Neuron, Vol.
21, 467–476, September, 1998, Copyright 1998 by Cell Press
Neuroscience of Addiction
George F. Koob,* Pietro Paolo Sanna,
and Floyd E. Bloom
Department of Neuropharmacology
The Scripps Research Institute
La Jolla, California 92037
Human addictions are chronically relapsing disorders
characterized by compulsive drug taking, an inability
to limit the intake of drugs, and the emergence of a
withdrawal syndrome during cessation of drug taking
(dependence). The development of an addiction impacts
on several separate neurobiological processes, and
these effects are both drug- and drug use–dependent.
In animal models of addiction, changes in specific neu-
rotransmitter systems within a highly limited band of
structures, including specific parts of the nucleus ac-
cumbens and amygdala, may underlie drug reward and
the motivational effects associated with dependence.
Changes in the signals mediated by several neurotrans-
mitters, including dopamine, opioid peptides, and corti-
cotropin-releasing factor (CRF), and in the regulation of
selected transcription factors within the neurons of this
reward circuit, may underlie the vulnerability to relapse
that characterizes addiction in humans.
Animal Models of Addiction
Addiction, also known as substance dependence (Amer-
ican Psychiatric Association, 1994), is a chronically re-
lapsing disorder that is characterized by three major
elements: (1) compulsion to seek and take the drug, (2)
loss of control in limiting intake, and (3) emergence of
a negative emotional state (e.g., dysphoria, anxiety, irri-
tability) when access to the drug is prevented (defined
here as dependence) (Koob and Le Moal, 1997). Both
clinically and in experimental animals, the occasional
use of an abusable drug is distinct from repeated drug
use and the emergence of chronic drug addiction. The
goal of current neuroscience research is to understand
the cellular and molecular mechanisms that mediate the
transition between occasional, controlled drug use and
the loss of behavioral control over drug seeking and
drug taking that defines chronic addiction (Koob and Le
Moal, 1997).
Much of the recent progress in understanding the
mechanisms of addiction has derived from the develop-
ment of animal models of addiction on specific drugs
such as opiates, stimulants, and ethanol. These animal
models have localized the synaptic sites and transduc-
tive mechanisms in the nervous system on which drugs
of abuse act initially and are beginning to be used to
explore how the nervous system adapts to drug use.
While no animal model of addiction fully emulates the
human condition, animal models do permit investigation
of elements of the process of drug addiction, and the
data derived from these models provides an empirical
* To whom correspondence should be addressed.
Review
framework for understanding the molecular basis of ad-
diction (Koob and Le Moal, 1997; Koob et al., 1998).
The underlying molecular and cellular changes that
occur with the transition from occasional drug use to
pathological abuse and addiction are only partially un-
derstood as yet. To reach a more complete understand-
ing, these events will have to be integrated with the
animal models of the different elements of the addiction
process, including models of the transition from simple
drug taking to compulsive use at the molecular, cellular,
and behavioral levels. In this review, we will focus in
particular on the factors that drive drug-seeking behav-
ior at different stages of the addiction cycle (Koob and
Le Moal, 1997), and we will place particular emphasis
on trying to identify what is currently known and what
remains to be elucidated.
Neurobiological Substrates for the Acute
Reinforcing Effects of Drugs of Abuse
Animals and humans will readily self-administer the
same classes of drugs, and such self-administration
defines these drugs as positive reinforcers (Headlee et
al., 1955). The powerful reinforcing properties of such
drugs are revealed by the efforts experimental animals
will perform to get them, such as pressing a lever multi-
ple times to receive an intravenous injection of the drug.
While early work focused on drug-taking behaviors in
dependent animals (largely primates), subsequent stud-
ies have replicated the same behaviors in nondepen-
dent animals (largely rodents) (Schuster and Thompson,
1969). Since rodents are more tractable experimentally,
they have provided much recent insight into the perti-
nent circuits and transductive mechanisms for the acute
reinforcing effects of drugs of abuse (Koob and Bloom,
1988).
Neuropharmacological studies have established an
important role for the dopaminergic system in the acute
reinforcing effects of cocaine (see Table 1) (Woolverton
and Johnson, 1992). The midbrain dopamine system is
composed of two major projections: the nigrostriatal
system, which projects from the substantia nigra to the
corpus striatum, and the mesocorticolimbic dopamine
system, which projects from the ventral tegmental area
(VTA) to the nucleus accumbens, olfactory tubercle,
frontal cortex, and amygdala. It is the mesocorticolimbic
system that has been primarily implicated in the reinforc-
ing actions of drugs of abuse. Psychostimulants such
as cocaine and d-amphetamine elevate extracellular do-
pamine by inhibiting reuptake of dopamine by the dopa-
mine transporter and, in the case of d-amphetamine,
also by promoting reverse transport of dopamine. Dur-
ing intravenous cocaine self-administration, increased
extracellular dopamine can be detected in the nucleus
accumbens using in vivo microdialysis (Pettit and Jus-
tice, 1989). In addition, selective destruction of meso-
corticolimbic dopamine neurons with the neurotoxin
6-hydroxydopamine (6-OHDA) eliminates cocaine self-
administration; similar decreases in cocaine self-admin-
istration occur when 6-OHDA lesions are restricted to
Neuron
468
Table 1. Neurobiological Substrates for the Acute Reinforcing
Effects of Drugs of Abuse
Drug of Abuse Neurotransmitter Sites
Cocaine and Dopamine Nucleus accumbens
amphetamines Serotonin Amygdala
Opiates Dopamine Ventral tegmental area
Opioid peptides Nucleus accumbens
Nicotine Dopamine Ventral tegmental area
Opioid peptides? Nucleus accumbens
Amygdala?
THC Dopamine Ventral tegmental area
Opioid peptides?
Ethanol Dopamine Ventral tegmental area
Opioid peptides Nucleus accumbens
Serotonin Amygdala
GABA
Glutamate
the dopamine fibers of the region of the nucleus accum-
bens (Roberts et al., 1980). Multiple receptor subtypes
exist for transducing the increase in extracellular dopa-
mine induced by psychomotor stimulants into behav-
ioral action. Antagonists for the dopamine D1, D2, and
D3 receptor subtypes all decrease the reinforcing prop-
erties of cocaine (Woolverton and Johnson, 1992; Caine
et al., 1995; Koob and Le Moal, 1997; Epping-Jordan et
al., 1998a).
The neuronal interaction responsible for cocaine rein-
forcement and the motivation to seek the drug appears
to reside within the nucleus accumbens (Chang et al.,
1994; Carelli and Deadwyler, 1996; Peoples et al., 1997).
Electrophysiological recordings in animals receiving in-
travenous cocaine by self-administration have identified
several patterns of neuronal responses in the nucleus
accumbens, all time-locked to the self-administered
drug infusion. One group of neurons fires just before
the lever press, and this anticipatory response may be
an initiation or trigger mechanism. A second group of
neurons appears to change firing rate only after the
cocaine infusion, and these neurons may represent the
direct effects of reinforcement (Carelli and Deadwyler,
1996). Other neurons fire in proportion to the interinfu-
sion interval between consecutive self-administration
responses (Peoples and West, 1996). However, there
appears to be a fourth type of neuronal firing pattern
unique to cocaine self-administration; these “cocaine-
specific cells” fire both before and after the cocaine-
reinforced response (Carelli and Deadwyler, 1996). Even
more intriguing is the observation that this subset of
neurons also fires to sensory stimuli (sounds or lights)
that have been experimentally paired with cocaine deliv-
ery. Nucleus accumbens neurons may therefore mediate
conditioned drug responses (Carelli and Deadwyler,
1996). Similarly, conditioned sensory stimuli are strong
elicitors of “craving” in cocaine-taking humans.
Recent studies using recombinant DNA techniques to
“knock out” specific genes involved in dopaminergic
neurotransmission may provide evidence of some re-
dundancy in the neurochemical basis of cocaine rein-
forcement. In a mouse strain in which the gene for the
dopamine transporter was disrupted by homologous re-
combination, psychostimulants failed to alter baseline
extracellular dopamine levels and failed to induce be-
havioral effects such as enhanced locomotor activity
and stereotypy (Giros et al., 1996). However, such dopa-
mine transporter–deficient mice still could be trained to
self-administer cocaine despite persistently high levels
of extracellular dopamine in dopaminergic terminal
fields, suggesting a more complex basis for the psycho-
stimulant reinforcement (Rocha et al., 1998). Besides
inhibiting the dopamine transporter, psychostimulants
also inhibit the reuptake of serotonin and noradrenaline,
which may contribute to their reinforcing actions possi-
bly in part by modulating dopamine neurotransmission
(Parsons et al., 1995; Tanda et al., 1997b).
Much like psychostimulants, opiate drugs are readily
self-administered intravenously by animals, and the sys-
temic and central administration of competitive opiate
antagonists will decrease opiate reinforcement (re-
viewed by Koob and Bloom, 1988; Di Chiara and North,
1992). The reinforcing actions of opiates appear to be
largely mediated by the m opioid receptor since selective
m antagonists decrease opioid reinforcement in a dose-
dependent manner (Negus et al., 1993). In addition, mor-
phine reinforcement is abolished in mice with a targeted
disruption of the m opioid receptor gene (Matthes et al.,
1996).
The reinforcing properties of opiates utilize the same
circuitry implicated in the actions of cocaine and am-
phetamine stimulants but may involve additional sites
of interaction (Koob and Bloom, 1988) (Table 1). Block-
ade of opioid receptors either in the VTA or the nucleus
accumbens will decrease heroin self-administration.
Furthermore, rats will lever press to administer opioid
peptides in their nucleus accumbens or VTA, and opiate
administration into these restricted brain regions will
reinforce drug-seeking behavior (reviewed by Di Chiara
and North, 1992; Shippenberg et al., 1992). Opiates, like
other drugs of abuse, increase dopamine release in the
nucleus accumbens (Di Chiara and Imperato, 1988; Pon-
tieri et al., 1995) (see below). However, the reinforcing
effect of opiates in the nucleus accumbens persists
when all dopamine projections there are destroyed, sug-
gesting that their reinforcing actions may involve both
a dopamine-dependent (VTA) and a dopamine-inde-
pendent (nucleus accumbens) mechanism (Koob and
Bloom, 1988).
Sedative-hypnotics, including ethanol, are thought to
produce their reinforcing actions through multiple neu-
rotransmitter systems (Engel et al., 1992) (Table 1). One
of the major sites proposed for ethanol reinforcement
is modulation of GABA receptors (Liljequist and Engel,
1982; Samson and Harris, 1992). GABA antagonists re-
verse many of the behavioral effects of ethanol (Liljequist
and Engel, 1982; Samson and Harris, 1992). Further-
more, the benzodiazepine R0 15–4513 (termed an in-
verse agonist because it produces effects opposite to
those of typical benzodiazepines) will reverse some of
the behavioral effects of ethanol, and dose-dependently
reduces oral ethanol self-administration in rats (Samson
and Harris, 1992). When potent GABA antagonists are
microinjected into the brain, the most effective site to
reduce ethanol consumption is the central nucleus of
the amygdala (Hyttia and Koob, 1995).
Ethanol reinforcement also appears to involve activa-
tion of brain dopamine systems. Acutely, ethanol con-
sumption or systemic injection reduces the firing rate
Review
469
of pars reticulata GABA neurons, which are thought to
exert an inhibitory control over VTA dopaminergic neu-
rons (Diana et al., 1993; Merue and Gessa, 1985), and
increases concentrations of dopamine in the extracellu-
lar compartment in the nucleus accumbens. Dopamine
receptor antagonists injected into the nucleus accum-
bens reduce lever pressing for ethanol in nondependent
rats, and basal extracellular dopamine levels are in-
creased in rats chronically consuming low doses of etha-
nol (reviewed by Koob et al., 1994b). However, virtually
complete 6-hydroxydopamine denervation of the dopa-
mine inputs to the nucleus accumbens fails to alter vol-
untary responding for ethanol, suggesting that nucleus
accumbens dopamine transmission may not be critical
for reinforcing actions of ethanol (Koob et al., 1994b).
Moreover, manipulations that change brain serotonin
synaptic availability can decrease ethanol intake (Sellers
et al., 1992). For example, serotonin reuptake blockers
decrease ethanol intake but so do antagonists of the
serotonin-3 and serotonin-2c receptors, suggesting a
complex interaction of serotonin function and ethanol
reinforcement (LeMarquand et al., 1994).
Opiate antagonists also reduce ethanol self-adminis-
tration in several animal models (Tabakoff and Hoffman,
1996). The central nucleus of the amygdala is particularly
sensitive to this opioid antagonism of oral ethanol self-
administration (Heyser, Roberts, and G. F. K., unpub-
lished data). Double-blind, placebo-controlled clinical
trials showed that naltrexone can significantly reduce
human ethanol consumption, frequency of relapse, and
“craving” for ethanol in detoxified alcoholics, suggesting
that the motivation to resume ethanol use may involve
opioid systems (Volpicelli et al., 1992). Finally, low doses
of ethanol will sensitively inhibit the NMDA subtype of
glutamate receptors; in drug discrimination studies, ani-
mals substitute glutamate antagonists for ethanol (Ta-
bakoff and Hoffman, 1996).
Intravenous self-administration of nicotine is also
blocked by dopamine antagonists and dopamine-selec-
tive lesions of the nucleus accumbens (Corrigall et al.,
1992; Dani and Heinemann, 1996). Nicotine withdrawal
can also be produced by opioid antagonists (Dani and
Heinemann, 1996). Nicotine thus may activate both the
dopamine system and some opioid peptide neurons in
the same neural circuitry associated with other drugs
of abuse (Corrigall et al., 1992) (see Table 1). Exactly
which type of nicotinic receptor subunit configuration
mediates the reinforcing effects of nicotine is not clear,
but neurons in the VTA and nucleus accumbens express
high levels of the a6, a2, and a3 subunits (Le Novere et
al., 1996).
Tetrahydrocannabinol (THC) shares effects in animal
models of drug reinforcement similar to those of other
drugs of abuse (Anthony et al., 1994). Upon acute admin-
istration, THC decreases reward thresholds in rats
(Gardner et al., 1988; Lepore et al., 1996), produces a
place preference in rats (Lepore et al., 1995), and, as a
synthetic THC analog, is intravenously self-administered
in mice (Fratta et al., 1997, Soc. Neurosci., abstract).
THC binds to the cannabinoid-1 receptor, which is
widely distributed throughout the brain but particularly
in the extrapyramidal motor system of the rat (Herken-
ham et al., 1990). THC activates the mesocorticolimbic
Table 2. Drug Effects on Thresholds for Rewarding Brain
Stimulation
Withdrawal
Acute from Chronic
Drug Class Administration Treatment
Psychostimulants ↓ ↑
(cocaine, amphetamines)
Opiates ↓ ↑
(morphine, heroin)
Nicotine ↓ ↑
Sedative-Hypnotics ↓ ↑
(ethanol)
THC ↓ ↑
dopamine system (Chen et al., 1991), and recent data
suggest that THC can selectively increase the release
of dopamine in the shell of the nucleus accumbens simi-
lar to other drugs of abuse (Tanda et al., 1997a).
Negative Reinforcement Associated with Addiction
Repeated drug use is thought to arise from the neuro-
chemical actions causing the positive reinforcing effects
of a drug. However, the transition from occasional drug
use to drug addiction has been thought to require an
additional source of reinforcement, the reduction of the
aversive (negative) emotional state arising from re-
peated use. Here, drug taking presumably removes the
dysphoria, anxiety, irritability, and other unpleasant feel-
ings produced by drug abstinence. Other somatic physi-
cal signs, such as tremor, temperature changes, and
sweating, which also reflect the state of dependence,
presumably have little if any motivating properties on
drug use (Koob, 1996; Koob and Le Moal, 1997). Indeed,
one of the defining features of drug addiction is thought
to be the establishment of such a negative emotional
state (Russell, 1976). Consistent with this hypothesis,
all major drugs of abuse have been found to produce a
negative emotional state in dependent humans during
acute abstinence. The combination of the positive rein-
forcing effects of the drugs with reduction of the nega-
tive emotional states of drug abstinence provides a pow-
erful motivational force for the compulsive drug taking
that characterizes addiction.
One likely mechanism for this negative emotional state
may be a reduction in brain reward function. In studies
employing intracranial self-stimulation behavior, to di-
rectly study brain reward circuits, animals that have
been made chronically dependent show increased re-
ward thresholds (i.e., decreased reward) during with-
drawal (reviewed by Koob et al., 1993; Koob, 1996; Koob
and Le Moal, 1997). These decreases in reward have
been observed following the withdrawal from psycho-
motor stimulants, opiates, ethanol, THC, and nicotine
and are dose related to the amount of drug that had
been administered before withdrawal (Epping-Jordan et
al., 1998b; Gardner and Vorel, 1998) (Table 2).
Common Neuropharmacological Elements
in Addiction
The molecular and cellular basis for these changes in
motivation to take drugs may reside in the neuroadapta-
tions arising in the same neural elements that mediate
Neuron
470
Table 3. Neurotransmitters Implicated in the Motivational
Effects of Withdrawal from Drugs of Abuse
↓ Dopamine
↓ Opioid peptides
↓ Serotonin
↓ GABA
↑ Corticotropin-releasing factor
the acute reinforcing actions of these drugs. Earlier work
suggested that these adaptations occur both within the
drug-sensitive reinforcement system and in additional
systems outside the drug-sensitive reward system (Koob
and Bloom, 1988). Several common elements have been
identified that appear to change with chronic drug ad-
ministration and could underlie (or mediate) the compul-
sive self-administration of all drugs (Koob, 1992; Koob,
1996; Koob and Le Moal, 1997) (Table 3).
Functional changes in the mesolimbic dopamine sys-
tem appear to be common to the chronic actions of all
drugs of abuse (Nestler, 1996). Drug discontinuation is
accompanied by biochemical and electrophysiological
evidence of decreases in dopamine function (Diana et
al., 1993; Nestler, 1996; Weiss et al., 1996). During with-
drawal from chronic ethanol administration, there is a
decrease in extracellular dopamine in the nucleus ac-
cumbens, and, concomitantly, neurons of the mesolim-
bic dopamine system show dramatic reductions in
spontaneous firing rates and patterns but not in the
number of spontaneously active neurons (Diana et al.,
1993; Weiss et al., 1996). These biochemical and electro-
physiological changes can be reversed by ethanol ad-
ministration. Another neurotransmitter system impli-
cated in the acute reinforcing actions of drugs of abuse
whose function is altered during the development of
dependence is the opioid peptide system, as seen with
either chronic opiate usage (Self and Nestler, 1995; Nest-
ler, 1996) or with other drugs of abuse (Di Chiara and
North, 1992). Here, the functional changes are manifest
by a dramatic increased sensitivity to opioid receptor
antagonists, probably mediated by alterations in opioid
receptor signal transduction (see below) (Nestler, 1996;
Widnell et al., 1996).
In addition, several neurotransmitter systems that are
not involved in the acute reinforcing effects of drugs of
abuse do appear to become involved following chronic
drug administration. These include several neuropep-
tides, notably dynorphin, neuropeptide FF (NPFF), and
CRF. Neurons containing the opioid peptide dynorphin
in the nucleus accumbens appear to be functionally
activated following chronic administration of cocaine
(Hyman, 1996), while anti-opioid neuropeptides such as
NPFF may be activated following chronic opiates (Malin
et al., 1990; Lake et al., 1992).
CRF, with its many actions on hormonal and behav-
ioral responses to stressors, may be a brain system en-
gaged during drug dependence by all drugs of abuse.
Activation of the pituitary adrenal axis long has been
recognized as a characteristic of drug dependence and
withdrawal in humans (Kreek, 1987). Rats treated re-
peatedly with cocaine, nicotine, and ethanol show signif-
icant anxiogenic-like responses following cessation of
chronic drug administration that are reversed with intra-
cerebroventricular administration of a CRF antagonist.
Microinjections into the central nucleus of the amygdala
of lower doses of a CRF antagonist also reversed the
anxiogenic-like effects of ethanol withdrawal. Similar
doses of the CRF antagonist injected into the amygdala
reversed the aversive effects of opiate withdrawal
(Koob, 1996). In vivo microdialysis studies have shown
an increase in extracellular CRF during ethanol, cocaine,
and THC withdrawal (Merlo-Pich et al., 1995; Rodriguez
de Fonseca et al., 1997; Richter and Weiss, personal
communication). Thus, CRF activation may be a com-
mon element in the development of drug dependence
and may contribute to motivational effects involving
such subjective symptoms as increased stress and neg-
ative affect (Koob, 1996).
The neurochemical changes highlighted above illus-
trate neuroadaptations common to all drugs of abuse
(Koob, 1996). However, even more intriguing is the pos-
sibility that a neuroanatomical entity termed the ex-
tended amygdala (Heimer and Alheid, 1991) (Figure 1)
may represent a common anatomical substrate for acute
drug reward and the negative effects of compulsive drug
administration on reward function. The extended amyg-
dala is comprised of the medial subregion of the nucleus
accumbens (termed the shell of the nucleus accum-
bens), the bed nucleus of the stria terminalis, and the
central nucleus of the amygdala (Heimer and Alheid,
1991). Heimer has noted that all of these regions share
certain cytoarchitectural and circuitry similarities (Heimer
and Alheid, 1991). The extended amygdala receives nu-
merous afferents from limbic structures, such as the
basolateral amygdala and hippocampus, and sends not
only efferents to the medial part of the ventral pallidum
but also a large projection to the lateral hypothalamus,
thus further defining the specific brain areas that inter-
face classical limbic (emotional) structures with the ex-
trapyramidal motor system.
The structures comprising the extended amygdala
may further define the neural substrates for the acute
reinforcing actions of drugs of abuse (Koob, 1992; Koob,
1996). Acute administration of all the major drugs of
abuse produces increases in extracellular levels of do-
pamine in the shell of the nucleus accumbens (Pontieri
et al., 1995). The ventromedial shell of the nucleus also
expresses high levels of the dopamine D3 receptor
mRNA (Diaz et al., 1995), and the shell of the nucleus
accumbens is particularly sensitive to the cocaine an-
tagonist activity of a dopamine D1 antagonist (Caine et
al., 1995). The central nucleus of the amygdala also has
a role in ethanol reinforcement. Microinjection of GABA
antagonists or opioid peptide antagonists into the cen-
tral nucleus can attenuate lever pressing for oral ethanol
(Hyttia and Koob, 1995; Heyser, Roberts, and G. F. K.,
unpublished data). Perhaps more intriguing is recent
evidence that the extended amygdala may be an impor-
tant substrate for the changes in the reward system
associated with dependence. In animals dependent on
ethanol, microinjections of previously ineffective doses
of a GABA agonist into the central nucleus of the amyg-
dala decreased ethanol self-administration (Roberts et
al., 1996), suggesting that the GABAergic system has
been altered to become more responsive to agonists
during the course of dependence. Thus, the same neuro-
chemical components in the extended amygdala in-
volved in acute drug actions may become compromised
Review
471
Figure 1. Horizontal Section of a Rat Brain Depicting the Principal Structures of the Extended Amygdala
These structures include the central nucleus of the amygdala, the shell part of the nucleus accumbens, and the bed nucleus of the stria
terminalis. The drugs listed below each structure refer to potential sites of action of drug reinforcement during the addiction cycle, either
positive or negative. Redrawn with permission from Heimer and Alheid (1991).
during the development of dependence. One may spec-
ulate that additional neurochemical systems also may
be engaged within the neurocircuitry of the extended
amygdala (Koob and Bloom, 1988), in an attempt to
overcome the chronic presence of the perturbing drug
and to restore normal function despite the presence of
drug. The changes in CRF on the central nucleus of
the amygdala observed in dependent animals during
withdrawal support this hypothesis (Koob et al., 1994a,
1994b) (see Table 1, Table 2, and Figure 1).
Molecular, Cellular, and System Adaptations
Associated with Brain Motivational Systems
In addition to the changes in neurochemical systems
that are common to the chronic administration of drugs
of abuse, there are molecular changes that may provide
not only the mechanism for the above neurochemical
changes but also the substrate for prolonged adaptation
to chronic drug administration. Differences in adaptive
responses at the molecular level may account for indi-
vidual differences in vulnerability to dependence to dif-
ferent drugs. The challenge for future research is to link
these changes in plasticity to motivational actions of
dependence (Figure 2).
Several molecular consequences of motivationally im-
portant adaptations to chronic cocaine or amphetamine
administration have been identified. For example, acti-
vation of D1-like receptors stimulates a cascade of
events, including activation of Gs proteins and increased
intracellular cyclic AMP formation that ultimately may
lead to phosphorylation of transcription factors such as
cyclic AMP response element–binding protein (CREB)
and to the induction of immediate-early genes (see Self
and Nestler, 1995; Hyman, 1996) (see Figure 3). An im-
portant role for the D1 receptor–cAMP–CREB pathway
in the neuroadaptation to chronic drug dependencies
is supported also by the recent evidence of effective
anti-cocaine actions from dopamine D1 antagonists and
effective anti-cocaine priming effects from D1 agonists
(i.e., ability of cocaine to reinstate responding in rats
that have stopped responding after cocaine has been
replaced by saline) (Self et al., 1996). Activation of Gi
proteins, linked to D2 receptors, also appears to be
involved in the acute effects of cocaine since pertussis
toxin, which inactivates Gi proteins, produces a dopa-
mine receptor antagonist-like effect on cocaine self-
administration (Self et al., 1994).
In the VTA, repeated administration of cocaine pro-
duces transient decreases in Gi proteins that may lead
to D2 receptor subsensitivity (Nestler, 1994; Self and
Nestler, 1995). More prolonged effects that persist up
to one month in the nucleus accumbens include a super-
sensitivity to D1-mediated responses (Henry and White,
1991), increased levels of adenylyl cyclase and protein
kinase A (PKA), decreased levels of Gi proteins (Self
and Nestler, 1995; Nestler, 1996), and a decrease in the
ability of cocaine to induce the immediate-early gene
c-fos, followed by the sustained expression of AP-1
transcription factor complexes with altered composition
(see below).
Chronic administration of cocaine also decreases the
levels of neurofilament proteins in the VTA (Self and
Nestler, 1995). Lower levels of neurofilament proteins
are associated with decreased axonal transport, and
this could decrease the amount of tyrosine hydroxylase
transported from the VTA to the dopamine nerve termi-
nals in the nucleus accumbens (Self and Nestler, 1995).
Neuron
472
Such changes could be responsible for short-term re-
ductions in extracellular levels of dopamine release dur-
ing drug withdrawal (Weiss et al., 1992), which then
could trigger upregulation of the cyclic AMP system
(Self and Nestler, 1995) (see above).
These adaptations within the mesolimbic dopamine
system, and its receptor systems, not only could change
the function of the dopamine system itself but also may
trigger a second major action, namely increased expres-
sion of protachykinin and prodynorphin mRNAs. Dynor-
phin peptides in the nucleus accumbens, in turn, may
decrease dopamine release via a presynaptic action on
k opioid receptors; k agonists are known to produce
aversive effects in rodents and humans (Hyman, 1996).
Thus, chronic administration of psychomotor stimulants
would induce prodynorphin gene expression in the nu-
cleus accumbens that opposes the effect of cocaine on
reward.
Chronic opiate administration is associated with lit-
tle evidence of changes in opioid peptide activity or
changes in the number of any of the known opioid recep-
tors (but see Mansour et al., 1995; Zadina et al., 1997).
However, there is strong evidence that a dramatic en-
hancement of sensitivity to the aversive effects of opioid
antagonists can occur in brain areas implicated in the
acute reinforcing effects of opiates (Stinus et al., 1990;
Self and Nestler, 1995; Nestler, 1996). The molecular
basis for this effect may be at the level of signal trans-
duction (see Figures 2 and 3). Acute administration of
morphine decreases adenylate cyclase activity in the
nucleus accumbens, whereas chronic morphine treat-
ment is associated with increases in second messenger
systems, including adenylate cyclase activity and PKA
(Nestler, 1996). The hypothesis that these effects con-
tribute to the motivational effects of opiates (e.g., toler-
ance and dependence) is supported by studies showing
that direct administration into the nucleus accumbens
of agents that inhibit Gai or activate protein kinase de-
creases the reinforcing effects of opiates (Self et al.,
1994). Chronic morphine administration also decreases
the level of the transcription factor CREB in the nucleus
accumbens. This provides a possible mechanism for
morphine to mediate the alterations in gene expression
that may underlie the long-term changes in motivational
systems associated with opioid dependence (Widnell et
al., 1996) (Figure 3). Genetic disruption of two of the
three types of CREB in mice dramatically reduced symp-
toms of morphine withdrawal and produced some re-
duction in tolerance to the analgesic effects of morphine
(Maldonado et al., 1996).
Figure 2. Schematic Diagram Showing the
Relationship between Different Levels of
Analysis in the Study of Addiction and the
Role of Neuroadaptive Processes
Neuroadaptive mechanisms have been hy-
pothesized to contribute to compulsive be-
havior and addiction by acting at different
levels of the spiraling cycle of the develop-
ment of dependence (Koob and Le Moal,
1997). Both sensitization and counteradapta-
tion may contribute to changes in hedonic
responsiveness and set-points.
Chronic ethanol exposure has been shown to compro-
mise dopamine and GABAergic systems that are linked
to the continued desire to consume ethanol (see above).
At the molecular level, chronic ethanol is associated
with decreases in the ability of ethanol to potentiate
GABA-stimulated Cl2 flux, decreased expression of the
a1 subunit of the GABA complex (Tabakoff and Hoffman,
1996) and other subunits, and increased expression of
the b subunit (Tabakoff and Hoffman, 1996). Chronic
ethanol also is associated with increases in specific
subunits (NR1 and NR2A) of NMDA receptors (Tabakoff
and Hoffman, 1996). The relationship of these molecular
changes in ethanol-receptive elements to the motiva-
tional substrates outlined above may ultimately involve
changes in the same transduction systems observed
for other drugs of abuse (Figures 2 and 3). A possible
anatomical substrate for the molecular action of GABA
on motivational aspects of ethanol dependence is the
amygdala, a brain region considered important in medi-
ating emotional behavior in general. GABAergic neurons
in the central amygdala also have been proposed to
mediate rapid changes in autonomic activity (Sun and
Cassell, 1993), and acute ethanol induces expression of
the immediate-early gene c-fos, a marker of neuronal
activation, in GABAergic neurons in the central amyg-
dala (Morales, Criado, and P. P. S., unpublished data).
Adaptation to chronic nicotine administration also
may occur in the same neuronal circuits associated with
its initial molecular actions, namely nicotine on nicotinic
acetylcholine receptors located in the brain mesolimbic
dopamine system. One recent hypothesis seeks to ex-
plain nicotine tolerance and dependence on the basis
of a prolonged stability in the desensitized state of the
nicotinic acetylcholine receptor following the initial stim-
ulation (or “activation”) (Dani and Heinemann, 1996).
Consistent with this desensitization hypothesis, long-
term nicotine exposure causes an increase in the actual
number of nicotinic acetylcholine receptors (Collins et
al., 1990).
Dani and Heinemann (Dani and Heinemann, 1996)
have proposed that nicotine stimulates the mesolimbic
dopamine system via activation of nicotinic acetylcho-
line receptors to produce the acute reinforcing effects
of nicotine; with continued use, the inactivation of these
receptors by desensitization would then lead to adaptive
tolerance. During abstinence, nicotine levels fall, and the
increased nicotinic acetylcholine receptors, throughout
the brain, begin to recover to a responsive state that
may be dependent on the receptor subtype. Engag-
ing nicotinic receptors in non-reward-related pathways
Review
473

Figure 3. Molecular Mechanisms of Neuro-

adaptation
Drugs of abuse, by acting on neurotransmit-
ter systems, affect the phenotypic and func-
tional properties of neurons through the gen-
eral mechanisms outlined in the diagram.
Shown are examples of ligand-gated ion
channels (1) such as the GABAA and the gluta-
mate NMDA receptor (NMR) and G protein–
coupled receptors (R) such as opioid, dopa-
mine (DA), or the cannabinoid CB1 receptors,
among others (2). The latter is activated by
endogenous cannabinoids such as anan-
damide. These receptors modulate the levels
of second messengers like cAMP and Ca21
(3), which in turn regulate the activity of pro-
tein kinase transducers (4). Such protein ki-
nases affect the functions of proteins located
in the cytoplasm, plasma membrane, and nu-
cleus (5–8). Among membrane proteins af-
fected are ligand-gated and voltage-gated
ion channels (6 and 7). Ethanol, for instance,
has been proposed to affect the GABAA re-
sponse via PKC phosphorylation and, at least
in Purkinje cells, via PKA phosphorylation
(Tabakoff and Hoffman, 1996). Gi and G o
proteins also can regulate potassium and cal-
cium channels directly through their bg sub-
units (9). Protein kinase transduction path-
ways also affect the activities of transcription
factors (8). Some of these factors, like CREB,
are regulated posttranslationally by phos-
phorylation; others, like Fos, are regulated
transcriptionally; still others, like Jun, are
regulated both posttranslationally and/or
transcriptionally. While membrane and cyto-
plasmic changes may be only local (e.g., den-
dritic domains or synaptic boutons), changes
in the activity of transcription factors may result in long-term functional changes. These may include changes in gene expression of proteins
involved in signal transduction (10) and/or neurotransmission (11–13), resulting in altered neuronal responses. For example, chronic exposure
to psychostimulants or opiates has been reported to increase levels of PKA (10) and adenylyl cyclase (11) in the nucleus accumbens and to
decrease levels of Gai (11) (Self and Nestler, 1995; Nestler, 1996). Moreover, chronic ethanol induces differential changes in subunit composition
in the GABAA and in the glutamate inotropic receptors (12) and increases expression of voltage-gated calcium channels (VGCC) (13) (Tabakoff
and Hoffman, 1996). Chronic exposure to drugs also alters the expression of transcription factors themselves (14). CREB expression, for
instance, is depressed in the nucleus accumbens and increased in the locus coeruleus by chronic morphine treatment (Nestler, 1996; Widnell
et al., 1996), while chronic cocaine and other chronic treatments induce a transition from Fos induction to the induction of the longer-lasting
Fos-related antigens (FRAs) (reviewed by Hyman, 1996). The receptor systems depicted in the figure may not coexist in the same cells. The
systematic elucidation of their fine distribution and colocalization as well as the definition of the role of pre- and postsynaptic mechanisms
are current and future challenges in the neuroscience of addiction.

could contribute to the aversive emotional states associ-
ated with nicotine withdrawal. According to this hypoth-
esis, smokers would be, in effect, ultimately medicating
themselves with nicotine to regulate the number of func-
tional nicotinic acetylcholine receptors (Dani and Heine-
mann, 1996). Recent work shows that such changes
may be evident in the a subunit family, with the a4,
a2, and a7 subunits showing inactivation with chronic
nicotine but the a3 subunit (which is similar in struc-
ture to the a6 subunit) showing resistance to desensiti-
zation (Olale et al., 1997). Thus, one could speculate
that there are two powerful forces for development of
nicotine addiction: desensitization of nicotinic receptors
in non-reward pathways (self-medication) and resis-
tance to desensitization in reward pathways (positive
reinforcement).
Marijuana is a drug of abuse and addiction in humans,
and chronic high dose administration of THC or THC
analogs produces a dependence syndrome as mea-
sured by behavioral and subjective signs of withdrawal
(Aceto et al., 1995). In animals, a withdrawal syndrome
is precipitated by administration of competitive THC
antagonists to animals treated chronically with canna-
binoids (Aceto et al., 1995), and this precipitated canna-
binoid withdrawal produced activation of limbic struc-
tures as measured by c-fos activation and an increase
in extracellular levels of CRF in the central nucleus of
the amygdala as seen by other major drugs of abuse
(Rodriguez de Fonseca et al., 1997). Whether chronic
THC produces cellular and molecular changes in the
extended amygdala similar to other drugs of abuse will
be a challenge for future research.
Craving and Relapse
Because drug addiction is a chronic relapsing disorder,
the study of vulnerability to relapse is a particularly com-
pelling challenge. However, the biological basis for the
states of craving and protracted abstinence are difficult
to define. Vulnerability to reinstatement of drug-taking
Neuron
474
behavior and ultimately a continuation of compulsive
drug use presumably reflects some underlying pro-
longed perturbation that emerges long after the with-
drawal reactions have subsided. A residual deficit state
in the reward system, or sensitization of the reward
system to stimuli that predict drug effects, or both, could
be responsible for this vulnerability (Koob and Le Moal,
1997) (Figure 2).
Animal models of drug craving and relapse continue
to be developed and refined and are based largely on
conditioned reinforcement where previously neutral
stimuli, such as drug injection paraphernalia or specific
locations where drugs are obtained, have become
paired with the drug state or the drug withdrawal state
(Koob, 1995). The neural substrates for such conditioned
positive reinforcement may involve elements of the ex-
tended amygdala and its afferent circuits from the baso-
lateral amygdala (Everitt et al., 1991) and the mesolimbic
dopamine system. The neural substrates for any hypo-
thetical conditioned negative reinforcement are largely
unknown. However, by definition, any conditioning that
occurs during chronic drug use also involves the forma-
tion of subtle associations, but it is unknown whether
the associations formed by the pairing of drugs with
previously neutral stimuli utilize the same circuits as
those for memories not related to drugs.
A possible molecular mechanism for such learned,
long-term neuroadaptation may involve alterations to
gene expression (Figure 3). For example, Nestler and
coworkers have observed that acute cocaine induces
the transcription factor complex AP-1 in the striatum
and nucleus accumbens. With continued exposure to
drugs, the composition of AP-1 gradually changes from
one in which the immediate early-gene product, c-fos,
is present, to one in which a longer-lived, Fos-related
antigen (termed “chronic FRA”) substitutes for c-fos
(Hope et al., 1994). Because of the longer half-life of
chronic FRA, repeated exposure to cocaine leads to its
progressive accumulation (Hope et al., 1994). Shifts of
AP-1 composition in relevant CNS regions may repre-
sent a general neuroadaptive process that contributes
to long-term functional changes (Rossetti and Carboni,
1995).
Nicotine or cocaine self-administration also has been
shown to elicit overlapping patterns of c-fos induction
followed by induction of FRA immunoreactivity in the
nucleus accumbens and other regions, but not in the
amygdala (Merlo-Pich et al., 1997). A recent study
showed c-fos induction in the nucleus accumbens, cen-
tral amygdala, and bed nucleus of the stria terminalis
following acute exposure to a synthetic cannabinoid
agonist (Rodriguez de Fonseca et al., 1997). The differ-
ences in activation patterns of immediate-early genes
and related proteins, such as the FRAs, by different
classes of drugs within neurons of the extended amyg-
dala may underlie their specific mechanisms of action.
Taken together, these observations support the po-
tential contribution of alteration in gene expression to
the long-term neuroadaptive changes associated with
the motivational aspects of drug dependence. However,
of critical importance will be the correlation over time
of the specific biochemical changes that account for
increased vulnerability for drug seeking in defined com-
ponents of all aspects of the addiction cycle: acquisition,
maintenance, withdrawal, and relapse. A particular chal-
lenge for future studies in the neuroscience of addiction
will be to elucidate the neuroadaptive changes pro-
duced by chronic drug use in animal models of pro-
tracted abstinence and relapse. Presumably, the an-
swers will be found not only in the same molecular and
cellular elements of the neurochemical systems and
neurocircuitry responsible for the positive and negative
reinforcement associated with chronic drug use, but
also in the multiple neuroadaptive mechanisms that es-
tablish long-term memories of the drug rewards (Bailey
et al., 1996).
Acknowledgments
This is publication number 11130-NP from The Scripps Research
Institute. Research was supported by National Institutes of Health
grants DA04043, DA04398, and DA08467 from the National Institute
on Drug Abuse and AA06420 and AA08459 from the National Insti-
tute on Alcohol Abuse and Alcoholism. The authors would like to
thank Mike Arends for his valuable assistance with manuscript prep-
aration.
References
Aceto, M.D., Scates, S.M., Lowe, J.A., and Martin, B.R. (1995). Can-
nabinoid precipitated withdrawal by the selective cannabinoid re-
ceptor antagonist, SR 141716A. Eur. J. Pharmacol. 282, R1–R2.
American Psychiatric Association (1994). Diagnostic and Statistical
Manual of Mental Disorders, Fourth Edition (Washington, D.C.:
American Psychiatric Press).
Anthony, J.C., Warner, L.A., and Kessler, R.C. (1994). Comparative
epidemiology of dependence on tobacco, ethanol, controlled sub-
stances, and inhalants: basic findings from the National Comorbidity
Survey. Exp. Clin. Psychopharmacol. 2, 244–268.
Bailey, C.H., Bartsch, D., and Kandel, E.R. (1996). Toward a molecu-
lar definition of long-term memory storage. Proc. Natl. Acad Sci.
USA 93, 13445–13452.
Caine, S.B., Heinrichs, S.C., Coffin, V.L., and Koob, G.F. (1995).
Effects of the dopamine D1 antagonist SCH 23390 microinjected
into the accumbens, amygdala or striatum on cocaine self-adminis-
tration in the rat. Brain Res. 692, 47–56.
Carelli, R.M., and Deadwyler, S.A. (1996). Dual factors controlling
activity of nucleus accumbens cell-firing during cocaine self-admin-
istration. Synapse 24, 308–311.
Chang, J.Y., Sawyer, S.F., Lee, R.-S., and Woodward, D.J. (1994).
Electrophysiological and pharmacological evidence for the role of
the nucleus accumbens in cocaine self-administration in freely mov-
ing rats. J. Neurosci. 14, 1224–1244.
Chen, J.P., Paredes, W., Lowinson, J.H., and Gardner, E.L. (1991).
Strain-specific facilitation of dopamine efflux by delta 9-tetrahydro-
cannabinol in the nucleus accumbens of rat: an in vivo microdialysis
study. Neurosci. Lett. 129, 136–180.
Collins, A.C., Bhat, R.V., Pauly, J.R., and Marks, M.J. (1990). Modula-
tion of nicotine receptors by chronic exposure to nicotinic agonists
and antagonists. In The Biology of Nicotine Dependence, G. Bock
and J. Marsh, eds. (New York: John Wiley), pp. 87–105.
Corrigall, W.A., Franklin, K.B.J., Coen, K.M., and Clarke, P.B.S.
(1992). The mesolimbic dopaminergic system is implicated in the
reinforcing effects of nicotine. Psychopharmacology 107, 285–289.
Dani, J.A., and Heinemann, S. (1996). Molecular and cellular aspects
of nicotine abuse. Neuron 16, 905–908.
Di Chiara, G., and Imperato, A. (1988). Drugs abused by humans
preferentially increase synaptic dopamine concentrations in the
mesolimbic system of freely moving rats. Proc. Natl. Acad. Sci. USA
85, 5274–5278.
Di Chiara, G., and North, R.A. (1992). Neurobiology of opiate abuse.
Trends Pharmacol. Sci. 13, 185–193.
Review
475
Diana, M., Pistis, M., Carboni, S., Gessa, G.L., and Rossetti, Z.L.
(1993). Profound decrement of mesolimbic dopaminergic neuronal
activity during ethanol withdrawal syndrome in rats: Electrophysio-
logical and biochemical evidence. Proc. Natl. Acad. Sci USA 90,
7966–7969.
Diaz, J., Lévesque, D., Lammers, C.H., Griffon, N., Martres, M.-P.,
Schwartz, J.-C., and Sokoloff, P. (1995). Phenotypical characteriza-
tion of neurons expressing the dopamine D3 receptor in the rat
brain. Neuroscience 65, 731–745.
Engel, J.A., Enerback, C., Fahlke, C., Hulthe, P., Hard, E., Johannes-
sen, K., Svensson, L., and Soderpalm, B. (1992). Serotonergic and
dopaminergic involvement in ethanol intake. In Novel Pharmacologi-
cal Interventions for Alcoholism, C.A. Naranjo and E.M. Sellers, eds.
(New York: Springer), pp. 68–82.
Epping-Jordan, M.P., Markou, A., and Koob, G.F. (1998a). The dopa-
mine D1 receptor antagonist SCH 23390 injected into the dorsolat-
eral bed nucleus of the stria terminalis decreased cocaine reward
in the rat. Brain Res. 784, 105–115.
Epping-Jordan, M.P., Watkins, S.S., Koob, G.F., and Markou, A.
(1998b). Dramatic decreases in brain reward function during nicotine
withdrawal. Nature 392, 869–870.
Everitt, B.J., Morris, K.A., O’Brien, A., and Robbins, T.W. (1991). The
basolateral amygdala–ventral striatal system and conditioned place
preference: further evidence of limbic–striatal interactions underly-
ing reward-related processes. Neuroscience 42, 1–18.
Gardner, E.L., and Vorel, S.R. (1998). Cannabinoid transmission and
reward-related events. Semin. Neurosci., in press.
Gardner, E.L., Paredes, W., Smith, D., Donner, A., Milling, C., Cohen,
D., and Morrison, D. (1988). Facilitation of brain stimulation reward
by delta-9-tetrahydrocannabinol. Psychopharmacology 96, 142–144.
Giros, B., Jaber, M., Jones, S.R., Wightman, R.M., and Caron, M.G.
(1996). Hyperlocomotion and indifference to cocaine and amphet-
amine in mice lacking the dopamine transporter. Nature 379,
606–612.
Headlee, C.P., Coppock, H.W., and Nichols, J.R. (1955). Apparatus
and technique involved in a laboratory method of detecting the
addictiveness of drugs. J. Am. Pharm. Assoc. Sci. Edn. 44, 229–231.
Heimer, L., and Alheid, G. (1991). Piecing together the puzzle of
basal forebrain anatomy. In The Basal Forebrain: Anatomy to Func-
tion, T.C. Napier, P.W. Kalivas, and I. Hanin, eds. (New York: Plenum
Press), pp. 1–42.
Henry, D.J., and White, F.J. (1991). Repeated cocaine administration
causes persistent enhancement of D1 dopamine receptor sensitivity
within the rat nucleus accumbens. J. Pharmacol. Exp. Ther. 258,
882–890.
Herkenham, M., Lynn, A.B., Little, M.D., Johnson, M.R., Melvin, L.S.,
de Costa, B.R., and Rice, K.C. (1990). Cannabinoid receptor localiza-
tion in brain. Proc. Natl. Acad. Sci. USA 87, 1932–1936.
Hope, B.T., Kelz, M.B., Duman, R.S., and Nestler, E.J. (1994). Chronic
electroconvulsive seizure (ECS) treatment results in expression of
a long-lasting AP-1 complex in brain with altered composition and
characteristics. J. Neurosci. 14, 4318–4328.
Hyman, S.E. (1996). Addiction to cocaine and amphetamine. Neuron
16, 901–904.
Hyttia, P., and Koob, G.F. (1995). GABAA receptor antagonism in the
extended amygdala decreases ethanol self-administration in rats.
Eur. J. Pharmacol. 283, 151–159.
Koob, G.F. (1992). Drugs of abuse: anatomy, pharmacology, and
function of reward pathways. Trends Pharmacol. Sci. 13, 177–184.
Koob, G.F. (1995). Animal models of drug addiction. In Psychophar-
macology, The Fourth Generation of Progress, F.E. Bloom and D.J.
Kupfer, eds. (New York: Raven Press), pp. 759–772.
Koob, G.F. (1996). Drug addiction: the yin and yang of hedonic
homeostasis. Neuron 16, 893–896.
Koob, G.F., and Bloom, F.E. (1988). Cellular and molecular mecha-
nisms of drug dependence. Science 242, 715–723.
Koob, G.F., and Le Moal, M. (1997). Drug abuse: hedonic homeo-
static dysregulation. Science 278, 52–58.
Koob, G.F., Markou, A., Weiss, F., and Schulteis, G. (1993). Opponent
process and drug dependence: neurobiological mechanisms.
Semin. Neurosci. 5, 351–358.
Koob, G.F., Heinrichs, S.C., Menzaghi, F., Merlo-Pich, E., and Brit-
ton, K.T. (1994a). Corticotropin releasing factor, stress and behavior.
Semin. Neurosci. 6, 221–229.
Koob, G.F., Rassnick, S., Heinrichs, S., and Weiss, F. (1994b). Alco-
hol, the reward system and dependence. In Toward a Molecular
Basis of Alcohol Use and Abuse (Series Title: Experientia Supple-
mentum, Volume 71), B. Jansson, H. Jörnvall, U. Rydberg, L. Tere-
nius, and B.L. Vallee, eds. (Boston: Birkhauser-Verlag), pp. 103–114.
Koob, G.F., Carrera, M.R.A., Gold, L.H., Heyser, C.J., Maldonado-
Irizarry, C., Markou, A., Parsons, L.H., Roberts, A.J., Schulteis, G.,
Stinus, L., et al. (1998). Substance dependence as a compulsive
behavior. J. Psychopharmacol. 12, 39–48.
Kreek, M.J. (1987). Multiple drug abuse patterns and medical conse-
quences. In Psychopharmacology: the Third Generation of Progress,
H.Y. Meltzer, ed. (New York: Raven Press), pp. 1597–1604.
Lake, J.R., Hebert, K.M., Payza, K., Deshotel, K.D., Hausam, D.D.,
Witherspoon, W.E., Arcangeli, K.A., and Malin, D.H. (1992). Analog
of neuropeptide FF attenuates morphine tolerance. Neurosci. Lett.
146, 203–206.
LeMarquand, D., Pihl, R.O., and Benkelfat, C. (1994). Serotonin and
ethanol intake, abuse and dependence: findings of animal studies.
Biol. Psychiatry 36, 395–421.
Le Novere, N., Zoli, M., and Changeux, J.-P. (1996). Neuronal nico-
tinic receptor a6 subunit mRNA is selectively concentrated in cate-
choliminergic nuclei of the rat brain. Eur. J. Neurosci. 8, 2428–2439.
Lepore, M., Vorel, S.R., Lowinson, J., and Gardner, E.L. (1995). Con-
ditioned place preference induced by delta 9-tetrahydrocannabinol:
comparison with cocaine, morphine, and food reward. Life Sci. 56,
2073–2080.
Lepore, M., Liu, X., Savage, V., Matalon, D., and Gardner, E.L. (1996).
Genetic differences in delta 9-tetrahydrocannabinol-induced facili-
tation of brain stimulation reward as measured by a rate-frequency
curve-shift electrical brain stimulation paradigm in three different
rat strains. Life Sci. 58, PL365–PL372.
Liljequist, S., and Engel, J. (1982). Effects of GABAergic agonists
and antagonists on various ethanol-induced behavioral changes.
Psychopharmacology 78, 71–75.
Maldonado, R., Blendy, J.A., Tzavara, E., Gass, P., Roques, B.P.,
Hanoune, J., and Schutz, G. (1996). Reduction of morphine absti-
nence in mice with a mutation in the gene encoding CREB. Science
273, 657–659.
Malin, D.H., Lake, J.R., Hammond, M.V., Fowler, D.E., Rogillio, R.B.,
Brown, S.L., Sims, J.L., Leecraft, B.M., and Yang, H.-Y.T. (1990).
FMRF-NH2-like mammalian octapeptide: possible role in opiate de-
pendence and abstinence. Peptides 11, 969–972.
Mansour, A., Watson, S.J., and Akil, H. (1995). Opioid receptors:
Past, present and future. Trends Neurosci. 18, 69–70.
Matthes, H.W.D., Maldonado, R., Simonin, F., Valverde, O., Slowe,
S., Kitchen, I., Befort, K., Dierich, A., Le Meur, M., Dolle, P., et
al. (1996). Loss of morphine-induced analgesia, reward effect and
withdrawal symptoms in mice lacking the mu-opioid-receptor gene.
Nature 383, 819–823.
Merlo-Pich, E., Lorang, M., Yeganeh, M., Rodriguez de Fonseca, F.,
Raber, J., Koob, G.F., and Weiss, F. (1995). Increase of extracellular
corticotropin-releasing factor-like immunoreactivity levels in the
amygdala of awake rats during restraint stress and ethanol with-
drawal as measured by microdialysis. J. Neurosci. 15, 5439–5447.
Merlo-Pich, E., Pagliusi, S.R., Tessari, M., Talabot-Ayer, D., Hooft
van Huijsduijnen, R., and Chiamulera, C. (1997). Common neural
substrates for the addictive properties of nicotine and cocaine. Sci-
ence 275, 83–86.
Merue, G., and Gessa, G.L. (1985). Low doses of ethanol inhibit the
firing of neurons in the substantia nigra, pars reticulata: a GABAergic
effect? Brain Res. 360, 325–330.
Negus, S.S., Henriksen, S.J., Mattox, A., Pasternak, G.W., Porto-
ghese, P.S., Takemori, A.E., Weinger, M.B., and Koob, G.F. (1993).
Effect of antagonists selective for mu, delta and k opioid receptors
on the reinforcing effects of heroin in rats. J. Pharmacol. Exp. Ther.
265, 1245–1252.
Neuron
476

Nestler, E.J. (1994). Molecular neurobiology of drug addiction. Neu- Tabakoff, B., and Hoffman, P.L. (1996). Alcohol addiction: an enigma
ropsychopharmacology 11, 77–87. among us. Neuron 16, 909–912.
Nestler, E.J. (1996). Under siege: the brain on opiates. Neuron 16, Tanda, G., Pontieri, F.E., and Di Chiara, G. (1997a). Cannabinoid
897–900. and heroin activation of mesolimbic dopamine transmission by a
Olale, F., Gerzanich, V., Kuryatov, R., Wang, F., and Lindstrom, J. common mu1 opioid receptor mechanism. Science 276, 2048–2050.
(1997). Chronic nicotine exposure differentially affects the function Tanda, G., Pontieri, F.E., Frau, R., and Di Chiara, G. (1997b). Contri-
of human a3, a4, and a7 neuronal nicotinic receptor subtypes. J. bution of the noradrenaline carrier to the increase of extracellular
Pharmacol. Exp. Ther. 283, 675–683. dopamine in the rat prefrontal cortex by amphetamine and cocaine.
Parsons, L.H., Koob, G.F., and Weiss, F. (1995). Serotonin dysfunc- Eur. J. Neurosci. 9, 2077–2085.
tion in the nucleus accumbens of rats during withdrawal after unlim- Volpicelli, J.R., Alterman, A.I., Hayashida, M., and O’Brien, C.P.
ited access to intravenous cocaine. J. Pharmacol. Exp. Ther. 274, (1992). Naltrexone in the treatment of ethanol dependence. Arch.
1182–1191. Gen. Psychiatry 49, 876–880.
Peoples, L.L., and West, M.O. (1996). Phasic firing of single neurons Weiss, F., Markou, A., Lorang, M.T., and Koob, G.F. (1992). Basal
in the rat nucleus accumbens correlated with the timing of intrave- extracellular dopamine levels in the nucleus accumbens are de-
nous cocaine self-administration. J. Neurosci. 16, 3459–3473. creased during cocaine withdrawal after unlimited-access self-
Peoples, L.L., Uzwiak, A.J., Gee, F., and West, M.O. (1997). Operant administration. Brain Res. 593, 314–318.
behavior during sessions of intravenous cocaine infusion is neces-
Weiss, F., Parsons, L.H., Schulteis, G., Hyytia, P., Lorang, M.T.,
sary and sufficient for phasic firing of single nucleus accumbens
Bloom, F.E., and Koob, G.F. (1996). Ethanol self-administration re-
neurons. Brain Res. 757, 280–284.
stores withdrawal-associated deficiencies in accumbal dopamine
Pettit, H.-O., and Justice, J.B., Jr. (1989). Dopamine in the nucleus and 5-hydroxytryptamine release in dependent rats. J. Neurosci. 16,
accumbens during cocaine self-administration as studied by in vivo 3474–3485.
microdialysis. Pharmacol. Biochem. Behav. 34, 899–904.
Widnell, K., Self, D.W., Lane, S.B., Russell, D.S., Vaidya, V., Miseren-
Pontieri, F.E., Tanda, G., and Di Chiara, G. (1995). Intravenous co- dino, M.J.D., Rubin, C.S., Duman, R.S., and Nestler, E.J. (1996).
caine, morphine, and amphetamine preferentially increase extracel- Regulation of CREB expression: in vivo evidence for a functional
lular dopamine in the “shell” as compared with the “core” of the rat role in morphine action in the nucleus accumbens. J. Pharmacol.
nucleus accumbens. Proc. Natl. Acad. Sci. USA 92, 12304–12308. Exp. Ther. 276, 306–315.
Roberts, A.J., Cole, M., and Koob, G.F. (1996). Intra-amygdala mus- Woolverton, W.L., and Johnson, K.M. (1992). Neurobiology of co-
cimol decreases operant ethanol self-administration in dependent
caine abuse. Trends Pharmacol. Sci. 13, 193–200.
rats. Alcohol Clin. Exp. Res. 20, 1289–1298.
Zadina, J.E., Hackler, L., Ge, L.J., and Kastin, A.J. (1997). A potent
Roberts, D.C.S., Koob, G.F., Klonoff, P., and Fibiger, H.C. (1980).
and selective endogenous agonist for the m opiate receptor. Nature
Extinction and recovery of cocaine self-administration following
386, 499–502.
6-hydroxydopamine lesions of the nucleus accumbens. Pharmacol.
Biochem. Behav. 12, 781–787.
Note Added in Proof
Rocha, B.A., Fumagalli, F., Gainetdinov, R.R., Jones, S.R., Ator,
R., Giros, B., Miller, G.W., and Caron, M.G. (1998). Cocaine self- The data referred to as “Richter and Weiss, personal communica-
administration in dopamine-transporter knockout mice. Nature Neu- tion,” are now in press: Richter, R.M., and Weiss, F. (1998). In vivo
rosci. 1, 132–137. CRF release in rat amygdala is increased during cocaine withdrawal
Rodriguez de Fonseca, F., Carrera, M.R.A., Navarro, M., Koob, G.F., in self-administering rats. Synapse, in press.
and Weiss, F. (1997). Activation of corticotropin-releasing factor
in the limbic system during cannabinoid withdrawal. Science 276,
2050–2054.
Rossetti, Z.L., and Carboni, S. (1995). Ethanol withdrawal is associ-
ated with increased extracellular glutamate in the rat striatum. Eur.
J. Pharmacol. 283, 177–183.
Russell, M.A.H. (1976). What is dependence? In Drugs and Drug
Dependence, G. Edwards, M.A.H. Russell, D. Hawks, and M. Mac-
Cafferty, eds. (Lexington: MA, Lexington Books), pp. 182–187.
Samson, H.H., and Harris, R.A. (1992). Neurobiology of ethanol
abuse. Trends Pharmacol. Sci. 13, 206–211.
Schuster, C.R., and Thompson, T. (1969). Self-administration of and
behavioral dependence on drugs. Annu. Rev. Pharmacol. 9,
483–502.
Self, D.W., and Nestler, E.J. (1995). Molecular mechanisms of drug
reinforcement and addiction. Annu. Rev. Neurosci. 18, 463–495.
Self, D.W., Terwilliger, R.Z., Nestler, E.J., and Stein, L. (1994). Inacti-
vation of Gi and Go proteins in nucleus accumbens reduces both
cocaine and heroin reinforcement. J. Neurosci. 14, 6239–6247.
Self, D.W., Barnhart, W.J., Lehman, D.A., and Nestler, E.J. (1996).
Opposite modulation of cocaine-seeking behavior by D1- and D2 -
like dopamine receptor agonists. Science 271, 1586–1589.
Sellers, E.M., Higgins, G.A., and Sobell, M.B. (1992). 5-HT and etha-
nol abuse. Trends Pharmacol. Sci. 13, 69–75.
Shippenberg, T.S., Herz, A., Spanagel, R., Bals-Kubik, R., and Stein,
C. (1992). Conditioning of opioid reinforcement: neuroanatomical
and neurochemical substrates. Ann. NY Acad. Sci. 654, 347–356.
Stinus, L., Le Moal, M., and Koob, G.F. (1990). Nucleus accumbens
and amygdala are possible substrates for the aversive stimulus
effects of opiate withdrawal. Neuroscience 37, 767–773.
Sun, N., and Cassell, M.D. (1993). Intrinsic GABAergic neurons in
the rat central extended amygdala. J. Comp. Neurol. 330, 381–404.