Neuropharmacology 47 (2004) 2432

www.elsevier.com/locate/neuropharm

Molecular mechanisms of drug addiction

Eric J. Nestler 
Department of Psychiatry and Center for Basic Neuroscience, The University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard,
Dallas, TX 75390-9070, USA

Received 13 April 2004; received in revised form 1 June 2004; accepted 30 June 2004

Abstract

Regulation of gene expression is one mechanism by which drugs of abuse can induce relatively long-lasting changes in the brain
to cause a state of addiction. Here, we focus on two transcription factors, CREB (cAMP response element binding protein) and
DFosB, which contribute to drug-induced changes in gene expression. Both are activated in the nucleus accumbens, a major brain
reward region, but mediate dierent aspects of the addicted state. CREB mediates a form of tolerance and dependence, which
dampens an individuals sensitivity to subsequent drug exposure and contributes to a negative emotional state during early phases
of withdrawal. In contrast, DFosB mediates a state of relatively prolonged sensitization to drug exposure and may contribute to
the increased drive and motivation for drug, which is a core symptom of addictive disorders. A major goal of current research is
to identify the many target genes through which CREB and DFosB mediate these behavioral states. In addition, future work
needs to understand how CREB and DFosB, acting in concert with numerous other drug-induced molecular changes in nucleus
accumbens and many other brain regions, interact with one another to produce the complex behavioral phenotype that denes
addiction.
# 2004 Elsevier Ltd. All rights reserved.

Keywords: Cocaine; Opiates; CREB; #8710; FosB; Drug abuse

1. Introduction operate. Ultimately, such neural circuit changes under-

Drug addiction is dened today solely on the basis
of behavioral abnormalities, for example, loss of con-
trol over drug intake and compulsive drug taking
despite horrendous adverse consequences. These beha-
vioral abnormalities develop gradually and progress-
ively during a course of repeated exposure to a drug of
abuse, and can persist for months or years after discon-
tinuation of drug use. As a result, drug addiction can
be considered a form of drug-induced neural plasticity
(Nestler et al., 1993). The stability of the behavioral
abnormalities that dene addiction suggests a role for
gene expression in this process. According to this view,
repeated exposure to a drug of abuse alters the
amounts, and even the types, of genes expressed in spe-
cic brain regions. Such altered expression of genes
then mediates altered function of individual neurons
and the larger neural circuits within which the neurons

Tel.: +1-214-648-1111; fax: +1-214-648-4947.
E-mail address: eric.nestler@utsouthwestern.edu (E.J. Nestler).
0028-3908/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.neuropharm.2004.06.031
lie the behavioral abnormalities seen in drug addicts
(Nestler et al., 1993; Nestler, 2001a; Hyman and
Malenka, 2001).
There are many mechanisms by which repeated
exposure to a drug of abuse could alter gene expression
in the brain. These include altered rates of transcrip-
tion of genes, altered processing of primary RNA tran-
scripts into mature mRNAs, altered translation of
these mRNAs into proteins, altered processing of pro-
teins, and altered tracking of mature proteins to their
intracellular sites of action (see Nestler et al., 2001a).
Of all these mechanisms, the best understood, and the
one which has received most study to date, is the regu-
lation of gene transcription. According to this scheme,
illustrated in Fig. 1, drug perturbation of synaptic
transmission causes changes in numerous intracellular
signaling pathways, which eventually signal to the cell
nucleus, where specic proteins, called transcription
factors, are altered. Transcription factors bind to short
sequences of DNA located in the regulatory regions of
genes and thereby control the rate of gene transcription.
 E.J. Nestler / Neuropharmacology 47 Supplement No. 1 (2004) 2432
Fig. 1. Scheme showing that initial eects of drugs of abuse on their
acute synaptic targets, which are located extracellularly, produce
physiological responses via perturbation of postreceptor intracellular
messenger pathways that mediate extracellular mechanisms. Among
these physiological responses are alterations in gene expression medi-
ated via drug-induced changes in transcription factors. From Nestler
(2001b).
Over the past decade, drugs of abuse have been shown
to alter many types of transcription factors in a variety
of brain regions (see ODonovan et al., 1999; Berke and
Hyman, 2000; Nestler, 2001a; Mackler et al., 2003).
This review focuses on two transcription factors,
CREB (cAMP response element binding protein) and
DFosB, which our laboratory has investigated. It is
important to state at the outset that these particular
proteins represent just a small part of the overall plasti-
city that drugs of abuse induce in the brain. Neverthe-
less, consideration of these proteins serves as an illus-
tration of how reductionistic changes (alterations in a
single transcription factor in a particular neuronal cell
type in the brain) can be related to something as com-
plex as the behavioral phenotype of addiction.
2. Role of CREB in drug addiction
CREB and related proteins were described originally
as transcription factors that mediate eects of the
cAMP second messenger pathway on gene expression
(Shaywitz and Greenberg, 1999; Mayr and Montminy,
2001). This occurs via the phosphorylation of CREB
on a single serine residue, Ser133, by protein kinase A
(a protein kinase activated by cAMP). Once phos-
phorylated, CREB dimers, bound to specic CRE
(cAMP response element) sites on target genes, can
25
interact with the basal transcriptional complex to regu-
late gene transcription. More recently, several other
protein kinases, including Ca2+ activated and growth
factor activated kinases, have also been shown to phos-
phorylate Ser133 of CREB and to thereby regulate the
transcription of target genes through CREB (Shaywitz
and Greenberg, 1999; Mayr and Montminy, 2001).
The rst evidence that CREB mediates aspects of
addiction came from studies of the locus coeruleus
(Guitart et al., 1992), the major noradrenergic nucleus
in brain. This brain region normally regulates atten-
tional states, and is thought to play an important role
in mediating physical opiate dependence and with-
drawal (Koob et al., 1992; Nestler and Aghajanian,
1997; Williams et al., 2001). One mechanism respon-
sible for opiate physical dependence is an upregulation
of the cAMP pathway, which occurs in these neurons
as a consequence of repeated opiate exposure (Nestler
and Aghajanian, 1997). Several lines of evidence
directly implicate opiate-induced activation of CREB
in mediating this upregulation of the cAMP pathway in
the locus coeruleus (Lane-Ladd et al., 1997; Shaw-
Lutchman et al., 2002; Chao et al., 2002). This is gener-
ally consistent with the observation that mice decient
in CREB in all tissues show attenuated opiate physical
dependence and withdrawal (Maldonado et al., 1996).
However, while the locus coeruleus is important for
physical opiate dependence and withdrawal, it does not
seem to be crucial for the eects of drugs on the brains
reward and motivational systems, which are more criti-
cal for drug addiction. These latter regions include
several overlapping limbic circuits, comprising the
nucleus accumbens (ventral striatum), ventral tegmen-
tal area, amygdala, hippocampus, lateral hypothala-
mus, and frontal cortex, to name a few. Interestingly,
upregulation of the cAMP pathway and activation
of CREB occurs in many of these other regions as
originally established for the locus coeruleus (see Cole
et al., 1995; Carlezon et al., 1998; Berke and Hyman,
2000; Nestler, 2001a; Shaw-Lutchman et al., 2002,
2003; Walters et al., 2003). In several of these regions,
activation of CREB occurs not only in response to
chronic opiate exposure, but also in response to
chronic administration of other drugs of abuse, in
particular, stimulants and alcohol. A leading hypoth-
esis is that drug-induced activation of CREB in these
motivation centers of the brain underlies some of the
common core features of drug addiction seen clinically.
Our best understanding of the precise role played by
CREB in motivational aspects of addiction concerns
the nucleus accumbens. We have used herpes simplex
viral (HSV) vectors to induce local changes in CREB
activity within the nucleus accumbens of adult animals.
One vector (HSV-CREB) encodes wildtype or normal
CREB; this vector increases CRE-mediated transcrip-
tion upon injection into the brain. A second vector
26 E.J. Nestler / Neuropharmacology 47 Supplement No. 1 (2004) 2432
(HSV-mCREB) encodes a mutant form of CREB,
called mCREB, which is identical to wildtype CREB
except that it has an alanine residue substituted for
Ser133. mCREB can dimerize with itself or with
endogenous CREB family proteins and these dimers
can bind to CRE sites within target genes. However,
because it cannot be phosphorylated on Ser133, dimers
containing even one copy of mCREB cannot interact
with the basal transcriptional apparatus. Consequently,
when expressed at high enough levels, mCREB can
block the action of endogenous CREB and CREB-like
proteins. Indeed, HSV-mCREB decreases CRE-medi-
ated transcription upon injection into the brain. Using
these viral vectors, we have found that increased
CREB function (via HSV-CREB injection) in the
nucleus accumbens decreases an animals sensitivity to
the rewarding eects of morphine and cocaine, whereas
decreased CREB function (via HSV-mCREB injection)
has the opposite eect (Carlezon et al., 1998; Barrot
et al., 2002). These results suggest that drug-induced
activation of CREB in the nucleus accumbens repre-
sents a homeostatic, negative feedback adaptation
which decreases an individuals sensitivity to sub-
sequent drug exposure. In this way, CREB activation
could mediate a form of tolerance to a drugs reward-
ing eects.
Interestingly, the consequences of CREB activation
in the nucleus accumbens go far beyond modulation of
drug reward. Increased CREB function in this region
also dampens an animals interest for natural rewards,
such as sucrose drinking, while decreased CREB func-
tion enhances sucrose drinking (Barrot et al., 2002).
These results suggest that drug-induced activation of
CREB in the nucleus accumbens could decrease an
individuals interest in natural rewards, as is seen in
many human addicts.
Moreover, exposure of an animal to stress causes a
similar activation of CREB in the nucleus accumbens
as seen with drugs of abuse. Accordingly, we found
that increased CREB function in the nucleus accum-
bens decreases an animals responsiveness to a variety
of aversive or negative emotional stimuli, including
anxiogenic, stressful, and nociceptive stimuli (Pliakas
et al., 2001; Barrot et al., 2002). Conversely, decreased
CREB function in this region increases the animals
sensitivity to these conditions. Hence, CREB activation
in the nucleus accumbens would appear to mediate a
more general behavioral syndrome, characterized by
reduced sensitivity to any type of emotional stimulus,
regardless of whether the stimulus is rewarding or aver-
sive. In a physiological context, CREB activation in
response to mild stress could be seen as part of a cop-
ing mechanism to diminish sensitivity to further stress.
However, under more extreme (pathological) circum-
stances, CREB activation could contribute to a beha-
vioral syndrome of emotional numbing and anhedonia,
as seen both during drug withdrawal syndromes and
certain cases of depression and post-traumatic stress
disorder (Nestler et al., 2002).
All the above conclusions concerning CREBs func-
tioning at the level of the nucleus accumbens are based
on the use of viral vectors, which could produce con-
founding results due to the fact that all animals receive
intracranial surgery and viral infections. However,
early work with inducible transgenic mice, in which
CREB or mCREB is expressed with some selectivity in
the nucleus accumbens, support these conclusions.
Thus, mice overexpressing CREB show reduced sensi-
tivity to drugs of abuse and stress, whereas mice
expressing mCREB show increased sensitivity (Sakai
et al., 2002; Newton et al., 2002; McClung and Nestler,
2003). The fact that the behavioral phenotype of
CREB is reproduced with a completely independent
method of locally controlling CREB function in adult
animals gives us greater condence in our current
working hypothesis of CREB action. Traditional
knockout mice lacking CREB globally from early
stages of development also partially support this
hypothesis (Walters and Blendy, 2001).
These same types of approaches are now needed to
determine the role played by drug-induced activation
of CREB in the other reward-related brain regions
mentioned above. Early work suggests that CREB in
certain of these regions, amygdala, lateral hypothala-
mus, and ventral tegmental area, for example, may
subserve very dierent aspects of the drug addiction
phenotype compared with the nucleus accumbens
(Josselyn et al., 2001; Olson et al., 2001; Jentsch et al.,
2002; Georgescu et al., 2003).
As a transcription factor, CREB produces its beha-
vioral eects via the regulation of other genes. Numer-
ous genes are known to contain CRE sites within their
promoter regions (see Mayr and Montminy, 2001).
However, genes that are regulated by CREB in the
nucleus accumbens or elsewhere remain largely
obscure. Work to date has implicated the opioid pep-
tide, dynorphin, as one relevant target gene for CREB
in nucleus accumbens. Chronic administration of
cocaine or other stimulants induces dynorphin
expression in this region, and this induction is depen-
dent on CREB (Cole et al., 1995; Carlezon et al.,
1998). Dynorphin is known to dampen reward mechan-
isms in the nucleus accumbens via activation of
j opioid receptors (Shippenberg and Rea, 1997),
suggesting that CREB induction of dynorphin could
mediate at least some of CREBs behavioral eects
in this region. Indeed, the ability of CREB activation
to dampen cocaine reward, as well as responses to
stress, can be blocked by j opioid receptor antagonists
(Carlezon et al., 1998; Pliakas et al., 2001), which is
consistent with this model (Fig. 2). A major focus of
current research is to identify additional target genes
 E.J. Nestler / Neuropharmacology 47 Supplement No. 1 (2004) 2432
Fig. 2. Regulation of CREB by drugs of abuse. The gure shows a
VTA dopamine (DA) neuron innervating a nucleus accumbens
GABAergic projection neuron, specically, a subtype of such neurons
that expresses dynorphin (dyn). Dynorphin serves a negative feed-
back mechanism in this circuit: dynorphin, released from terminals of
the nucleus accumbens neurons, acts on j opioid receptors located on
nerve terminals and cell bodies of the DA neurons to inhibit their
functioning. Chronic exposure to cocaine or opiates upregulates the
activity of this negative feedback loop via upregulation of the cAMP
pathway, activation of CREB, and induction of dynorphin. From
Nestler (2001b).
through which CREB also acts in this and other brain
regions (McClung and Nestler, 2003).
3. Role of DFosB in drug addiction
DFosB is a member of the Fos family of transcrip-
tion factors. These proteins dimerize with a Jun family
member to form activator protein-1 (AP-1) transcrip-
tion factor complexes, which bind to AP-1 sites present
within the regulatory regions of certain genes (Morgan
and Curran, 1995). Acute administration of most types
of drugs of abuse causes the rapid and transient induc-
tion of several Fos and Jun proteins in the nucleus
accumbens and dorsal striatum (see Graybiel et al.,
1990; Young et al., 1991; Nestler et al., 2001b). The
transient nature of this induction is due to the fact that
the mRNAs for these proteins and the proteins them-
selves are highly unstable.
In contrast, DFosB is unique among these proteins
in that it is induced to only a small degree in response
to initial drug exposures. In addition, it is a highly
stable protein. As a result, during a course of repeated
drug administration, DFosB gradually accumulates,
and after a time becomes the predominant Fos protein
present in the nucleus accumbens and dorsal striatum
(Hope et al., 1994; Chen et al., 1995, 1997; Moratalla
et al., 1996; Hiroi et al., 1997) (Fig. 3). Also, because of
its stability, once induced DFosB persists in these
regions for weeks or months after discontinuation of
drug exposure. The accumulation of DFosB therefore
represents a novel mechanism by which chronic drug
exposure can lead to changes in gene expression that
27
Fig. 3. Scheme for the gradual accumulation of FosB versus the
rapid and transient induction of acute Fos family proteins in brain.
(A) Several waves of Fos-like proteins are induced in neurons by
acute stimuli. c-Fos is induced rapidly and degraded within several
hours of the acute stimulus, whereas other acute FRAs (Fos-
related antigens; e.g., FosB, DFosB, FRA-1, FRA-2) are induced
somewhat later and persist somewhat longer than c-Fos. The
chronic FRAs are biochemically modied isoforms of DFosB; they,
too, are induced (although at low levels) following a single acute
stimulus but persist in brain for long periods. In a complex with Jun-
like proteins, these waves of FRAs form AP-1 binding complexes
with shifting composition over time. (B) With repeated (e.g., twice
daily) stimulation, each acute stimulus induces a low level of DFosB.
This is indicated by the lower set of overlapping lines, which indicate
DFosB induced by each acute stimulus. The result is a gradual
increase in the total levels of DFosB with repeated stimuli during a
course of chronic treatment. This is indicated by the increasing step-
ped line in the graph. The increasing levels of DFosB with repeated
stimulation would result in the gradual induction of signicant levels
of a long-lasting AP-1 complex, which is hypothesized to underlie
persisting forms of neural plasticity in the brain. From Nestler
(2001b).
persist long after drug taking ceases (Nestler et al.,
2001a,b).
Recent evidence indicates that DFosBs unique stab-
ility is based on two mechanisms. First, DFosB lacks
the C-terminal domain of FosB. The terminal 21 amino
acids of this domain are also present in all other Fos
family proteins, and are partly responsible for the
instability of these proteins. Hence, the absence of this
motif from DFosB renders the protein more stable
(Carle et al., 2003). Second, DFosB is phosphorylated
on several serine residues by casein kinase II and
certain other protein kinases and such phosphorylation
further stabilizes the protein (Ulery et al., 2003).
Insight into the function of DFosB in the nucleus
accumbens and dorsal striatum comes from studies of
28 E.J. Nestler / Neuropharmacology 47 Supplement No. 1 (2004) 2432
inducible transgenic mice, where DFosB can be induced
selectively within these regions of adult mice (Kelz
et al., 1999). Studies to date indicate three major phe-
notypes in mice that overexpress DFosB. First, the ani-
mals show greater sensitivity to the behavioral eects
of cocaine in locomotor activity, conditioned place
preference, and self-administration assays (Kelz et al.,
1999; Nestler et al., 2001a,b; Colby et al., 2003).
Second, the mice seem to show enhanced incentive
motivation for cocaine. This conclusion is based on
studies in progressive ratio tests and in a model
of relapse which involves non-reinforced responding
(Nestler et al., 2001a,b; Colby et al., 2003). Third, the
mice show greater sensitivity to the behavioral eects
of morphine, and also show greater responding to
naturally reinforcing behaviors, including running and
eating (Nestler et al., 2001a, b; Werme et al., 2002;
Olausson et al., 2002; Zachariou et al., 2003).
Together, these ndings suggest that drug-induced
accumulation of DFosB in nucleus accumbens and dor-
sal striatum could be a mechanism of relatively pro-
longed sensitization to drug exposure. Moreover, the
ndings suggest that DFosB could be part of a sus-
tained molecular switch, which functions to rst induce
and later maintain a state of heightened incentive
motivation, perhaps even compulsiveness, toward rein-
forced behaviors (Nestler et al., 2001a,b). The obser-
vation that drugs of abuse more readily induce DFosB
in adolescent animals compared with adults raises the
possibility that DFosB induction may provide a mech-
anism for the greater vulnerability of younger animals
to develop addiction (Ehrlich et al., 2002).
Recent work, in a dierent mouse model, has pro-
vided further support for an important role for DFosB
in drug addiction. These mice enable the expression of
a dominant negative mutant of c-Jun, which functions
as an antagonist of DFosB and other Fos family
proteins, in nucleus accumbens and dorsal striatum.
These mice show reduced sensitivity to the rewarding
eects of cocaine and of morphine in conditioned place
preference assays (Peakman et al., 2003; Zachariou
et al., 2003). Along with the ndings from DFosB over-
expressing mice, these data suggest that DFosB is both
necessary and sucient for induction of a sensitized
behavioral state after drug exposure. Analysis of this
mouse in other behavioral assays is underway.
As with CREB, a major interest is to identify target
genes through which DFosB produces these behavioral
eects in nucleus accumbens and dorsal striatum.
Several putative target genes have been found. GluR2
is an AMPA glutamate receptor subunit. Over-
expression of DFosB induces GluR2 in nucleus accum-
bens, and cocaine induction of GluR2 in this region is
blocked upon expression of dominant negative c-Jun
(Kelz et al., 1999; Peakman et al., 2003). Moreover,
overexpression of GluR2 in nucleus accumbens, by use
of HSV vectors, causes increased sensitivity to the
behavioral eects of cocaine, indicating that induction
of GluR2 could account for at least some of DFosBs
behavioral phenotype. Since AMPA receptors contain-
ing a GluR2 subunit show reduced conductance com-
pared to those not containing GluR2 subunits, it is
possible that cocaine- and DFosB-mediated induction
of GluR2 could account for the reduced sensitivity of
nucleus accumbens neurons to glutamate, which has
been observed in electrophysiological studies (White
et al., 1995; Thomas et al., 2001). Regulation of GluR2
is likely to be just one of the many ways in which
chronic drug exposure alters glutamatergic trans-
mission in the nucleus accumbens (see Baker et al.,
2003; Yao et al., 2004).
Another putative target for DFosB is the cyclin-
dependent kinase, Cdk5. Cdk5 is enriched in neural
tissues, where it is implicated in the regulation of neu-
ronal growth and survival. The identication of Cdk5
as a target for DFosB came from DNA microarray stu-
dies (Chen et al., 2000). Subsequently, chronic cocaine
administration was shown to induce Cdk5 mRNA and
protein expression, as well as Cdk5 catalytic activity, in
nucleus accumbens and dorsal striatum (Bibb et al.,
2001). Based on Cdk5s hypothesized role in neural
growth, we considered the possibility that cocaine- and
DFosB-induction of Cdk5 could be involved in
cocaines demonstrated ability to induce dendritic
spines in nucleus accumbens neurons (Robinson and
Kolb, 1997). Indeed, infusion of roscovitine, a Cdk5
inhibitor, directly into the nucleus accumbens was
found to prevent the ability of cocaine to induce den-
dritic spines in this region (Norrholm et al., 2003). This
nding raises the notion that DFosB may be respon-
sible for this morphological plasticity and, moreover,
that some of DFosBs impact on the nucleus accum-
bens may be even more long-lasting than the protein
itself. One major question in the eld is what is reec-
ted by the cocaine-induced increase in dendritic spine
density on nucleus accumbens neurons. Such an
increase in spine density could be associated with the
strengthening of synapses and perhaps sensitized beha-
vioral responses to cocaine. However, it is just as poss-
ible that the increase in spine density could represent a
neuropathologic change in these neurons. Further work
is needed to answer this critical question.
In a similar way, we have little knowledge of the
behavioral consequences of Cdk5 induction in the
nucleus accumbens. Acute inhibition of Cdk5, by
infusion of an inhibitor such as roscovitine, increases
locomotor responses to cocaine (Bibb et al., 2001).
This suggests that activation of Cdk5 may normally
dampen sensitivity to cocaine. However, these ndings
must be viewed with caution, since they involve acute
disruption of Cdk5 and not the prolonged enhance-
ment of Cdk5 function as would be expected after
 E.J. Nestler / Neuropharmacology 47 Supplement No. 1 (2004) 2432 29

Fig. 4. Regulation of gene expression by cocaine: role of CREB and DFosB. Mice were treated with cocaine for ve days (A) or four weeks (B),
and RNA isolated from the nucleus accumbens was subjected to DNA microarray analysis. The Venn diagrams show the number of genes whose
regulation by short- or longer-term cocaine administration (blue circles) could be accounted for by DFosB (red) or CREB (yellow). The gure
illustrates an important role for CREB in earlier phases of cocaine exposure, and an increasingly dominant role for DFosB during more prolonged
periods of drug administration. From McClung and Nestler (2003) (For interpretation of the references to colour in this gure legend, the reader
is referred to the web version of this article).

chronic cocaine exposure. We are now in the process of
developing new experimental systems to test the func-
tional eects of such prolonged induction of Cdk5 in
this brain region.
Still other putative targets for DFosB have been
identied. DFosB induces nuclear factor-jB (NF-jB),
another transcription factor, in nucleus accumbens
(Ang et al., 2001). DFosB has also been shown to
inhibit the expression of the opioid peptide, dynorphin,
in this brain region. Many additional targets for
DFosB have been identied on DNA microarrays,
which now warrant further investigation (Chen et al.,
2000; McClung and Nestler, 2003).
A recent comparison of the gene expression changes
induced in nucleus accumbens by CREB or by DFosB
revealed interesting patterns of gene regulation
(McClung and Nestler, 2003). In general, CREB and
DFosB were found to induce opposite changes in the
expression of numerous genes, which is consistent with
the generally opposite behavioral phenotypes produced

Fig. 5. Scheme showing the opposite functional eects, and distinct temporal properties, of drug-induced adaptations mediated via CREB versus
DFosB. CREB appears to mediate an aversive or dysphoric state during early phases of withdrawal, whereas DFosB appears to mediate sensitiza-
tion to subsequent drug exposures at more distant withdrawal times. From Nestler (2001b).
30 E.J. Nestler / Neuropharmacology 47 Supplement No. 1 (2004) 2432
by these two transcription factors. In addition, while
CREBs contribution to the transcriptional eects of
cocaine predominated during shorter-term cocaine
administration, the contribution of DFosB increased
over time, such that after four weeks of cocaine more
than 25% of all the gene expression changes observed
could be accounted for by DFosB (Fig. 4). These nd-
ings support the dominant role played by DFosB in
mediated the eects of chronic cocaine exposure on
gene expression.
4. Conclusions
This review focuses on two transcription factors,
CREB and DFosB, as targets for drugs of abuse. Both
are induced in nucleus accumbens, among other brain
regions, but seem to function very dierently even in
that one region (Fig. 5). Chronic exposure to drugs of
abuse causes the activation of CREB in this region, but
this activation dissipates within a few days of coming
o the drug. In contrast, DFosB persists in brain for up
to two months, meaning that it mediates a much more
long-lived signal compared to CREB. In addition,
while CREB activation seems to mediate a state of
reduced reward and reduced emotional reactivity,
DFosB accumulation mediates a state of heightened
drug sensitivity and increased compulsion for reward-
ing behavior. We are still in relatively early stages of
identifying the many target genes in nucleus accumbens
through which CREB and DFosB produce these beha-
vioral eects. Moreover, we are just beginning to ana-
lyze the behavioral consequences of drug-induced
activation of CREB and DFosB in other brain regions
where these proteins are also induced by drug
exposure. Finally, these two proteins are among a very
large number of molecular changes that drugs of abuse
are known to cause in reward-related brain regions
after chronic administration. Ultimately, it will be criti-
cal to integrate this growing body of information to
understand how drugs, through plastic changes involv-
ing numerous proteins, induce a complex behavioral
phenotype that denes the addicted state.
Acknowledgements
Preparation of this review was supported by grants
from the National Institute on Drug Abuse. This
review is based, with permission, on Nestler, E.J., 2003.
Molecular mechanisms of drug addiction. In: D.S.
Charney (Ed.), Molecular Neurobiology for the Clin-
ician. American Psychiatric Press, pp. 107121.
References
Ang, E., Chen, J.S., Zagouras, P., Magna, H., Holland, J., Schaeer,
E., Nestler, E.J., 2001. Induction of NFjB in nucleus accumbens
by chronic cocaine administration. J. Neurochem. 79, 221224.
Baker, D.A., McFarland, K., Lake, R.W., Shen, H., Tang, X.C.,
Toda, S., Kalivas, P.W., 2003. Neuroadaptations in cystine-gluta-
mate exchange underlie cocaine relapse. Nat. Neurosci. 6,
743749.
Barrot, M., Olivier, J.D.A., Perrotti, L.I., Impey, S., Storm, D.R.,
Neve, R.L., Zachariou, V., Nestler, E.J., 2002. CREB activity in
the nucleus accumbens shell controls gating of behavioral respon-
ses to emotional stimuli. Proc. Natl. Acad. Sci. USA 99, 11435
11440.
Berke, J.D., Hyman, S.E., 2000. Addiction, dopamine, and the mole-
cular mechanisms of memory. Neuron 25, 515532.
Bibb, J.A., Chen, J.S., Taylor, J.R., Svenningsson, P., Nishi, A.,
Snyder, G.L., Yan, Z., Sagawa, Z.K., Ouimet, C.C., Nairn, A.C.,
Nestler, E.J., Greengard, P., 2001. Cdk5 regulates action of
chronic cocaine. Nature 410, 376380.
Carle, T., Ulery, P.G., Nestler, E.J., 2003. Absence of conserved Fos
family C-terminal domain in DFosB may contribute to its unique
stability. Soc. Neurosci. Abs. 29, 856.
Carlezon Jr., W.A., Thome, J., Olson, V.G., Lane-Ladd, S.B., Brod-
kin, E.S., Hiroi, N., Duman, R.S., Neve, R.L., Nestler, E.J., 1998.
Regulation of cocaine reward by CREB. Science 282, 22722275.
Chao, J.R., Ni, Y.G., Bolanos, C.A., Chen, J.S., Rahman, Z.,
DiLeone, R.J., Olson, V.G., Nestler, E.J., 2002. Characterization
of the mouse adenylyl cyclase type VIII gene promoter: regulation
by cAMP and CREB. Eur. J. Neurosci. 16, 12841294.
Chen, J.S., Nye, H.E., Kelz, M.B., Hiroi, N., Nakabeppu, Y., Hope,
B.T., Nestler, E.J., 1995. Regulation of DFosB and FosB-like pro-
teins by electroconvulsive seizure (ECS) and cocaine treatments.
Mol. Pharmacol. 48, 880889.
Chen, J., Kelz, M.B., Hope, B.T., Nakabeppu, Y., Nestler, E.J.,
1997. Chronic FRAs: stable variants of DFosB induced in brain
by chronic treatments. J. Neurosci. 17, 49334941.
Chen, J.S., Zhang, Y.J., Kelz, M.B., Steen, C., Ang, E.S., Zeng, L.,
Nestler, E.J., 2000. Induction of cyclin-dependent kinase 5 in hip-
pocampus by chronic electroconvulsive seizures: role of DFosB. J.
Neurosci. 20, 89658971.
Colby, C.R., Whisler, K., Steen, C., Nestler, E.J., Self, D.W., 2003.
DFosB enhances incentive for cocaine. J. Neurosci. 23, 24882493.
Cole, R.L., Konradi, C., Douglass, J., Hyman, S.E., 1995. Neuronal
adaptation to amphetamine and dopamine: molecular mechan-
isms of prodynorphin gene regulation in rat striatum. Neuron 14,
813823.
Ehrlich, M.E., Sommer, J., Canas, E., Unterwald, E.M., 2002. Peria-
dolescent mice show enhanced DeltaFosB upregulation in
response to cocaine and amphetamine. J. Neurosci. 22, 9155
9159.
Georgescu, D., Zachariou, V., Barrot, M., Mieda, M., Willie, J.T.,
Eisch, A.J., Yanagisawa, M., Nestler, E.J., DiLeone, R.J., 2003.
Involvement of the lateral hypothalamic peptide orexin in mor-
phine dependence and withdrawal. J. Neurosci. 23, 31063111.
Graybiel, A.M., Moratalla, R., Robertson, H.A., 1990. Ampheta-
mine and cocaine induce drug-specic activation of the c-fos gene
in striosome-matrix compartments and limbic subdivisions of the
striatum. Proc. Natl. Acad. Sci. USA 87, 69126916.
Guitart, X., Thompson, M.A., Mirante, C.K., Greenberg, M.E.,
Nestler, E.J., 1992. Regulation of CREB phosphorylation by
acute and chronic morphine in the rat locus coeruleus. J.
Neurochem. 58, 11681171.
Hiroi, N., Brown, J., Haile, C., Ye, H., Greenberg, M.E., Nestler,
E.J., 1997. FosB mutant mice: loss of chronic cocaine induction
of Fos-related proteins and heightened sensitivity to cocaines
 E.J. Nestler / Neuropharmacology 47 Supplement No. 1 (2004) 2432
psychomotor and rewarding eects. Proc. Natl. Acad. Sci. USA
94, 1039710402.
Hope, B.T., Nye, H.E., Kelz, M.B., Self, D.W., Iadarola, M.J.,
Nakabeppu, Y., Duman, R.S., Nestler, E.J., 1994. Induction of a
long-lasting AP-1 complex composed of altered Fos-like proteins
in brain by chronic cocaine and other chronic treatments. Neuron
13, 12351244.
Hyman, S.E., Malenka, R.C., 2001. Addiction and the brain: the
neurobiology of compulsion and its persistence. Nat. Rev. Neu-
rosci. 2, 695703.
Jentsch, J.D., Olausson, P., Nestler, E.J., Taylor, J.R., 2002. Stimu-
lation of protein kinase A activity in the rat amygdala enhances
reward-related learning. Biol. Psychiatry 52, 111118.
Josselyn, S.A., Shi, C.J., Carlezon Jr., ., W.A., Neve, R.L., Nestler,
E.J., Davis, M., 2001. Long-term memory is facilitated by CREB
overexpression in the amygdala. J. Neurosci. 21, 24042412.
Kelz, M.B., Chen, J.S., Carlezon, W.A., Whisler, K., Gilden, L.,
Beckmann, A.M., Steen, C., Zhang, Y.-J., Marotti, L., Self,
D.W., Tkatch, R., Baranauskas, G., Surmeier, D.J., Neve, R.L.,
Duman, R.S., Picciotto, M.R., Nestler, E.J., 1999. Expression of
the transcription factor DFosB in the brain controls sensitivity to
cocaine. Nature 401, 272276.
Koob, G.F., Maldonado, R., Stimus, L., 1992. Neural substrates of
opiate withdrawal. Trends Neurosci. 15, 186191.
Lane-Ladd, S.B., Pineda, J., Boundy, V., Pfeuer, T., Krupinski, J.,
Aghajanian, G.K., Nestler, E.J., 1997. CREB (cAMP response
element binding protein) in the locus coeruleus: biochemical,
physiological, and behavioral evidence for a role in opiate depen-
dence. J. Neurosci. 17, 78907901.
Mackler, S.A., Homan, Y.X., Korutla, L., Conti, A.C., Blendy, J.A.,
2003. The mouse nac1 gene, encoding a cocaine-regulated Bric-a-
brac Tramtrac Broad complex/Pox virus and zinc nger protein,
is regulated by AP1. Neuroscience 121, 355361.
Maldonado, R., Blendy, J.A., Tzavara, E., Gass, P., Roques, B.P.,
Hanoune, J., Schutz, G., 1996. Reduction of morphine abstinence
in mice with a mutation in the gene encoding CREB. Science 273,
657659.
Mayr, B., Montminy, M., 2001. Transcriptional regulation by the
phosphorylation-dependent factor CREB. Nat. Rev. Mol. Cell
Biol. 2, 599609.
McClung, C.A., Nestler, E.J., 2003. Regulation of gene expression
and cocaine reward by CREB and DFosB. Nat. Neurosci. 6,
12081215.
Moratalla, R., Elibol, B., Vallejo, M., Graybiel, A.M., 1996. Net-
work-level changes in expression of inducible FosJun proteins in
the striatum during chronic cocaine treatment and withdrawal.
Neuron 17, 147156.
Morgan, J.I., Curran, T., 1995. Immediate-early genes: ten years on.
Trends Neurosci. 18, 6667.
Nestler, E.J., 2001a. Molecular basis of long-lived neural plasticity to
drugs of abuse. Nat. Rev. Neurosci. 2, 119128.
Nestler, E.J., 2001b. Molecular neurobiology of addiction. Am. J.
Addict. 201, 201217.
Nestler, E.J., Aghajanian, G.K., 1997. Molecular and cellular basis of
addiction. Science 278, 5863.
Nestler, E.J., Hope, B.T., Widnell, K.L., 1993. Drug addiction: a
model for the molecular basis of neural plasticity. Neuron 11,
9951006.
Nestler, E.J., Hyman, S.E., Malenka, R.C., 2001a. Molecular Basis
of Neuropharmacology. McGraw-Hill, New York.
Nestler, E.J., Barrot, M., Self, D.W., 2001b. DFosB: a molecular
switch for addiction. Proc. Natl. Acad. Sci. USA 98, 11042
11046.
Nestler, E.J., Barrot, M., DiLeone, R.J., Eisch, A.J., Gold, S.J.,
Monteggia, L.M., 2002. Neurobiology of depression. Neuron 34,
1325.
31
Newton, S.S., Thome, J., Wallace, T., Shirayama, Y., Dow, A.,
Schlesinger, L., Duman, C.H., Sakai, N., Chen, J.S., Neve, R.,
Nestler, E.J., Duman, R.S., 2002. Inhibition of CREB or dynor-
phin in the nucleus accumbens produces an antidepressant-like
eect. J. Neurosci. 22, 1088310890.
Norrholm, S.D., Bibb, J.A., Nestler, E.J., Ouimet, C.C., Taylor, J.R.,
Greengard, P., 2003. Cocaine-induced proliferation of dendritic
spines in nucleus accumbens is dependent on the activity of the
neuronal kinase Cdk5. Neuroscience 116, 1922.
ODonovan, K.J., Tourtellotte, W.G., Millbrandt, J., Baraban, J.M.,
1999. The EGR family of transcription-regulatory factors: pro-
gress at the interface of molecular and systems neuroscience.
Trends Neurosci. 22, 167173.
Olausson, P., Jentsch, J.D., Nestler, E.J., Taylor, J.R., 2002.
Enhanced incentive motivation for food in DFosB overexpressing
mice. Soc. Neurosci. Abs. 28, 502.
Olson, V.G., Wolf, D., Russell, D., Hughes, T., Neve, R.L., Nestler,
E.J., 2001. Regulation of drug reward by CREB in ventral mid-
brain. Soc. Neurosci. Abs. 27, 977.
Peakman, M.C., Colby, C., Perrotti, L.I., Tekumalla, P., Carle, T.,
Ulery, P., Chao, J., Duman, C., Steen, C., Monteggia, L., Allen,
M.R., Stock, J.L., Duman, R.S., McNeish, J.D., Barrot, M., Self,
D.W., Nestler, E.J., Schaeer, E., 2003. Inducible, brain region
specic expression of a dominant negative mutant of c-Jun in
transgenic mice decreases sensitivity to cocaine. Brain Res. 970,
7386.
Pliakas, A.M., Carlson, R.R., Neve, R.L., Konradi, C., Nestler, E.J.,
Carlezon, Jr.., W.A., 2001. Altered responsiveness to cocaine and
increased immobility in the forced swim test associated with elev-
ated CREB expression in the nucleus accumbens. J. Neurosci. 21,
73977403.
Robinson, T.E., Kolb, B., 1997. Persistent structural modications in
nucleus accumbens and prefrontal cortex neurons produced by
previous experience with amphetamine. J. Neurosci. 17, 8491
8497.
Sakai, N., Thome, J., Chen, J.S., Kelz, M.B., Steen, C., Nestler,
E.J., Duman, R.S., 2002. Inducible and brain region specic
CREB transgenic mice. Mol. Pharmacol. 61, 14531464.
Shaw-Lutchman, T.Z., Barrot, M., Wallace, T., Gilden, L., Zachar-
iou, V., Impey, S., Duman, R.S., Storm, D., Nestler, E.J., 2002.
Regional and cellular mapping of CRE-mediated transcription
during naltrexone-precipitated morphine withdrawal. J. Neurosci.
22, 36633672.
Shaw-Lutchman, S.Z., Impey, S., Storm, D., Nestler, E.J., 2003.
Regulation of CRE-mediated transcription in mouse brain by
amphetamine. Synapse 48, 1017.
Shaywitz, A.J., Greenberg, M.E., 1999. CREB: a stimulus-induced
transcription factor activated by a diverse array of extracellular
signals. Annu. Rev. Biochem. 68, 821861.
Shippenberg, T.S., Rea, W., 1997. Sensitization to the behavioral
eects of cocaine: modulation by dynorphin and kappa-opioid
receptor agonists. Pharmacol. Biochem. Behav. 57, 449455.
Thomas, M.J., Beurrier, C., Bonci, A., Malenka, R.C., 2001. Long-
term depression in the nucleus accumbens: a neural correlate of
behavioral sensitization to cocaine. Nat. Neurosci. 4, 12171223.
Ulery, P.G., Carle, T., Nestler, E.J., 2003. Role of phosphorylation
on DFosB protein stability. Soc. Neurosci. Abs. 29, 777.
Walters, C.L., Blendy, J.A., 2001. Dierent requirements for cAMP
response element binding protein in positive and negative reinfor-
cing properties of drugs of abuse. J. Neurosci. 21, 94389444.
Walters, C.L., Kuo, Y.C., Blendy, J.A., 2003. Dierential distribution
of CREB in the mesolimbic dopamine reward pathway. J. Neu-
rochem. 87, 12371244.
Werme, M., Messer, C., Olson, L., Gilden, L., Thoren, P., Nestler,
E.J., Brene, S., 2002. DFosB regulates wheel running. J. Neurosci.
22, 81338138.
32 E.J. Nestler / Neuropharmacology 47 Supplement No. 1 (2004) 2432
White, F.J., Hu, X.T., Zhang, X.-F., Wolf, M.E., 1995. Repeated
administration of cocaine or amphetamine alters neuronal respon-
ses to glutamate in the mesoaccumbens dopamine system. J. Phar-
macol. Exp. Ther. 273, 445454.
Williams, J.T., Christie, M.J., Manzoni, O., 2001. Cellular and synaptic
adaptations mediating opioid dependence. Physiol. Rev. 81, 299
343.
Young, S.T., Porrino, L.J., Iadarola, M.J., 1991. Cocaine induces
striatal c-fos-immunoreactive proteins via dopaminergic D1 recep-
tors. Proc. Natl. Acad. Sci. USA 88, 12911295.
Yao, W.D., Gainetdinov, R.R., Arbuckle, M.I., Sotnikova, T.D.,
Cyr, M., Beaulieu, J.M., Torres, G.E., Grant, S.G., Caron,
M.G., 2004. Identication of PSD-95 as a regulator of dop-
amine-mediated synaptic and behavioral plasticity. Neuron 41,
625638.
Zachariou, V., Berton, O., Perrotti, L.I., Nestler, E.J., 2003. DFosB
in the nucleus accumbens modulates morphine reward, tolerance,
and dependence. Soc. Neurosci. Abs. 29, 112.