Behavioural Brain Research 220 (2011) 126–131
Contents lists available at ScienceDirect
Behavioural Brain Research
journal homepage: www.elsevier.com/locate/bbr
Memantine abolishes the formation of cocaine-induced conditioned place
preference possibly via its IL-6-modulating effect in medial prefrontal cortex
Kuei-Ying Lina, Chianfang G. Cherngb, Fu-Rong Yanga, Li-Ching Linc, Ru-Band Lud, Lung Yua,c,∗
aInstitute of Behavioral Medicine, National Cheng Kung University College of Medicine, Tainan 701, Taiwan, ROC
bDepartment of Health Psychology, Chang Jung Christian University, Tainan 711, Taiwan, ROC
cDepartment of Physiology, National Cheng Kung University College of Medicine, Tainan 701, Taiwan, ROC
dDepartment of Psychiatry, National Cheng Kung University College of Medicine, Tainan 701, Taiwan, ROC
a r t i c l ei n f o
Received 14 July 2010
Received in revised form
15 December 2010
Accepted 19 January 2011
a b s t r a c t
function in cocaine-supported conditioning. Cocaine-induced conditioned place preference (CPP) was
used to assess the hedonic value and/or reinforcing efficacy of cocaine and cocaine-supported condi-
tioning. Systemic pretreatment with memantine (20, 2.0, 0.2, and 0.02mg/kg/injection) 30min before
each cocaine and saline conditioning trial abolished the acquisition of cocaine-induced CPP in mice.
Even a total of 0.12mg/kg memantine pretreatment in three days was effective in diminishing cocaine-
tumor necrosis factor (TNF-?) levels in medial prefrontal cortex (mPFC) and nucleus accumbens (Acb).
Interestingly, pretreatment with memantine at the lowest effective dose (0.02mg/kg/injection) reversed
cocaine conditioning-enhanced IL-6 and -decreased TNF-? levels in these brain regions. Neverthe-
less, such a memantine dosing regimen did not affect dopamine metabolism in mPFC and Acb. Single
memantine (0.02mg/kg) injection did not acutely affect mouse locomotor activity or cocaine-increased
locomotor activity. Similar memantine dosing regimen was ineffective to affect the maintenance of
cocaine-induced CPP. Finally, intra-mPFC infusion of recombinant IL-6, but not thalidomide, reversed
memantine (0.02mg/kg/injection×6)-decreased cocaine-induced CPP. These results, taken together,
suggest that cocaine conditioning-enhanced IL-6 in mPFC may be, in part, involved in the acquisition
of cocaine-induced CPP. Moreover, an extremely low dose of memantine may decrease the acquisition
of cocaine-induced CPP by reversing cocaine conditioning-increased IL-6 levels in mPFC.
© 2011 Elsevier B.V. All rights reserved.
Accrued evidence supports the notion that glutamatergic neu-
rotransmission plays a significant role in the neurobiology of
cocaine addiction [1–5]. Several glutamate receptor antagonists
have been reported to decrease cocaine-induced behavioral sen-
sitization [6,7]. Memantine, a non-competitive glutamate NMDA
antagonist, has shown a potential therapeutic effect in treat-
ment of cocaine addiction in animal models [8,9]. Specifically,
memantine pretreatment was found to abolish the establish-
ment of cocaine-induced conditioned place preference (CPP) .
∗Corresponding author at: Behavioral Neuropharmacology Laboratory, Institute
of Behavioral Medicine, National Cheng Kung University College of Medicine,
1 University Rd., Tainan 70101, Taiwan, ROC. Tel.: +886 6 2353535x5114x5106;
fax: +886 6 2095616.
E-mail address: firstname.lastname@example.org (L. Yu).
statement following the extinction of cocaine-induced CPP .
Furthermore, memantine was found to attenuate the reinforcing
effect of cocaine-conditioned stimuli in Rhesus monkeys . These
to decrease the hedonic effect of cocaine.
However, conflicting results obtained from other studies com-
promised clinical use of memantine. For example, memantine
altering the choice to self-administer cocaine in six cocaine addicts
. Although memantine was well tolerated by humans and did
not appear to have abuse potential , memantine occasionally
increased levels of cocaine self-administration in Rhesus monkeys
. Paradoxically, these results suggest that subjects afflicted with
cocaine abuse or dependence may not gain beneficial effect on
memantine treatment. More importantly, memantine doses used
effect on cocaine addiction.
Many studies reported the effects of memantine on cocaine
addiction by assuming the antagonistic effect of memantine on
0166-4328/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
K.-Y. Lin et al. / Behavioural Brain Research 220 (2011) 126–131
glutamate NMDA receptor. However, a recent study indicated that
memantine attenuated lipopolysaccharide-produced central neu-
rotoxicity via modulating brain immune system . In addition
to its actions at NMDA receptors, memantine was known to block
5-HT3receptors, nicotinic receptors, voltage-dependent sodium
channels with lower affinity . In an attempt to avoid the
confound that derived from memantine’s antagonistic effect on
various types of receptor and to investigate the plausible immune-
a series of experiments to assess low doses of memantine treat-
frequently-used cocaine-supported conditioning, in C57BL/6 mice.
2. Materials and methods
imental Animal Center (Taipei, Taiwan, ROC). Mice were housed four per plastic
cage. Mice had free access to food (Purina mouse chow, Richmond, IN, USA) and
tap water in a temperature- (24±1◦C) and humidity (60%)-controlled colony room
maintained on a 12h light-dark cycle (lights on at 07:00) at National Cheng Kung
performed in accordance with the National Institutes of Health Guide for the Care
and Use of Laboratory Animals (NIH Publications No. 80-23) revised in 1996. All
procedures were approved by the local Animal Care Committee at NCKUCM.
Cocaine hydrochloride, memantine hydrochloride, thalidomide were obtained
from Sigma Chem. (St. Louis, MO, USA) and dissolved in saline before use. Recombi-
nant IL-6 from mice was obtained from Santa Cruz Biotechnology (Santa Cruz, CA,
2.3. Pretest, cocaine-induced CPP training, test, and retest
Pretest, cocaine-induced CPP training, test, and retest were conducted in com-
mercial chambers designed for mice (MedAssociates Inc., Georgia, VT, USA). Each
chamber consisted of a center and two side compartments. The side compart-
ments provided three sets of distinctive cues: medium vs. dim light illumination,
black vs. white walls and ceilings, and steel grid bar vs. wire-mesh floors. The cen-
ter was bright-lit with gray walls and ceiling and a gray platform floor. Passages
between the center and the side compartments were controlled by automatic guil-
lotine doors. Mouse location in chambers was monitored by photocell detectors
aligned 1.5cm above the floor across the three compartments, each 3cm apart and
connected via interface cards to IBM-compatible PCs. The time spent in each com-
partment was recorded and analyzed by MED-PC for Windows. Chambers were
deodorized by a thorough cleaning with an isopropyl alcohol (70%)-rinsed paper
towel wiping and followed by a period of drying. One day before the beginning of
cocaine-induced CPP training, mice were placed in the center of any randomly cho-
sen chamber and the time spent in each compartment was measured for 15min
as the pretest for their unconditioned preference. For cocaine-induced CPP train-
ing, mice receiving an intraperitoneal saline injection were immediately confined
in their preferred compartment for 30min in the morning (around 08:00). Approx-
imately 8h later, mice receiving cocaine hydrochloride (20mg/kg) injection were
confined in their non-preferred compartment for 30min. These procedures were
after the last training trial, mice were placed in the center compartment with guil-
lotine doors open for a 15-min cocaine-free, preference test. Three days after the
test, mice were placed in their corresponding chambers for a 15-min retest. Dura-
tions (s) for mice exploring each of the three compartments were recorded. In the
pretest, test and retest, cocaine-induced CPP magnitude was represented by sub-
spent in non-preferred/cocaine conditioning compartment. Reliable acquisition of
magnitude in the test as compared to it in the pretest. Maintenance of cocaine-
induced CPP was determined by comparable cocaine-induced CPP magnitudes in
the test and the retest.
In order to assess whether repeated low-dose memantine injections alter the
maintenance of cocaine-induced CPP, mice received 3 consecutive days of meman-
tine treatment (0.04mg/kg/day for 0.12mg/kg in total) after the acquisition of
cocaine-induced CPP. Approximately 20h (around 12:00) after the last memantine
injection, mice were placed in their respective chambers for a 15-min retest. Mice
receiving saline injections at a similar protocol served as the control group in this
2.4. ELISAs for TNF-˛, IL-1 ˇ and IL-6 assay in brain tissue
Some mice were killed by rapid decapitation approximately 20h after the last
conditioning trial of the 3-day cocaine conditioning with an omission of preference
test. Tissue samples from medial prefrontal cortex (mPFC) and nucleus accumbens
(Acb) were dissected and prepared according to our established procedure . In
first coronal section taken at the center of the olfactory tubercle. The second coronal
section was made at the anterior border of the hypothalamus. Nucleus accumbens
sample was punched out (a circle surrounding the anterior commissure) from the
slice between these two coronal sections with a 1-ml pipetman tip cutting to an
internal diameter of 1.1mm. Tissue sample was homogenized with a glass homog-
enizer in an ice-cold extracting buffer, consisting of 0.1% Igepal CA-630 nonionic
of tissue. After incubation for at least 10min on ice, the mixture was transferred to
a microcentrifuge tube and centrifuged for 10min at 20,000×g at 4◦C. The levels
of IL-1?, IL-6, and TNF-? in supernatants were measured using their correspond-
ing ELISA kits. Measurement of TNF-?, IL-1? and IL-6 were performed based on the
protocol booklet of ELISA kit. Total protein was determined by Bradford’s method.
2.5. Locomotor activity
In an attempt to rule out the possibility that memantine treatment may affect
mouse activity levels or cocaine-increased locomotor activity, four groups of mice
were used for monitoring their locomotor activity approximately 30min after
memantine injection. Locomotor activity (ambulatory activity and vertical rearing
in combination) was monitored in a custom-made transparent Plexiglas chamber
(41cm×41cm×30cm) inside the Optovarimax (Columbus Instrument, Columbus,
OH, USA). Mice were individually placed in the center of the chamber and allowed
free navigation for 30min. The vertical IR beam break number was used as an index
for vertical rearing and the horizontal distance traveled was recorded as an indi-
cator of ambulatory activity. Locomotor activity stood for the sum of rearing and
2.6. Dopamine (DA) and 3,4-dihydroxyphenylacetic acid (DOPAC) contents
the acquisition of cocaine-induced CPP , one group of mice was used to examine
plausible modulating effects of memantine treatment on DA metabolism in these
brain regions. Mice were killed by rapid decapitation and their FCx and Acb tissues
were dissected. Tissue samples were stored in liquid nitrogen until assayed by HPLC
with an LC-4C amperometric detector (BAS, West Lafayette, IN, USA) for measuring
DA and its primary metabolite, DOPAC levels. All tissue samples were homogenized
in 0.4N perchloric acid and centrifuged at 16,000×g for 20min at 4◦C. The super-
natant was filtered and delivered through a high-pressure valve fitted with a 20-?l
loop onto a Phase-II ODS column (3?m, 3.2mm i.d.×10cm), and oxidized with a
+0.72-V potential between the glassy carbon electrode and the Ag/AgCl reference
electrode. The mobile phase consisted of 0.1M sodium phosphate dibasic, 0.1M
citric acid, 5mg EDTA, and 7% methanol delivered at a 0.6-ml/min-flow rate.
2.7. Intra-mPFC infusion of recombinant IL-6 and thalidomide
Although a chemokine and its receptor activation in Acb were involved in mod-
ulating cocaine-stimulated ambulatory activity in rats , role of mPFC immune
that memantine modulated cocaine conditioning-altered cytokine levels in both
mPFC and Acb, we decided to study such a memantine effect in mPFC. In order to
examine the effect of memantine-altered cytokine concentrations on modulating
the acquisition of cocaine-induced CPP, intra-mPFC infusion of mouse recombinant
IL-6 or thalidomide, a TNF-? inhibitor, was done 30min before each condition-
ing trial. Likewise, memantine (0.02mg/kg/injection) was given, i.p., 30min before
each conditioning trial. Stereotaxic surgery and bilateral 26-gauge guide cannula
implantation (Coordinates: anteroposterior, +1.1mm; lateral, ±0.5mm; dorsoven-
before the beginning of the experiment. Bregma and the skull surface served as the
stereotaxic zero point. Clearance through the guide cannula was maintained with
dummy cannulas. The infusion cannula, a 33-gauge dental needle, was inserted
into the guide cannula and was lowered 1.0mm below the guide cannula. A 0.5-
?l volume of Recombinant IL-6 (1ng/?l/side) and thalidomide (10?g/?l/side) was
infused bilaterally with a Hamilton 10-?l microsyringe driven by a microdialysis
the infusion cannulas were left for an additional 2min before withdrawal.
2.8. Statistical analysis
All results were indicated as the mean±standard error of mean. t-Tests were
employed to examine the acquisition and maintenance of cocaine-induced CPP for
all groups of mice. Likewise, t-tests were employed to examine group differences in
DA, DOPAC levels of mFCX and Acb. One-way ANOVAs were used to examine treat-
ment effect on cocaine-induced CPP magnitude in pretest, preference test, retest,
K.-Y. Lin et al. / Behavioural Brain Research 220 (2011) 126–131
ing 20–0mg/kg, 2 injections per day for 3 consecutive days) on cocaine-induced
CPP magnitude in test and retest. Memantine was given before each conditioning
trial in the cocaine-induced CPP training. Cocaine-induced CPP (s) was represented
by subtracting the time spent in preferred/saline conditioning compartment from
the time spent in non-preferred/cocaine conditioning compartment. *Significantly
higher than the other memantine-pretreated (20–0.02mg/kg/injection) groups in
the test and retest.
and locomotor activity followed by Bonferroni tests if appropriate. The levels of
statistical significance were set at p<0.05.
3.1. Memantine pretreatment abolishes cocaine-induced CPP
One-way ANOVA revealed that six groups of mice exhibited
comparable magnitude of unconditioned preference in the pretest
nitude were different in the test [F(5,61)=8.557, p<0.0001] and
retest [F(5,61)=10.07, p<0.0001] (Fig. 1). Saline- and meman-
tine (0.002mg/kg each for 6 injections in total)-treated groups
displayed reliable cocaine-induced CPP in the test (t26=5.526,
p<0.0001; t16=4.965, p=0.0001). Their cocaine-induced CPP mag-
nitudes remained to be unaltered in the retest [t26=0.1153,
p=0.9091 for the saline-treated group; t16=0.1678, p=0.8689
for the memantine (0.002mg/kg)-treated group]. In contrast,
memantine (20, 2, 0.2, 0.02mg/kg each for 6 injections in total)-
treated groups did not display obvious cocaine-induced CPP in
the test (t16=1.168, p=0.2598; t18=1.589, p=0.1294; t22=2.041,
p=0.0535; t24=2.051, p=0.0514) or in the retest (t16=1.652,
3.2. Low-dose (0.02mg/kg/injection×6) memantine
pretreatment reverses cocaine conditioning-altered interleukin-6
(IL-6) and tumor necrosis factor-alpha (TNF-˛) level in mPFC and
the IL-1 levels in the Acb [F(3,21)=1.48, p=0.2488] or mPFC
[F(3,21)=0.6102, p=0.6158] (Fig. 2A and B). Memantine treatment
ing significantly increased IL-6 levels in the mPFC [F(3,21)=8.073,
p=0.0009] (Fig. 2C). It was of interest to note that meman-
tine pretreatment abolished cocaine conditioning-enhanced IL-6
levels in the mPFC (Fig. 2C). Memantine treatment alone did
not affect IL-6 levels in the Acb, while cocaine conditioning
significantly increased IL-6 levels in this region [F(3,21)=8.0,
p=0.001] (Fig. 2D). Memantine pretreatment abolished cocaine
conditioning-enhanced IL-6 levels in the Acb (Fig. 2D). Meman-
tine treatment did not affect TNF-? levels in the mPFC or Acb,
while cocaine conditioning significantly decreased TNF-? levels in
these regions [F(3,21)=15.77, p<0.0001 for mPFC; F(3,21)=18.2,
p<0.0001 for Acb] (Fig. 2E and F). Memantine treatment reversed
cocaine conditioning-decreased TNF-? levels in both mPFC and
3.3. A low-dose memantine injection does not affect mouse
locomotor activity or cocaine-increased locomotor activity
In an attempt to rule out the possible confound that meman-
tine pretreatment may abolish cocaine-induced CPP by decreasing
mouse locomotor activity or cocaine-increased locomotor activity,
pretreatment with memantine or saline was done 30min before
their locomotor activity immediately following cocaine or saline
tion did not alter mouse locomotor activity or cocaine-increased
locomotor activity [F(3,44)=10.83, p<0.0001] (Fig. 3).
3.4. A low-dose memantine injection do not affect DA or DOPAC
levels in the mPFC and Acb
Twenty-two mice were used to examine plausible modulating
effects of memantine treatment (0.04mg/kg/day for 3 consecu-
tive days) on DA metabolism in the mPFC and Acb. An unpaired
t-test revealed that memantine treatment (0.12mg/kg in total) did
not alter DA or DOPAC levels in the mPFC (t20=0.8373, p=0.4123;
ment (0.12mg/kg in total) did not alter DA or DOPAC levels in the
Acb (t20=0.753, p=0.4602; t20=0.6262, p=0.5382) (Fig. 4B).
3.5. Low-dose memantine treatment does not affect the
maintenance of cocaine-induced CPP
Both groups of mice reliably acquired cocaine-induced CPP in
the test (t28=6.195, p<0.0001 for the saline-treated group and
t28=9.364, p<0.0001 for the memantine-treated group) (Fig. 5).
It was of importance to notice that 3 consecutive days of meman-
tine treatment (0.04mg/kg/day) did not affect the maintenance of
cocaine-induced CPP in retest (t28=0.6497, p=0.5212 for the dif-
ference between test and retest results in the saline-treated group;
t28=0.4018, p=0.6909 for the difference between test and retest
results in the memantine-treated group).
3.6. Intra-mPFC infusion of recombinant IL-6 reversed
memantine-decreased cocaine-induced CPP
Mice were divided into two groups. One group received i.p.
saline injection 30min before each conditioning trial, while the
other group received i.p. memantine injection before each trial.
Each group was subdivided into three groups (n=9 for each group).
nant IL-6, or thalidomide infusion 30min before each CPP training
trial. One-way ANOVA results revealed that six groups of mice
exhibited comparable magnitude of unconditioned preference
in the pretest [F(5,48)=0.013, p=0.9999], while their cocaine-
induced CPP magnitude were different in the test [F(5,48)=6.907,
p<0.0001] and retest [F(5,48)=4.728, p=0.0014] (Fig. 6). Saline-
treated groups all displayed reliable cocaine-induced CPP in the
test (t16=5.162, p<0.0001; t16=4.725, p=0.0002; t16=4.965,
p=0.0001). However, mice receiving memantine injection and
intra-mPFC PBS or thalidomide infusion did not display reli-
able cocaine-induced CPP in the test (t16=1.001, p=0.3317;
t16=0.8149, p=0.4271). Mice receiving systemic memantine injec-
tion and intra-mPFC recombinant IL-6 infusion exhibited reliable
cocaine-induced CPP in the test (t16=3.967, p=0.0011). Their
cocaine-induced CPP magnitude in the test remained to be unal-
tered in the retest for all six groups (t16=1.093, p=0.2906;
K.-Y. Lin et al. / Behavioural Brain Research 220 (2011) 126–131
Fig. 2. Brain cytokine levels following cocaine conditioning (20mg/kg/trial for 3 trials in total) and/or low-dose memantine dosing regimen (0.02mg/kg/injection for 6
injections in total in 3 days). Approximately 20h following the conclusion of cocaine conditioning, IL-1 levels were measured in mouse (A) medial prefrontal cortex (mPFC)
in mouse (E) mPFC and (F) Acb samples. *Significantly lower than the other groups. Mem and Sal are short form for memantine and saline treatment. N=8 for each group.
p=0.7053; t16=0.1268, p=0.9007; t16=0.1391, p=0.8911) (Fig. 6).
cocaine-induced CPP magnitude in both test and retest. Such
a memantine-decreased CPP performance could be due to
memantine-exerted inhibitory effect against the hedonic value
and/or reinforcing effect of cocaine. Another possibility is that
memantine pretreatment may abolish the formation of cocaine
conditioning. The latter possibility implies that memantine pre-
treatment could impair mouse sensory processing (environmental
cue encoding) . In fact, few reports addressed that memantine
caused severe fluctuations in awareness and delirium in patients
[18,19]. It was of interest to note that acute memantine treatment
did not affect cocaine-increased locomotor activity. This finding
favors the hypothesis that a low-dose memantine pretreatment
may diminish the acquisition of cocaine-induced CPP by impairing
environmental cue processing.
Although a low-dose memantine treatment regimen abolished
the acquisition of cocaine-induced CPP, a similar treatment reg-
imen did not affect the maintenance of cocaine-induced CPP.
ity and cocaine-increased locomotor activity. Cocaine (20mg/kg) or an equivalent
volume of saline was given 30min after memantine or saline pretreatment. Imme-
diately after cocaine or saline injection, mouse locomotor activity was measured for
30min. *Significantly higher then saline-treated group. N=12 for each group.
ment may be beneficial in curbing initial attempt of cocaine taking,
while such a memantine treatment could be ineffective in altering
treatment could be ineffective in facilitating a dissociation learning
K.-Y. Lin et al. / Behavioural Brain Research 220 (2011) 126–131
Fig. 4. Effect of a low-dose memantine dosing regimen (0.02mg/kg/injection for 6 injections in 3 days) on dopamine (DA) metabolism. Approximately 20h after the last
memantine injection, mouse DA and DOPAC (one primary metabolite of DA) contents in mPFC and Acb were measured by HPLC. Mem and Sal stand for memantine and saline
treatment, respectively. N=11 for each group.
of conditioned cues and omission of cocaine. Nevertheless, effec-
to be employed in experimental design before ineffectiveness of
memantine on facilitating the extinction of cocaine-taking habit in
cocaine addicts can be concluded .
Systemic memantine treatment at an extremely low dose
diminished cocaine-induced CPP, suggesting that this low dose
of memantine may decrease cocaine-induced CPP via a mecha-
nism that is less related to the known pharmacological property of
Fig. 5. Effect of a low-dose memantine dosing regimen (0.02mg/kg/injection for 6
test, mice were introduced to their respective chambers for measuring the cocaine-
induced CPP magnitude.
Fig. 6. Effect of intra-mPFC recombinant IL-6 and thalidomide infusion on
memantine-decreased cocaine-induced CPP. Systemic saline-treated groups dis-
played cocaine-induced CPP. However, mice receiving memantine injection and
intra-mPFC PBS or thalidomide infusion did not display reliable cocaine-induced
CPP. Mice receiving memantine injection and intra-mPFC recombinant IL-6 infu-
sion exhibited cocaine-induced CPP in the test. *Significantly higher than respective
CPP magnitude in pretest.
tioning significantly increased interleukin-6 but decreased TNF-?
levels in mouse mPFC and Acb. More interestingly, pretreatment
with memantine at this dose (0.02mg/kg/injection for 6 injections
in total) was found to effectively reverse cocaine conditioning-
enhanced interleukin-6 and -decreased TNF-? levels in these brain
regions. It was of importance to emphasize that single meman-
tine injection (0.02mg/kg) did not acutely affect mouse locomotor
activity or cocaine-increased locomotor activity. Therefore, it was
reasonable to suspect that this low dose of memantine treatment
regimen may disrupt the acquisition of cocaine-induced CPP by
reversing cocaine conditioning-altered cytokines in these brain
involved in repeated psychostimulant-produced dendritic mor-
and hippocampus [5,21]. For example, TNF-?, a pro-imflammatory
cytokine, was reported to attenuate rewarding effects and loco-
motor sensitization induced by methamphetamine and morphine
[22–24]. Intra-mPFC infusion with mouse recombinant IL-6 or
thalidomide alone did not affect the acquisition of cocaine-induced
CPP, while intra-mPFC recombinant IL-6, not thalidomide, infu-
sion effectively reversed memantine-decreased cocaine-induced
CPP. These findings, taken together, support a notion that a low-
dose memantine may diminish cocaine-induced CPP by reversing
cocaine conditioning-increased IL-6 levels in mPFC. However,
memantine-decreased cocaine-induced CPP may not be attributed
to its effect on modulating TNF-? levels in mPFC. A previous study
revealed that IL-6 receptors were abundant in C57BL/6J mouse cor-
tex . Interestingly, activation of IL-6 receptor signaling was
found to effectively reduce glutamate-mediated excitotoxic effects
through the enhancement of endogenous adenosinergic signal-
ing . In this study, an extremely low dose of memantine
could abolish cocaine-induced CPP, in part, by reversing cocaine
conditioning-provoked IL-6 receptor signaling cascades in mPFC.
Studies should be extended to verify the roles of mPFC IL-6 in
mediating the acquisition of cocaine-induced CPP and memantine-
decreased CPP in IL-6 knockout mice.
Medial prefrontal cortex and Acb are two primary targets
of mesolimbic dopamine projections. Synaptic dopamine levels
have been known to play a critical role in the hedonic value
and/or reinforcing efficacy of varieties of illicit drugs . There-
fore, it was reasonable to suspect that low-dose memantine
treatment abolished the formation of cocaine-induced CPP by
attenuating cocaine-enhanced dopamine release in these brain
regions. Nevertheless, we found that low-dose memantine treat-
ment (0.02mg/kg/injection for 6 injections in total in 3 days) did
not affect DA and its primary metabolite, DOPAC, contents in mPFC
K.-Y. Lin et al. / Behavioural Brain Research 220 (2011) 126–131
and Acb. Since tissue DA and DOPAC contents both are indirect
indices of synaptic DA level, whether low-dose memantine can
curb cocaine-stimulated DA release in these regions remains to be
Memantine has been used for treating severely demented
[28,29]. In addition to its effect on decreasing care dependence,
memantine has been claimed to diminish emotional distress in
these patients. In this study, we found that an extremely low dose
of memantine affected cocaine-provoked cytokine concentrations
in central nervous system. These findings prompted us to hypoth-
esize that memantine treatment-associated brain immunological
responses could be involved in its treatment effects in the above-
mentioned clinical observations.
This research is supported by ROC National Science Council
grants 97-2321-B-006-010 and 97-2410-H-006-074-MY3 to L.Y.
 Boudreau AC, Wolf ME. Behavioral sensitization to cocaine is associated with
increased AMPA receptor surface expression in the nucleus accumbens. J Neu-
 Cornish JL, Kalivas PW. Glutamate transmission in the nucleus accumbens
mediates relapse in cocaine addiction. J Neurosci 2000;20:RC89.
 Gass JT, Olive MF. Glutamatergic substrates of drug addiction and alcoholism.
Biochem Pharmacol 2008;75:218–65.
 Uys JD, LaLumiere RT. Glutamate: the new frontier in pharmacother-
apy for cocaine addiction. CNS Neurol Disord Drug Targets 2008;7:
 William JM, Steketee JD. Cocaine increases medial prefrontal cortical gluta-
mate overflow in cocaine-sensitized rats: a time course study. Eur J Neurosci
 Bespalov AY, Dravolina OA, Zvartau EE, Beardsley PM, Balster RL. Effects of
NMDA receptor antagonists on cocaine-conditioned motor activity in rats. Eur
J Pharmacol 2000;390:303–11.
Dizocilpine infusion has a different effect in the development of morphine and
cocaine sensitization: behavioral and neurochemical aspects. Neuroscience
 Maldonado C, Rodríguez-Arias M, Castillo A, Aguilar MA, Mi˜ narro J. Effect of
memantine and CNQX in the acquisition, expression and reinstatement of
cocaine-induced conditioned place preference. Prog Neuropsychopharm Biol
 Newman JL, Beardsley PM. Effects of memantine, haloperidol, and cocaine
on primary and conditioned reinforcement associated with cocaine in rhesus
monkeys. Psychopharmacology (Berl) 2006;185:142–9.
 Collins ED, Vosberg SK, Ward AS, Haney M, Foltin RW. The effects of acute
effects of cocaine in humans. Exp Clin Psychopharmacol 2007;15:228–37.
 Vosburg SK, Hart CL, Haney M, Foltin RW. An evaluation of the reinforcing
effects of memantine in cocaine-dependent humans. Drug Alcohol Depend
 Wu HM, Tzeng NS, Qian L, Wei SJ, Hu X, Chen SH. Novel neuroprotective mech-
anisms of memantine: increase in neurotrophic factor release from astroglia
and anti-inflammation by preventing microglial activation. Neuropsychophar-
 Blanchard AP, Guillemette G, Boulay G. Memantine potentiates agonist-
of central neurotoxicity induced by ketamine followed by methamphetamine.
Toxicol Appl Pharmacol 2008;227:239–47.
 Kuhar MJ, Ritz MC, Boja JW. The dopamine hypothesis of the reinforcing prop-
erties of cocaine. Trends Neurosci 1991;14:299–302.
 Trecki J, Unterwald EM. Modulation of cocaine0induced activity by intracere-
bral administration of CXCL12. Neuroscience 2009;161:13–22.
 Cherng CG, Tsai C-W, Tsai Y-P, Ho M-C, Kao S-F, Yu L. Methamphetamine-
disrupted sensory processing mediates conditioned place preference perfor-
mance. Behav Brain Res 2007;182:103–8.
 Menendez-Gonzalez M, Calatayud MT, Blazquez-Menes B. Exacerbation of
Lewy bodies dementia due to memantine. J Alzheimers Dis 2005;8:289–91.
 Moellentin D, Picone C, Leadbetter E. Memantine-induced myclonus and delir-
ium exacerbated by trimethoprim. Ann Pharmacother 2008;42:443–7.
 Fan H-Y, Cherng CG, Yang F-Y, Cheng L-Y, Tsai C-J, Lin L-C, et al. Systemic treat-
ment with protein synthesis inhibitors attenuates the expression of cocaine
memory. Behav Brain Res 2010;208:522–7.
 Robinson TE, Kolb B. Alterations in the morphology of dentrites and dendritic
spines in the nucleus accumbens and prefrontal cortex following repeated
treatment amphetamine and cocaine. Eur J Neurosci 1999;11:1598–604.
 Nakajima A, Yamada K, Nagai T, Uchiyama T, Miyamoto Y, Mamiya T, et al.
dence and neurotoxicity. J Neurosci 2004;24:2212–25.
rotrophic factor, tumor necrosis factor-alpha, and an inducer of these factors
in drug dependence. J Pharmacol Sci 2007;104:116–21.
sis factor-alpha and its inducer inhibit morphine-induced rewarding effects
and sensitization. Biol Psychiatry 2007;62:658–68.
in the brain after stimulation with LPS in young and old mice. Exp Gerontol
 Biber K, Pinto-Duarte A, Wittendorp MC, Dolga AM, Fernandes CC, von Fri-
jtaq Drabbe Kunzel, et al. Interleukin-6 upregulates neuronal adenosine A1
receptors: implications for neuromodulation and neuroprotection. Neuropsy-
 Stahl SM. Stahl’s essential psychopharmacology. 3rd ed. New York, USA: Cam-
bridge University Press; 2008.
 Winblad B, Poritis N. Memantine in severe dementia: results of the 9M-best
study (benefit and efficacy in severely demented patients during treatment
with memantine. Int J Geriatr Psychiatry 1999;14:135–46.
 Reisberg B, Doody R, Stoffler A, Schmitt F, Ferris S, Mobius HJ. Memantine in
moderate-to-severe Alzheimer’s disease. New Eng J Med 2003;348:1333–41.
293 cells. Cell Physiol Biochem