Essay

Neuron

Report

Relapse Induced by Cues Predicting Cocaine Depends on Rapid, Transient Synaptic Potentiation Cassandra D. Gipson,1,3,* Yonatan M. Kupchik,1,3 Haowei Shen,2,3 Kathryn J. Reissner,1 Charles A. Thomas,1

and Peter W. Kalivas1,* 1Department of Neurosciences, Medical University of South Carolina, Charleston, SC 29425, USA 2National Institute on Drug Dependence, Peking University, Beijing 100191, China 3These authors contributed equally to this work

*Correspondence: [email protected] (C.D.G.), [email protected] (P.W.K.)

http://dx.doi.org/10.1016/j.neuron.2013.01.005

SUMMARY

Cocaine addiction is characterized by long-lasting vulnerability to relapse arising because neutral envi- ronmental stimuli become associated with drug use and then act as cues that induce relapse. It is not known how cues elicit cocaine seeking, and why cocaine seeking is more difficult to regulate than seeking a natural reward. We found that cocaine- associated cues initiate cocaine seeking by inducing a rapid, transient increase in dendritic spine size and synaptic strength in the nucleus accumbens. These changes required neural activity in the prefrontal cortex. This is not the case when identical cues were associated with obtaining sucrose, which did not elicit changes in spine size or synaptic strength. The marked cue-induced synaptic changes in the ac- cumbens were correlated with the intensity of cocaine, but not sucrose seeking, and may explain the difficulty addicts experience in managing relapse to cocaine use.

INTRODUCTION

Understanding the neurobiology of relapse to drug use will facil-

itate the development of pharmacotherapies to treat addiction

(Kalivas and Volkow, 2011; Vocci and Ling, 2005). An important

feature of the enduring vulnerability to relapse is that neutral

environmental stimuli become associated with drug use and

act as cues that initiate relapse (Goldstein and Volkow, 2002;

See, 2002; Wilson et al., 2004). Presenting cues previously

paired with cocaine use initiates craving and drug seeking, which

are associated with activating the glutamatergic projection from

the prefrontal cortex to the nucleus accumbens (Kalivas, 2009;

Koob and Volkow, 2010; Wilson et al., 2004). Given the well-

established role of the corticostriatal projection in regulating

motivated behavior (Balleine et al., 2007; Lüscher and Malenka,

2011; Miller and Marshall, 2004), it is thought that cocaine-

induced changes in this glutamatergic projection enable environ-

mental stimuli associated with cocaine use to act as conditioned

cues that elicit uncontrollable motivation to relapse to drug use

compared with the more manageable motivation to obtain

natural reward (Garavan et al., 2000; Levy et al., 2007).

The neurobiology of relapse to cocaine use is most frequently

studied in animal models by measuring long-lasting changes in

brain structure and function after experimenter-injected or self-

administered cocaine followed by varying periods of withdrawal.

A key observation using this approach is that excitatory

synapses on medium spiny neurons (MSNs) in the accumbens

show evidence of long-term potentiation (LTP), including

increased dendrite spine head diameter, elevated AMPA gluta-

mate receptor-mediated synaptic currents, and AMPA receptor

surface expression (Boudreau et al., 2007; Conrad et al., 2008;

Kourrich and Thomas, 2009; Moussawi et al., 2009; Shen

et al., 2009; Wolf and Ferrario, 2010). This LTP-like state is sug-

gested to mediate the enhanced motivation underlying relapse

to drug use compared to natural reward (Wolf, 2010). However,

it remains unknown how initiating relapse with cocaine-condi-

tioned cues (i.e., absent the pharmacological effects of the

drug) affects synaptic physiology and morphology, if synaptic

changes are important for initiating relapse, or if cues initiating

cocaine seeking produce distinct synaptic changes compared

with the same cues initiating seeking of a natural reward.

In order to investigate these synaptic mechanisms contrib-

uting to cocaine relapse, we used a ‘‘short-access’’ model of

cocaine self-administration and examined the reinstatement

of cocaine seeking by a light/tone cue previously associated

with cocaine delivery. While this paradigm may not model

compulsive drug self-administration (Koob, 2012), it allows

investigation of cue-induced cocaine seeking after a period

of withdrawal (Epstein et al., 2006; Shaham et al., 2003) and

elicits enduring physiological and neurochemical changes in

the projection from the prefrontal cortex to nucleus accumbens

(Kalivas, 2009; Wolf, 2010). Using this model, we show that

presenting cocaine-associated cues simultaneously initiated

cocaine seeking and a rapid, transient increase in dendritic spine

size and synaptic strength in the nucleus accumbens. The

synaptic changes were positively correlated with the intensity

of reinstated cocaine seeking and required activity in the

prefrontal cortex. Importantly, the increase in spine size and

synaptic response did not occur when the same seeking behavior

was induced by identical cues paired with a natural reward

(sucrose). Our data demonstrate that associating cocaine, but

not a natural reward, with environmental cues confers a capacity

for these cues to transiently potentiate accumbens excitatory

Neuron 77, 867–872, March 6, 2013 ª2013 Elsevier Inc. 867

Figure 1. Cue-Induced Cocaine Seeking Rapidly Enlarges Spine

Head Diameter in NAcore MSNs

(A) Cue-induced reinstatement of cocaine seeking increased active lever

pressing over the 15, 45, or 120 min prior to euthanizing rats for morphological

or A/N measurements (F(7,92) = 39.93, p < 0.0001). (B) Time course of active

lever pressing during the cue reinstatement session. (C) Sample dendrites

from NAcore MSNs in yoked-saline (dh = 0.329 mm) or cocaine-trained rats at

T = 0 (0.422 mm); T = 15 (0.528 mm) min after initiating cue-induced rein-

statement. (D) Cumulative distribution of spine head diameter reveals changes

in dh between treatment groups (group: F(5,1698) = 8769, p < 0.0001; dh F(4,1698) = 115.2, p < 0.0001; interaction: F(20,1698) = 21, p < 0.0001). (E) Cocaine

self-administration increased mean dh (F(6,326) = 43.98, p < 0.0001). Spine dh was elevated at T = 15, decreased below prereinstatement levels at T = 45, and

returned to prereinstatement levels at T = 120. (F) The increase in dh at 15 min

was significantly correlated with active lever pressing. N in (D) is the number of

rats, and N shown as the number in bars corresponds to either the number of

animals (A) or the number of neurons quantified (E). Five to twelve neurons

were measured from each rat. Data are shown as mean ± SEM. *p < 0.05,

compared to yoked-saline or extinction lever presses; #p < 0.05, compared to

T = 0 cocaine.

Neuron

Cued Relapse and Synaptic Plasticity

synaptic transmission. This distinction between cocaine and

sucrose may explain the relatively uncontrollable motivation to

relapse to cocaine use compared with the more manageable

desire for natural reward.

RESULTS

Rats were trained to self-administer cocaine by pressing one of

two levers to receive an intravenous cocaine injection. Rats self-

868 Neuron 77, 867–872, March 6, 2013 ª2013 Elsevier Inc.

administered cocaine for 2 hr a day over 10 days to achieve

stable daily cocaine use, and lever pressing was then extin-

guished over another 14 days of 2 hr sessions (see Figure S1

available online). A light/tone compound stimulus was paired

with cocaine infusions during the self-administration sessions

and lever pressing during extinction training yielded neither

cocaine nor the light/tone stimulus. A parallel yoked-saline

control group was included consisting of rats administered an

intravenous infusion of saline when a paired rat self-administered

cocaine. Once rats achieved a stable extinguished baseline of

lever pressing (Figure S1), the light/tone cue was presented

with each press of the lever that previously provided cocaine

(active lever), but no cocaine was delivered. Returning the

conditioned light/tone cue resulted in a marked reinstatement

of lever pressing (Figures 1A and 1B), which was used to model

cue-induced relapse (Epstein et al., 2006).

As described above, cocaine self-administration causes

stable LTP-like synaptic potentiation in the core subcompart-

ment of the nucleus accumbens (NAcore) that endures for

months after discontinuing cocaine use and is proposed to

contribute to cocaine relapse. To test whether synaptic alter-

ations initiated in the NAcore by presentation of cocaine-condi-

tioned cues contribute to relapse, we examined animals just

prior to beginning cue-induced reinstatement (time [T] = 0) or

at 15, 45, or 120 min after beginning the reinstatement trial for

two measures of synaptic plasticity. To quantify spine density

and dh, we made three-dimensional (3D) confocal images of

neurons in the NAcore that were diolistically labeled with the lipo-

philic dye DiI (Figure 1C) (Shen et al., 2011). Synaptic strength

was also estimated by calculating the ratio of AMPA to NMDA

currents (A/N) (Malenka and Bear, 2004) using whole-cell patch

recordings from MSNs in NAcore tissue slices (Moussawi

et al., 2011; Shen et al., 2011).

Conditioned cues reinstated robust active lever pressing com-

pared to inactive lever pressing or to active lever pressing during

extinction (Figure 1A). The increase in lever pressing was

maximal during the first 10 min of the reinstatement session

and progressively decreased thereafter for the remainder of

the session (Figure 1B). At T = 0, dh was increased in rats extin-

guished from cocaine self-administration (0.415 ± 0.007 mm)

compared to yoked-saline controls (0.327 ± 0.004 mm) (Figures

1C–1E). Cue-induced reinstatement further increased dh at

15 min after the cue was presented (0.491 ± 0.009 mm). By

45 min after initiating reinstatement, dh decreased below the

resting (T = 0) cocaine levels (0.361 ± 0.009 mm) and had returned

to prereinstatement levels by 120 min after initiating the rein-

statement session (0.427 ± 0.011 mm). Importantly, the amount

of reinstated active lever pressing at 15 min was positively corre-

lated with the increase in dh (Figure 1F). No difference in spine

density was found between groups (Figure S2A).

To test whether cue-induced morphological plasticity was

specific for reinstating lever pressing for cocaine and not for

a natural reward, we trained rats to self-administer sucrose

pellets and we paired pellet delivery with the same light/tone

stimulus used for cocaine training. While the sucrose-trained

rats showed robust cue-induced reinstatement of lever pressing

(Figure 2A), no change in mean dh (Figure 2B) or spine density

(Figure S2B) was measured at 15 or 45 min after the cue

Figure 2. Sucrose-Trained Rats Do Not

Show Increased Spine Head Diameter

during Cue-Induced Reinstatement

(A) Sucrose-trained rats show significant cue-

induced reinstatement (F(3,31) = 86.981, p < 0.001).

(B) Lever responding for contingent cues was

maximal during the first 10 min of the session and

decreased thereafter. (C) Sucrose reinstatement

was not accompanied by a change in dh. N is

the number in bars and corresponds to either the

number of animals (A) or the number of neurons

quantified (C). Five to twelve neurons were

measured from each rat. Data are shown as

mean ± SEM. *p < 0.05, compared to extinction

lever presses.

Neuron

Cued Relapse and Synaptic Plasticity

(Figure 2C). To further test that the cue-induced increase in dh depended on a contingent association between the cocaine-

paired lever and the light/tone cue, we exposed rats to the

chamber either without presenting cues or when cues were pre-

sented independent of lever pressing, and these animals also

showed no increase in dh, spine density, or lever pressing at

15 min after beginning the session (Figures S3A–S3C). Taken

together, the morphological measurements show that with-

drawal from daily cocaine self-administration causes a resting

enlargement of dh and that cue-induced reinstatement of

cocaine, not sucrose, seeking is accompanied by a further rapid,

transient enlargement of the spines that is significantly corre-

lated with reinstated behavior.

Whole-cell patch recordings revealed parallel evidence for

rapid synaptic potentiation during cue-induced reinstatement

(Figures 3A and 3B). Withdrawal from self-administered cocaine

increased A/N compared to yoked-saline rats. The A/N was

further increased 15 min after initiating cue-induced reinstate-

ment and returned to prereinstatement levels after 120 min. In

contrast with the decrease in dh (Figure 1F), the A/N remained

elevated at T = 45. Similar to dh, the increase of A/N at 15 min

was significantly correlated with the number of reinstated active

lever presses (Figure 3C), and cue-induced reinstatement of

sucrose did not show a change in A/N at 15 or 45 min after initi-

ating the reinstatement session (Figure 3D).

The NAcore receives glutamatergic input from various sour-

ces, and increased release of glutamate from the prelimbic

cortex (PL) into the NAcore is required for reinstating drug

seeking (LaLumiere and Kalivas, 2008; McFarland et al., 2003).

We found that neural activity in the PL is also critical for the

cue-induced synaptic changes in the NAcore. Inhibiting the PL

by microinjecting GABA agonists (baclofen plus muscimol) prior

to the reinstatement session prevented cue-induced increases in

dh and A/N in NAcore MSNs, as well as blocked reinstated active

lever pressing (Figure 4; Figure S4 for histology).

DISCUSSION

We show here that the reinstatement of cocaine seeking by

conditioned cues, but not the reinstatement of seeking a natural

reward, was accompanied by rapid, transient synaptic potentia-

tion in NAcore MSNs. The rapid potentiation contrasts with

previous reports showing that cocaine use reduces the ability

of prefrontal input to induce classical forms of synaptic plasticity,

such as LTP and LTD (Martin et al., 2006; Moussawi et al.,

2009). Thus, while cocaine use diminishes the capacity of stimuli

not associated with drug use to induce synaptic plasticity,

LTP-like plasticity is readily induced by stimuli paired with

cocaine use. The importance in relapse of synaptic plasticity

selectively coded by cocaine-associated cues was supported

by a significant correlation between the intensity of cocaine

seeking and both morphological and electrophysiological

measures of synaptic potentiation.

Changes in spine density and/or head diameter (dh) are a struc-

tural substrate for synaptic plasticity, with larger dh being asso-

ciated with LTP and reduced dh with LTD (Carlisle and Kennedy,

2005; De Roo et al., 2008; Yang and Zhou, 2009). Consistent with

previous reports (Kourrich et al., 2007; Moussawi et al., 2011;

Shen et al., 2009), withdrawal from investigator- or self-adminis-

tered cocaine increased dh and A/N compared to yoked-saline

rats. The dh and A/N were further increased 15 min after initiating

cue-induced reinstatement and returned to prereinstatement

levels after 120 min. Although there was a decrease in dh at

T = 45, the A/N remained elevated. The slower normalization of

the A/N is consistent with previous in vitro studies indicating

that although both dh and A/N are reliable markers of synaptic

plasticity, they are regulated in part by distinct signaling path-

ways (Fukazawa et al., 2003; Henley et al., 2011). For example,

inhibiting protein phosphatase 1 prevents electrophysiological

measures of LTP without affecting enlargement of dendritic

spines (Zhou et al., 2004). Conversely, inhibiting actin polymeri-

zation reduces spine size in cultured neurons (Gu et al., 2010)

but inhibits only enduring LTP (>1 hr), leaving intact short-term

synaptic potentiation that is akin to what we show here being

initiated by cocaine-conditioned cues (Fukazawa et al., 2003;

Krucker et al., 2000; Ramachandran and Frey, 2009).

A link between reinstated cocaine seeking and the rapid LTP-

like plasticity was also indicated by inactivating the PL and

showing necessary involvement of this region of the PFC in

cue-induced increases in dh and A/N. It is likely that the glutama-

tergic projection from the PL to the NAcore is contributing to the

effects of inactivation since double-dissociation pharmacolog-

ical inactivation and more selective optogenetic inhibition show

that this pathway is necessary for reinstated cocaine seeking

(McFarland and Kalivas, 2001; Stefanik et al., 2013). This mech-

anism is also consistent with in vivo recordings showing

Neuron 77, 867–872, March 6, 2013 ª2013 Elsevier Inc. 869

Figure 3. Synaptic Potentiation Initiated by Cue-Induced Cocaine

Seeking

(A) Sample AMPA and NMDA current traces from each group. (B) AMPA to

NMDA ratios (A/N) were significantly elevated in animals withdrawn with

extinction training from cocaine self-administration (1.423 ± 0.075) compared

to yoked-saline animals (1.064 ± 0.050). In addition, the initiation of cue-

induced reinstatement further elevated A/N at T = 15 (1.780 ± 0.060). Ratios

remained elevated at T = 45 (1.815 ± 0.122) and returned to prereinstatement

levels by T = 120 (1.538 ± 0.103) (F(4,101) = 14.45, p < 0.001). (C) The increase in

A/N at 15 min was significantly correlated with the number of active lever

presses. (D) Cue-induced reinstatement of sucrose seeking did not alter A/N.

Two to five neurons were recorded from each animal. Data are shown as

mean ± SEM. *p < 0.05, compared to yoked-saline animals at T = 0 (white bar);

#p < 0.01, compared to T = 0 (black bar).

Figure 4. Inactivation of the PL Prevents Cue-Induced Reinstate-

ment and the Increase in dh and A/N in NAcore

(A) B/M infusions into PL inhibited cue-induced reinstatement (T = 15; F(5,59) =

11.971, p < 0.001; N is shown in bars). (B) Sample dendrites of animals

receiving either aCSF or B/M into PL prior to initiating cue-induced rein-

statement and sacrificed 15 min later. (C) B/M into PL inhibited the mean

increase in dh (t(82) = 7.504, p < 0.001; see inset) and shifted the cumulative

distribution to the left. (D) B/M into PL inhibited the increase in A/N (t(17) = 2.554,

p = 0.021). Data are shown as mean ± SEM. *p < 0.05, comparing aCSF to B/M.

Neuron

Cued Relapse and Synaptic Plasticity

increased activation of NAcore neurons in response to cocaine-

conditioned cues after a period of extinction training (Hollander

and Carelli, 2007) and with neuron culture studies indicating

that glutamate induces LTP-like synaptic changes (Shepherd

and Huganir, 2007). In addition, the lack of rapid LTP-like

plasticity accompanying reinstated sucrose seeking supports

a role for PL glutamatergic input, since cocaine reinstatement

requires a marked rise in the release of synaptic glutamate

from the PL into the NAcore, but reinstated sucrose seeking

does not induce measurable glutamate release (McFarland

et al., 2003). However, it is possible that PL projections to other

brain regions innervating the NAcore known to regulate rein-

stated behavior may also play a role, such as dopamine projec-

tions from the ventral tegmental area or glutamatergic input from

the basolateral amygdala (Koob and Volkow, 2010). A role for

dopaminergic afferents is supported by the fact that in cocul-

tured prefrontal and accumbens neurons, D1 receptor stimula-

tion facilitates trafficking of AMPA receptors to the surface and

costimulation of NMDA receptors promotes D1 synaptic inser-

tion (Sun et al., 2008). In this regard, it will be of interest in future

studies to determine whether the changes identified here are

selective for D1 or D2 receptor-expressing MSNs.

Cocaine addiction is defined in part by the unmanageable

motivation to take cocaine and differs markedly from relative

870 Neuron 77, 867–872, March 6, 2013 ª2013 Elsevier Inc.

control over engaging natural reward. The lack of change in dh and A/N after cue-induced sucrose seeking indicates that asso-

ciating cues with cocaine delivery is conferring neuroadaptations

that are not occurring when the identical cues are associated

with sucrose delivery. This supports the possibility that the rapid,

transient synaptic potentiation may be a biomarker for a cocaine

seeking neuropathology and poses the possibility that counter-

manding the synaptic potentiation may selectively disrupt the

vulnerability to relapse to cocaine use without affecting the

motivation to seek natural reward.

EXPERIMENTAL PROCEDURES

Animal Housing and Surgery

Male Sprague-Dawley rats (250 g; Charles River Laboratories) were individ-

ually housed with a 12:12 hr dark/light cycle. All experimentation occurred

in the dark cycle. Rats received food ad libitum until the day prior to

behavioral training, after which food restriction (20 g of rat chow per day)

was implemented and maintained throughout the experiment. Rats were

allowed 1 week to acclimate to the vivarium before inducing anesthesia

and implanting indwelling jugular catheters, and in some experiments,

microinjection guide cannula were also implanted in the PL (surgical details

in Supplemental Experimental Procedures). All procedures were in accor-

dance with the National Institutes of Health Guide for the Care and Use

of Laboratory Animals and the Assessment and Accreditation of Laboratory

Animal Care.

Neuron

Cued Relapse and Synaptic Plasticity

Cocaine Self-Administration Procedures

Seven days after surgery, rats began daily 2 hr cocaine self-administration

sessions, in which one response on the active lever yielded one intravenous

cocaine infusion (0.2 mg/infusion, followed by a 20 s timeout period), paired

with a white cue light above the active lever and a discrete tone cue. An inac-

tive lever was also available throughout each session. After ten consecutive

sessions of self-administration (greater than or equal to ten infusions a day),

rats were placed into daily extinction training sessions (no cocaine delivery

or cues) for at least 14 sessions or until extinction criteria were met (%25 active

lever responses for a minimum of two sessions). Reinstatement was elicited by

cues (tone plus light delivery after an active lever press).

Microinfusion Procedures and Histology

Rats were stereotaxically implanted immediately after catheterization with

bilateral guide cannulae aimed above PL (see Supplemental Experimental

Procedures for surgical details). Obturators were placed into the guide

cannulae and were removed during bilateral injection of 0.3 ml baclofen/

muscimol cocktail (0.3/0.03 nmol, GABAB/GABAA receptor agonists, respec-

tively) over 1 min (McFarland and Kalivas, 2001). Rats were placed in the

operant chamber 10 min after removal of injection cannulae and replacement

of the obturators. Rats were sacrificed at various times for either dendritic

spine or electrophysiological quantification. When appropriate, coronal slices

(100 mm thick) of PL were mounted and stained via cresyl violet to verify guide

cannulae placement (Figure S4).

Quantification of Dendritic Spines

All dendritic spine quantification procedures have been described previously

(Shen et al., 2009). Briefly, a confocal microscope was used to image

DiI-labeled sections, and DiI was excited using the Helium/Neon 543 nm laser

line. Images of DiI-labeled dendrites (see Figure 1C) were acquired via optical

sectioning using a 633 oil immersion objective (Plan-Apochromat, Zeiss; NA =

1.4, WD = 90 mm) with pixel size 0.07 mm at xy plane and 0.1 mm intervals along

the z axis. Images were deconvoluted prior to analysis, and a 3D perspective

was rendered by the Surpass module of Imaris software package (Bitplane).

Only spines on dendrites beginning at >75 mm and ending at %200 mm distal

to the soma and after the first branch point were quantified from cells localized

to the NA core (see Table S1). The length of quantified dendrites was

45–55 mm. Five to twelve neurons were analyzed from each animal, and the

minimum end segment diameter (spine head) was set at R0.143 mm.

Slice Preparation and Whole-Cell Recordings

Rats were anesthetized with ketamine and decapitated, and coronal accum-

bens brain slices were collected into a vial containing artificial cerebrospinal

fluid (aCSF). All recordings were collected at 32�C in the dorsomedial NAcore, where the prefrontal inputs are most dense (Gorelova and Yang, 1997). Inhib-

itory synaptic transmission was blocked with picrotoxin (50 mM), and AMPA

and NMDA currents were recorded in whole-cell patch-clamp configuration.

Glass microelectrodes (1–2 MU) were filled with cesium-based internal

solution. To evoke postsynaptic currents, we placed a bipolar stimulating

electrode �300 mm dorsomedial of the recorded cell to maximize chances of stimulating PL afferents. The stimulation intensity chosen evoked an

�50% of maximal AMPA current. Recordings were collected every 20 s and begun >10 min after the cell membrane was ruptured to allow diffusion of

the internal solution into the cell. AMPA currents were first measured

at �80 mV to ensure stability of response. Then the membrane potential was gradually increased until +40 mV. Recording of currents was resumed

5 min after reaching +40 mV to allow stabilization of cell parameters. Currents

composed of both AMPA and NMDA components were then obtained. Then

D-AP5 was bath applied (50 mM) to block NMDA currents and recording of

AMPA currents at +40 mV was started after 2 min. NMDA currents were ob-

tained by subtracting the AMPA currents from the total current at +40 mV.

Statistics

All spine density and dh data were statistically analyzed after averaging the

values for all the neurons in each animal. The number of determinations in

each group was established using an analysis of statistical power based on

previous morphological data from our laboratory (Shen et al., 2009). A/N

data were analyzed using ANOVA. Behavioral data were analyzed using

repeated-measures ANOVA, and t tests were used to compare dh and A/N

in animals receiving aCSF or B/M. Additionally, linear regression was used

to determine the association between magnitude of reinstated lever pressing

and dh or A/N. Post hoc comparisons were conducted using Bonferroni-

corrected t tests.

SUPPLEMENTAL INFORMATION

Supplemental Information includes four figures, one table, and Supplemental

Experimental Procedures and can be found with this article online at http://dx.

doi.org/10.1016/j.neuron.2013.01.005.

ACKNOWLEDGMENTS

We thank Dr. Rachel Smith, Dr. Joshua Beckmann, Megan Hensley, Brenton

Mahaffey, Rebecca Szer, and Phong Do for technical assistance. This work

was supported by DA007288, DA033690 (C.D.G.), DA003906, DA012513,

and DA015369 (P.W.K.) grants from the National Institutes of Health.

Accepted: January 2, 2013

Published: March 6, 2013

REFERENCES

Balleine, B.W., Delgado, M.R., and Hikosaka, O. (2007). The role of the dorsal

striatum in reward and decision-making. J. Neurosci. 27, 8161–8165.

Boudreau, A.C., Reimers, J.M., Milovanovic, M., and Wolf, M.E. (2007). Cell

surface AMPA receptors in the rat nucleus accumbens increase during

cocaine withdrawal but internalize after cocaine challenge in association

with altered activation of mitogen-activated protein kinases. J. Neurosci. 27,

10621–10635.

Carlisle, H.J., and Kennedy, M.B. (2005). Spine architecture and synaptic

plasticity. Trends Neurosci. 28, 182–187.

Conrad, K.L., Tseng, K.Y., Uejima, J.L., Reimers, J.M., Heng, L.J., Shaham, Y.,

Marinelli, M., and Wolf, M.E. (2008). Formation of accumbens GluR2-lacking

AMPA receptors mediates incubation of cocaine craving. Nature 454, 118–121.

De Roo, M., Klauser, P., Garcia, P.M., Poglia, L., and Muller, D. (2008). Spine

dynamics and synapse remodeling during LTP and memory processes. Prog.

Brain Res. 169, 199–207.

Epstein, D.H., Preston, K.L., Stewart, J., and Shaham, Y. (2006). Toward

a model of drug relapse: an assessment of the validity of the reinstatement

procedure. Psychopharmacology (Berl.) 189, 1–16.

Fukazawa, Y., Saitoh, Y., Ozawa, F., Ohta, Y., Mizuno, K., and Inokuchi, K.

(2003). Hippocampal LTP is accompanied by enhanced F-actin content within

the dendritic spine that is essential for late LTP maintenance in vivo. Neuron

38, 447–460.

Garavan, H., Pankiewicz, J., Bloom, A., Cho, J.K., Sperry, L., Ross, T.J.,

Salmeron, B.J., Risinger, R., Kelley, D., and Stein, E.A. (2000). Cue-induced

cocaine craving: neuroanatomical specificity for drug users and drug stimuli.

Am. J. Psychiatry 157, 1789–1798.

Goldstein, R.Z., and Volkow, N.D. (2002). Drug addiction and its underlying

neurobiological basis: neuroimaging evidence for the involvement of the

frontal cortex. Am. J. Psychiatry 159, 1642–1652.

Gorelova, N., and Yang, C.R. (1997). The course of neural projection from the

prefrontal cortex to the nucleus accumbens in the rat. Neuroscience 76,

689–706.

Gu, J., Lee, C.W., Fan, Y., Komlos, D., Tang, X., Sun, C., Yu, K., Hartzell, H.C.,

Chen, G., Bamburg, J.R., and Zheng, J.Q. (2010). ADF/cofilin-mediated actin

dynamics regulate AMPA receptor trafficking during synaptic plasticity. Nat.

Neurosci. 13, 1208–1215.

Henley, J.M., Barker, E.A., and Glebov, O.O. (2011). Routes, destinations and

delays: recent advances in AMPA receptor trafficking. Trends Neurosci. 34,

258–268.

Neuron 77, 867–872, March 6, 2013 ª2013 Elsevier Inc. 871

Neuron

Cued Relapse and Synaptic Plasticity

Hollander, J.A., and Carelli, R.M. (2007). Cocaine-associated stimuli increase

cocaine seeking and activate accumbens core neurons after abstinence.

J. Neurosci. 27, 3535–3539.

Kalivas, P.W. (2009). The glutamate homeostasis hypothesis of addiction. Nat.

Rev. Neurosci. 10, 561–572.

Kalivas, P.W., and Volkow, N.D. (2011). New medications for drug addiction

hiding in glutamatergic neuroplasticity. Mol. Psychiatry 16, 974–986.

Koob, G.F. (2012). Animal models of psychiatric disorders. Handb. Clin.

Neurol. 106, 137–166.

Koob, G.F., and Volkow, N.D. (2010). Neurocircuitry of addiction.

Neuropsychopharmacology 35, 217–238.

Kourrich, S., and Thomas, M.J. (2009). Similar neurons, opposite adaptations:

psychostimulant experience differentially alters firing properties in accumbens

core versus shell. J. Neurosci. 29, 12275–12283.

Kourrich, S., Rothwell, P.E., Klug, J.R., and Thomas, M.J. (2007). Cocaine

experience controls bidirectional synaptic plasticity in the nucleus accum-

bens. J. Neurosci. 27, 7921–7928.

Krucker, T., Siggins, G.R., and Halpain, S. (2000). Dynamic actin filaments are

required for stable long-term potentiation (LTP) in area CA1 of the hippo-

campus. Proc. Natl. Acad. Sci. USA 97, 6856–6861.

LaLumiere, R.T., and Kalivas, P.W. (2008). Glutamate release in the nucleus

accumbens core is necessary for heroin seeking. J. Neurosci. 28, 3170–3177.

Levy, D., Shabat-Simon, M., Shalev, U., Barnea-Ygael, N., Cooper, A., and

Zangen, A. (2007). Repeated electrical stimulation of reward-related brain

regions affects cocaine but not ‘‘natural’’ reinforcement. J. Neurosci. 27,

14179–14189.

Lüscher, C., and Malenka, R.C. (2011). Drug-evoked synaptic plasticity in

addiction: from molecular changes to circuit remodeling. Neuron 69, 650–663.

Malenka, R.C., and Bear, M.F. (2004). LTP and LTD: an embarrassment of

riches. Neuron 44, 5–21.

Martin, M., Chen, B.T., Hopf, F.W., Bowers, M.S., and Bonci, A. (2006).

Cocaine self-administration selectively abolishes LTD in the core of the

nucleus accumbens. Nat. Neurosci. 9, 868–869.

McFarland, K., and Kalivas, P.W. (2001). The circuitry mediating cocaine-

induced reinstatement of drug-seeking behavior. J. Neurosci. 21, 8655–8663.

McFarland, K., Lapish, C.C., and Kalivas, P.W. (2003). Prefrontal glutamate

release into the core of the nucleus accumbens mediates cocaine-induced

reinstatement of drug-seeking behavior. J. Neurosci. 23, 3531–3537.

Miller, C.A., and Marshall, J.F. (2004). Altered prelimbic cortex output during

cue-elicited drug seeking. J. Neurosci. 24, 6889–6897.

Moussawi, K., Pacchioni, A., Moran, M., Olive, M.F., Gass, J.T., Lavin, A., and

Kalivas, P.W. (2009). N-Acetylcysteine reverses cocaine-induced meta-

plasticity. Nat. Neurosci. 12, 182–189.

872 Neuron 77, 867–872, March 6, 2013 ª2013 Elsevier Inc.

Moussawi, K., Zhou, W., Shen, H., Reichel, C.M., See, R.E., Carr, D.B., and

Kalivas, P.W. (2011). Reversing cocaine-induced synaptic potentiation

provides enduring protection from relapse. Proc. Natl. Acad. Sci. USA 108,

385–390.

Ramachandran, B., and Frey, J.U. (2009). Interfering with the actin network

and its effect on long-term potentiation and synaptic tagging in hippocampal

CA1 neurons in slices in vitro. J. Neurosci. 29, 12167–12173.

See, R.E. (2002). Neural substrates of conditioned-cued relapse to drug-

seeking behavior. Pharmacol. Biochem. Behav. 71, 517–529.

Shaham, Y., Shalev, U., Lu, L., De Wit, H., and Stewart, J. (2003). The reinstate-

ment model of drug relapse: history, methodology and major findings.

Psychopharmacology (Berl.) 168, 3–20.

Shen, H.W., Toda, S., Moussawi, K., Bouknight, A., Zahm, D.S., and Kalivas,

P.W. (2009). Altered dendritic spine plasticity in cocaine-withdrawn rats.

J. Neurosci. 29, 2876–2884.

Shen, H., Moussawi, K., Zhou, W., Toda, S., and Kalivas, P.W. (2011). Heroin

relapse requires long-term potentiation-like plasticity mediated by NMDA2b-

containing receptors. Proc. Natl. Acad. Sci. USA 108, 19407–19412.

Shepherd, J.D., and Huganir, R.L. (2007). The cell biology of synaptic

plasticity: AMPA receptor trafficking. Annu. Rev. Cell Dev. Biol. 23, 613–643.

Stefanik, M.T., Moussawi, K., Kupchik, Y.M., Smith, K.C., Miller, R.L., Huff,

M.L., Deisseroth, K., Kalivas, P.W., and Lalumiere, R.T. (2013). Optogenetic

inhibition of cocaine seeking in rats. Addict. Biol. 18, 50–53.

Sun, X., Milovanovic, M., Zhao, Y., and Wolf, M.E. (2008). Acute and chronic

dopamine receptor stimulation modulates AMPA receptor trafficking in

nucleus accumbens neurons cocultured with prefrontal cortex neurons.

J. Neurosci. 28, 4216–4230.

Vocci, F., and Ling, W. (2005). Medications development: successes and

challenges. Pharmacol. Ther. 108, 94–108.

Wilson, S.J., Sayette, M.A., and Fiez, J.A. (2004). Prefrontal responses to drug

cues: a neurocognitive analysis. Nat. Neurosci. 7, 211–214.

Wolf, M.E. (2010). The Bermuda Triangle of cocaine-induced neuroadapta-

tions. Trends Neurosci. 33, 391–398.

Wolf, M.E., and Ferrario, C.R. (2010). AMPA receptor plasticity in the nucleus

accumbens after repeated exposure to cocaine. Neurosci. Biobehav. Rev. 35,

185–211.

Yang, Y., and Zhou, Q. (2009). Spine modifications associated with long-term

potentiation. Neuroscientist 15, 464–476.

Zhou, Q., Homma, K.J., and Poo, M.M. (2004). Shrinkage of dendritic spines

associated with long-term depression of hippocampal synapses. Neuron 44,

749–757.