Brain-derived neurotrophic factor mediatesthe suppression of alcohol self-administrationby memantine
Jérôme Jeanblanc*, Fabien Coune*, Béatrice Botia & Mickaël NaassilaGroupe de Recherche sur l’Alcool et les Pharmacodépendances – INSERM ERI 24, UFR de Pharmacie, Université de Picardie Jules Verne, SFR CAP Santé, France
ABSTRACT
Brain-derived neurotrophic factor (BDNF) within the striatum is part of a homeostatic pathway regulating alcoholconsumption. Memantine, a non-competitive antagonist of N-methyl-D-aspartate receptors, induces expression ofBDNF in several brain regions including the striatum. We hypothesized that memantine could decrease ethanol (EtOH)consumption via activation of the BNDF signalling pathway. Effects of memantine were evaluated in Long-Evans ratsself-administering moderate or high amounts of EtOH 6, 30 and 54 hours after an acute injection (12.5 and 25 mg/kg). Motivation to consume alcohol was investigated through a progressive ratio paradigm. The possible role for BDNFin the memantine effect was tested by blockade of the TrkB receptor using the pharmacological agent K252a and by theBDNF scavenger TrkB-Fc. Candidate genes expression was also assessed by polymerase chain reaction array 4 and 28hours after memantine injection. We found that memantine decreased EtOH self-administration and motivation toconsume EtOH 6 and 30 hours post-injection. In addition, we found that inhibition or blockade of the BDNF signallingpathway prevented the early, but not the delayed decrease in EtOH consumption induced by memantine. Finally, Bdnfexpression was differentially regulated between the early and delayed timepoints. These results demonstrate that anacute injection of memantine specifically reduces EtOH self-administration and motivation to consume EtOH for atleast 30 hours. Moreover, we showed that BDNF was responsible for the early effect, but that the delayed effect wasBDNF-independent.
Correspondence to: Jérôme Jeanblanc, UFR Pharmacie, INSERM ERI24 – GRAP, 1 rue des Louvels, 80000 Amiens, France. E-mail:[email protected]
INTRODUCTION
Alcohol abuse and alcoholism are serious medical andsocial problems (Spanagel 2009). Currently availablealcoholism pharmacotherapies are only moderately suc-cessful and there are thus high expectancies for findingnovel treatments to improve clinical outcomes and help toreduce the devastating consequences of the disease.Emerging evidence from alcoholism research suggeststhat neuroadaptations in glutamatergic transmissionplay a central role, so medications that reverse thesechanges could be therapeutically beneficial (Krystal et al.2003; Olive 2009).
Several studies have tested the potential of glutama-tergic antagonists such as MK-801, LY-274614 and
phencyclidine as pharmacotherapies for alcohol-use dis-orders; however, their therapeutic efficacy may be tem-pered by adverse side effects (McMillen et al. 2004; Lipton2005). An alternative molecule is memantine (1-amino-3,5-dimethyladamantane), a non-competitive N-methyl-D-aspartate (NMDA) antagonist that has not beenextensively studied so far in the alcohol field. However,both preclinical and clinical studies have revealed apotential action of memantine on ethanol (EtOH) con-sumption with fewer side effects than other antagonists.For example, multiple injections of memantine overseveral days reduce voluntary EtOH consumption(Malpass, Williams & McMillen 2010). In addition,memantine suppressed the alcohol deprivation effect inWistar rats, suggesting that it produces an anti-craving
*Jérôme Jeanblanc and Fabien Coune equally participated to this work.
effect for EtOH and may be effective in relapse prevention(Holter, Danysz & Spanagel 1996). In alcohol-dependentpatients, memantine reduces craving (Krupitsky et al.2007), but the results of another pilot study with asmall sample size did not support the use of memantinefor the treatment of actively drinking alcohol-dependentpatients (Evans et al. 2007). Therefore, the interest ofmemantine for the treatment of alcoholism remains con-troversial, in part because of the small sizes of thesamples used in these clinical studies.
Together, both preclinical and clinical studies havesuggested that memantine may have some promise asan alcoholism treatment, but these studies have notaddressed its mechanism of action. Recent reports statedthat intra-peritoneal (i.p.) injections of memantineincrease the expression of brain-derived neurotrophicfactor (BDNF) in different brain regions such as the dorsalstriatum (Meisner et al. 2008) or in different cortices pro-jecting to the dorsal striatum such as the medial prefron-tal cortex (mPFC) and limbic cortex (Marvanova et al.2001). The interest for BDNF in the alcohol field isgrowing, and some strong evidence supporting a role forBDNF in the regulation of EtOH consumption hasemerged these recent years (Hensler, Ladenheim & Lyons2003; McGough et al. 2004; Jeanblanc et al. 2006;Logrip, Janak & Ron 2008). More precisely, endogenousBDNF within the dorsolateral striatum (DLS) decreasesEtOH consumption (Jeanblanc et al. 2009). Together,these results suggest that BDNF is part of a homeostaticpathway responsible for the regulation of EtOH consump-tion after its own EtOH-induced stimulation.
Based on memantine’s ability to enhance BDNFmRNA expression within brain regions involved inEtOH consumption, we examined whether memantinecould alter EtOH self-administration and motivation toconsume EtOH in moderate- and high-drinking rats.Finally, we tested the hypothesis that the effect ofmemantine on EtOH intake is mediated by BNDF andfurther explore the molecular mechanisms involved.
MATERIALS AND METHODS
Animals
Male Long-Evans rats (320–345 g at the beginning ofthe experiment) were obtained from Charles River(L’Arbresle, France). Animals were individually housedunder a light/dark cycle of 12 hours (lights on at7:00 am) with food and water available ad libitum. Experi-ments were carried out in accordance to the guidelinesfor Care and Use of Laboratory Animals (National Insti-tutes of Health) and the European community regula-tions for animal use in research (CEE no. 86/609) andwere approved by the local ethics committee.
Reagents
Memantine was purchased from Tocris (Bristol, UK) andwas dissolved in NaCl (0.9%). K252a and paraformalde-hyde (PFA) were purchased from Sigma-Aldrich (Stein-heim, Germany). K252a was dissolved to a concentrationof 100 mM in a 50% dimethylsulfoxide (DMSO, Prolabo,Fontenay-sous-bois, France) – 50% artificial cerebrospi-nal fluid (aCSF) solution (CMA Microdialysis, Solna,Sweden) as described in Whitfield et al. (2011). TrkB-Fcand immunoglobulin G (IgG) were purchased from R&DSystems Europe (Lille, France). EtOH 96% was purchasedfrom VWR (Fontenay-sous-bois, France). D-(+)-glucosewas purchased from Sigma-Aldrich (Saint Quentin,France).
EtOH self-administration
Following 3 weeks of exposure to EtOH in the home cage(using a sucrose-fading procedure adapted from Samson1986) consisting of a two-bottle choice paradigm with 1week exposure to a 10% EtOH – 10% sucrose solution, 1week with 10% EtOH – 5% sucrose, and the last weekwith a solution of 10% EtOH in addition to a tap waterbottle), rats were trained to self-administer a solution of10% EtOH (v/v). The self-administration chambers con-tained two levers: an active lever for which pressesresulted in delivery of a 0.1 ml of fluid reward (a 10%EtOH solution), and an inactive lever, for which presseswere counted, but no programmed events occurred. After3 days under a fixed ratio 1 (FR1, 1 press delivers 1reward) schedule, the rats were trained on an FR3 sched-ule (3 presses are required to receive 1 reward) duringdaily 30-minute sessions 5 days per week. Animals weretrained for at least 5 weeks before the beginning of theexperimental manipulations. The animals exhibiting lessthan 40 presses on the active lever at the end of the 5weeks of training for five consecutive sessions wereexcluded from the study because of their low EtOH con-sumption. During the self-administration sessions,number of lever presses and number of EtOH deliverieswere recorded using PackWin software (Bioseb, Vitrolles,France).
Self-administration of high levels of EtOH
The induction of EtOH consumption in the home cagepreviously described leads to moderate amounts of self-administered EtOH. In order to obtain larger amounts ofself-administered EtOH, we performed an induction ofEtOH consumption in the home cage based on the 20%intermittent access (IA) protocol. Naïve rats were exposedto a 20% IA paradigm with access to two bottles onecontaining tap water and the other a 20% EtOH solution,every other day in the homecage. This procedure inducesescalation of EtOH consumption in several strains of
rats (Wise 1973; Carnicella, Amamoto & Ron 2009;Bito-Onon et al. 2011). After 4–5 weeks of 20% IA, ratswere trained to self-administer EtOH as described earlier.However, for this experiment, reward was 0.1 ml of a20% EtOH solution. This paradigm allows us to observeEtOH intake reaching a level near 1.5 g/kg/30 min.According to Carnicella et al., this level of EtOH con-sumption leads to blood EtOH concentration rangingfrom 60 to 80 mg% (Carnicella et al. 2009).
Glucose self-administration
Naïve rats were initially trained in three overnightsessions under an FR1 schedule using 0.1 ml of a 5%glucose solution as the reinforcer. Subsequently, rats weretrained 5 days a week in 30-minute sessions, with the FRschedule progressively increased to FR3 and the glucoseconcentration progressively decreased to 1.5%.
Memantine injection
The different doses of memantine (12.5 and 25 mg/kg,i.p.) were chosen accordingly to the literature (Marvanovaet al. 2001). The delay of 6 hours between the injectionand the beginning of the self-administration session waschosen based on our hypothesis that memantine wouldstimulate the expression of BDNF and in turn, induce theexpression of specific effectors. The half-life of memantinehas been shown to be between 126 and 205 minutes(Spanagel, Eilbacher & Wilke 1994). These solutions andthe vehicle were injected once a week, and each week theinjections were counterbalanced following a Latin squaredesign.
Progressive ratio
After rats were trained to self-administer EtOH (FR3—10% EtOH) a progressive ratio session was conducted6 or 30 hours post-memantine injection. In the pro-gressive ratio session, the effort needed to get 1 reward(i.e. number of presses on the active lever) was continu-ously increased after each reward delivery (3, 4, 5, 7, 9,12, 15, 17, 20, 22, 25, 28, 30, 33, 35). First meman-tine was injected 6 hours prior to the test session. Aweek later a second injection occurred and the ratshad a regular session 6 hours post-injection andwere submitted to the progressive ratio test 30 hourspost-injection.
Surgery
Stereotaxic surgery was used to implant each rat with acannula into the lateral ventricle. Rats were continuouslyanesthetised with isoflurane during the surgery. Fourholes were drilled for screws, one other hole was drilledfor the placement of cannulae (Lateral ventricle: single
cannula (26GA, 12 mm, Phymep). The coordinates forthe lateral ventricle were +1.4 mm lateral to the medialsuture, -0.8 mm posterior to bregma and -3 mm fromthe skull surface. The cannulae were fixed with dentalcement. Subject weights were monitored daily after thesurgery to ensure recovery. One week after recovery, sub-jects returned to self-administration training and werehabituated to the microinjection procedure with twosham injections. The experimental micro-injectionsbegan upon acquisition of stable responding for EtOH.The injectors used for each group extended 1 mm belowthe tip of the cannula.
Intra-cerebroventricular (i.c.v.) micro-infusions
K252a or its vehicle (50% DMSO – 50% aCSF) weremicro-infused directly into the lateral ventricle 2 or 26hours after the i.p. injection of memantine (25 mg/kg) orits vehicle. Micro-infusions were performed with anHarvard pump 11 Plus Advanced (Phymep, France) and25-ml Hamilton syringes (1702N, Bonaduz, Switzerland).Injectors were built using stainless steel tubing (outerdiameter: 0.229 mm, inner diameter: 0.127 mm,Phymep) and had a length of 13 mm (1 mm more thanthe tip of the guide cannula implanted in the rat brain).The delay of 2 hours between the memantine and theK252a injections was chosen to maximize the inhibitoryeffect of K252a on TrkB receptors, while the expression ofBDNF is increased. TrkB-Fc or its control IgG were micro-infused i.c.v. at a concentration of 0.1 mg/ml simultane-ously with memantine injection. The injections of 3 ml ofK252a, TrkB-Fc or their respective controls were per-formed at a speed of 1 ml/min and the injector remainedin the guide for an additional 2 minutes.
Striatal BDNF protein level evaluation
Our hypothesis is based on the finding that BDNF withinthe dorsal striatum reduces EtOH consumption. There-fore, we evaluated BDNF protein levels 6 and 30 hourspost-memantine injection in EtOH-naïve rats. Striatawere dissected, collected and frozen until protein determi-nation and evaluation of BDNF level by enzyme-linkedimmunosorbent assay (ELISA). After adding boilinglysis buffer (pH 9.5, 10 mM sodium orthovanadate) thesamples were sonicated, heated 5 minutes at 100°C andthen centrifuged for 10 minutes at 8000 g. BDNF levelswere evaluated using a commercial Rat BDNF ELISA kitaccording to the manufacturer’s instructions (MyBio-Source, San Diego, CA, USA).
EtOH-naïve rats were used for this particular experiment.Because of translation mechanisms, which require timeto produce the protein from the mRNA, we decided to
evaluate the effect of memantine on gene expression 2hours prior to the evaluation of the protein levelsdescribed earlier. Therefore, 4 or 28 hours after i.p. injec-tion of memantine (25 mg/kg), rats were euthanized,brain removed and prefrontal cortices collected andfrozen until the real-time PCR experiment as describedpreviously (Botia et al. 2012). Total RNAs were harvestedfrom brain tissues using the Trizol method (Invitrogen,Carlsbad, CA, USA). After further purification using theRNeasy Lipid Tissue Mini kit (Qiagen, Valencia, CA, USA),contaminating genomic DNA was eliminated by treat-ment with DNAse I (Qiagen) and cDNA were synthesizedfrom 1 mg of RNA using RT2 profiler PCR array firststrand kit (SABiosciences, by way of Qiagen, Frederick,MD, USA). Custom RT2 Profiler PCR arrays (SABio-sciences, see complete list of genes in SupportingInformation Table S1) were performed according to themanufacturer’s instructions in the presence of SYBRgreen master mix (SABiosciences) using the ABI PrismStep One Plus detection system (Applied Biosystems,Courtabœuf, France). Each custom array contains apanel of 44 primer sets for a set of 38 genes (see completelist in Supporting Information Table S1 and Table 1),three RNA and PCR quality controls and three referencegenes for variations in amounts of input mRNA (gusb,gapdh, actb). The normalization to the reference geneswas performed as defined by SABiosciences. The meanCt for the three reference genes was calculated over thefive replicates. Genes of interest were normalized to thismean (21.3 < mean Ct reference genes < 21.9). Datawere analysed with a PCR array data analysis templatedownloaded from the Superarray Web site (http://www.sabiosciences.com) using the 2-DDCt method.
Histology
Rats implanted with cannulae were perfused transcar-dially with the fixative (4% PFA), and then 75-mm coronalslices were cut and examined for cannula placementsafter thionine staining. No animals were excluded formisplacements of the cannula.
Statistical analysis
Biochemical and behavioural data were analysed by one-or two-way (ANOVA) with repeated measures, dependingon the experiment, followed by the Student–Newman–Keuls test when indicated by significant effects of treat-ments or interactions. For simple comparisons, data wereanalysed by a Student’s t-test. Significance for all testswas set at P < 0.05. A Bonferroni correction was appliedfor the gene expression analyses. The significance for thisexperiment was set at P < 0.000658.
RESULTS
Memantine decreases EtOH self-administration
First we evaluated the levels of self-administration of a10% EtOH solution 6, 30 and 54 hours after acute i.p.injections of 0, 12.5 and 25 mg/kg of memantine. Wefound that a single injection of memantine dose-dependently decreased the number of presses on theactive lever (delivering the 10% EtOH solution) wheninjected 6 hours prior to the test session (Fig. 1a,F(2,24) = 4.10, P < 0.05). Post hoc analysis showed a sig-nificant difference between the 0 and 25 mg/kg groups(P < 0.001), and between the 12.5 and 25 mg/kg groups(P = 0.01). Similarly, the amount of EtOH consumed wasreduced after memantine injection (Fig. 1b, F(2,24) = 4.93,P < 0.05) and post hoc analysis indicated significant dif-ferences between the 25 mg/kg group and both control(0 mg/kg) and 12.5 mg/kg groups (P’s < 0.05).
EtOH self-administration was still reduced 30 hoursafter the memantine treatment (Fig. 1c, F(2,24) = 4.097,P < 0.05). Post hoc analysis revealed a significant differ-ence between the 0 and 25 mg/kg groups (P < 0.001),and between the 12.5 and 25 mg/kg groups (P < 0.01).The amount of EtOH consumed is thus also reduced bymemantine (Fig. 1d). A one-way ANOVA-RM analysisrevealed a main effect of the Treament (F(2,24) = 5.14,P < 0.05) and post hoc analysis showed significant differ-ences between the 25 mg/kg group and the groups 0and 12.5 mg/kg (P’s < 0.05). No effect of memantinewas observed 54 hours after the injection (Fig. 1e,F(2,24) = 0.47, P = 0.63).
Differential effect of memantine on the patternof drinking
The pattern of drinking was analysed using the cumula-tive active lever presses (Fig. 2a,b). Six hours post-injection, memantine postpones the beginning of thedrinking episode but does not alter the profile of EtOHdrinking whereas at the +30 hours timepoint rats undermemantine started drinking immediately at the begin-ning of the session but stopped rapidly. The two-wayANOVA-RM conducted on the data obtained 6 hourspost-injection revealed main effect of both factors (treat-ment: F(2,120) = 9.18, P < 0.001, time: F(5,120) = 40,88,P < 0.001), but no interaction between both factors(F(10,120) = 1.62, P = 0.11). The group memantine25 mg/kg is significantly different from the control groupall along the session (P’s < 0.01). Regarding the meman-tine effect +30 hours post-injection the analysis revealeda marginal main effect of the treatment (F(2,120) = 3.23,P = 0.057), a main effect of the time (F(5,120) = 31.43,P < 0.001) and an interaction between these two fac-tors (F(10,120) = 4.96, P < 0.001). The post hoc analysis
revealed that the group memantine 25 mg/kg starts todiffer from the controls at the intervals 10–15 until theend of the session (P’s < 0.05).
Memantine decreases motivation to consume EtOH
Motivation to consume EtOH was assessed by means of aprogressive ratio protocol in which increasing numbers oflever presses are required to get a reward after eachreward delivery. We chose to inject the most effectivedose of memantine (25 mg/kg). The breakpoint, or themaximum number of lever presses emitted by theanimals to obtain only 0.1 ml of the EtOH solution, is anaccepted index of the animal’s motivation to obtain thereward. We found that the breakpoint was significantlyreduced 6 hours (Fig. 3a, P < 0.01) and 30 hours(Fig. 3b, P < 0.05) after the memantine injection.
Memantine effect in ‘high-drinking’ rats on EtOHself-administration and on glucose self-administration
The rats used in the first experiments were ‘moderatedrinkers’ with an average of 0.34 � 0.07 g/kg EtOHconsumed in a 30-minute self-administration session. Wethen tested the effects of memantine on ‘high drinkers’
consuming an average of 1.27 � 0.31 g/kg/30 minduring each self-administration session. We found thatmemantine had a similar effect on these ‘high drinkers’compared with moderate drinkers, thus confirming itsefficacy on the reduction of EtOH consumption evenwhen the consumption is elevated. (Fig. 4a, P’s < 0.05).
To test whether memantine has a general effect onself-administration of rewarding substances, we trainednaive rats to self-administer the natural reinforcerglucose (1.5% in tap water). We found that memantinedid not alter glucose self-administration at 6 or 30 hourspost-injection (Fig. 4b, +6 hours: F(1,13) = 0.78, P = 0.39;+30 hours: F(1,13) = 1.63, P = 0.22).
Taken together, these results demonstrate thatmemantine decreases EtOH self-administration in both‘moderate’ and ‘high drinkers’ for at least 30 hours.
Involvement of BDNF in the early and delayed effect ofmemantine on EtOH self-administration
BDNF levels were measured in dorsal striatum samplescollected 6 or 30 hours post-memantine injection. Wefound that BDNF protein levels were significantlyincreased 6 hours after the memantine injection whereas
0
20
40
60
80
100
0 12.5 25Memantine (mg/kg)
Nu
mb
er
of
pre
ss
es Active
lever
Inactive
lever
*****
(a) (b)
(c) (d)
Nu
mb
er
of
pre
ss
es
0
20
40
60
80
100
0 12.5 25
*****
Memantine (mg/kg)
0.0
0.1
0.2
0.3
0.4
0.5
EtO
H c
on
su
mp
tio
n
(g/k
g/3
0 m
in)
0 12.5 25Memantine (mg/kg)
0.0
0.1
0.2
0.3
0.4
0.5
EtO
H c
on
su
mp
tio
n
(g/k
g/3
0 m
in)
0 12.5 25Memantine (mg/kg)
**
**Active
lever
Inactive
lever
0
20
40
60
80
0 12.5 25Memantine (mg/kg)
Nu
mb
er
of
pre
ss
es
Active
lever
Inactive
lever
(e)
Figure 1 A single injection of memantinedecreases ethanol (EtOH) self-administration.Memantine was injected i.p. and EtOH self-administration was evaluated 6 hours (a, b)and 30 hours (c, d) post-injection. (a, c)Memantine decreased the number of presseson the active lever. Results are expressedas mean � standard error of measurement(SEM) number of presses. (b, d) Memantinedecreased EtOH consumed during the self-administration sessions. Results are expressedas mean � SEM EtOH consumed in g/kg/30 min. (e) Fifty-four hours post-injection,memantine did not affect EtOH self-administration. Results are expressed asmean � SEM number of presses. n = 13,*P < 0.05, **P < 0.01, ***P < 0.001
the BDNF levels went back to baseline at the delayed time-point [Fig. 5a, Memantine: F(1,24) = 5.97, P < 0.05; Time:F(1,24) = 4.62, P < 0.05; Memantine ¥ Time: F(1,24) =2.23, P = 0.15]. The post hoc analysis indicated a signifi-cant effect of memantine at 6 hours compared with thevehicle (P < 0.01) and a significant difference betweenthe two timepoints (P < 0.05). Then we inhibited TrkBreceptors 2 hours after the memantine injection by i.c.v.micro-infusion of the Trk inhibitor K252a (3 ml at100 mM). We found that co-treatment with K252a
resulted in the total blockade of the memantine effectwhile K252a had no effect by itself (Fig. 5b, K252a:F(1,9) = 111.01, P < 0.001; Memantine: F(1,9) = 2.88,P = 0.12; K252a ¥ Memantine: F(1,9) = 9.94, P < 0.05).The post hoc analysis showed a significant differencebetween the saline–vehicle/memantine–vehicle groups(P < 0.01) and between the memantine–vehicle/memantine–K252a groups (P < 0.001). We found that,despite the reversal of the memantine-decreasing effectby K252a 6 hours post-memantine injection, EtOHself-administration was again decreased 30 hours afterthe memantine injection (Fig. 5c, K252a: F(1,9) = 0.24,P = 0.64; Memantine: F(1,9) = 8.44, P < 0.05; K252a ¥Memantine: F(1,9) = 0.19, P = 0.67). The post hoc testrevealed a significant difference between the saline –vehicle and memantine – vehicle groups (P < 0.05).
To ensure that the delayed effect of memantine wasBDNF-independent, K252a was micro-infused 4 hoursbefore the +30 hours self-administration session. Wefound that K252a was unable to block the delayedmemantine effect on EtOH consumption (Fig. 5d, K252a:F(1,7) = 0.02, P = 0.9; Memantine: F(1,7) = 11.73, P <0.05; K252a ¥ Memantine: F(1,7) = 0.09, P = 0.76).
Then, in a separate group of rats we micro-infusedTrkB-Fc which is a specific BDNF scavenger and hasa longer half-life than K252a. We found that TrkB-Fc blocked the memantine effect on EtOH self-administration 6 hours after the micro-infusion (Fig. 5e,Memantine: F(1,10) = 12.0, P < 0.01, TrkB-Fc: F(1,10) =11,29, P < 0.01; Memantine ¥ TrkB-Fc: F(1,10) = 19.03,P < 0.001), but not 30 hours after the micro-infusion(Fig. 5f, Memantine: F(1,10) = 23.18, P < 0.001; TrkB-Fc:F(1,10) = 3.25, P = 0.10; Memantine ¥ TrkB-Fc: F(1,10) =0.08, P = 0.78) confirming the K252a results. The posthoc analysis conducted on the +6 hours data revealed asignificant difference between the groups Saline-IgG/Memantine-IgG and between the groups Memantine-IgG/Memantine-TrkB-Fc (P’s < 0.001). However the posthoc analysis on the +30 hours data revealed differencesbetween only the Saline-IgG/Memantine-IgG groupsand the Memantine-TrkB-Fc/Saline-TrkB-Fc groups(P’s < 0.01).
Memantine dynamically regulates specific genesexpression within the mPFC
We performed a PCR array analysis on 38 candidategenes (Supporting Information Table S1 and Table 1)within the mPFC. Brain samples were collected 4 or 28hours post-memantine injection. Specifically, variationsin the expression of genes coding neurotrophic factorsand their receptors were investigated (Fig. 6). Asexpected, we found an increase in Bdnf expression 4hours post-memantine injection (P = 0.0001), but not at28 hours, when we observed a small and not significant
0–5 5–10 10–15 15–20 20–25 25–30
0–5 5–10 10–15 15–20 20–25 25–30
Saline
25 mg/kg12.5 mg/kg
0
20
40
60
80
Cu
mm
ula
tive
lever
pre
ss
es
Time (blocks of 5 minutes)
Time (blocks of 5 minutes)
0
20
40
60
80
Cu
mm
ula
tive
lever
pre
ss
es
(a)
(b)
+6 hours
+30 hours
*****
*
*
*** *** *****
*
* * ** **
Saline
25 mg/kg12.5 mg/kg
Figure 2 Memantine postpones the initiation of the drinkingepisode at 6 hours, but stops it at 30 hours. (a) Memantine did notalter the profile of the pattern of drinking when injected 6 hoursprior to the self-administration session. (b) Memantine altered theprofile of the pattern of drinking when injected 30 hours prior to theself-administration session. Results are expressed as mean cumulativelever presses. n = 13, *P < 0.05, **P < 0.01, ***P < 0.001 comparedwith the saline group. Errors bars have been omitted for claritypurpose
(a)
0
Bre
ak
po
int
25
**
0
5
10
15
20
25
Memantine (mg/kg)
0
5
10
15
20
25
Bre
ak
po
int
0
*
25Memantine (mg/kg)
(b)
Figure 3 Memantine reduces the motivation to consume ethanol(EtOH). At both timepoints, +6 hours (a) and +30 hours (b) post-memantine injection, motivation to consume EtOH evaluated in aprogressive ratio paradigm was reduced by memantine (25 mg/kg).Results are expressed as mean � standard error of measurementmaximum number of consecutive presses made to obtain 1 reward(i.e. breakpoint). n = 12, *P < 0.05, **P < 0.01
Figure 4 Memantine decreases high levels of ethanol (EtOH) consumption, but does not alter glucose self-administration. (a) Naïve rats weretrained to self-administer a 20% EtOH solution after 5 weeks of two-bottle choice 20% intermittent access.The amounts of EtOH consumedwere threefold those obtained with moderate drinkers. In this model of ‘high drinking’, a single 25 mg/kg memantine injection still decreasedEtOH self-administration at 6 and 30 hours post-injection. Results are expressed as mean � standard error of measurement (SEM) EtOHconsumed in g/kg/30 min. n = 5, *P < 0.05. (b) Memantine (25 mg/kg) was tested on a separate group of EtOH-naïve rats trained toself-administer glucose (1.5%). Memantine did not alter glucose self-administration at 6 nor 30 hours post-injection. Results are expressed asmean � SEM number of presses. n = 14
(b)
Sal.
Veh.
Mem.
Veh.
Mem.
K252a
Sal.
K252a
*****
0
20
40
60
80
Nu
mb
er
of
pre
ss
es
(c)
0
20
40
60
80
Nu
mb
er
of
pre
ss
es
*
Sal.
Veh.
Mem.
Veh.
Mem.
K252a
Sal.
K252a
0
20
40
60
80
100
Nu
mb
er
of
pre
ss
es
*** *
Sal.
Veh.
Mem.
Veh.
Mem.
K252a
Sal.
K252a
0
20
40
60
80
100
Nu
mb
er
of
pre
ss
es
*** ***
(d)
0204060
80100120
Nu
mb
er
of
pre
ss
es
** **(e)
(a)
–10
0
10
20
30
40
50
60
6 hours 30 hours
% c
ha
ng
e v
ers
us c
on
tro
ls
**#
(f)
Figure 5 The early decrease in ethanol (EtOH) self-administration induced by memantine is brain-derived neurotrophic factor (BDNF)-dependent. (a) Memantine increased BDNF protein levels within the dorsal striatum 6, but not 30 hours post-injection. n = 6–8. **P < 0.01versus controls, #P < 0.05 versus memantine +30 hours. (b) Rats received a micro-infusion of the inhibitor of the Trk receptors, K252a 2 hoursafter the memantine injection. The co-injection of K252a with memantine fully reversed the memantine-induced decrease in EtOHself-administration. Results are expressed as mean � SEM number of presses on the active lever. n = 10, **P < 0.01, ***P < 0.001. (c) On thefollowing day (+30 hours post-memantine injection), K252a micro-infusion had no effect on the memantine-induced reduction of EtOHself-administration. Results are expressed as mean � standard error of measurement (SEM) number of presses on the active lever. n = 10,*P < 0.05. (d) Micro-infusion of K252a 2 hours prior to the +30 hours post-memantine injection had no effect on the reduction of EtOHself-administration induced by memantine. Results are expressed as mean � SEM number of presses on the active lever. n = 8, *P < 0.05,***P < 0.001. (e, f ) Rats received a micro-infusion of the BDNF scavenger TrkB-Fc simultaneously with the memantine injection. Theco-injection of TrkB-Fc with memantine fully reversed the memantine-induced decrease in EtOH self-administration. Results are expressed asmean � SEM number of presses on the active lever. n = 11, *** P < 0.001. (f ) On the following day (+30 hours post-memantine and TrkB-Fcinjections), TrkB-Fc had no effect on the memantine-induced reduction of EtOH self-administration. Results are expressed as mean � SEMnumber of presses on the active lever. n = 11, **P < 0.01
decrease (-34%). The gene coding the receptor TrkB(Ntrk2) was not affected by memantine at either time-point. Expression of the genes coding the two other neu-rotrophins (Ngf and Ntf3) was unaltered by memantine atboth timepoints.
With respect to the glutamatergic system (seeTable 1), several receptors or sub-units of glutamatergicreceptors were down-regulated 4 hours post-injection(Grin3a and the metabotropic receptors Grm1 and 2).Twenty-eight hours post-injection, none of the genesstudied showed significant regulation after Bonferroni’scorrection. However, there was a trend toward anincrease for two genes, Npy (coding neuropeptide Y) andTh (coding tyrosine hydroxylase).
DISCUSSION
Here, we report a dose-dependent effect of memantine onEtOH self-administration 6 and 30 hours post-injection inrats self-administering moderate amounts of EtOH butalso in a model of ‘high drinking rats’ consuming 3 timesmore EtOH and leading to high blood EtOH concentra-tions near 70 mg% (Carnicella et al. 2009). Our resultsalso reveal that memantine reduces motivation toconsume EtOH. We also show that a single injection issufficient to observe a long lasting effect on EtOH intakeand motivation whereas memantine is typically testedchronically on cognitive functions. Because the maximalconcentrations of memantine are reached in the brain60–80 minutes after the injection and that its eliminationhalf-life is 126–205 minutes (Spanagel et al. 1994), boththe early and delayed effects of memantine are likely to bedue to protein and gene expression alterations. Meman-tine injected 6 hours prior to the self-administrationsession dose-dependently postponed the initiation of the
drinking episode but did not altered its termination. It isnoteworthy that this retardment effect is not observed 30hours post-injection but instead memantine induced anearly termination of the drinking episode. This differencesuggests distinct mechanisms between the short termand long term effects of memantine. We also show thatmemantine has no effect on glucose self-administrationdemonstrating that its effect seems selective to alcoholand is not linked to a general impact on reward system. Inaddition, it is unlikely that the effect of memantine couldbe due to motor and memory impairments since meman-tine had no effect on glucose self-administration.
Our results on EtOH intake are in line with findingsfrom different groups (Holter et al. 1996; Krupitsky et al.2007; Kuzmin, Stenback & Liljequist 2008; Malpass et al.2010). These results in addition to those described in thisreport suggest that memantine is able to decrease EtOHcraving in the absence of alcohol. It is of interest to notethat in humans memantine is reported to be effective oncraving when administered 4 hours before EtOH but notafter EtOH was given (Bisaga & Evans 2004). Together,these results suggest that memantine may have thera-peutic value in the treatment of alcohol-use disorders.
Emerging evidence suggests that BDNF regulatesEtOH drinking (McGough et al. 2004; Logrip et al. 2008;Jeanblanc et al. 2009) and BDNF is produced after block-ade of NMDA receptors (Marvanova, Lakso & Wong2004; Jeon et al. 2006; Kalinichev et al. 2008). We con-firmed that Bdnf expression is rapidly increased aftermemantine injection (Marvanova et al. 2001) and thatBDNF protein levels within the dorsal striatum are alsoincreased but only at the early timepoint. Therefore wetested whether the memantine effects are mediated viaactivation of the BDNF signalling pathway using first theK252a Trk inhibitor and then the BDNF scavenger TrkB-Fc-IgG chimera fusion protein. The blockade of the earlymemantine effect with the co-injection of either K252aor TrkB-Fc shows that the effect of memantine at 6 hoursis BDNF-dependent, but its delayed effect at 30 hoursremains unaltered by these treatments. In addition,despite the blockade of the early effect, the delayed effectof memantine is still present suggesting that these twosimilar effects are unrelated. Thus, instead of having asingle mechanism lasting for 30 hours we identified bothan initial effect of memantine mediated by the BDNF sig-nalling pathway and a delayed memantine effect that isBDNF-independent. The absence of regulation of Bdnfexpression 28 hours after memantine may explain thelack of efficacy of K252a at this timepoint. Accordingly,K252a has no effect on EtOH consumption in hetero-zygous BDNF knockout mice (Jeanblanc et al. 2006);however, K252a and TrkB-Fc have no effect wheninjected alone. This lack of effect can be explained by thefact the K252a has a short half-life and thus is not active
–100
–50
0
50
100
150
200
250
Bdnf Ntrk2 Ngf Ngfr Ntf3% c
ha
ng
e v
ers
us
co
ntr
ols
Memantine +4 hours
Memantine +28 hours
***
Figure 6 Dynamic regulation of gene expression within the medialprefrontal cortex (mPFC) by memantine 4 and 28 hours post-injection. Ethanol (EtOH)-naïve rats were injected with memantine(25 mg/kg) and 4 or 28 hours later, brains were removed and mPFCsamples were collected. Results are expressed as mean � standarderror of measurement percentage of change compared with theircontrols (+4 and +28 hours post-saline injection). Bdnf was the soleneurotrophin-related coding gene up-regulated by memantine +4hours post-injection. Ngf and Ngfr were unaltered by memantinewhile Ntf3 was down-regulated at both timepoints. n = 5,***P < 0.001
oprm1 NM_013071 Opioid receptor, mu 1 1.04 0.4864 1.08 0.4914Pak1 NM_017198 P21 protein (Cdc42/Rac)-activated kinase 1 0.96 0.4520 0.89 0.0485Ppp2cb NM_017040 Protein phosphatase 2, catalytic subunit, beta isoform 1.07 0.1671 0.96 0.3148Creb1 NM_031017 CAMP responsive element binding protein 1 0.79 0.0065 0.99 0.9415
Genomic profiles were investigated by polymerase chain reaction array in control (saline injection) and memantine injected rats 4 and 28 hours post-injections. Foreach gene, simple comparison was performed between the memantine group (4 or 28 hours) versus its own saline group (4 or 28 hours). n = 5 per group. P-valuesdepicted in the Table 1 are uncorrected. Grey shading represents and gene name in bold P-values < 0.05 before Bonferroni correction and the asterisk signalssignificant difference after Bonferroni’s correction.
during the self-administration session (4 hours after themicro-infusion) and does not inhibit anymore the TrkBreceptors. With respects to both blockers of the BDNFsignalling pathway, we suggest that the BDNF systemneeds to be stimulated first by EtOH or memantine tobe able to observe their effect on memantine-induceddecrease in EtOH self-administration. In other word, onbaseline levels of BDNF, the signalling pathway might notbe activated enough to observe behavioural changes fol-lowing its inhibition. The absence of up-regulation ofother neurotrophins or their receptors in addition to theeffect of K252a strongly suggest that the observed block-ade is due to the inhibition of the TrkB receptors and notother Trk receptors. These results are in line with the factthat the BDNF-reducing effect on cocaine seeking withinthe prefrontal cortex is mediated through TrkB receptoractivation (Whitfield et al. 2011).
To identify putative mediator(s) of the delayedmemantine effect we investigated the effect of memantineon the expression of 38 candidate genes, selected eitherbecause of their interaction with the glutamatergicsystem or based on their known involvement in addic-tion. We chose to perform this experiment on mPFCsamples because this brain region is recognized as acrucial area responsible for the control of the action andin EtOH-related behaviours (Alexander, DeLong & Strick1986; Balleine & O’Doherty 2010). In addition, alteredsynaptic plasticity of the glutamatergic system within thecortico-limbic involving the nucleus accumbens and thePFC after drugs of abuse exposure has been largely dem-onstrated (Kalivas 2009; Luscher & Malenka 2011). Thegene expression analysis reveals that memantine potentlydown-regulated expression of numerous genes codingglutamatergic receptors or their sub-units and most ofthe gene regulations occurred at the early timepoint.Several genes such as those coding the mglu2, three areknown to be regulators of glutamate release (Moussawi &Kalivas 2010). Thus, it is not surprising to observe a com-pensatory adaptation of the glutamatergic system afterthe inhibition of one of its components.
We also found a trend to an increase of CRH mRNAlevels in the mPFC. This finding is in line with previousobservations in other brain regions (Zhou et al. 1998;Marvanova et al. 2004), associated with a decrease inexpression of the gene coding the main CRH receptor,Crhr1, as previously shown (Zhou et al. 1998). Both i.c.v.microinfusion of CRH and lack of expression of the Crhr1gene in knockout mice reduce EtOH intake (Sillaber et al.2002; Thorsell, Slawecki & Ehlers 2005; Kaur et al.2012). However, crhr1 expression was increased specifi-cally within the basolateral and medial nuclei of the amy-gdala in rats with history of dependence (Sommer et al.2008). In EtOH-dependent animals, it is now well estab-lished that the CRH system is a promising therapeutic
target because CRH levels are increased during acutewithdrawal and CHRH1 antagonists block EtOH seeking(Koob & Zorrilla 2010). Only 2 genes show a trend to anup-regulation at the delayed timepoint. The first one is thegene coding the NPY and numerous studies have sug-gested a role for NPY in EtOH-related behaviours (Thieleet al. 2004) with a specific role within the amygdala,which receive direct projections from the mPFC. Thesecond gene, Th, codes the enzyme responsible for thesynthesis of dopamine, or Tyrosine hydroxylase. Anincrease in dopamine release within the mesocorticolim-bic system can be responsible for a decrease in the interestfor alcohol (Gibb et al. 2011) and such an increase in DArelease has been observed within the PFC after meman-tine injection (Spanagel et al. 1994). Thus, both genes arepotential effectors responsible for the reduction of EtOHself-administration observed 30 hours after the meman-tine injection but further investigation is required toconfirm these results.
It is noteworthy that ketamine, another non-competitive NMDA antagonist, has been shown to haverapid antidepressant effect through the activation of theBDNF signalling pathway (Autry et al. 2011; Monteggia,Gideons & Kavalali 2012). Knowing the link between lowlevels of BDNF and depression and/or alcohol-use disor-ders, it is of particular interest to unravel the molecularmechanisms of action of such non-competitive NMDAantagonists.
In conclusion, our results demonstrate that BDNFplays a key role in the suppression of both EtOH self-administration and motivation to drink EtOH by meman-tine and that the long-lasting effect of memantine may bemediated by BNDF-independent mechanisms or down-stream events in the BDNF-signalling cascade. Our geneexpression analysis identified candidate genes potentiallyinvolved in the sustained decrease in EtOH intake. Moregenerally, it is possible that memantine could substitutefor EtOH because EtOH also increases BDNF expressionwithin the DLS and hippocampus (McGough et al. 2004;Jeanblanc et al. 2009), which in turn leads to a control ofEtOH consumption (Jeanblanc et al. 2009) and thatmemantine could produce EtOH-like effects in humans(Krupitsky et al. 2007), and this point deserves furtherinvestigation. Our results further clarify the mechanismby which glutamatergic antagonist may reduce alcoholcraving and prolong abstinence duration in alcohol-dependent patients.
Acknowledgements
This study was supported by the Conseil Régional de Picar-die (CRP) and the National Institute of Health and MedicalResearch (INSERM). JJ and BB are supported by a post-doctoral Fellowship from the CRP. We thank Professor
Margaret Martinetti for helpful discussions and carefulrevision of the manuscript. We thank Ludovic Didier andVirginie Jeanblanc for their technical assistance.
Authors Contribution
JJ designed, performed and analysed the data and wrotethe manuscript. FC performed and analysed the behav-ioural experiments. BB performed and analysed themolecular biology experiments. MN designed the experi-ments and wrote the manuscript. All authors havecritically reviewed content and approved final versionsubmitted for publication.
References
Alexander GE, DeLong MR, Strick PL (1986) Parallel organiza-tion of functionally segregated circuits linking basal gangliaand cortex. Annu Rev Neurosci 9:357–381.
Autry AE, Adachi M, Nosyreva E, Na ES, Los MF, Cheng PF,Kavalali ET, Monteggia LM (2011) NMDA receptor blockadeat rest triggers rapid behavioural antidepressant responses.Nature 475:91–95.
Balleine BW, O’Doherty JP (2010) Human and rodent homolo-gies in action control: corticostriatal determinants of goal-directed and habitual action. Neuropsychopharmacology35:48–69.
Bisaga A, Evans SM (2004) Acute effects of memantine in com-bination with alcohol in moderate drinkers. Psychopharma-cology (Berl) 172:16–24.
Bito-Onon JJ, Simms JA, Chatterjee S, Holgate J, Bartlett SE(2011) Varenicline, a partial agonist at neuronal nicotinicacetylcholine receptors, reduces nicotine-induced increases in20% ethanol operant self-administration in Sprague-Dawleyrats. Addict Biol 16:440–449.
Botia B, Legastelois R, Alaux-Cantin S, Naassila M (2012)Expression of ethanol-induced behavioral sensitization isassociated with alteration of chromatin remodeling in mice.PLoS ONE 7:e47527.
Carnicella S, Amamoto R, Ron D (2009) Excessive alcohol con-sumption is blocked by glial cell line-derived neurotrophicfactor. Alcohol 43:35–43.
Evans SM, Levin FR, Brooks DJ, Garawi F (2007) A pilot double-blind treatment trial of memantine for alcohol dependence.Alcohol Clin Exp Res 31:775–782.
Gibb SL, Jeanblanc J, Barak S, Yowell QV, Yaka R, Ron D (2011)Lyn kinase regulates mesolimbic dopamine release: implica-tion for alcohol reward. J Neurosci 31:2180–2187.
Hensler JG, Ladenheim EE, Lyons WE (2003) Ethanol con-sumption and serotonin-1A (5-HT1A) receptor function inheterozygous BDNF (+/-) mice. J Neurochem 85:1139–1147.
Holter SM, Danysz W, Spanagel R (1996) Evidence for alcoholanti-craving properties of memantine. Eur J Pharmacol314:R1–R2.
Jeanblanc J, He DY, Carnicella S, Kharazia V, Janak PH, Ron D(2009) Endogenous BDNF in the dorsolateral striatum gatesalcohol drinking. J Neurosci 29:13494–13502.
Jeanblanc J, He DY, McGough NN, Logrip ML, Phamluong K,Janak PH, Ron D (2006) The dopamine D3 receptor is part ofa homeostatic pathway regulating ethanol consumption.J Neurosci 26:1457–1464.
Jeon GS, Park SH, Lee KJ, Lee MS, Chun BG, Shin KH (2006)Valproate prevents MK801-induced changes in brain-derivedneurotrophic factor mRNA in the rat brain. Eur J Pharmacol545:142–146.
Kalinichev M, Robbins MJ, Hartfield EM, Maycox PR, Moore SH,Savage KM, Austin NE, Jones DN (2008) Comparison betweenintraperitoneal and subcutaneous phencyclidine administra-tion in Sprague-Dawley rats: a locomotor activity and geneinduction study. Prog Neuropsychopharmacol Biol Psychiatry32:414–422.
Kaur S, Li J, Stenzel-Poore MP, Ryabinin AE (2012)Corticotropin-releasing factor acting on corticotropin-releasing factor receptor type 1 is critical for binge alcoholdrinking in mice. Alcohol Clin Exp Res 36:369–376.
Koob GF, Zorrilla EP (2010) Neurobiological mechanisms ofaddiction: focus on corticotropin-releasing factor. Curr OpinInvestig Drugs 11:63–71.
Krupitsky EM, Neznanova O, Masalov D, Burakov AM, DidenkoT, Romanova T, Tsoy M, Bespalov A, Slavina TY, GrinenkoAA, Petrakis IL, Pittman B, Gueorguieva R, Zvartau EE,Krystal JH (2007) Effect of memantine on cue-inducedalcohol craving in recovering alcohol-dependent patients. AmJ Psychiatry 164:519–523.
Krystal JH, Petrakis IL, Krupitsky E, Schutz C, Trevisan L,D’Souza DC (2003) NMDA receptor antagonism and theethanol intoxication signal: from alcoholism risk to pharma-cotherapy. Ann N Y Acad Sci 1003:176–184.
Kuzmin A, Stenback T, Liljequist S (2008) Memantine enhancesthe inhibitory effects of naltrexone on ethanol consumption.Eur J Pharmacol 584:352–356.
Lipton SA (2005) The molecular basis of memantine actionin Alzheimer’s disease and other neurologic disorders: low-affinity, uncompetitive antagonism. Curr Alzheimer Res 2:155–165.
Logrip ML, Janak PH, Ron D (2008) Dynorphin is a downstreameffector of striatal BDNF regulation of ethanol intake. FASEB J22:2393–2404.
Luscher C, Malenka RC (2011) Drug-evoked synaptic plasticityin addiction: from molecular changes to circuit remodeling.Neuron 69:650–663.
Malpass GE, Williams HL, McMillen BA (2010) Effects of thenon-competitive NMDA receptor antagonist memantine onthe volitional consumption of ethanol by alcohol-preferringrats. Basic Clin Pharmacol Toxicol 106:435–444.
Marvanova M, Lakso M, Pirhonen J, Nawa H, Wong G,Castren E (2001) The neuroprotective agent memantineinduces brain-derived neurotrophic factor and trkB recep-tor expression in rat brain. Mol Cell Neurosci 18:247–258.
Marvanova M, Lakso M, Wong G (2004) Identification of genesregulated by memantine and MK-801 in adult rat brain bycDNA microarray analysis. Neuropsychopharmacology 29:1070–1079.
McGough NN, He DY, Logrip ML, Jeanblanc J, Phamluong K,Luong K, Kharazia V, Janak PH, Ron D (2004) RACK1 andbrain-derived neurotrophic factor: a homeostatic pathwaythat regulates alcohol addiction. J Neurosci 24:10542–10552.
McMillen BA, Joyner PW, Parmar CA, Tyer WE, Williams HL(2004) Effects of NMDA glutamate receptor antagonist drugson the volitional consumption of ethanol by a genetic drink-ing rat. Brain Res Bull 64:279–284.
Meisner F, Scheller C, Kneitz S, Sopper S, Neuen-Jacob E,Riederer P, ter Meulen V, Koutsilieri E (2008) Memantineupregulates BDNF and prevents dopamine deficits in SIV-infected macaques: a novel pharmacological action ofmemantine. Neuropsychopharmacology 33:2228–2236.
Monteggia LM, Gideons E, Kavalali ET (2012) The role ofeukaryotic elongation factor 2 kinase in rapid antidepressantaction of ketamine. Biol Psychiatry DOI: doi.org/10.1016/j.biopsych.2012.09.006.
Moussawi K, Kalivas PW (2010) Group II metabotropic gluta-mate receptors (mGlu2/3) in drug addiction. Eur J Pharmacol639:115–122.
Olive MF (2009) Metabotropic glutamate receptor ligands aspotential therapeutics for addiction. Curr Drug Abuse Rev2:83–98.
Samson HH (1986) Initiation of ethanol reinforcement using asucrose-substitution procedure in food- and water-sated rats.Alcohol Clin Exp Res 10:436–442.
Sillaber I, Rammes G, Zimmermann S, Mahal B, ZieglgansbergerW, Wurst W, Holsboer F, Spanagel R (2002) Enhanced anddelayed stress-induced alcohol drinking in mice lacking func-tional CRH1 receptors. Science 296:931–933.
Sommer WH, Rimondini R, Hansson AC, Hipskind PA, GehlertDR, Barr CS, Heilig MA (2008) Upregulation of voluntaryalcohol intake, behavioral sensitivity to stress, and amygdalacrhr1 expression following a history of dependence. BiolPsychiatry 63:139–145.
Spanagel R (2009) Alcoholism: a systems approach from molecu-lar physiology to addictive behavior. Physiol Rev 89:649–705.
Spanagel R, Eilbacher B, Wilke R (1994) Memantine-induceddopamine release in the prefrontal cortex and striatum of the
Thiele TE, Sparta DR, Hayes DM, Fee JR (2004) A role for neu-ropeptide Y in neurobiological responses to ethanol and drugsof abuse. Neuropeptides 38:235–243.
Thorsell A, Slawecki CJ, Ehlers CL (2005) Effects of neuropeptideY and corticotropin-releasing factor on ethanol intake inWistar rats: interaction with chronic ethanol exposure. BehavBrain Res 161:133–140.
Whitfield TW, Jr, Shi X, Sun WL, McGinty JF (2011) The sup-pressive effect of an intra-prefrontal cortical infusion of BDNFon cocaine-seeking is Trk receptor and extracellular signal-regulated protein kinase mitogen-activated protein kinasedependent. J Neurosci 31:834–842.
Wise RA (1973) Voluntary ethanol intake in rats following expo-sure to ethanol on various schedules. Psychopharmacologia29:203–210.
Zhou Y, Yuferov VP, Spangler R, Maggos CE, Ho A, Kreek MJ(1998) Effects of memantine alone and with acute ‘binge’cocaine on hypothalamic-pituitary-adrenal activity in the rat.Eur J Pharmacol 352:65–71.
SUPPORTING INFORMATION
Additional Supporting Information may be found in theonline version of this article at the publisher’s web-site:
Table S1 List of the genes studied within mPFC samples 4and 28 hours after memantine injection in our array andtheir Ct values
![Turbine suppression[2]](https://static.documents.pub/doc/80x56/546d3cafb4af9f662c8b53ba/turbine-suppression2.jpg)
