53:618–632.

amens HM, Phillips TJ (2008) A role for neuronal nicotinic acetylcho-line receptors in ethanol-induced stimulation, but not cocaine- ormethamphetamine-induced stimulation. Psychopharmacology 196:377–387.

im J, Nakajima K, Oomura Y, Wayner MJ, Sasaki K (2009) Electro-physiological effects of ghrelin on pedunculopontine tegmentalneurons in rats: an in vitro study. Peptides 30:745–757.

lesges RC, Meyers AW, Klesges LM, La Vasque ME (1989) Smok-ing, body weight, and their effects on smoking behavior: a com-prehensive review of the literature. Psychol Bull 106:204–230.

anca AJ, Adamson KL, Coen KM, Chow BL, Corrigall WA (2000) Thepedunculopontine tegmental nucleus and the role of cholinergicneurons in nicotine self-administration in the rat: a correlative

neuroanatomical and behavioral study. Neuroscience 96:735–742.

arsson A, Engel JA (2004) Neurochemical and behavioral studies onethanol and nicotine interactions. Neurosci Biobehav Rev 27:713–720.

alik S, McGlone F, Bedrossian D, Dagher A (2008) Ghrelin modu-lates brain activity in areas that control appetitive behavior. CellMetab 7:400–409.

ansvelder HD, Keath JR, McGehee DS (2002) Synaptic mecha-nisms underlie nicotine-induced excitability of brain reward areas.Neuron 33:905–919.

aleid AM, Grace MK, Cummings DE, Levine AS (2005) Ghrelininduces feeding in the mesolimbic reward pathway between theventral tegmental area and the nucleus accumbens. Peptides26:2274–2279.

’Dell TJ, Christensen BN (1988) Mecamylamine is a selective non-competitive antagonist of N-methyl-D-aspartate- and aspartate-in-duced currents in horizontal cells dissociated from the catfishretina. Neurosci Lett 94:93–98.

apke RL, Sanberg PR, Shytle RD (2001) Analysis of mecamylaminestereoisomers on human nicotinic receptor subtypes. J PharmacolExp Ther 297:646–656.

axinos G, Watson C (1986) The rat brain in stereotaxic coordinates.San Diego, CA: Academic Press.

ada PV, Mark GP, Yeomans JJ, Hoebel BG (2000) Acetylcholinerelease in ventral tegmental area by hypothalamic self-stimulation,eating, and drinking. Pharmacol Biochem Behav 65:375–379.

alomé N, Haage D, Perrissoud D, Moulin A, Demange L, EgeciogluE, Fehrentz JA, Martinez J, Dickson SL (2009) Anorexigenic andelectrophysiological actions of novel ghrelin receptor (GHS-R1A)antagonists in rats. Eur J Pharmacol 612:167–173.

ershen H, Hashim A, Lajtha A (2010) Differences between nicotineand cocaine-induced conditioned place preferences. Brain ResBull 81:120–124.

ani Y, Saito K, Tsuneyoshi A, Imoto M, Ohno T (1997) Nicotinicacetylcholine receptor (nACh-R) agonist-induced changes in brainmonoamine turnover in mice. Psychopharmacology 129:225–232.

apper AR, Nashim R, Lester HA (2006) Neuronal nicotinic acetylcho-line receptor and nicotine dependence. In: Cell biology of addiction(Madras BK, et al, eds) Pasadena, CA: Cold Spring Harbor Labo-ratory Press.

schop M, Smiley DL, Heiman ML (2000) Ghrelin induces adiposity inrodents. Nature 407:908–913.

ooltorton JR, Pidoplichko VI, Broide RS, Dani JA (2003) Differentialdesensitization and distribution of nicotinic acetylcholine receptorsubtypes in midbrain dopamine areas. J Neurosci 23:3176–3185.

ren AM, Seal LJ, Cohen MA, Brynes AE, Frost GS, Murphy KG,Dhillo WS, Ghatei MA, Bloom SR (2001) Ghrelin enhances appe-tite and increases food intake in humans. J Clin Endocrinol Metab86:5992.

eomans JS, Mathur A, Tampakeras M (1993) Rewarding brain stim-ulation: role of tegmental cholinergic neurons that activate dopa-mine neurons. Behav Neurosci 107:1077–1087.

arrindast MR, Fattahi Z, Rostami P, Rezayof A (2005) Role of thecholinergic system in the rat basolateral amygdala on morphine-induced conditioned place preference. Pharmacol Biochem Behav82:1–10.

igman JM, Jones JE, Lee CE, Saper CB, Elmquist JK (2006) Expres-sion of ghrelin receptor mRNA in the rat and the mouse brain.

J Comp Neurol 494:528–548.

(Accepted 2 October 2010)(Available online 8 October 2010)