Reversing cocaine-induced synaptic potentiation
provides enduring protection from relapse
Khaled Moussawia, Wenhua Zhoua,b, Haowei Shena, Carmela M. Reichela, Ronald E. Seea, David B. Carra,
and Peter W. Kalivasa,c,1
Departments ofaNeurosciences andcPsychiatry and Behavioral Sciences, Medical University of South Carolina, Charleston, SC 29425; andbLaboratory of
Behavioral Neuroscience, Ningbo Addiction Research and Treatment Center, Ningbo University, Ningbo 315000, People’s Republic of China
Edited by Leslie Lars Iversen, University of Oxford, Oxford, United Kingdom, and approved November 30, 2010 (received for review August 10, 2010)
Cocaine addiction remains without an effective pharmacotherapy
and is characterized by an inability of addicts to inhibit relapse to
drug use. Vulnerability to relapse arises from an enduring impair-
by dysregulated synaptic potentiation and extracellular glutamate
homeostasis in the projection from the prefrontal cortex to the
nucleus accumbens. Here weshow in rats trainedto self-administer
cocaine that the enduring cocaine-induced changes in synaptic
potentiation and glutamate homeostasis are mechanistically linked
through group II metabotropic glutamate receptor signaling. The
enduring cocaine-induced changes in measures of cortico-accum-
bens synaptic and glial transmission were restored to predrug
parameters for at least 2 wk after discontinuing chronic treatment
with the cystine prodrug, N-acetylcysteine. N-acetylcysteine pro-
duced these changes by inducing an enduring restoration of nonsy-
naptic glutamatergic tone onto metabotropic glutamate receptors.
The long-lasting pharmacological restoration of cocaine-induced
glutamatergic adaptations by chronic N-acetylcysteine also caused
These data mechanistically link nonsynaptic glutamate to cocaine-
induced adaptations in excitatory transmission and demonstrate a
mechanism to chronically restore prefrontal to accumbens transmis-
sion and thereby inhibit relapse in an animal model.
inhibit relapse to drug use. An enduring cocaine-induced im-
pairment in cognitive control of motivated behavior contributes
to the vulnerability to relapse (1, 2). Projections from the frontal
cortex to the basal ganglia constitute a primary brain substrate
for regulating motivated behavior (3). Cocaine-induced neuro-
pathologies in the projection from the prefrontal cortex to the
nucleus accumbens are implicated in cocaine addiction (4), in-
cluding impaired neuroplasticity and synaptic communication
(4, 5). For example, prefrontal synapses in the accumbens un-
dergo enduring potentiation (6–9), and the ability of these
synapses to increase or decrease synaptic strength is impaired
(7, 10). In addition, withdrawal from chronic cocaine use
increases both presynaptic release estimated by elevated fre-
quency of miniature excitatory postsynaptic currents (mEPSC)
and by postsynaptic strength measured as increases in the surface
expression of AMPA glutamate receptors and the ratio of
AMPA/NMDA currents at glutamatergic synapses (6, 8, 11).
Neuroimaging in cocaine addicts reveals reduced activity in
prefrontal cortex under baseline conditions, but marked hyper-
responsiveness in the prefrontal cortex and accumbens that is
correlated with a desire for drug upon exposure to drug-associ-
ated stimuli (12). Similarly, animal models of relapse show that
activation of this pathway is necessary and sufficient to reinstate
cocaine-seeking behavior (13) and that neuronal activity in py-
ramidal neurons in prefrontal cortex projecting to medium spiny
neurons (MSNs) in the nucleus accumbens core (NAcore) region
is correlated with cocaine-seeking (14, 15). Accordingly, it is
thought that prefrontal-to-accumbens synapses undergo endur-
ing potentiation after chronic cocaine use that contributes to
relapse vulnerability (4, 6, 16).
ocaine addiction remains without an effective pharmaco-
therapy and is characterized by an inability of addicts to
Possibly linked to a cocaine synaptic pathology is a long-lasting
impairment in nonsynaptic glutamate release and transport,
a process termed ‘glutamate homeostasis’ (4). Cystine–glutamate
exchange regulates extrasynaptic glutamate levels by providing
a 1:1 stoichiometric exchange of extracellular cystine for in-
it is down-regulated following repeated cocaine administration
(20, 21). In this way, chronic cocaine use reduces basal levels of
extracellular glutamate, thereby decreasing tone upon peri-
synaptic metabotropic glutamate receptors (mGluR) known to
regulate glutamatergic synaptic transmission (22–24). Supporting
involvement of this adaptation in relapse, administering cystine
prodrugs to acutely activate the exchanger inhibits reinstated co-
caine-seeking in animal models (20, 25, 26).
Despite the value of acutely administered pharmacotherapies,
neuropathologies and enduring protection from relapse would
have greater therapeutic utility. Thus, we sought to determine (i)
repair of cocaine-induced neuropathologies in prefrontal-to-
accumbens glutamatergic synapses, (ii) how the repair of the
extrasynaptic and synaptic neuropathologies are mechanistically
linked, and (iii) whether synaptic restoration translates into re-
duced cocaine-seeking in an animal model of relapse.
Enduring Relapse Protection by Chronic N-Acetylcysteine. We adap-
ted the extinction–reinstatement model of relapse to cocaine-
seeking (27). Rats were trained to lever press for i.v. cocaine for
12 d, followed by 12 d of extinction training, during which lever
pressing no longer delivered cocaine (Fig. 1A). Two hours before
each extinction session, animals received the cystine prodrug N-
2–3 additional weeks of cocaine withdrawal in the absence of ex-
tinction training or N-acetylcysteine treatments to evaluate long-
lasting effects of N-acetylcysteine on glutamate homeostasis,
synaptic physiology, and conditioned cue- or cocaine-induced re-
instatement of lever pressing (cocaine-seeking behavior). Control
animals received an i.v. saline infusion that was yoked to cocaine
infusions in rats self-administering cocaine. This protocol of co-
caine self-administration, N-acetylcysteine treatment, and with-
drawal was used throughout the study.
In the first experiment, reinstatement trials were conducted to
estimate the effect of N-acetylcysteine on reinstated cocaine-
seeking 2 h after N-acetylcysteine or vehicle on days 25 and 29
and again on days 43 and 47 after leaving the rats in their home
Author contributions: K.M., R.E.S., D.B.C., and P.W.K. designed research; K.M., W.Z., H.S.,
and C.M.R. performed research; K.M., W.Z., H.S., R.E.S., and P.W.K. analyzed data; and
K.M. and P.W.K. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
1To whom correspondence should be addressed. E-mail: firstname.lastname@example.org.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
| January 4, 2011
| vol. 108
| no. 1
cages for 2–3 wk without drug treatments (Fig. 1A). Animals
trained to self-administer cocaine were divided into two groups
such that levels of active lever pressing across the last 3 d of
cocaine self-administration were equivalent in each group (Fig.
1b). The N-acetylcysteine-treated animals showed an initial re-
duction in lever pressing during extinction training, although by 2
wk the levels of extinction pressing were equivalent between the
vehicle and N-acetylcysteine groups (Fig. 1B). Reduced extinc-
tion lever pressing unlikely resulted from nonspecific motor in-
hibition since acute N-acetylcysteine does not inhibit locomotor
activity or operant responding (25).
Cocaine-seeking was elicited by presenting cues (tone and
light) previously paired with cocaine delivery during self-admin-
istration or by injecting cocaine (10 mg/kg, i.p) just before the
cue-reinstatement session. Reinstated lever pressing induced by
either modality was inhibited in the N-acetylcysteine pretreated
group (days 25 and 29; Fig. 1C) compared with the vehicle-
treated group. More importantly, 2–3 wk after discontinuing
N-acetylcysteine or vehicle treatments, both cue-induced and
cocaine- and cue-induced reinstatement remained inhibited in
the N-acetylcysteine group (Fig. 1D).
Enduring Restoration of Glutamate Homeostasis in the NAcore. The
reduction in basal extracellular concentrations of glutamate in
the NAcore by chronic cocaine use is a marker of disrupted
glutamate homeostasis, arising from an enduring reduction in
cystine–glutamate exchange (20, 21, 25). Basal extracellular
glutamate concentrations assayed in the NAcore by in vivo
microdialysis arise largely from cystine–glutamate exchange
rather than from synaptic activity (19). In the next experiment,
we measured basal glutamate levels after the daily treatment
regimen of N-acetylcysteine to determine whether the reduction
in cocaine-seeking was associated with restoration of nonsynaptic
glutamate in the extracellular space (Fig. 2A). Animals trained to
self-administer cocaine and treated with saline vehicle during
extinction training showed reduced basal levels of glutamate
compared with yoked-saline controls using no-net-flux micro-
dialysis (Fig. 2B). However, basal extracellular glutamate levels
in cocaine-trained animals were restored to control levels when
measured 2–3 wk after the last daily injection of N-acetylcysteine.
In contrast, the N-acetylcysteine regimen did not alter glutamate
in yoked-saline rats. Fig. 2C shows the location of the active zone
of the microdialysis probes in the NAcore.
Enduring Restoration of Synaptic Strength and Role of mGluR2/3. The
long-lasting normalization of extracellular glutamate by N-
acetylcysteine (Fig. 2B) may increase glutamatergic tone onto
perisynaptic mGluR2/3 (28) that negatively regulate synaptic
glutamate release (23, 29–31). To examine this possibility,
stimulating electrodes were placed into the prefrontal cortex and
duced long-lasting protection from cue-induced and cocaine- and cue-
induced reinstatement of lever pressing. (A) Treatment protocol for cocaine
and NAC. (B) Active lever pressing during self-administration and extinction
sessions. A two-way ANOVA over extinction (days 13–24) revealed a signifi-
cant interaction between treatment group and time [F(11, 407) = 3.08, P <
0.001]. *P < 0.05, comparing vehicle to NAC using a Bonferroni multiple
comparisons test. (C and D) Reinstatement of active lever pressing by cues
(tone and light) or cocaine and cues (10 mg/kg, i.p.) was significantly
inhibited both in the presence of NAC (days 25 and 29) (C) and 2–3 wk after
discontinuing NAC treatment (days 43 and 47) (D). Student’s t tests revealed
Cue25: t(37) = 2.67, P = 0.011; Cocaine29: t(37) = 4.12, P < 0.001; Cue43:
t(37) = 3.13, P = 0.003; Cocaine47: t(37) = 3.14, P = 0.003. *P < 0.05 comparing
NAC to saline-treated groups.
N-acetylcysteine (NAC) treatment during extinction training pro-
tylcysteine (NAC) restores the reduced levels of extracellular glutamate eli-
cited by cocaine self-administration. (A) Treatment protocol for cocaine and
NAC used for data shown in Figs. 2, 3 and 5. (B) Plots from the no-net-flux in
vivo microdialysis experiment. Net flux (y axis) is the difference between the
glutamate concentration dialyzed into and recovered from the brain. The x-
intercept corresponding to the extracellular glutamate concentration was
extrapolated from fitted regression lines and differed between treatment
groups: cocaine, saline 1.16 ± 0.17 μM; cocaine, NAC 1.98 ± 0.22 μM; saline,
saline 2.19 ± 0.63 μM; saline, NAC 2.09 ± 0.61 μM [F(3, 115) = 3.90; P = 0.011].
There was no difference in the slope of the regression line between the four
groups. *P < 0.05 for (cocaine, saline) compared with the other three groups
using a Bonferroni test for multiple comparisons. (C) Summary and micro-
graph example of the location of the active membrane of the dialysis probe
in the NAcore according to Paxinos and Watson (47). ac, anterior commis-
sure; arrows, dorsal and ventral extent of the active dialysis membrane.
No-net-flux microdialysis in the NAcore reveals that chronic N-ace-
| www.pnas.org/cgi/doi/10.1073/pnas.1011265108Moussawi et al.
recording electrodes into the NAcore to elicit field potentials as
an estimate of the overall strength of prefrontal to NAcore
synapses (Fig. 3B; ref. 7). This protocol permits selective acti-
vation of prefrontal afferents to the NAcore in an in vivo envi-
ronment where glutamate homeostasis is not disrupted (32). We
previously found a long-term potentiation (LTP)-like shift to
the left of the input–output curve 3 wk after the last cocaine
self-administration session (7). Similarly, 4–5 wk after the last
cocaine exposure, the input–output curve for rats trained to self-
administer cocaine was potentiated compared with yoked-saline
controls (Fig. 3A). This LTP-like shift was abolished 2–3 wk after
completing daily N-acetylcysteine. In contrast, N-acetylcysteine
did not alter the input–output curve in yoked-saline animals.
The importance of mGluR2/3 in N-acetylcysteine-induced
restoration of synaptic strength was examined by infusing the
mGluR2/3 antagonist LY341495 through a microdialysis probe
attached to the recording electrode (Fig. 3B). To validate this
methodology, reverse-dialysis of the AMPA receptor antagonist,
CNQX, or the mGluR2/3 agonist, LY379268, to drug-naive
animals reduced field potential amplitude (Fig. S1). Infusing
receptor-selective concentrations (33) of LY341495 into the
NAcore potentiated evoked field potentials in yoked-saline rats
(Fig. 3C). The increase in field amplitude was absent in cocaine-
trained rats treated with vehicle during extinction training,
indicating decreased tone on mGluR2/3 commensurate with
reduced basal extracellular concentrations of glutamate (Fig. 1B).
However, similar to controls, LY341495 potentiated prefrontal-
to-NAcore synaptic transmission in rats receiving N-acetylcysteine
during extinction training. Therefore, the capacity of LY341495
to block inhibitory mGluR2/3 tone and potentiate field poten-
tials was restored by N-acetylcysteine administration during ex-
tinction training, corroborating the enduring restoration of
extracellular glutamate levels (Fig. 2B).
Enduring Relapse Protection by Restoring Glutamatergic Tone to
mGluR2/3. Up to this point the data show an enduring normali-
zation of glutamate homeostasis by N-acetylcysteine, and thereby
a mGluR2/3-dependent depotentiation of prefrontal-to-NAcore
excitatory synapses. Given the postulated role by heightened
accumbens glutamatergic transmission in cocaine relapse (4), we
asked whether the long-lasting protection from reinstated co-
caine-seeking provided by the chronic N-acetylcysteine treatment
regimen was also mGluR2/3-dependent. In this experiment, ani-
mals treated with N-acetylcysteine during extinction were micro-
injected with the mGluR2/3 antagonist LY341495 into the
NAcore (Fig. 4A). Blockade of mGluR2/3 prevented N-ace-
tylcysteine-induced inhibition of reinstated cocaine-seeking in
the presence of N-acetylcysteine (days 25 and 29) and 2–3 wk
after the last N-acetylcysteine injection (days 43 and 47) (Fig. 4B).
In contrast, LY341495 did not alter reinstated responding in
control animals (Fig. 4C). These data demonstrate that N-ace-
tylcysteine induced a long-lasting restoration of glutamatergic
tone onto mGluR2/3, thereby normalizing synaptic strength and
causing an enduring reduction in reinstated cocaine-seeking.
Enduring Restoration of Both Presynaptic and Postsynaptic Gluta-
mate Transmission. The experiments so far demonstrate that N-
acetylcysteine produced an enduring increase in nonsynaptic
glutamate (Fig. 2B) that impinges on mGluR2/3 (Fig. 3D) to
normalize synaptic strength (Fig. 3A). Since the measure of
output curve measured in the NAcore following in vivo stimulation of the prefrontal cortex. The increase in field amplitude in rats with a history of cocaine
self-administration was reduced by NAC treatment during extinction training; two-way ANOVA [treatment group F(3, 150) = 3.83; P = 0.022; stimulation
intensity F(6, 18) = 409.3; P < 0.001; interaction F(18, 150) = 1.77; P = 0.034]. *P < 0.05, for comparing cocaine, saline to other groups using a Bonferroni post
hoc test. (B) Illustration of the experimental preparation where a dialysis probe and adjacent recording electrode were implanted into the NAcore. (C) Field
potentials (normalized to levels following aCSF) were increased in yoked-saline control and NAC, but not vehicle-treated animals trained to self-administer
cocaine after two concentrations of the mGluR2/3 antagonist LY341495 were delivered through a dialysis probe. Two-way repeated-measures ANOVA
revealed significant group and time effects of LY341495 [group: F (2, 493) = 5.07, P = 0.019; time: F(29, 493) = 1.59, P = 0.027]. (Inset) Mean ± SEM normalized
field amplitude averaged over the last 10 min (last 5 data samples were averaged) after 20 μm LY341495 [one-way ANOVA: F(2, 19) = 7.04, P = 0.006]. *P <
0.05 for (cocaine, saline) compared with other groups using a Bonferroni multiple comparisons test.
N-acetylcysteine (NAC) restores synaptic strength in prefrontal cortex projections to the NAcore through modulating mGluR2/3 receptors. (A) Input–
Moussawi et al. PNAS
| January 4, 2011
| vol. 108
| no. 1
synaptic strength was via field potentials in vivo, it is unclear if N-
acetylcysteine’s effects were primarily presynaptic, postsynaptic,
or both. To further examine the role of glutamate homeostasis in
presynaptic regulation of excitatory transmission in the NAcore,
we measured mEPSC in MSNs using whole-cell patch recordings
in tissue slices obtained from rats treated as in Fig. 2A. Pre-
synaptic release probability was estimated by measuring the
frequency of mEPSC after eliminating action potential-evoked
glutamate release with tetrodotoxin (TTX) (29, 34, 35). The
frequency of mEPSC was elevated in NAcore MSNs recorded in
cocaine-trained subjects compared with yoked-saline controls,
and this increase was abolished when measured 2–3 wk after the
last N-acetylcysteine injection (Fig. 5 A and B). In contrast to
mEPSC frequency, there was no effect by either cocaine or N-
acetylcysteine on mEPSC amplitude (Fig. 5C). Also, spontane-
ous EPSC (sEPSC) recorded in the absence of TTX showed an
increased frequency in cocaine rats treated with vehicle, but not
in cocaine rats that received N-acetylcysteine treatment, and no
treatment group differed in sEPSC amplitude (Fig. S2). These
results indicate that by restoring extracellular glutamate (Fig.
2B) and glutamatergic tone on presynaptic release-regulating
mGluR2/3s (Fig. 3 C and D), chronic N-acetylcysteine normal-
izes presynaptic glutamate release probability.
To estimate the postsynaptic effects of cocaine and N-
the ratio of peak AMPA to NMDA receptor-mediated currents.
Recordings were restricted to the dorsomedial NAcore, and
stimulation electrodes were placed dorsomedial to the recording
electrode to preferentially stimulate prefrontal afferents (36).
Akin to a previous report with chronic cocaine followed by with-
drawal (8), cocaine-trained rats showed an increase in AMPA:
NMDA relative to yoked-saline controls (Fig. 5D). The AMPA:
NMDA ratio elevation was reversed in animals treated with N-
acetylcysteine during extinction training, supporting an enduring
normalization of postsynaptic strength by N-acetylcysteine. Con-
sidering the observed changes in AMPA:NMDA ratio, the lack of
effect on mEPSC or sEPSC amplitude was surprising. This dis-
crepancy between mEPSC amplitude and AMPA:NMDA ratio is
diverse excitatory synapses onto a neuron, whereas the evoked
that inhibits cocaine-seeking initiated by either cocaine-condi-
tioned cues or a noncontingent injection of cocaine in animals
withdrawn from self-administered cocaine. Importantly, N-
acetylcysteine inhibition of cocaine-seeking endured for at least 2
wk after the last daily N-acetylcysteine injection (Fig. 1). The in-
hibition of cocaine-seeking by N-acetylcysteine was then linked to
its ability to restore neuroadaptations produced by cocaine self-
administration on glutamate homeostasis and synaptic glutamate
transmission (see Fig. S3 for qualitative model). Chronic N-
acetylcysteine increases cystine–glutamate exchange and the ex-
result is an increase in extracellular, nonsynaptic glutamate
(Fig. 2B) that restored tone onto presynaptic release-regulating
mGluR2/3 (Fig. 3C), thereby normalizing synaptic strength in
prefrontal to NAcore afferents (Fig. 3A). The normalization of
synaptic strength in the NAcore arose from restoring both pre-
synaptic release probability (Fig. 5B) and postsynaptic potentia-
tion as estimated by the AMPA:NMDA current ratio (Fig. 5D).
The importance of these mechanisms in cocaine relapse was
shown by blocking the N-acetylcysteine-induced inhibition of re-
instated cocaine-seeking with a microinjection of a mGluR2/3
antagonist into the NAcore (Fig. 4B). Importantly, akin to inhib-
iting reinstated cocaine-seeking, the N-acetylcysteine-induced
amelioration of the cocaine-induced glutamatergic neuroadap-
tations endured for at least 2 wk after discontinuing daily N-
behavioral impairments in an animal model of relapse supports
the potential clinical utility of a strategy to restore glutamate
homeostasis in treating cocaine addiction.
The profile of physiological responses normalized by N-
acetylcysteine is remarkable by ranging from extracellular, non-
synaptic glutamate concentration to measures of both presyn-
aptic and postsynaptic glutamate transmission. Although this
regulation supports the concept that homeostasis between syn-
aptic and nonsynaptic glutamate release is important in cocaine-
induced pathophysiology (4), some underlying mechanisms link-
ing glutamate homeostasis to synaptic plasticity remain unclear.
For example, although the data in Fig. 4 show that normalizing
function at mGluR2/3 is critical in the relapse protection afforded
by chronic N-acetylcysteine, it is unclear how stimulation of pre-
synaptic mGluR2/3 function and corresponding reduction in
synaptic glutamate release probability also restores postsynaptic
measures, such as the AMPA:NMDA ratio.
Multiple synaptic mechanisms can be considered. First, ele-
vating extrasynaptic glutamate concentrations also stimulates
group I mGluRs (mGluR1/5) (7). The stimulation of mGluR1/5
induces long-term depression that has been associated with an
endocannabinoid-mediated reduction in presynaptic glutamate
acetylcysteine is mGluR2/3-dependent. (A) Treatment protocol for data in B.
(B) The inhibition of reinstated cocaine-seeking in the N-acetylcysteine treat-
ment group was reversed by intra-NAcore microinjection of LY341495; paired
Student’s t test: days 25, 29 [t(7) = 2.57, P = 0.037]; days 43, 47 [t(7) = 3.92,
P = 0.006]. (C) LY341495 did not alter reinstatement in the saline treatment
group. (D) Illustration and example micrograph showing the location of mi-
croinjection cannula tips in the NAcore. ac, anterior commissure; arrow, injec-
tion site in the dorsomedial NAcore. *P < 0.05, comparing LY341495 to aCSF.
4. Enduring protection from relapse to cocaine-seeking by N-
| www.pnas.org/cgi/doi/10.1073/pnas.1011265108 Moussawi et al.
release (38) and a rise in intracellular Ca2+that internalizes
AMPA receptors and/or substitutes high-conductance AMPA
receptor subunits with the low-conductance GluR2 AMPA re-
ceptor subunit (39). Such a mechanism within the NAcore may
explain the reduction of AMPA:NMDA ratio after N-acetylcys-
teine treatment. However, behavioral studies indicate that acti-
vation of mGluR5 likely promotes the reinstatement of drug-
seeking in animal models (7, 40), indicating that either this
mechanism does not contribute to decreasing postsynaptic
strength or that decreasing postsynaptic strength is not critical to
the suppression of cocaine-seeking by N-acetylcysteine.
Second, although regulating cystine-glutamate exchange in the
NAcore modulates reinstated cocaine-seeking (20), it is impor-
tant to consider that systemically administered N-acetylcysteine
will have broader actions that could indirectly influence gluta-
mate homeostasis and synaptic plasticity in the NAcore. For
example, basal activity in the prefrontal cortex is blunted in
human addicts and in animal models (12, 15), and by restoring
activity to the prefrontal cortex, N-acetylcysteine could decrease
action potential-mediated synaptic glutamate release and elicit
homeostatic down-regulation of AMPA receptors (41).
Finally, the possibility exists that the reduction in AMPA:
NMDA ratio does not represent a decrease in excitatory trans-
mission due to reducing AMPA receptors, but rather a potentia-
tion due to an increase in NMDA currents. It has been shown that
mGluR1/5 stimulation potentiates striatal NMDA receptor-
5, N-acetylcysteine may enhance NMDA currents and reduce the
AMPA:NMDA ratio. However, this explanation seems unlikely
because cocaine does not alter NMDA currents in the accumbens
(8), and we found that the evoked NMDA current amplitude and
decay time-constants did not differ across treatment groups (Fig.
S4), suggesting that there is no change in kinetics or subunit com-
position of NMDA receptors after N-acetylcysteine treatment (44).
Together, this series of experiments indicates that pharma-
cological restoration of glutamate homeostasis with daily N-
acetylcysteine may provide a mechanism for treating cocaine de-
pendence, apossibility also indicatedby a recentpilotclinical trial
with N-acetylcysteine in nicotine dependence (21). Moreover,
3 was identified as a critical link between restoring glutamate
homeostasis and synaptic potentiation, stimulating mGluR2/3
withdirect orallosteric agonists maybe therapeutically beneficial.
Indeed, mGluR2/3 agonist administration systemically or into the
NAcore inhibits cue- or cocaine-reinstated drug seeking (45, 46).
Thus, N-acetylcysteine alone or in combination with drugs tar-
geting other key proteins contributing to glutamate homeostasis,
including metabotropic glutamate receptors and glutamate
transporters, may proveeffective treatmentsfor inhibitingrelapse
in addiction to cocaine and perhaps other drugs of abuse.
Animal Housing and Surgery. All experiments were conducted in accordance
with the National Institutes of Health Guidelines for the Care and Use of
Laboratory Animals, and the Institutional Animal Care and Use Committee at
the Medical University of South Carolina approved all procedures. Male
Sprague-Dawley rats were single housed, and food intake was maintained at
∼25 g/d. Rats were anesthetized with ketamine/xylazine and implanted with
i.v. catheters. For in vivo intracranial experiments, rats were stereotaxically
implanted with bilateral guide cannulae aimed at the NAcore according to
Paxinos and Watson (47).
Self-Administration, Extinction, and Reinstatement Procedures. Daily self-
administration consisted of 12 d at >10 infusions/2-h session (lever press = 0.2
mg in 0.05 mL cocaine infusion over 2 s and a 5-s presentation of light- and
tone-conditioned cues). Paired yoked saline controls were used. Extinction
sessions began on the day after the final self-administration session and
continued for 2 wk, after which rats were left in their home cages for 2–3
wk. Following the extinction period, animals underwent cue and cocaine-
and cue-primed (10 mg/kg) reinstatement trials (Fig. 4A). For the microin-
jection reinstatement study, a counterbalanced cross-over design was used,
and LY341495 (2 μg/0.5 μL per side) or artificial cerebrospinal fluid (aCSF)
were infused over 2 min.
msec; 25.6 pA. (B) Cumulative probability and mean values showing mEPSC frequency is increased by cocaine self-administration and was normalized by NAC.
n = no. of cells, no. of animals; one-way ANOVA: F(2, 27) = 6.61, P = 0.005; *P < 0.05, compared with saline, vehicle. (C) Cumulative probability and mean
values indicating mEPSC amplitude did not differ between treatment groups. (D) (Upper) Sample EPSC from the three different treatment groups illustrating
total-, AMPA-, and NMDA-mediated currents. Calibration: 100 ms; 100 pA. (Lower) The increase in AMPA:NMDA ratio after withdrawal from self-adminis-
tered cocaine is normalized to control values by NAC. One-way ANOVA: F(2, 40) = 6.57, P = 0.004. *P < 0.05 for (cocaine, saline) compared with other groups
using a Bonferroni multiple comparisons test.
N-acetylcysteine (NAC) restores mEPSC frequency and the AMPA:NMDA ratio. (A) Example of mEPSC from each treatment group. Calibration: 500
Moussawi et al. PNAS
| January 4, 2011
| vol. 108
| no. 1
In Vivo Microdialysis and Glutamate Quantification. Probes were constructed
as described (48). The night before the experiment, a probe was inserted
through an indwelling guide cannula, and the following morning 20-min
samples were collected. Glutamate was measured using HPLC with electro-
chemical detection (48).
In Vivo Field Potentials. Ratswereanesthetizedwithurethaneandmountedin
in the ventral prelimbic prefrontal cortex, glass recording electrodes were
aimed at the dorsomedial region of the accumbens, and field potential am-
plitude was recorded (7). In some field recordings, microdialysis probes were
the NAcore to deliver an mGluR2/3 agonist or antagonist (2 μL/min flow rate).
Slice Preparation and In Vitro Whole Cell Recording. Rats were administered
ketamine/xylazine and transcardially perfused with aCSF. Coronal brain slices
were collected and incubated in recording aCSF at 35°. Recordings targeted
the dorsomedial NAcore where in vivo recordings were made and prefrontal
inputs are most dense (36, 49). MSNs were visualized using video microscopy.
Picrotoxin was used to block inhibitory synaptic transmission. EPSC were
recorded under −80 mV voltage clamp whole-cell configuration using glass
microelectrodes and were evoked using bipolar stimulation at ∼100–200 μm
dorsomedial of the recorded cell. AMPA currents were isolated by applying
the NMDA receptor antagonist D-AP5 at +40 mV, and NMDA currents were
obtained by subtracting the AMPA current from the total current (8). mEPSC
were recorded in presence of bath applied TTX, with detection criteria set
at >7 pA (37). Cumulative probability plots were generated using 400–600
Histology and Statistics. For the microinjection and microdialysis studies, fixed
brains were sectioned (100 μm thick) and stained with a Nissl stain. For field
potential studies, pontamine sky blue was used to mark the recording site at
the end of each recording session (7). Parametric statistical analysis were
used for behavioral and electrophysiology data as indicated, and for greater
than two groups, a Bonferroni multiple comparison post hoc comparison
was used. For the no-net-flux analysis, linear regression with a Bonferroni
correction for multiple comparisons was used. F values are reported in the
figure legends when significant (P < 0.05).
expertise. This research was funded in part by Grant T32007288 (to C.M.R.),
Grant DA015369 (to P.W.K.), Grant DA012513 (to P.W.K.), Grant DA003906
(to P.W.K.), and Grant RR015455 from the National Institutes of Health.
1. Hyman SE, Malenka RC, Nestler EJ (2006) Neural mechanisms of addiction: the role of
reward-related learning and memory. Annu Rev Neurosci 29:565–598.
2. Kelley AE (2004) Memory and addiction: shared neural circuitry and molecular
mechanisms. Neuron 44:161–179.
3. Graybiel AM (2008) Habits, rituals, and the evaluative brain. Annu Rev Neurosci 31:
4. Kalivas PW (2009) The glutamate homeostasis hypothesis of addiction. Nat Rev
5. Kauer JA, Malenka RC (2007) Synaptic plasticity and addiction. Nat Rev Neurosci 8:
6. Conrad KL, et al. (2008) Formation of accumbens GluR2-lacking AMPA receptors
mediates incubation of cocaine craving. Nature 454:118–121.
7. Moussawi K, et al. (2009) N-Acetylcysteine reverses cocaine-induced metaplasticity.
Nat Neurosci 12:182–189.
8. Kourrich S, Rothwell PE, Klug JR, Thomas MJ (2007) Cocaine experience controls
bidirectional synaptic plasticity in the nucleus accumbens. J Neurosci 27:7921–7928.
9. Anderson SM, et al. (2008) CaMKII: a biochemical bridge linking accumbens dopamine
and glutamate systems in cocaine seeking. Nat Neurosci 11:344–353.
10. Martin M, Chen BT, Hopf FW, Bowers MS, Bonci A (2006) Cocaine self-administration
selectively abolishes LTD in the core of the nucleus accumbens. Nat Neurosci 9:
11. Boudreau AC, Wolf ME (2005) Behavioral sensitization to cocaine is associated with
increased AMPA receptor surface expression in the nucleus accumbens. J Neurosci 25:
12. Goldstein RZ, Volkow ND (2002) Drug addiction and its underlying neurobiological
basis: neuroimaging evidence for the involvement of the frontal cortex. Am J
13. Kalivas PW, Volkow ND (2005) The neural basis of addiction: a pathology of
motivation and choice. Am J Psychiatry 162:1403–1413.
14. Carelli R, Ijames S (2000) Nucleus accumbens cell firing during maintenance,
extinction, and reinstatement of cocaine self-administration behavior in rats. Brain
15. Sun W, Rebec GV (2006) Repeated cocaine self-administration alters processing of
cocaine-related information in rat prefrontal cortex. J Neurosci 26:8004–8008.
16. Wolf ME (2010) The Bermuda Triangle of cocaine-induced neuroadaptations. Trends
17. Danbolt NC (2001) Glutamate uptake. Prog Neurobiol 65:1–105.
18. McBean GJ (2002) Cerebral cystine uptake: a tale of two transporters. Trends
Pharmacol Sci 23:299–302.
19. Baker DA, Xi ZX, Shen H, Swanson CJ, Kalivas PW (2002) The origin and neuronal
function of in vivo nonsynaptic glutamate. J Neurosci 22:9134–9141.
20. Kau KS, et al. (2008) Blunted cystine-glutamate antiporter function in the nucleus
accumbens promotes cocaine-induced drug seeking. Neuroscience 155:530–537.
21. Knackstedt LA, Melendez RI, Kalivas PW (2010) Ceftriaxone restores glutamate
homeostasis and prevents relapse to cocaine seeking. Biol Psychiatry 67:81–84.
22. Fourgeaud L, et al. (2004) A single in vivo exposure to cocaine abolishes
endocannabinoid-mediated long-term depression in the nucleus accumbens. J
23. Moussawi K, Kalivas PW (2010) Group II metabotropic glutamate receptors (mGlu2/3)
in drug addiction. Eur J Pharmacol 639:115–122.
24. Anwyl R (2009) Metabotropic glutamate receptor-dependent long-term potentiation.
25. Baker DA, et al. (2003) Neuroadaptations in cystine-glutamate exchange underlie
cocaine relapse. Nat Neurosci 6:743–749.
26. Madayag A, et al. (2007) Repeated N-acetylcysteine administration alters plasticity-
dependent effects of cocaine. J Neurosci 27:13968–13976.
27. Shaham Y, Shalev U, Lu L, De Wit H, Stewart J (2003) The reinstatement model of drug
28. Alagarsamy S, Sorensen SD, Conn PJ (2001) Coordinate regulation of metabotropic
glutamate receptors. Curr Opin Neurobiol 11:357–362.
exchange regulates metabotropic glutamate receptor presynaptic inhibition of
excitatory transmission and vulnerability to cocaine seeking. J Neurosci 25:6389–6393.
30. Huang CC, Yang PC, Lin HJ, Hsu KS (2007) Repeated cocaine administration impairs
group II metabotropic glutamate receptor-mediated long-term depression in rat
medial prefrontal cortex. J Neurosci 27:2958–2968.
31. Losonczy A, Somogyi P, Nusser Z (2003) Reduction of excitatory postsynaptic responses
by persistently active metabotropic glutamate receptors in the hippocampus. J
32. Sah P, Hestrin S, Nicoll RA (1989) Tonic activation of NMDA receptors by ambient
glutamate enhances excitability of neurons. Science 246:815–818.
33. Wright RA, Arnold MB, Wheeler WJ, Ornstein PL, Schoepp DD (2001) [3H]LY341495
binding to group II metabotropic glutamate receptors in rat brain. J Pharmacol Exp
34. Dietrich D, Kral T, Clusmann H, Friedl M, Schramm J (2002) Presynaptic group II
metabotropic glutamate receptors reduce stimulated and spontaneous transmitter
release in human dentate gyrus. Neuropharmacology 42:297–305.
35. Thomas MJ, Beurrier C, Bonci A, Malenka RC (2001) Long-term depression in the
nucleus accumbens: a neural correlate of behavioral sensitization to cocaine. Nat
a spin on the dorsal-ventral divide of the striatum. Trends Neurosci 27:468–474.
37. Chen BT, et al. (2008) Cocaine but not natural reward self-administration nor passive
cocaine infusion produces persistent LTP in the VTA. Neuron 59:288–297.
38. Kreitzer AC, Malenka RC (2007) Endocannabinoid-mediated rescue of striatal LTD and
motor deficits in Parkinson’s disease models. Nature 445:643–647.
39. Gubellini P, Pisani A, Centonze D, Bernardi G, Calabresi P (2004) Metabotropic
glutamate receptors and striatal synaptic plasticity: implications for neurological
diseases. Prog Neurobiol 74:271–300.
40. Gass JT, Olive MF (2008) Glutamatergic substrates of drug addiction and alcoholism.
Biochem Pharmacol 75:218–265.
41. Turrigiano GG (2008) The self-tuning neuron: synaptic scaling of excitatory synapses.
42. Collett VJ, Collingridge GL (2004) Interactions between NMDA receptors and mGlu5
receptors expressed in HEK293 cells. Br J Pharmacol 142:991–1001.
43. Homayoun H, Stefani MR, Adams BW, Tamagan GD, Moghaddam B (2004) Functional
interaction between NMDA and mGlu5 receptors: effects on working memory, instru-
mental learning, motor behaviors, and dopamine release. Neuropsychopharmacology
44. Cull-Candy SG, Leszkiewicz DN (2004) Role of distinct NMDA receptor subtypes at
central synapses. Sci STKE 2004:re16.
45. Baptista MA, Martin-Fardon R, Weiss F (2004) Preferential effects of the metabotropic
glutamate 2/3 receptor agonist LY379268 on conditioned reinstatement versus
primary reinforcement: comparison between cocaine and a potent conventional
reinforcer. J Neurosci 24:4723–4727.
46. Peters J, Kalivas PW (2006) The group II metabotropic glutamate receptor agonist,
LY379268, inhibits both cocaine- and food-seeking behavior in rats. Psychophar-
macology (Berl) 186:143–149.
47. Paxinos G, Watson C (1986) The Rat Brain in Stereotaxic Coordinates (Academic, San
48. LaLumiere RT, Kalivas PW (2008) Glutamate release in the nucleus accumbens core is
necessary for heroin seeking. J Neurosci 28:3170–3177.
49. Gorelova N, Yang CR (1997) The course of neural projection from the prefrontal
cortex to the nucleus accumbens in the rat. Neuroscience 76:689–706.
| www.pnas.org/cgi/doi/10.1073/pnas.1011265108Moussawi et al.