Neuron, Vol. 16, 905±908, May, 1996, Copyright 1996 by Cell Press
Molecular and Cellular Aspects
of Nicotine Abuse
J ohn A. Dani*and Steve Heinemann²
*Division of Neuroscience
Baylor College of Medicine
Houston, Texas 77030-3498
²Molecular Neurobiology Laboratory
The Salk Institute
San Diego, California 92186-5800
conducting cations, and are in desensitized or inactive
states while unresponsive to agonist. The likelihood of
being in a particular state depends on many factors,
including the nAChR subtype, the agonist concentra-
tion, and the rate of agonist application. A rapid pulse
of agonist causes synchronized activation of nAChRs,
but long-term exposure to an agonist causes desensiti-
zation.A slow applicationof a lowagonistconcentration
can cause some desensitization without activation be-
cause the desensitized receptorhas a higheraffinity for
agonist than the resting or open receptor. In addition,
there is evidence that neuronal nAChRs can exist on
the cell surface as nonfunctional receptors (Margiotta
et al., 1987) or can enter long-lived inactivated states
(Lester and Dani, 1994).
The higher affinity of the desensitized receptor for
agonistand the changing distribution of nAChRs among
the various functional states must be considered to un-
derstand what takes place during sustained nicotine
use. A knowledgeof long-termforms of inactivationmay
be especially important for understanding the phases
of withdrawal symptoms and the development of toler-
ance to nicotine. Aspects of tolerance and withdrawal
could be explained by nicotinic receptors slowly recov-
ering to functional states from various levels of desensi-
tization and inactivation.
Pathways of Reward in Nicotine Abuse
A multiplicity ofpsychopharmacological effects contrib-
ute to the reinforcing actions of drugs. A widely ac-
cepted hypothesis is that drugs of abuse commandeer
existing reward pathways that are normally essential for
survival. Themesolimbic dopaminergic systemis known
to have an important role in mediating reward and contri-
butes to the rewarding effects of cocaine and d-amphet-
amine (Koob, 1992). Cocaine, for instance, is thought
to act by inhibiting the DA transporter; knockout mice
lacking the DA transporter are unaffected by the admin-
istration of cocaine (Giros et al., 1996). The most impor-
tant dopamine (DA) pathway originates in the ventral
tegmental area (VTA) of the midbrain and projects to
forebrain structures including the prefrontal cortex and
to limbic areas such as the olfactory tubercle, the amyg-
dala, the septal region, and the nucleus accumbens. A
range of studies using DA agonists and antagonists and
behavioral studies on the self-administration of drugs
after destruction of mesolimbic neurons have led to the
conclusion that DA release in the nucleus accumbens
is ªrewardingº or represents an encounter with reward
from the environment.
There is direct evidence that nicotine acts upon the
mesolimbic pathways. Autoradiography and in situ hy-
bridization indicate that multiple nAChR ? and ? sub-
units are present throughout these areas (Marks et al.,
1992; McGehee and Role, 1995; Wada et al., 1989). VTA
neurons have nAChRs located on their cell bodies and
on their terminals in the nucleus accumbens. Intermit-
tently administered nicotinic receptor agonists directly
excite VTA neurons, but during long exposure, the influ-
ence of the nicotinic agonists decreases (Calabresi et
al., 1989). Nicotine stimulates the release of DA in the
Tobaccouse indevelopedcountries has beenestimated
to cause nearly 20% of all deaths, making it the largest
single cause of premature death (Peto et al., 1992). The
drive for tobacco by humans is clear. The majority of
smokers havetried repeatedlyto quitand failed.Inabout
80% of the attempts to quit, smokers return to tobacco
in less than 2 years (Schelling, 1992). Although the un-
derlying mechanisms that cause tobacco abuse are not
wellunderstood, theaccumulation ofevidenceindicates
that nicotine is the primary component of tobacco that
motivates continued use despite harmful effects (Schel-
ling, 1992; Stolerman and Shoaib, 1991).
Nicotine alone, free of smoke or associated factors,
can elicit drug-seeking behavior in animal studies as
demonstrated by self-administration and place prefer-
ence experiments (Stolerman and Shoaib, 1991; Corri-
gall and Coen, 1989). Intravenous self-administration of
nicotine is best demonstrated under conditions of lim-
ited availability; rats have higher lever-pressing rates
whennicotineis delivered intermittently ratherthancon-
tinuously (Goldberg and Henningfield, 1988). The re-
sponding rate to nicotine is dose dependent, falling off
at both lower and higher concentrations. Responding
rates sometimes continue until rats experience toxic
effects. At these and higher concentrations, nicotine
causes vomiting, tremors, convulsions, and death at
extreme doses. The onset of aversive effects can com-
plicate the reinforcing effectiveness of nicotine when
compared with other drugs, which serve as reinforcers
over a wider range of test situations. The evidence is
clear, however, that like other addictive drugs, nicotine
reinforces self-administration, increases locomotor ac-
tivity, enhances reward from brainstimulation, and rein-
forces place preference (Clarke, 1991; Goldberg and
Henningfield, 1988; Stolerman and Shoaib, 1991).
Nicotinic Acetylcholine Receptors as the
Primary Site of Nicotine Action
Itislikely thatnicotinic acetylcholinereceptors (nAChRs)
are the initial sites of action for nicotine obtained from
tobacco. Understanding nicotine abuse will require
some knowledge of how these receptors functionwithin
the neuronal pathways that are relevant to addiction. A
nAChR normally binds acetylcholine (ACh) and under-
goes a conformational change that opens a cation-se-
lective channel for several milliseconds. Subsequently,
the ion channel closes, and the receptor may be refrac-
tory to agonist for many milliseconds or more.
Evidencefrom avariety of sources indicates that nico-
tinic receptors can exist in many different functional
states (Figure 1; Changeux etal., 1984). Nicotinic recep-
tors are largely in a closed (resting) state before agonist
arrives, are briefly in an open state while the channel is
increase. An increased number of nAChRs is not a re-
sponse that might at first be expected because chronic
exposure to an agonist usually produces excessive re-
ceptor activation; homeostasis is then achieved by
down-regulation of the receptors. Likewise, chronic ex-
posure to an antagonist produces receptor up-regula-
tionin many systems. These forms of self-regulation are
presumably mechanisms to maintain relatively normal
of abnormal receptor activity induced by endogenous
or exogenous agonists or antagonists.
A reasonable explanationforthe unexpected increase
in nAChRs is that low levels of nicotine cause signifi-
cant receptor desensitization, and over the long term,
nAChRs enter long-lasting inactive states (Lester and
Dani, 1994; Peng et al., 1994; Wonnacott, 1990). These
changes would enable some cholinergic systems to
move toward their initial levels of excitability even as
the numberof nAChRs increases dueto chronic nicotine
exposure. Direct support for this idea is provided by
the finding that high doses of the nAChR antagonist
mecamylamine also cause an increased number of
nAChRs. There is evidence that the number of surface
receptors increases whennAChRs enterparticularunre-
sponsive states. Interestingly, the number of nicotinic
receptors seems to be regulated by a posttranscrip-
tional mechanism that decreases nicotinic receptor turn-
over (Peng et al., 1994); the level of nAChR mRNA does
not seem to change (Marks et al., 1992).
In summary, there is support for the following model:
chronic exposure to low levels of nicotine induces
inactivation of some nAChRs, which then turn over
more slowly (Peng et al., 1994). Consequently, the num-
ber of nicotinic receptors on the surface of the mem-
brane increases. Depending on cholinergic activity and
changes in nicotine concentration in the brain, these
nAChRs will distribute among the various functional
states: resting,open, short-termdesensitized, andlong-
terminactivated. DifferentnAChR subtypes and particu-
lar cholinergic systems would be expected to recover
from inactivation to responsive states at different rates.
Although nAChR desensitization and inactivation
may underlie the increase in nAChRs, the cholinergic
systems are probably not relaxing back to the initial
condition present before nicotine exposure. In some
cases, cholinergic sensitivity has been shown to in-
crease after the number of nicotinic sites increases.
For example, in rats after chronic treatment, nicotine
induces greater magnitudes of conditioned placed pref-
erence and evokes greater DA release from striatal syn-
aptosomes (Shoaib et al., 1994; Wonnacott, 1990). In
other cases, however, cholinergic efficacy decreases.
Afterchronic nicotine, a singlepulse of nicotine induces
less prolactin release (Hulihan-Gublin et al., 1990),
evokes a smaller behavioral response in mice, and
evokes less DA release from mouse striatal synapto-
somes (Marks etal., 1993).These differences couldarise
from various factors. One factor might be multiphasic
recovery from inactivation by distinct nAChR subtypes
in separate areas of the brain. There are many subtypes
(McGehee and Role, 1995). Theoretically, there could
be a range of relaxation times as the various nAChR
Figure 1. A Representation of Functional States of a nAChR
The diagram represents a kinetic model, in which A is the agonist,
R is the resting or responsive state of the receptor, O is the ion-
conducting open state of the channel, D is a desensitized state, D?
is a deeper-level desensitized state, and I is an inactivated state.
Chronic exposure to the agonist nicotine could induce additional,
longer-lasting inactivated states. The arrows represent possible
transitions between states. The states in the top row leading to O
are visited for short times, and the open state lasts a few millisec-
onds. The transitions between the rows of states can be much
slower. Good estimates are not available for the lifetimes of the
inactivestates. Those states could last minutes orhours; the deeper
states of inactivation could last for days or could be essentially
irreversible. The population distribution of the states depends on
many factors, especially the agonist concentration and the rate of
application. The model is not intended to represent all the possible
modifications or kinetic properties or subtype diversity of nAChRs.
nucleus accumbens of freely moving rats andstimulates
release from synaptosomes isolated from various areas
including the nucleus accumbens (Clarke, 1991). Finally,
the role of the mesolimbic system in nicotine abuse is
supportedby the findings thatDA antagonists orlesions
ofthe nucleus accumbens reducenicotine self-adminis-
tration in rats (Corrigall et al., 1992; Stolerman and
It must be kept in mind, however, that there are other
reward pathways and that other compounds in tobacco
may affect the reward. For instance, monoamine oxi-
dase B (MAOB), which participates in the degradation
of DA, is partially inhibited in the brains of smokers
(Fowler et al., 1996). Although inhibitors of MAOB do
notseemto have addictivepotency, the increasedavail-
ability of DA to chronic smokers arising from MAOB
inhibitioncould enhancethe addictivepowerof nicotine.
Furthermore,otherpathways involvedinreward, inaddi-
tion to the mesolimbic pathway, could be affected di-
rectly and indirectly by nicotine, possibly contributing
to a myriad of reinforcing effects and learned behaviors.
Chronic Nicotine Use Increases
the Number of nAChRs
In addition to the possibility of an immediate effect on
the functional states of nAChRs, long-term nicotine ex-
posure causes an increase in the actual number of
nAChRs in humans, mice, and rats (Marks et al., 1992;
Wonnacott, 1990). This increase is specific to nicotinic
AChRs. Muscarinic ACh receptors, for instance, do not
subtypes distribute among functionalstates inresponse
to the changing concentration of nicotine.
Effect of Nicotine Delivered By Smoking
A basic question in the study of nicotine abuse is how
much nAChR activation and inactivation is caused by
a smoker's level of nicotine (Clarke, 1991; Wonnacott,
1990). A smoker can deliver small pulses of nicotine
into the arterial blood in the range of about 0.5 ?M
(Henningfield et al., 1993). It is appealing to speculate
that nicotine may be abused because the small peaks
of nicotine associated with each cigarette can activate
nAChRs and cause DA release. This activity leading to
DA release and an associated reward could be the main
mechanism that initiates nicotine abuse.
It also must be considered, however, that the peaks
of nicotine associated with each cigarette are superim-
posedonasteady-state nicotinelevelof?0.1?M, which
increases with repeated cigarette consumption during
the day because nicotine has a long half-life of about 2
hr (Benowitz et al., 1989; Russell, 1989). There is evi-
dence that steady levels of nicotine can cause signifi-
cant desensitization because, after one nicotine dose,
there develops an acute tolerance to a second dose
following within an hour.
Hypothesis for Sustained Nicotine Use
It is possible that nicotine-induced release of DA drives
tobacco usage, while inactivation of nAChRs by low
levels of nicotine may play a role in the processes of
tolerance and withdrawal. Presumably a regularsmoker
has an excess numberof nAChRs, butat the same time,
the smoker maintains a low level of nicotine that may
inactivate many of the nAChRs. After many hours of
abstinence (such as overnight), a smoker's nicotine lev-
els fall and the inactivated nAChRs begin to recover to
a responsive state with a rate that may be dependent
on the receptor subtype. As an excessive number of
nAChRs become responsive, there might be heightened
or abnormal potentiation of ordinary synaptic activity
in some nonrewarding cholinergic pathways that could
contribute to theagitation anddiscomfort (orwithdrawal
symptoms) that drive the smoker to the next cigarette.
That next dose of nicotine would have at least two
effects. First, after a night time of abstinence, a dose of
nicotine could be more rewarding than normal, either
by directly causing DA release within the mesolimbic
system or by acting elsewhere on other pathways that
provide reward or by indirectly activating reward path-
ways. This hypothesis is supported by reports from
smokers that they receive the most pleasurable impact
from the first cigarette of the day (Russell, 1989). Sec-
ond, afterthe initialreward (nicotine dose), a longerterm
effect of subsequent cigarettes (nicotine doses) could
be to desensitize the excess number of nAChRs back
to their usual state of inactivation for a regular smoker.
Thus, smokers report a relief from agitation and tension
afterthey have consumed nicotine. This relief fromwith-
drawal symptoms could be explained as follows: nico-
tine desensitizes the excess number of responsive
nAChRs in nonreward pathways back to a lower more
appropriate number of functional receptors for the
smoker. This hypothesis is supported by the finding that
nAChR antagonists can suppress drug-seeking behav-
ior(Corrigalland Coen, 1989;Corrigall etal., 1992; Gold-
berg and Henningfield, 1988), possibly decreasing the
Figure 2. A Hypothetical Cycle for Perpetuating Nicotine Use
The increased number of nAChRs and the subsequent pathology
of nicotinic cholinergic function is hypothesized to develop after
chronic use of nicotine. The simplified scheme is described in
drive for nicotine because the antagonist would be ex-
pected to inactivate the excess pool of functional
This modelofnicotine addictioncanbe tested inmore
detail by studying the activation and desensitization
mechanisms induced by nicotine. A number of other
questions need further investigation. What is the role
of the many nicotinic receptor subtypes? Do specific
nAChR subtypes mediate addiction? Is nicotine addic-
tion mediated directly by the same reward pathways
involvedinotherdrug addictions? Does chronic nicotine
induce long-term changes in the mesolimbic dopamin-
ergic system beyond the increased number of nAChRs
that havebeenseeninmany areas of thebrain, including
the ventral tegmentalarea and the nucleus accumbens?
A simplistic hypothesis can be put forward as a working
basis forresearch (Figure 2). Upon smoking a cigarette,
a small pulse of nicotine activates nAChRs that directly
or indirectly induce DA release that provides a pleasur-
able effect. It is likely that the mesolimbic dopaminergic
systemmediates at leastpartofthis reward.Withcontin-
ued use, nicotine builds up to a low steady-state con-
centration that causes significant nAChR desensitiza-
tion and (over time) longer-term inactivation. There is
evidence that nicotinic receptor turnoverdecreases fol-
lowing inactivation, leading to an increased number of
nAChRs, which subsequently may lead to nicotinic cho-
linergic systems that are pathological. In between ciga-
rettes, during sleep, or under conditions of abstinence
while attempting to stop smoking, nicotine levels drop
and a portion of the inactive nAChRs recover to a re-
sponsive state. Because of the increased number of
nAChRs thathavenow becomeresponsiveinthis patho-
logical condition, some cholinergic systems other than
the rewardpathways becomehyperexcitableto synapti-
cally released ACh, contributing to the drive forthe next
cigarette. Thus, smokers medicate themselves withnic-
otine to regulate the number of functional nAChRs.
Superimposed on this simplistic cycle of nicotine ex-
posure, there may be long-term synaptic changes that
result in the learned behaviors that are associated with
smoking and with the context in which smoking takes
place. Because these behaviors are reinforced by re-
peatedvariable rewards fromcigarettes (especially after
abstinence)and by associated sensory cues, the desire
for cigarettes extinguishes slowly and sometimes in-
completely. These factors coupled to the easy access
of cigarettes and constant advertising contribute to the
difficulty in breaking the nicotine habit.
Benowitz, N.L., Porchet, H., and J acob, P., III (1989). Prog. Brain
Res. 79, 279±287.
Calabresi, P., Lacey, M.G., and North, R.A. (1989). Br. J . Pharmacol.
Changeux, J .-P., Devillers-Thiery, A., and Chemouilli, P. (1984). Sci-
ence 225, 1335±1345.
Clarke, P.B.S. (1991). In Effects of Nicotine on Biological Systems,
F. Adlkofer and K. Thurau, eds. (Basel, Switzerland: Birkha Èuser Ver-
lag), pp. 285±294.
Corrigall, W.A., and Coen, K.M. (1989). Psychopharmacology 99,
Corrigall, W.A., Franklin, K.B., Coen, K.M., and Clarke, P.B. (1992).
Psychopharmacology 107, 285±289.
Fowler, J .S., Volkow, N.D., Wang, G.-J ., Pappas, N., Logan, J ., Mac-
Gregor, R., Alexoff, D., Shea, C., Schlyer, D., Wolf, A.P., Warner, D.,
Zezulkova, I., and Cilento, R. (1996). Nature 379, 733±736.
Giros, B., J aber, M., J ones, S.R., Wightman, R.M., and Caron, M.G.
(1996). Nature 379, 606±612.
Goldberg, S.R., and Henningfield, J .E. (1988). Pharmacol. Biochem.
Behav. 30, 227±234.
Henningfield, J .E., Stapleton, J .M., Benowitz, N.L., Grayson, R.F.,
and London, E.D. (1993). Drug Alcohol Depend. 33, 23±29.
Hulihan-Giblin, B.A., Lumpkin, M.D., and Kellar, K.J . (1990). J . Phar-
macol. Exp. Ther. 252, 21±25.
Koob, G.F. (1992). Trends Pharmacol. Sci. 13, 177±184.
Lester, R.A.J ., and Dani, J .A.(1994). Neuropharmacology 33, 27±34.
Margiotta,J .F., Berg, D.K., and Dionne, V.E. (1987). Proc. Natl.Acad.
Sci. USA 84, 8155±8159.
Marks, M.J ., Grady, S.R., and Collins, A.C. (1993). J . Pharmacol.
Exp. Ther. 266, 1268±1276.
Marks, M.J ., Pauly, J .R., Gross, S.D., Deneris, E.S., Hermans-Borg-
meyer, I., Heinemann, S.F., and Collins, A.C. (1992). J . Neurosci. 12,
McGehee, D.S., and Role, L.W. (1995). Annu. Rev. Physiol. 57,
Peng, X., Gerzanich, V., Anand, R., Whiting, P.J ., and Lindstrom, J .
(1994). Mol. Pharmacol. 46, 523±530.
Peto, R., Lopez, A.D., Boreham, J ., Thun, M., and Heath, C. (1992).
Lancet 339, 1268±1278.
Russell, M.A.H. (1989). Prog. Brain Res. 79, 289±302.
Schelling, T.C. (1992). Science 255, 430±433.
Shoaib,M., Stolerman, I.P., and Kumar, R.C. (1994).Psychopharma-
cology 113, 445±452.
Stolerman, I.P., and Shoaib, M. (1991). Trends Pharmacol. Sci. 12,
Wada, E., Waka, K., Boulter, J ., Deneris, E., Heinemann, S., Patrick,
J ., and Swanson, L.W. (1989). J . Comp. Neurol. 284, 314±335.
Wonnacott, S. (1990). Trends Pharmacol. Sci. 11, 216±218.