ilable at ScienceDirect

Alcohol 48 (2014) 677e685

Contents lists ava

Alcohol

journal homepage: http: / /www.alcohol journal .org/

Ethanol intake and ethanol-conditioned place preference arereduced in mice treated with the bioflavonoid agent naringin

Amine Bahi a,*, Syed M. Nurulain b, Shreesh Ojha b

aDepartment of Anatomy, College of Medicine & Health Sciences, United Arab Emirates University, P.O. Box 17666, Al Ain, United Arab EmiratesbDepartment of Pharmacology & Therapeutics, College of Medicine & Health Sciences, United Arab Emirates University, Al Ain, United Arab Emirates

Keywords:Conditioned place preferenceGW9662EthanolNaringinPeroxisome proliferator-activated receptorsTwo-bottle choice

Abbreviations: CPP, conditioned-place preference;line receptors; NAR, naringin; PPAR, peroxisome prTZD, thiazolidinedione.

Disclosure/Conflict of interest: The authors havemight be perceived to influence the results or the discu* Corresponding author. Tel.: þ971 3 7137 516; fax:

E-mail address: amine.bahi@uaeu.ac.ae (A. Bahi).

http://dx.doi.org/10.1016/j.alcohol.2014.06.0080741-8329/� 2014 Elsevier Inc. All rights reserved.

a b s t r a c t

Recently, PPAR-g activation has emerged as a potential treatment for alcoholism. However, the adverseeffects of synthetic PPAR-g activators, despite being effective drugs, prompted the need for novel PPAR-gagonists that retain efficacy and potency with a lower potential of side effects. Hence, naringin, a bio-flavonoid isolated from citrus fruits and recently identified as a natural ligand of PPAR-g, has begun to beevaluated for treatment of alcoholism. It is well known to possess several therapeutic benefits in additionto its anti-anxiety and antidepressant properties. In the present study, we assessed whether naringintreatment possesses anti-ethanol reward properties in C57BL/6 mice. We used the two-bottle choicedrinking paradigm and ethanol-induced conditioned place preference (CPP) to examine the effect ofnaringin treatment on ethanol drinking. Results have shown that, compared with vehicle, naringin (10e100 mg/kg) significantly and dose-dependently decreased voluntary ethanol intake and preference in atwo-bottle choice drinking paradigm [3e15% (v/v) escalating over 2 weeks], with no significant effectobserved on saccharin [0.02e0.08% (w/v)] or on quinine [15e60 mM (w/v)] intake. In addition, there wasno significant difference in blood ethanol concentration (BEC) between groups following naringinadministration of 3 g of ethanol/kg body weight. Interestingly, when mice were treated with vehicle ornaringin (30 mg/kg) before injection of ethanol (1.5 g/kg) during conditioning days, naringin inhibitedthe acquisition of ethanol-CPP. More importantly, these effects were significantly attenuated when micewere pre-injected with the peroxisome proliferator-activated receptor-g (PPAR-g) antagonist, GW9662.Taken together, the present findings are the first to implicate naringin and PPAR-g receptors in thebehavioral and reward-related effects of ethanol and raise the question of whether specific drugs thattarget PPAR-g receptors could potentially reduce excessive ethanol consumption and preference.

� 2014 Elsevier Inc. All rights reserved.

Introduction

The World Health Organization (WHO) estimates that approxi-mately 2.5 million people die every year from ethanol use. Whileseveral therapeutic options are available for the treatment of ethanoldependence, all existing therapies have only modest efficacy. One ofthe most exciting developments in the treatment of alcoholism wasthe advent of medications such as naltrexone and acamprosate(Rezvani, Lawrence, Arolfo, Levin, & Overstreet, 2012; Soyka &Rösner, 2010). The development of these medications provided a

nAChRs, nicotinic acetylcho-oliferator-activated receptor;

no financial interests thatssion reported in this article.þ971 3 7672 033.

proof-of-concept for the feasibility of integrated, pharmacologicallyaided treatment of alcoholism, but also pointed to an urgent need foridentifying additional pharmacological approaches, because only aminority of patients benefitted from these drug treatments (Heilig,Goldman, Berrettini, & O’Brien, 2011; Myrick & Anton, 1998).

Very recently, peroxisome proliferator-activated receptors(PPARs), which are ligand-activated transcription factors of the nu-clear hormone receptor superfamily, have appeared to be involved inethanol addiction. PPARs were identified in the neurons, the oligo-dendrocytes, and the astrocytes of the central nervous system (CNS)(Gofflot et al., 2007; Sarruf et al., 2009). Within the 3 distinct PPARisoforms (PPARa, PPARb/d, and PPAR-g), in particular, PPAR-g ishighly expressed in the lateral hypothalamus (LH), the para-ventricular nucleus (PVN) of the hypothalamus, the arcuate nucleus(ARC), and the ventral tegmental area (VTA). In LH and ARC, PPAR-g isexpressed in the cells producing a-melanocyte-stimulating hormone(a-MSH), agouti-related protein (AgRP), and pro-opiomelanocortin

A. Bahi et al. / Alcohol 48 (2014) 677e685678

(POMC). In the VTA, PPAR-g co-localizes with tyrosine hydroxylase(TH)-positive cells suggesting the expression of PPAR-g in dopami-nergic cells (Sarruf et al., 2009). The activation of PPAR-g has beenshown to protect neurons against NMDA-mediated excitotoxicityand inflammatory damage (Tontonoz & Spiegelman, 2008). Also, theactivation of PPAR-g reducedmethamphetamine-induced locomotorsensitization in mice (Maeda et al., 2007).

A recent report demonstrated that activation of PPAR-g by pio-glitazone, a prototype thiazolidinedione (TZD) approved for thetreatment of type 2 diabetes and insulin resistance, has been shownto reduce ethanol drinking and stress-induced relapse to ethanol-seeking behavior, and prevented withdrawal signs in alcohol-dependent rats (Stopponi et al., 2011). The emergence of PPAR-gactivation as a potential treatment for alcoholism has opened theinvestigation of using PPAR-g agonists in pharmacotherapy ofaddiction-related conditions, including ethanol addiction. However,the PPAR-g activators, despite being effective drugs, present seriousconcerns due to increased risk of fluid retention, weight gain, boneloss, bladder cancer, and congestive heart failure. These concernsand the potential for side effects led to a decrease in the clinical useof TZDs and further prompted the need for novel PPAR-g agonists,which would retain efficacy and potency like TZDs but have a lowerpotential of side effects (Cariou, Charbonnel, & Staels, 2012).

Over the past few decades, in the search for better medications,naturally derived bioactive phytochemical agents have attractedmuch attention as potential therapeutic agents for alcoholism, dueto their multiple targets and fewer side effects. Recently, numerousnaturally derived phytochemicals have been identified as ligands ofPPARs and offer a substitute for TZDs (Sharma et al., 2011). Amongthe several bioactive phytochemicals, naringin has been shown toactivate PPAR-g (Sharma et al., 2011). Naringin is a widely distrib-uted polyphenolic bioflavonoid compound predominantly found ingrapefruits and related citrus herbs (Chanet et al., 2012; Shin et al.,2013). It possesses a wide range of biological and pharmacologicalactivities including anti-inflammatory (Golechha, Chaudhry, Bhatia,Saluja, & Arya, 2011), anti-anxiety (Fernandez et al., 2009), anti-hyperglycemic (Mahmoud, Ashour, Abdel-Moneim, & Ahmed,2012), anti-hyperlipidemic (Chanet et al., 2012), cardioprotective(Rani et al., 2013), neuroprotective (Gopinath & Sudhandiran, 2012),and hepatoprotective (Pari & Amudha, 2011) activities. Additionally,because of its PPAR-g agonist activity, naringin confers neuro-protection from neurodegenerative diseases, affective disorders, andneuroinflammation (Golechha et al., 2011; Gopinath & Sudhandiran,2012; Sharma et al., 2011). There is an increasing research interest innaringin to evaluate its protective activity against different diseases.Due to the emerging role of PPAR-g agonists in alcohol consumptionand preference as well as reward and other associated illnesses, itwould be interesting to investigate the effect of naringin on alcoholconsumption and preference as well as ethanol-induced condi-tioned reward, when the activity of naringin against ethanol con-sumption is lacking. In order to elucidate the PPAR-g receptor-mediated mechanism and assess its role in the anti-addictive ac-tion of naringin, we administered GW9662, a pharmacologicalantagonist of PPAR-g, prior to naringin treatment. The aim of thisstudy was to assess the involvement of the PPAR-g in modulatingethanol-related behaviors by comparing naringin-treated mice tocontrols, using a 24-h continuous-access two-bottle choice drinkingparadigm and ethanol-induced conditioned place preference (CPP).

Materials and methods

Subjects

Adult male C57BL/6 mice (12e15 weeks old) obtained from thecentral breeding facility of College of Medicine and Health Sciences

(CMHS) of the United Arab Emirates University (UAEU) were used.Themicewere kept in clear plastic cageswith 5 per cage under a 12/12-h light/dark cycle (lights on at 06:00 h), with a room tempera-ture of approximately 22 �C and relative humidity of about 50%. Tapwater and rodent pellets, obtained from the National Feed and FlourProduction and Marketing Company LLC (Abu Dhabi, UAE) wereavailable ad libitum at all times except where specified. The animalswere acclimatized for 7 days before the experiment. The protocolswere approved by the Institutional Animal Care and Use Committeeof the CMHS (Protocol #: A41-13).

Drugs

Ethanol solutions (3, 6, 9, 12, and 15%; v/v) were prepared fromabsolute ethanol (Panreac Quimica SAU, Barcelona, Spain) anddiluted using tap water. For taste sensitivity, saccharin sodium saltdehydrate (0.02, 0.04, and 0.08%;w/v) and quinine hemisulphate (15,30, and 60 mM; w/v) were purchased from Sigma-Aldrich (St. Louis,MO, USA) and were dissolved in tap water. For the CPP experiments,1.5 g/kg ethanol was diluted in isotonic saline (0.9% sodium chloride)(10%; v/v). Naringin (0, 10, 30, and 100 mg/kg) was dissolved inisotonic saline. However, the PPAR-g antagonist GW9662 (1.5mg/kg;based on previously published studies [Paterniti et al., 2013; Shiotaet al., 2012]) was dissolved in 2.5% DMSO. GW9662 is a selectiveantagonist for PPAR-g (IC50 value 3.3 nM) with 1000-fold functionalselectivity for PPAR-g, and is used to demonstrate the PPAR-gdependent mechanism (Goyal et al., 2011). Both of the drugs wereobtained from Sigma-Aldrich and administered at a volume of10 mL/kg. The control treatments consisted of an equal volume ofvehicle solutions. For the two-bottle choice drinking paradigm, thedrugs were injected 30 min before the lights were turned off.

Voluntary ethanol drinking and preference, 24-h access

The two-bottle choice drinking procedure was carried out aspreviously described, with minor modifications (Bahi, 2012, 2013a,2013b). In brief, the mice were allowed to acclimate for 1 week tosingle housing in standard Plexiglas� cages equipped with two 10-mL graduated drinking pipettes containing water. On thefollowing days, one of the pipettes was filled with tap water and theother one with an ethanol solution. This way, mice were offered 3%ethanol (v/v) vs. water for 3 days. Immediately following 3% ethanol,a choice between 6% (v/v) ethanol and water was offered for 3 days,then 9% (v/v) ethanol for 3 days, then 12% (v/v) ethanol for 3 days,and finally 15% (v/v) ethanol for 3 days. The volume of ethanol andwater consumed were measured every day. The ethanol solutionswere replaced daily and the drinking pipettes were rotated everyday to control for position preference. Throughout the experiment,evaporation/spillage estimates were calculated daily from 2 pipettesplaced in an empty cage, one containing drinking water and theother containing the appropriate ethanol solution. The quantity ofethanol consumed (g/kg body weight/24 h) was calculated for eachmouse and these values were averaged for every concentration ofethanol. The ethanol preference for each mouse was calculated ac-cording to the formula: [ethanol intake/total fluid intake] � 100%.The total fluid consumed (mL/kg body weight/24 h) was calculatedfor each mouse. All values were averaged for every concentration ofthe ethanol. In the experiments using the PPAR-g antagonist,GW9662 was injected 15 min before naringinwas administered andethanol intake was measured 24 h later as described above.

Voluntary saccharin and quinine drinking and preference, 24-h access

The consumption and preference for the non-ethanol tastantswere measured 1 week after the ethanol preference study. For this

A. Bahi et al. / Alcohol 48 (2014) 677e685 679

aim, the same mice were tested for saccharin (sweet) and quinine(bitter) intake and preference using a two-bottle choice drinkingparadigm as described above. Briefly, the mice continued to drinkfrom two 10-mL graduated pipettes, one containing tap water andthe other containing increasing concentrations of saccharin orquinine. The consumption and the taste-preference testing wereperformed as described for ethanol preference, except that thesecond pipette contained the tastant solution. The mice wereserially offered saccharin sodium salt dehydrate (0.02, 0.04, and0.08%; w/v) and quinine hemisulphate (15, 30, and 60 mM; w/v).Each concentration of the tastant solution was offered for 3 days,with the pipette position reversed every day to control for sidepreference. For each tastant, the lower concentration was alwayspresented first, followed by the higher concentration.

Ethanol-induced conditioned place preference (CPP)

The conditioned reward was induced by ethanol in a battery of 8place-preference stations partly customized in our laboratory. Eachstation comprised 2 equally sized wooden chambers (30 � 30 cmsurface� 30 cmheight) separated by a smaller removable guillotinedoor (10 cm). In one of the chambers the inside walls were white,whereas the inside walls of the other chamber were painted blackthroughout. To provide tactile cues, the floor of the white chamberconsisted of a large grid; however the floor in the black chamberwas a narrow grid. The testing room’s luminance was adjusted sothat the environmental (visual and tactile) cues would not producea significant baseline preference for a specific chamber as describedpreviously (Bahi & Dreyer, 2014; Bahi, Sadek, Schwed, Walter, &Stark, 2013; Bahi, Tolle, et al., 2013).

The CPP procedure was divided into the 3 following successivephases: i) a single pre-CPP or pre-acquisition session (1st day), ii) 5sessions each day of acquisition interspersed with 5 control ses-sions (2nd to 6th days), and iii) a CPP test session (7th day). Spe-cifically, on the first day (the pre-acquisition [Pre-CPP] session), allmice were injected with saline and placed into the apparatus withfree access to all compartments for 20 min, and the time spent ineach compartment was manually recorded. During this session,mice exhibiting unconditioned preference (more than 800 s) oraversion (less than 400 s) for any compartment were discardedfrom the conditioning session (Bahi, 2012, 2013a, 2013b; Bahi &Dreyer, 2012). Therefore, 10.8% of the total mice were eliminatedfrom the study.

During the second phase, beginning 24 h later, mice weretreated over 5 consecutive twice-daily sessions with alternate in-jections of ethanol (1.5 g/kg from 10% ethanol solution diluted inisotonic saline) or the same volume of saline. The mice were alwaysinjected immediately before being placed into the conditioningchamber. During each conditioning day, a mouse was conditionedwith one vehicle-environment and one drug-environment pairingseparated by at least 6 h. Groups were counterbalanced for drugorder (morning or evening), drug side, and drug chamber associa-tion. This procedure was chosen to avoid circadian (morning/eve-ning) variability. This pretreatmentetreatment combinationcreated 6test groups: [saline conditioning: DMSO-saline (n ¼ 9),DMSO-naringin (n ¼ 9), GW9662-naringin (n ¼ 9)]; [ethanol con-ditioning: DMSO-saline (n ¼ 11), DMSO-naringin (n ¼ 12),GW9662-naringin (n ¼ 10)]. The PPAR-g antagonist GW9662 wasinjected 15 min before naringin. Saline or ethanol were injected30 min later.

During the third phase, which took place 24 h after the lastconditioning session, mice from all groups were injected with sa-line, placed between the 2 chambers and given free access to allcompartments for 20min. The time spent in each boxwasmanuallyscored as for the Pre-CPP session. The ethanol-induced CPP was

measured and analyzed as time spent in the ethanol-pairedchamber during Post-CPP minus time spent in the ethanol-pairedchamber during Pre-CPP (Post minus Pre).

Blood ethanol concentration (BEC)

In this experiment, ethanol-naïve male adult mice were used totest the effect of naringin (30 mg/kg) on ethanol pharmacokinetics.We followed a well standardized method, published previously(Bahi, Sadek, et al., 2013). Briefly, mice were injected with vehicle ornaringin (n¼ 4 each), and 30min later theywere treated with a 3 g/kg dose of ethanol. Three and 6 h later, trunk blood samples werecollected following a rapid decapitation intomicro-centrifuge tubescontaining the anti-coagulant EDTA. The blood samples werecentrifuged at 3000 rpm for 10 min and the supernatant plasmawas collected and frozen until analysis. The BECs were determinedin the serum using an ethanol dehydrogenase assay kit (BioVisionResearch Products, CA, USA). A standard calibration curve repre-senting the ethanol concentration (0e10 nmol/mL) was madefollowing the manufacturer’s instructions.

Statistical analysis

For statistical comparisons, the software package IBM SPSSStatistics 21 was used. Data were expressed as means � SEM. Thedata representing the effects of naringin on ethanol (or tastants)consumption and preference were analyzed using a mixed two-repeated-measure ANOVA with drug (vehicle or naringin) andethanol (or tastants) concentration as the between-subjects factors.The data representing the effects of naringin doses on ethanolconsumption and preference as well as BEC data were analyzedusing a one-repeated-measure ANOVA with drug as the between-subjects factor. The data representing the effect of naringin on theethanol-induced CPP acquisition was analyzed using a two-wayANOVA with drug and conditioning (saline or ethanol) as thebetween-subjects factors. Following a significant F value, post hocanalyses (Bonferroni’s test) were performed for assessing specificgroup comparisons. The level of statistical significance was set atp � 0.05.

Results

Naringin attenuated ethanol consumption and preference

In this experiment, mice were injected once a day with eithervehicle (n ¼ 6) or naringin (30 mg/kg; n ¼ 7) and given access toescalating concentrations of ethanol over a 15-day test period in atwo-bottle choice drinking paradigm. Ethanol consumption (g/kg/24 h) results are depicted in Fig. 1A. The two-way ANOVA(treatment � concentration) repeated measures indicated that therewas a main effect of ethanol as intake increased with increasingethanol concentrations [F(4,55)¼ 110.010, p< 0.0001]. Interestingly,the main effect of treatment was found significant [F(1,55) ¼ 36.739,p < 0.0001]. Most importantly, the interaction between the 2 factors[F(4,55) ¼ 4.770, p ¼ 0.002] resulted from naringin-treated miceconsuming less ethanol than vehicle-control mice. Post hoc evalua-tions revealed that when mice had access to a 3% ethanol solution,therewas no significant difference between the 2 groups (p¼ 0.257).Similarly, Fig. 1B shows the marked effect of treatment on the pref-erence for ethanol solution. In fact, the two-way ANOVA analysisrevealed a main effect of concentration [F(4,55)¼ 22.563, p < 0.001]as well as a main effect of treatment [F(1,55) ¼ 25.074, p < 0.001].Most importantly, the treatment � concentration interaction wassignificant [F(4,55)¼ 2.820, p¼ 0.034]. Post hoc analysis showed thatnaringin-injected mice had a reduced ethanol preference at all

Fig. 1. Effect of naringin (30 mg/kg) on ethanol (3, 6, 9, 12, and 15%) consumption and preference. A) Ethanol consumption calculated as grams of ethanol consumed per kilogram ofbody weight in male C57BL/6 mice. B) Ethanol preference expressed as ethanol consumed/total fluid consumed. C) Water consumption expressed in milliliters and D) Average totalfluid (water þ ethanol) intake. Data are expressed as mean � SEM. The p values refer to the difference between vehicle and naringin groups; vehicle (VEH, n ¼ 6) and naringin (NAR,n ¼ 7).

A. Bahi et al. / Alcohol 48 (2014) 677e685680

ethanol concentrations tested except at 3% (p ¼ 0.481). We also re-ported water consumption across the five ethanol concentrations. Asdepicted in Fig. 1C, naringin-injected mice consumed more waterthan vehicle controls. The two-way ANOVA analysis revealed a maineffect of ethanol concentration [F(4,55)¼ 9.070, p< 0.001] as well asamain effect of treatment [F(1,55)¼ 13.944, p< 0.001]. However, thetreatment � concentration interaction did not reach significance[F(4,55) ¼ 2.112, p ¼ 0.092]. Finally, to determine whether total fluidintake was affected by drug treatment, thus possibly mediating thedifferences in ethanol consumption and preference, we analyzedtotal fluid intake. The two-way ANOVA revealed that there were notreatment differences on any measure of total fluid intake; maineffect of treatment: [F(1,55) ¼ 0.094, p ¼ 0.760]; main effect ofethanol concentration: [F(4,55) ¼ 1.963, p ¼ 0.113]; treatment �concentration interaction: [F(4,55) ¼ 1.008, p ¼ 0.411] (Fig. 1D).

Next to this, we carried out a dose-ranging study to investigatethe effects of escalating doses of naringin (0, 10, 30, and 100 mg/kg)on ethanol intake and preference. Therefore, naïve C57BL/6 adultmalemice (n¼ 8) were trained to drink 10% ethanol vs. water under

Fig. 2. Effect of naringin (0, 10, 30, and 100 mg/kg) on ethanol (10%) consumption and preferbody weight in male C57BL/6 mice. B) Ethanol preference expressed as ethanol consumeexpressed as mean � SEM. *p < 0.05, **p < 0.01 vs. 0 mg/kg (n ¼ 8).

a two-bottle choice, home-cage, continuous access to ethanolparadigm. Results are presented in Fig. 2. The one-way repeated-measures ANOVA with dose as the between-subject factor indi-cated that naringin significantly and dose-dependently reducedethanol consumption measured after 24 h access to ethanol[F(3,28) ¼ 19.509, p < 0.001]. As depicted in Fig. 2A, post hocevaluation indicated that, as observed in the previous set of ex-periments, naringin (30 mg/kg) significantly reduced ethanol con-sumption (approximately 40% of control; p ¼ 0.004). In additionand compared to control animals, naringin at 100 mg/kg signifi-cantly reduced ethanol consumption by approximately 70%(p < 0.0001). However, no significant effect on ethanol intake wasobserved with a low dose of naringin (10 mg/kg) (approximately 5%reduction compared to control animals, p ¼ 1.000). Similarly,naringin dose-dependently reduced ethanol preference asdisplayed in Fig. 2B. The one-way ANOVA revealed a main effect ofdose [F(3,28) ¼ 64.374, p < 0.001]. Specifically and compared tosaline-dosed mice, ethanol preferences in 30 and 100 mg/kgnaringin-injected mice were approximately 35% and 68% lower,

ence. A) Ethanol consumption calculated as grams of ethanol consumed per kilogram ofd/total fluid consumed and C) Average total fluid (water þ ethanol) intake. Data are

Fig. 3. Effect of naringin (30 mg/kg) on saccharin (0.02, 0.04, and 0.08%) consumption and preference. A) Saccharin consumption calculated as grams of saccharin consumed perkilogram of body weight in male C57BL/6 mice. B) Saccharin preference expressed as saccharin consumed/total fluid consumed and C) Average total fluid (water þ saccharin) intake.Data are expressed as mean � SEM. Vehicle (VEH, n ¼ 6) and naringin (NAR, n ¼ 7).

A. Bahi et al. / Alcohol 48 (2014) 677e685 681

respectively (p < 0.0001). In contrast, and as for ethanol con-sumption, no significant effect on ethanol preference was observedwith a lowdose of naringin (10mg/kg) compared to control animals(p ¼ 1.000). More importantly, naringin treatment had no effect ontotal fluid intake as revealed by a one-way repeated-measureANOVA [F(3,28) ¼ 0.841, p ¼ 0.483] (Fig. 2C). Taken together, nar-ingin effectively reduced ethanol consumption in C57BL/6 mice,with no effect on total fluid intake, suggesting that naringin effectsoccurred through a specific reduction of ethanol intake rather thannon-specific motor effects. Therefore, the naringin effects areconsidered to have certain selectivity and specificity, because thereduction of voluntary ethanol intake coincided with a reduction ofethanol preference.

Naringin did not affect tastants intake and preference

To rule out differences in taste preference, the same mice weretested for their preference for sweet (saccharin) and bitter (quinine)solutions following injections of naringin (30 mg/kg). The two-wayANOVA revealed that for saccharin intake there was a significant ef-fect of concentration [F(2,33)¼ 137.369, p< 0.001], with the highestconsumption score for the 2 groups observed with the higher con-centrations. However, there was no significant treatment effect[F(1,33) ¼ 0.807, p ¼ 0.375] and the interaction termwas not signifi-cant [F(2,33)¼ 0.067, p¼ 0.935], as depicted in Fig. 3A. Therewere nodifferences between the vehicle and naringin animals in saccharin

Fig. 4. Effect of naringin (30 mg/kg) on quinine (15, 30, and 60 mM) consumption and prkilogram of body weight in male C57BL/6 mice. B) Quinine preference expressed as quinine care expressed as mean � SEM. Vehicle (VEH, n ¼ 6) and naringin (NAR, n ¼ 7).

preference. The two-way ANOVA indicated a non-significant treat-ment [F(1,33) ¼ 0.037, p ¼ 0.849] and concentration effect[F(2,33) ¼ 1.680, p ¼ 0.202]. Also, the treatment � concentrationinteraction [F(2,33)¼0.084,p¼0.920]wasnot significant asdepictedin Fig. 3B. Finally, the ANOVAdid not showanybetween-treatment orbetween-concentrationdifferences in termsof totalfluid intake (fromboth tubes); treatment effect: [F(1,33) ¼ 0.526, p ¼ 0.474]; concen-tration effect: [F(2,33)¼ 1.514, p¼ 0.235]; treatment� concentrationinteraction effect: [F(2,33) ¼ 0.237, p ¼ 0.790]. Therefore, total fluidintake did not differ between the 2 groups regardless of the solutionavailable in the second bottle (Fig. 3C).

Fig. 4 presents themean (�SEM)quinine intakeandpreference forthe2groups (vehicle and30mg/kgnaringin) across the choice test. Asfor saccharin, a two-wayANOVAwasconducted to test the interactionbetween the treatment and the concentrations on quinine con-sumption. As depicted in Fig. 4A, no significant treatment effect wasobserved [F(1,33) ¼ 0.088, p ¼ 0.768]. However, the concentrationeffect was significant [F(2,33) ¼ 14.238, p < 0.001], but not theinteraction term [F(2,33) ¼ 0.017, p ¼ 0.983]. The quinine preferenceresults are displayed in Fig. 4B. A two-way ANOVA with repeatedmeasures to compare quinine preference revealed no significantdifference between vehicle- and naringin-injected mice; treatmenteffect: [F(1,33) ¼ 0.065, p ¼ 0.800]; concentration effect:[F(2,33) ¼ 2.547, p ¼ 0.094]; treatment � concentration interactioneffect: [F(2,33) ¼ 0.696, p ¼ 0.506]. Finally, the ANOVA indicated anon-significant effect of treatment on total fluid intake

eference. A) Quinine consumption calculated as milligrams of quinine consumed peronsumed/total fluid consumed and C) Average total fluid (water þ quinine) intake. Data

Fig. 5. Effect of vehicle, naringin (30 mg/kg), GW9662 (1.5 mg/kg), and GW9662-naringin on ethanol (10%) consumption and preference. A) Ethanol consumption calculated asgrams of ethanol consumed per kilogram of body weight in male C57BL/6 mice. B) Ethanol preference expressed as ethanol consumed/total fluid consumed and C) Average totalfluid (water þ ethanol) intake. Data are expressed as mean � SEM. *p < 0.05 vs. DMSO-SAL. #p < 0.05 vs. DMSO-NAR. DMSO-SAL (n ¼ 7); DMSO-NAR (n ¼ 7); GW-NAR (n ¼ 7).

Fig. 6. Effect of vehicle, naringin (30 mg/kg), GW9662 (1.5 mg/kg), and GW9662-naringin on ethanol-elicited conditioned place preference (CPP) and BEC. A) Salineand ethanol-elicited CPP scores are expressed as Post minus Pre of time spent in theethanol-paired chamber. C57BL/6 mice were conditioned with saline or ethanol (1.5 g/kg; i.p.). Before conditioning, mice were previously injected with vehicle, NAR,GW9662, or GW9662-NAR. Data are expressed as mean � SEM. *p < 0.05, **p < 0.01 vs.saline conditioning. #p < 0.05 vs. DMSO-NAR; n ¼ 9e12. B) Blood alcohol concentra-tions 3 and 6 h after an acute injection of 3 g/kg ethanol. n ¼ 4 in each group.$p < 0.0001 for 6 vs. 3 h.

A. Bahi et al. / Alcohol 48 (2014) 677e685682

[F(1,33) ¼ 0.066, p ¼ 0.798], a non-significant concentration effect[F(2,33) ¼ 0.038, p ¼ 0.963], and a non-significanttreatment � concentration interaction [F(2,33) ¼ 0.519, p ¼ 0.600],depicted in Fig. 4C.

The PPAR-g antagonist GW9662 reversed naringin-reduced ethanolintake

In this experiment, we tested whether naringin-induced reduc-tion of ethanol consumption and preference can be blocked by pre-treatment with the PPAR-g selective antagonist, GW9662 (1.5 mg/kg). For this purpose, mice were injected with GW9662 and 15 minlater, naringin (30 mg/kg) was administered (n ¼ 7). The controlanimals received daily DMSO-saline (n ¼ 7) or DMSO-naringin(n ¼ 7) injections with 24-h access to a 10% ethanol solution.

As depicted in Fig. 5A, a one-way repeated-measure ANOVAshowed a significant effect of treatment on ethanol consumption[F(2,18) ¼ 10.458, p ¼ 0.001]. Post hoc analysis revealed that, asexpected, naringin decreased ethanol consumption (p ¼ 0.001DMSO-SAL vs. DMSO-NAR). However, GW9662 pre-injectionreversed the reduced ethanol consumption caused by naringin(p ¼ 0.035 DMSO-NAR vs. GW-NAR) to control levels (p ¼ 0.307DMSO-SAL vs. GW-NAR). Similarly, ethanol preference was alsoaffected following drug treatment [F(2,18) ¼ 15.269, p < 0.001]. Asdisplayed in Fig. 5B, a one-way ANOVA revealed that naringinreduced ethanol preference (p < 0.001 DMSO-SAL vs. DMSO-NAR).In contrast, GW9662 nullified the effect of naringin on ethanolpreference (p ¼ 0.006 DMSO-NAR vs. GW-NAR). Interestingly, nodifference was found between DMSO-SAL and GW-NAR groups(p¼ 0.264). Finally, combined naringin and GW9662 treatment hadno effect on total fluid intake, which provides strong evidence forthe specific PPAR-g agonism effect as the main cause for the changein voluntary ethanol consumption and preference [F(2,18) ¼ 0.971,p ¼ 0.398] as represented in Fig. 5C.

Naringin reduced ethanol-CPP acquisition

The effect of systemic injection of naringin with or withoutGW9662 on acquisition of ethanol-induced CPP is depicted inFig. 6A. It should be emphasized that before conditioning, micedemonstrated no significant preference for any of the conditioningchambers: main effect of treatment, [F(2,54) ¼ 0.635, p ¼ 0.534];main effect of conditioning, [F(1,54) ¼ 0.465, p ¼ 0.498];treatment � conditioning interaction effect, [F(2,54) ¼ 0.013,p ¼ 0.987] (data not shown). However, after conditioning, results

have shown that naringin administered 30 min prior to the ethanoltreatment decreased the effect of ethanol-CPP. In fact, application ofa two-way ANOVA showed that when conditioned with ethanol butnot saline, mice showed a robust increase in time spent in theethanol-paired chamber [main effect of conditioning:F(1,54) ¼ 136.179, p < 0.0001]. Interestingly, drug treatmentaffected ethanol-CPP [F(2,54)¼ 4.408, p¼ 0.017]. Most importantly,the treatment � conditioning interaction was found significant:[F(2,54) ¼ 6.528, p ¼ 0.003]. Post hoc Bonferroni’s multiple com-parisons on ethanol-conditioning data showed that naringinsignificantly decreased the effect of ethanol on CPP as compared tothe vehicle-treated mice (p < 0.001, DMSO-SAL vs. DMSO-NAR).However, a lower dose of GW9662 (1.5 mg/kg, intra-peritoneally[i.p.]) was effective in reversing naringin-decreased ethanol CPP(p ¼ 0.022, DMSO-NAR vs. GW-NAR). Interestingly, no differencewas found between DMSO-SAL and GW-NAR groups (p ¼ 0.114).

Naringin did not affect blood ethanol concentration (BEC)

Finally, the BECs in vehicle- and naringin-injected mice wereassessed in order to test whether the naringin dose utilized in thisstudy elicited its effect by altering ethanol metabolism. Animals

A. Bahi et al. / Alcohol 48 (2014) 677e685 683

were injectedwith either vehicle or naringin (30mg/kg, i.p.) 30minbefore ethanol administration (3 g/kg). BECs were determined bycollecting blood 3 and 6 h after ethanol injection. Results are dis-played in Fig. 6B. The one-way ANOVA with repeated measure anddrug as the between-subject factor and time as the within-subjectfactor revealed amain effect of time [F(1,14)¼ 603.494, p< 0.0001].Thus, in vehicle and naringin-injected mice, BEC decreased signif-icantly over time. However, the rate of ethanol metabolism, asmeasured by the slope of the lines, was not significantly differentbetween the 2 groups [F(1,14) ¼ 0.098, p ¼ 0.759]. Therefore, thechanges in ethanol-rewarding properties did not result from po-tential differences in clearance of ethanol.

Discussion

Alcoholism is a complex heterogenous disease, largelycontrolled by the CNS, and involving numerous neurotransmitterand neuromodulator systems. Recently, PPAReg modulation in theCNS has been shown to play an important and pervasive role in therewarding effects of ethanol (Kane et al., 2011; Panlilio, Justinova, &Goldberg, 2013; Stopponi et al., 2011). The aim of the current studywas to examine the effects of the pharmacological modulation ofPPAR-g using a naturally available PPAR-g agonist, naringin, onvoluntary ethanol consumption and ethanol-elicited CPP. The pre-sent study demonstrates for the first time that naringin is effectivein reducing voluntary ethanol intake and decreasing CPP scores byaltering its rewarding effects, suggesting an additional role of nar-ingin in regulation of alcoholism mediated by PPAR-g. The PPAR-g-dependent mechanism was further supported by the findings thatGW9662, an antagonist of PPAR-g, nullified all the effects of nar-ingin. Given no effect of naringin on ethanol metabolism, the pre-sent study clearly demonstrates that PPAR-g plays a role in ethanoldisorders and that naringin could be a promising novel agent ofnatural origin in the pharmacotherapy of alcoholism.

Very recently, the members of the TZD drugs (pioglitazone androsiglitazone) known as PPAR-g agonists were reported effective insuppressing ethanol drinking and relapse to ethanol-seekingbehavior in rodent models of alcoholism. This effect was blockedby a PPAR-g antagonist, GW9662, indicating that the effects weremediated by activation of PPAR-g receptors in the brain (Stopponiet al., 2011). In agreement with this, in the present study, nar-ingin has been observed to reduce ethanol consumption and pref-erence and these effects were not associated with differences intotal fluid intake between the 2 groups. Accordingly, other factorsalso known to contribute to ethanol-drinking behaviors wereexamined. Considering the possibility that the taste “palatability” ofethanol differs between vehicle- and naringin-injected mice,however, in the present study experiments, we did not observedifferences in the consumption and preference of saccharin orquinine solutions, indicating that there were no differences in tastepreference between the 2 groups. In addition to the pharmacologicreinforcing properties, ethanol is also a caloric substance. This isfurther substantiated by the fact that naringin not only reduced theethanol consumption but has also been reported to ameliorateglucose intolerance, fat accumulation, obesity, and food intake(Alam, Kauter, & Brown, 2013; Shin et al., 2013). In contrast tonaringin, the synthetic analog pioglitazone also altered ethanol’srewarding properties. However, it increased food intake, promotedweight gain and caused adverse effects such as congestive heartfailure, bone fractures and possibly bladder cancer. These featuresare of serious concern and not favored pioglitazone for long termtreatment (Cariou et al., 2012). Additionally, naringin supplemen-tation ameliorated glucose intolerance and liver mitochondrialdysfunction, lowered plasma lipid concentrations, and improvedthe structure and function of the heart and liver without decreasing

total body weight and altering food intake (Alam et al., 2013). Theseadditional beneficial effects of naringin are incorporated into itsusefulness in ethanol addiction.

In order to support the behavioral finding with ethanol intakeand preference, we compared the ethanol pharmacodynamics invehicle- and naringin-treated mice to assess whether or not thedissimilarities in voluntary ethanol consumption and preferencecould be due to a reduction in the proportion of ethanol metabolismin naringin-treated mice. Results showed that there was no differ-ence in the time-matched BECs between the 2 groups, suggestingno divergence in ethanol metabolism levels. These results furtherdemonstrate the benefits of naringin in ethanol intake and con-sumption and substantiate the main outcome of the current study,namely, that activation of PPAR-g is associated with attenuatedvoluntary ethanol consumption in mice.

Furthermore, we carried out a CPP test for the measurement ofreward phenomenon for ethanol in mice to determine whether theactivation of PPAR-g produces a decrease or an increase in ethanol’srewarding properties. Although all groups acquired ethanol-CPP,vehicle-treated mice developed a reliable place preference whilenaringin-injected mice showed reduced preference for thecompartment paired with ethanol during conditioning. It should beemphasized that naringin had no aversive effects on its own and didnot cause any place aversion that would appear to reverse phar-macological ethanol’s rewarding effect, but simply because possibleaversive effects of naringin negatively summatewith the rewardingeffect of ethanol. The CPP datawere very vital in suggesting that thePPAR-g played a crucial role in ethanol’s rewarding properties,suggesting that PPAR-g might be involved in the saliency of theenvironmental cues associated with ethanol. Therefore, reducedvoluntary ethanol intake in naringin-injected mice reflected adecrease rather than an increase in ethanol reward. Nullification ofall the effects of naringin by prior administration of a PPAR-gantagonist, GW9662, clearly demonstrates the role of PPAR-g inmediating the reward phenomenon, similar to the results of aprevious study, in which the activation of PPAR-g by TZDs showedsimilar effects (Stopponi et al., 2011). Moreover, PPAR-g activationalso provides protection against adverse effects of fetal ethanolspectrum disorder (Kane et al., 2011) as well as stress-inducedrelapse (Stopponi et al., 2011), a major obstacle to abstinence.

The actions of naringin on peripheral PPAR-g are not to beexcluded, as naringin has been reported to provide protection fromethanol-induced liver injury by up-regulation of Sirt1 and PGC-1a,rather than by up-regulation of PPAR-g (Oliva et al., 2008). Hence,the alternative hypothesis garners attention focusing on brainPPAR-g and how naringin in reducing ethanol consumption andpreference and related behaviors. The activation of PPAR-gexpressed in the CNS, including microglia, has been shown toreduce ethanol consumption (He et al., 2005) by inhibiting theactivation of microglia and suppressing pro-inflammatory cyto-kines and chemokines through a process termed “receptor-dependent trans-repression” (Diab et al., 2004). Additionally, inhi-bition of pro-inflammatory cytokines reduces the vulnerability todrug abuse and consequent neurodegeneration (Marshall et al.,2013; Sharrett-Field, Butler, Berry, Reynolds, & Prendergast, 2013).Moreover, PPAR-g activation has been shown to contribute signif-icantly to the reduction of drug abuse including alcoholism as wellas potentiating the effect of the existing medication, naltrexone(Maeda et al., 2007; Stopponi et al., 2011, 2013). Thus, it is reason-able to speculate that the PPAR-g agonist property of naringin notonly reduces ethanol consumption and preference but also maypotentiate the efficacy of existing medications for the pharmaco-therapy of alcoholism.

The other alternate supportive hypothesis suggesting the pos-sibility of involvement of PPAR-g concerns its presence in brain

A. Bahi et al. / Alcohol 48 (2014) 677e685684

parts, namely in LH, PVN, ARC, and VTA, where PPAR-g co-localizeswith a-MSH, AgRP, and POMC (Sarruf et al., 2009). A closely inter-connected hypothalamic neurocircuit plays an important role in thecontrol of energy expenditure, locomotor activity, and ethanoladdiction (Schwartz, Woods, Porte, Seeley, & Baskin, 2000; Shelkaret al., 2014). There are close interconnections between the ARC, theLH, and the ventromedial hypothalamus, which interconnect withdopaminergic circuits and may contribute to the hedonic aspects ofethanol consumption. Moreover, PPAR-g is also expressed in theVTA and co-localizes with tyrosine hydroxylase, suggesting theexpression of this receptor in dopaminergic cells (Sarruf et al.,2009). The mesolimbic dopamine system originates in the VTAand projects to forebrain regions, which include the nucleusaccumbens and the prefrontal cortex (PFC) and are thought to bethe neurocircuits governing the rewarding and reinforcing prop-erties of ethanol (Deehan, Hauser, Wilden, Truitt, & Rodd, 2013). Agrowing body of evidence indicates that a-MSH and AgRP systemsregulate the functions associated with positive reinforcement anddrug reward (Shelkar et al., 2014).

Integrating the facts all together, the attenuated ethanol con-sumption and preference in naringin-treated mice could bereasonably linked to several PPAR-g-associated mechanismsinvolving hypothalamic, dopaminergic, and cholinergic neuro-circuits. Additionally, our two-bottle choice drinking paradigm andethanol-elicited CPP studies showing the effect of the PPAR-gantagonist, GW9662, on naringin-induced decrease of ethanol-related behaviors, lead to an important conclusion, namely, thatthe changes in the voluntary ethanol intake and CPP scores of miceinduced by naringin were dependent on PPAR-g activation. This isconsistent with previous reports which showed that GW9662significantly reverse the activities of the PPAR-g agonist, naringin(Sharma et al., 2011).

To summarize, the results demonstrate that naringin impedesethanol consumption and the reward phenomenon mediated byactivation of PPAR-g. The roles of PPAR-g receptors were evidencedby abolition of the effects of naringin by pretreatment withGW9662, a PPAReg receptor antagonist. The lack of effect of nar-ingin on ethanol metabolism further supports the notion that theeffects of naringin are mediated by brain PPAR-g. Taken with thepresent study findings and previous reports of its benefits as well asits safety and tolerability profiles, it appears that naringin could be apotential promising novel agent of natural origin in the treatmentof ethanol addiction, either alone or as an adjunct to improve theefficacy of existing medications in pharmacotherapy.

Acknowledgments

The authors would like to acknowledge Mr. Mohamed Elwasilaand Mr. Mohamed Shafiullah for their technical assistance and Dr.Mahmoud Hag Ali from the Animal Research Facility for his sug-gestions on animal care and welfare.

Role of the funding source: The research grant support from theUnited Arab Emirates University and the National Research foun-dation, UAE to AB (grant No. 31M082) and SO (grant No. 31M099)are duly acknowledged. The funders had no further role in studydesign, in the collection, analysis and interpretation of data, in thewriting of the report, and in the decision to submit the paper forpublication.

Author’s contribution: AB was responsible for the study conceptand design. AB and SMN contributed to the collection of animaldata. AB performed the data analysis and SO helped in the inter-pretation of findings. AB and SO drafted the manuscript. All theauthors provided critical revision of the manuscript for important

intellectual content, reviewed content, and approved the finalversion for publication.

References

Alam, M. A., Kauter, K., & Brown, L. (2013). Naringin improves diet-induced car-diovascular dysfunction and obesity in high carbohydrate, high fat diet-fed rats.Nutrients, 5, 637e650.

Bahi, A. (2012). The selective metabotropic glutamate receptor 7 allosteric agonistAMN082 prevents reinstatement of extinguished ethanol-induced conditionedplace preference in mice. Pharmacology, Biochemistry, and Behavior, 101,193e200.

Bahi, A. (2013a). Increased anxiety, voluntary alcohol consumption and ethanol-induced place preference in mice following chronic psychosocial stress. Stress,16, 441e451.

Bahi, A. (2013b). Individual differences in elevated plus-maze exploration predictedhigher ethanol consumption and preference in outbred mice. Pharmacology,Biochemistry, and Behavior, 105, 83e88.

Bahi, A., & Dreyer, J. L. (2012). Involvement of nucleus accumbens dopamine D1receptors in ethanol drinking, ethanol-induced conditioned place preference,and ethanol-induced psychomotor sensitization in mice. Psychopharmacology,222, 141e153.

Bahi, A., & Dreyer, J. L. (2014). Chronic psychosocial stress causes delayed extinctionand exacerbates reinstatement of ethanol-induced conditioned place prefer-ence in mice. Psychopharmacology, 231, 367e377.

Bahi, A., Sadek, B., Schwed, S. J., Walter, M., & Stark, H. (2013). Influence of the novelhistamine H3 receptor antagonist ST1283 on voluntary alcohol consumptionand ethanol-induced place preference in mice. Psychopharmacology, 228,85e95.

Bahi, A., Tolle, V., Fehrentz, J. A., Brunel, L., Martinez, J., Tomasetto, C. L., et al. (2013).Ghrelin knockout mice show decreased voluntary alcohol consumption andreduced ethanol-induced conditioned place preference. Peptides, 43, 48e55.

Cariou, B., Charbonnel, B., & Staels, B. (2012). Thiazolidinediones and PPARg ago-nists: time for a reassessment. Trends in Endocrinology and Metabolism, 23,205e215.

Chanet, A., Milenkovic, D., Deval, C., Potier, M., Constans, J., Mazur, A., et al. (2012).Naringin, the major grapefruit flavonoid, specifically affects atherosclerosisdevelopment in diet-induced hypercholesterolemia in mice. The Journal ofNutritional Biochemistry, 23, 469e477.

Deehan, G. A., Jr., Hauser, S. R., Wilden, J. A., Truitt, W. A., & Rodd, Z. A. (2013).Elucidating the biological basis for the reinforcing actions of alcohol in themesolimbic dopamine system: the role of active metabolites of alcohol. Fron-tiers in Behavioral Neuroscience, 7, 104.

Diab, A., Hussain, R. Z., Lovett-Racke, A. E., Chavis, J. A., Drew, P. D., & Racke, M. K.(2004). Ligands for the peroxisome proliferator-activated receptor-gamma andthe retinoid X receptor exert additive anti-inflammatory effects on experi-mental autoimmune encephalomyelitis. Journal of Neuroimmunology, 148,116e126.

Fernandez, S. P., Nguyen, M., Yow, T. T., Chu, C., Johnston, G. A., Hanrahan, J. R., et al.(2009). The flavonoid glycosides, myricitrin, gossypin and naringin exert anxi-olytic action in mice. Neurochemical Research, 34, 1867e1875.

Gofflot, F., Chartoire, N., Vasseur, L., Heikkinen, S., Dembele, D., Le Merrer, J., et al.(2007). Systematic gene expression mapping clusters nuclear receptors ac-cording to their function in the brain. Cell, 131, 405e418.

Golechha, M., Chaudhry, U., Bhatia, J., Saluja, D., & Arya, D. S. (2011). Naringinprotects against kainic acid-induced status epilepticus in rats: evidence for anantioxidant, anti-inflammatory and neuroprotective intervention. Biological &Pharmaceutical Bulletin, 34, 360e365.

Gopinath, K., & Sudhandiran, G. (2012). Naringin modulates oxidative stress andinflammation in 3-nitropropionic acid-induced neurodegeneration through theactivation of nuclear factor-erythroid 2-related factor-2 signalling pathway.Neuroscience, 227, 134e143.

Goyal, S. N., Bharti, S., Bhatia, J., Nag, T. C., Ray, R., & Arya, D. S. (2011). Telmisartan, adual ARB/partial PPAR-g agonist, protects myocardium from ischaemic reper-fusion injury in experimental diabetes. Diabetes, Obesity & Metabolism, 13,533e541.

He, D. Y., McGough, N. N., Ravindranathan, A., Jeanblanc, J., Logrip, M. L.,Phamluong, K., et al. (2005). Glial cell line-derived neurotrophic factor mediatesthe desirable actions of the anti-addiction drug ibogaine against alcohol con-sumption. The Journal of Neuroscience, 25, 619e628.

Heilig, M., Goldman, D., Berrettini, W., & O’Brien, C. P. (2011). Pharmacogeneticapproaches to the treatment of alcohol addiction. Nature Reviews. Neuroscience,12, 670e684.

Kane, C. J., Phelan, K. D., Han, L., Smith, R. R., Xie, J., Douglas, J. C., et al. (2011).Protection of neurons and microglia against ethanol in a mouse model of fetalalcohol spectrum disorders by peroxisome proliferator-activated receptor-gagonists. Brain, Behavior, and Immunity, 25(Suppl. 1), S137eS145.

Maeda, T., Kiguchi, N., Fukazawa, Y., Yamamoto, A., Ozaki, M., & Kishioka, S. (2007).Peroxisome proliferator-activated receptor gamma activation relieves expres-sion of behavioral sensitization to methamphetamine in mice. Neuro-psychopharmacology, 32, 1133e1140.

Mahmoud, A. M., Ashour, M. B., Abdel-Moneim, A., & Ahmed, O. M. (2012). Hes-peridin and naringin attenuate hyperglycemia-mediated oxidative stress and

A. Bahi et al. / Alcohol 48 (2014) 677e685 685

proinflammatory cytokine production in high fat fed/streptozotocin-inducedtype 2 diabetic rats. Journal of Diabetes and Its Complications, 26, 483e490.

Marshall, S. A., McClain, J. A., Kelso, M. L., Hopkins, D. M., Pauly, J. R., & Nixon, K.(2013). Microglial activation is not equivalent to neuroinflammation in alcohol-induced neurodegeneration: the importance of microglia phenotype. Neurobi-ology of Disease, 54, 239e251.

Myrick, H., & Anton, R. F. (1998). Treatment of alcohol withdrawal. Alcohol Healthand Research World, 22, 38e43.

Oliva, J., French, B. A., Li, J., Bardag-Gorce, F., Fu, P., & French, S. W. (2008). Sirt1 isinvolved in energy metabolism: the role of chronic ethanol feeding andresveratrol. Experimental and Molecular Pathology, 85, 155e159.

Panlilio, L. V., Justinova, Z., & Goldberg, S. R. (2013). Inhibition of FAAH and acti-vation of PPAR: new approaches to the treatment of cognitive dysfunction anddrug addiction. Pharmacology & Therapeutics, 138, 84e102.

Pari, L., & Amudha, K. (2011). Hepatoprotective role of naringin on nickel-inducedtoxicity in male Wistar rats. European Journal of Pharmacology, 650, 364e370.

Paterniti, I., Impellizzeri, D., Crupi, R., Morabito, R., Campolo, M., Esposito, E., et al.(2013). Molecular evidence for the involvement of PPAR-d and PPAR-g in anti-inflammatory and neuroprotective activities of palmitoylethanolamide afterspinal cord trauma. Journal of Neuroinflammation, 10, 20.

Rani, N., Bharti, S., Manchanda, M., Nag, T. C., Ray, R., Chauhan, S. S., et al. (2013).Regulation of heat shock proteins 27 and 70, p-Akt/p-eNOS and MAPKs bynaringin dampens myocardial injury and dysfunction in vivo after ischemia/reperfusion. PLoS One, 8, e82577.

Rezvani, A. H., Lawrence, A. J., Arolfo, M. P., Levin, E. D., & Overstreet, D. H. (2012).Novel medication targets for the treatment of alcoholism: preclinical studies.Recent Patents on CNS Drug Discovery, 7, 151e162.

Sarruf, D. A., Yu, F., Nguyen, H. T., Williams, D. L., Printz, R. L., Niswender, K. D., et al.(2009). Expression of peroxisome proliferator-activated receptor-gamma in keyneuronal subsets regulating glucose metabolism and energy homeostasis.Endocrinology, 150, 707e712.

Schwartz, M. W., Woods, S. C., Porte, D., Jr., Seeley, R. J., & Baskin, D. G. (2000).Central nervous system control of food intake. Nature, 404, 661e671.

Sharma, A. K., Bharti, S., Ojha, S., Bhatia, J., Kumar, N., Ray, R., et al. (2011). Up-regulation of PPARg, heat shock protein-27 and -72 by naringin attenuatesinsulin resistance, b-cell dysfunction, hepatic steatosis and kidney damagein a rat model of type 2 diabetes. The British Journal of Nutrition, 106,1713e1723.

Sharrett-Field, L., Butler, T. R., Berry, J. N., Reynolds, A. R., & Prendergast, M. A.(2013). Mifepristone pretreatment reduces ethanol withdrawal severity in vivo.Alcoholism: Clinical and Experimental Research, 37, 1417e1423.

Shelkar, G. P., Kale, A. D., Singh, U., Singru, P. S., Subhedar, N. K., & Kokare, D. M.(2014). Alpha-melanocyte stimulating hormone modulates ethanol self-administration in posterior ventral tegmental area through melanocortin-4receptors. Addiction Biology. http://dx.doi.org/10.1111/adb.12126, (Epub aheadof print).

Shin, E. J., Hur, H. J., Sung, M. J., Park, J. H., Yang, H. J., Kim, M. S., et al. (2013). Ethanolextract of the Prunus mume fruits stimulates glucose uptake by regulatingPPAR-g in C2C12 myotubes and ameliorates glucose intolerance and fat accu-mulation in mice fed a high-fat diet. Food Chemistry, 141, 4115e4121.

Shiota, A., Shimabukuro, M., Fukuda, D., Soeki, T., Sato, H., Uematsu, E., et al. (2012).Telmisartan ameliorates insulin sensitivity by activating the AMPK/SIRT1pathway in skeletal muscle of obese db/db mice. Cardiovascular Diabetology, 11,139.

Soyka, M., & Rösner, S. (2010). Emerging drugs to treat alcoholism. Expert Opinion onEmerging Drugs, 15, 695e711.

Stopponi, S., de Guglielmo, G., Somaini, L., Cippitelli, A., Cannella, N., Kallupi, M.,et al. (2013). Activation of PPARg by pioglitazone potentiates the effects ofnaltrexone on alcohol drinking and relapse in msP rats. Alcoholism: Clinical andExperimental Research, 37, 1351e1360.

Stopponi, S., Somaini, L., Cippitelli, A., Cannella, N., Braconi, S., Kallupi, M., et al.(2011). Activation of nuclear PPARg receptors by the antidiabetic agent piogli-tazone suppresses alcohol drinking and relapse to alcohol seeking. BiologicalPsychiatry, 69, 642e649.

Tontonoz, P., & Spiegelman, B. M. (2008). Fat and beyond: the diverse biology ofPPARgamma. Annual Review of Biochemistry, 77, 289e312.