CNS Drugs 2004; 18 (15): 1105-1118

REVIEW ARTICLE 1172-7047/04/0015-1105/$31.00/0

 2004 Adis Data Information BV. All rights reserved.

Role of the Serotonergic System in the

Neurobiology of Alcoholism
Implications for Treatment
Bankole A. Johnson
University of Virginia Health System, Charlottesville, Virginia, USA

Contents
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1105
1. Serotonin Pathways: Anatomical Distribution and Functional Correlates . . . . . . . . . . . . . . . . . . . . . . 1106
2. SSRIs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1107
3. 5-HT1 Partial Agonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1108
4. 5-HT2 Antagonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1109
5. 5-HT3 Antagonists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1109
6. Other 5-HT Receptor Subtypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1112
7. Combining Serotonergic Agents with Other Types of Medication . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1112
8. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1114

Abstract Preclinical studies have contributed greatly to our understanding of the

neurochemical pathways associated with the development and maintenance of
alcohol-seeking behaviour. These studies have demonstrated the important role of
serotonin pathways, particularly as they relate to dopaminergic function, which
mediates alcohol-induced reward associated with its abuse liability. Naturally, this
has led to the study of serotonergic agents as treatments for alcoholism.
SSRIs do not appear to be effective treatment for a heterogeneous alcoholic
group. However, they may be useful as treatment for late-onset alcoholics, or
alcoholism complicated by comorbid major depression. Buspirone, a serotonin
5-HT1A partial agonist, does not appear to be an effective treatment for alcoholics
without comorbid disease. Buspirone may, however, have some utility for treating
alcoholics with comorbid anxiety disorder. The 5-HT2 antagonist ritanserin, at
pharmacologically relevant clinical doses, does not appear to be an effective
treatment for alcoholism. Ondansetron, a 5-HT3 antagonist, is an efficacious and
promising medication for the treatment of early-onset alcoholism. Preliminary
evidence suggests that combining the mu antagonist naltrexone with the 5-HT3
antagonist ondansetron promises to be more effective for treating alcoholism than
either alone.
The differential treatment effect of SSRIs and ondansetron among various
subtypes of alcoholic is intriguing. Future research is needed to understand more
clearly the molecular genetic differences and the interactions of such differences
with the environment that typify a particular alcoholic subtype. Such an under-
standing could enable us to make comfortable predictions as to which alcoholic
1106 Johnson

subtype might respond best to a particular serotonergic agent, which could then be
provided.

Serotonergic agents have been studied for more tion of the cortico-midbrain regions. Coarse, thick
than 20 years as potential treatments for alcoholism. beaded fibres with varicosed projections from the
Preclinical studies have implicated the serotonin median raphe innervate the hippocampus and septal
system, and especially its interaction with midbrain regions. Fine dorsal raphe projections typically in-
and cortical dopamine pathways, as being central to nervate the limbic (including the nucleus accum-
the expression and appreciation of the rewarding bens), striatal and cortical regions.[1,2]
effects of alcohol. Such studies have increased This organisational delineation of ascending pro-
knowledge about the bio-behavioural basis of alco- jections from serotonin cell bodies is interesting in
holism and have provided the foundation for testing terms of morphology and topological distribution.[3]
medications that interact with specific serotonin re- That is, not only do these projections differ morpho-
ceptor subtypes as treatment for particular types of logically, but also they vary in sensitivity and re-
alcoholic, with or without comorbid psychiatric dis- sponsiveness to various neurochemicals, an effect
order. that may be due to their possessing different distri-
The sections of this review are organised first to butions of serotonin receptor subtypes. Although
provide the scientific rationale for testing a particu- identification of serotonin receptor subtypes is in-
lar class of serotonergic medication for treating al- creasing, this review focuses on those that have been
coholism and, subsequently, to evaluate the reported the targets of medications for altering alcohol-seek-
efficacy of these agents in clinical studies. Finally, ing behaviour in both animals and humans. These
the rationale of ongoing studies that combine sero- include the serotonin 5-HT1 class (divided into four
tonergic medications with other agents to provide an additional subclasses A–D), which consists of both
added or synergistic therapeutic treatment effect is postsynaptic receptors and the presynaptic
discussed. These augmentation strategies hold the autoreceptor.
promise of developing more efficacious treatments
for alcoholism. The 5-HT1A receptor type is seen both post-
synaptically and as the autoreceptor. Postsynaptic
1. Serotonin Pathways: Anatomical 5-HT1 receptors (subclasses A, B and D) express
Distribution and Functional Correlates their function by increasing cyclic adenosine mono-
phosphate (cAMP) turnover; in contrast, the 5-HT1C
Mesocorticolimbic dopamine pathways do not receptor is linked with phosphatidyl inositol. 5-HT2
mediate the rewarding effects of alcohol in isolation. and 5-HT3 class receptors are all postsynaptic; how-
Importantly, there is solid evidence that midbrain ever, they are functionally different. 5-HT2 recep-
dopamine is itself modulated by a variety of neuro- tors are mediated by increasing turnover of the
transmitters, including serotonin. Indeed, these mid- second messenger phosphatidyl inositol. Serotonin
brain dopamine fibres are under the tonic inhibitory projections from the raphe synapse on dopamine
control of serotonin. Research has dramatically in- cells,[4,5] and 5-HT2 receptors are tonic inhibitors of
creased our understanding of serotonin receptors dopamine burst firing. 5-HT3 receptors exert their
and pathways and has improved knowledge of how effects through ligand-gated ion channels and are
the serotonin system mediates the expression of positive modulators of dopamine function. 5-HT3
numerous physiological and behavioural conditions, receptors also express their activity through co-lo-
including alcohol dependence. Of the nine serotonin calised neurotransmitters such as GABA and chole-
nuclei, the median and dorsal raphe and the centralis cystokinin, both of which also have indirect effects
posterior provide the most dense serotonin innerva- on dopamine function.[6] Finally, 5-HT4 receptors

 2004 Adis Data Information BV. All rights reserved. CNS Drugs 2004; 18 (15)
Alcoholism and the Serotonergic System 1107

are coupled positively to adenylate cyclase[7-9] and ginning on day 1 of treatment and increasing on
are distributed densely in the limbic system, includ- subsequent days of the 7-day treatment regimen.
ing the nucleus accumbens.[10-13] 5-HT4 receptors Levels of responding for ethanol returned to base-
facilitate serotonin-associated neurotransmission in line when fluoxetine treatment was stopped. Initial
hippocampal pyramidal cells,[11,14,15] thereby modu- suppression of food intake also returned to baseline
lating striatal dopamine release.[16,17] on subsequent treatment days. Thus, in animal
drinking models, fluoxetine reduces both alcohol
2. SSRIs reinforcement and general consummatory beha-
viour.
Lesions or pharmacological manipulations that
Despite these positive preclinical results, the
result in serotonin depletion have been associated
with decreased ethanol preference.[18,19] Preference clinical data on SSRIs as single agents for the treat-
and operant ethanol-reinforced behaviour paradigms ment of alcoholics without major depressive disor-
have been used to examine the effects of SSRIs in der have been inconsistent. Early studies with small
animal drinking models. sample sizes showed that SSRIs can produce short-
term (1–4 weeks) reductions in alcohol consumption
Generally, SSRIs decrease voluntary ethanol
among problem drinkers.[41-45] However, these stud-
consumption in preference paradigms.[20-24] Al-
ies had three additional limitations. Firstly, they
though SSRIs suppress ethanol consumption, there
were conducted predominantly in men,[41-43] thereby
is debate as to whether suppression of ethanol intake
reducing the generalisability of the results. Second-
is due to a reduction in the reinforcing effects of
ly, standardised psychosocial treatments, which can
ethanol, or to a generalised decrease in consumma-
tory behaviour.[25,26] The impact of SSRIs on food reduce the apparent effectiveness of putative thera-
intake and fluid consumption is complex.[27] Typi- peutic medications, were used infrequently; hence,
cally, SSRIs decrease food intake,[28,29] fluid con- nonspecific factors may be responsible for at least
sumption,[27] and palatability;[30] they may also de- some of the clinical improvement. Thirdly, the short
crease motivation toward consummatory behav- treatment period did not allow for determination of
iours.[31] Effects of SSRIs on food intake also appear whether SSRI-related clinical improvement in prob-
to be greater in well nourished compared with lem drinkers can be sustained. With longer SSRI
malnourished rats.[32] Notably, not only do SSRIs treatment periods and an alcohol-dependent popula-
enhance satiety,[26] but also they selectively reduce tion, there has been no clear evidence of efficacy.
preference for sweet items and carbohydrates.[33-35] For example, while Gorelick and Paredes[46] ob-
Food is, therefore, perceived as being less palatable served an initial dramatic decrease in alcohol con-
and rewarding.[36-38] Decreased general consumma- sumption (approximately 15%) during the first of 4
tory behaviour may, at least partially, mediate the weeks in a clinical trial, the overall treatment effect
effects of SSRIs on ethanol consumption in animals. with fluoxetine was no better than that with placebo.
Operant studies have also shown that SSRIs exert Naranjo et al.[47] were unable to demonstrate differ-
their effects on ethanol consumption through a com- ences in drinking between patients taking the SSRI
bination of CNS and peripheral effects. Haraguchi citalopram 40 mg/day or placebo after 12 weeks of
and colleagues[39] showed that same-day pretreat- treatment. Kabel and Petty[48] also did not find
ments with fluoxetine dose-dependently decreased fluoxetine 60 mg/day treatment for 12 weeks to be
ethanol responding. Thus, in animals for which etha- superior to placebo for reducing drinking among 28
nol serves as a reinforcer, classical extinction pat- alcohol-dependent men. Furthermore, Kranzler and
terns should be evident with abstinence. Corre- colleagues,[49] in a well executed 12-week study, did
spondingly, Murphy and colleagues[40] showed that not find fluoxetine 60 mg/day to be superior to
single-day fluoxetine infusion produced a signif- placebo in treating alcoholism. Interestingly, predi-
icant decrease in ethanol-reinforced responding be- cated on the findings of human laboratory studies

 2004 Adis Data Information BV. All rights reserved. CNS Drugs 2004; 18 (15)
1108 Johnson

that alcoholics with an early onset of disease may considerable inter-individual variation. Privette and
have reduced serotonin levels,[50-53] Kranzler and colleagues[63] also have reported that buspirone sig-
colleagues[54] re-examined their data. Presumably, nificantly reduced ethanol intake in Sprague-Daw-
their premise was that if type B (i.e. early-onset) ley rats induced to drink by repeated brain stem
alcoholics were deficient in serotonin, then SSRIs injection of tetrahydropapaveroline. In a group of
would be expected to be particularly beneficial for medium alcohol-preferring rats, buspirone
this alcoholic subtype. Paradoxically, they found (0.0025–0.63 mg/kg) suppressed, while buspirone
that fluoxetine worsened rather than improved the (>2.5 mg/kg) increased, alcohol consumption with-
clinical benefit of adjunctive cognitive behavioural out affecting water consumption.[64] Finally, bus-
therapy, and was certainly no better than placebo. In pirone may reduce stimulus-conditioned responding
contrast, Pettinati and colleagues[55] have shown that for ethanol.[65] Mechanistically, these effects are
sertraline appears to benefit type A (i.e. late-onset) supported by the finding that the relatively seroto-
alcoholics. Early-onset alcoholics differ from their nin-deficient fawn-hooded rat is predisposed to eth-
late-onset counterparts by having a stronger family anol consumption;[66] correspondingly, alcohol-pre-
history of alcoholism and greater propensity for ferring rats, compared with their non-alcohol-prefer-
impulse-dyscontrol, serotonergic dysfunction and ring counterparts, have reduced cortical and
antisocial behaviours.[56,57] It is therefore tempting subcortical grey matter serotonin content (see
to speculate that the relationship between serotonin Gongwer et al.[67] and McBride et al.[68] compared
dysfunction and the onset of alcoholism might not with Korpi et al.[69]). Thus, enhancing serotonin
be a simple deficiency state but rather due to differ- neurotransmission should decrease ethanol con-
ential expression of genotypic differences in the sumption. 5-HT1A receptors exhibit differential sen-
serotonin system. This topic is revisited (see below) sitivity, the greater effect being seen at the post-
in section 5 on 5-HT3 receptor antagonists. synaptic receptor compared with the autoreceptor.
Although outside the scope of this review article, Therefore, long-term buspirone administration aug-
SSRIs may be useful in treating alcoholics with ments serotonin function and down-regulates
suicidal tendencies and severe comorbid depres- autoreceptor function.[70] For these reasons, and de-
sion;[58] however, further studies are needed to con- spite the paucity of operant dose-response studies
firm these results. Among alcoholics with more examining the effects of buspirone on ethanol drink-
moderate levels of comorbid depression, TCAs re- ing, there appears to be some foundation for testing
duce dysphoric symptoms significantly; neverthe- its efficacy as a treatment agent for alcoholism.
less, these decreases in dysphoric symptoms are not
accompanied by similar or significant rates of dimi- Generally, clinical studies have not found bus-
nution of drinking behaviour.[59-61] pirone to be efficacious at treating alcoholics with-
In conclusion, SSRIs do not appear to be effec- out concurrent comorbid anxiety disorder. For ex-
tive treatment for a heterogeneous alcoholic group. ample, Malcolm and colleagues[71] did not find that
SSRIs may, however, be useful as treatment for late- buspirone was superior to placebo at reducing either
onset alcoholics, or for alcoholism complicated by the symptoms of anxiety or drinking among anxious
comorbid major depression. alcoholics. In a similarly well executed trial (n =
61), Kranzler and colleagues[72] did not find a signif-
3. 5-HT1 Partial Agonists icant effect of buspirone for reducing the amount of
alcohol consumed, the number of drinks per day, or
Buspirone, a 5-HT1A partial agonist, reduces eth- the number of drinking days; however, there was an
anol consumption in a variety of animal paradigms. effect of buspirone to increase significantly the in-
Collins and Myers[62] have shown that buspirone terval to the first heavy drinking day following
decreases volitional alcohol consumption by randomisation. In a review of five published studies,
30–60% in macaque monkeys; however, there was buspirone was found to have no convincing effect in

 2004 Adis Data Information BV. All rights reserved. CNS Drugs 2004; 18 (15)
Alcoholism and the Serotonergic System 1109

alcoholics; however, alcoholics with comorbid anxi- alcohol consumption were seen at study end in all
ety appeared to derive some benefit.[73,74] Finally, treatment groups. Ritanserin was, however, not su-
George et al.[75] showed that the younger the age of perior to placebo at improving drinking outcomes. A
onset, the lower the CSF level of 5-hydroxyindole- later study using a similar methodology demonstra-
acetic acid (5-HIAA), the major metabolite of sero- ted that even using a higher ritanserin dose (10 mg/
tonin. Despite the likelihood of serotonin facilitation day) did not improve drinking outcome compared
by long-term buspirone treatment, buspirone was with placebo.[87] Although it is plausible that even
not superior to placebo at improving drinking out- higher ritanserin doses may show effectiveness in
comes in early-onset alcoholics. Again, this demon- treating alcoholism, the potential to test this para-
strates that the apparent serotonin dysfunction digm is limited by the ability of ritanserin to produce
among early-onset alcoholics is not a simple defi- dose-dependent prolongation of the corrected QT
ciency but rather may be the result of other complex (QTc) interval on the ECG,[86] thereby increasing the
interactions within the serotonin system (see section potential for cardiac arrhythmias and the risk of
5 on 5-HT3 antagonists below for additional details). sudden death.
In sum, buspirone does not appear to be an effec- In conclusion, ritanserin at pharmacologically
tive treatment for alcoholics without comorbid dis- relevant clinical doses does not appear to be an
ease. Buspirone may, however, have some utility for effective treatment for alcoholism. There are, at
treating alcoholics with comorbid anxiety disorder. present, no clinical data on the use of other 5-HT2
antagonists as a treatment for alcoholism.
4. 5-HT2 Antagonists
5. 5-HT3 Antagonists
In a series of animal studies, ritanserin, a 5-HT2
receptor antagonist, has been shown to attenuate Of the numerous serotonergic agents that have
ethanol consumption (see Meert et al.[76] and Myers been suggested as pharmacotherapies for alcohol-
et al.[77] compared with Svensson et al.[78]). Also, the ism, the 5-HT3 antagonist ondansetron appears to be
5-HT2 antagonists amperozide[79-81] and FG particularly promising.[88] Preclinical studies using
5974[82,83] significantly suppress ethanol intake cloned cells and animals show that 5-HT3 receptors
without affecting water consumption. The magni- modulate important neurochemical and behavioural
tude of amperozide-induced ethanol suppression ap- effects of alcohol associated with its abuse liability.
peared relatively larger than that reported previously These studies laid the foundation for testing 5-HT3
for naltrexone.[81] Both amperozide and naltrexone antagonists for the treatment of alcoholism. For
diminish food palatability; hence, these medications instance, neurophysiological experiments show that
may, at least in part, also cause general decreases in ethanol potentiates 5-HT3 receptor-mediated ion
consummatory behaviour. Mechanistically, 5-HT2 currents in NCB-20 neuroblastoma cells[89,90] and
receptor antagonist-mediated acute anti-drinking be- human embryonic kidney 293 cells transfected with
haviour may be due to substitution for the 5-HT3RA complementary DNA.[91] 5-HT3 receptor
pharmacobehavioural effects of alcohol by in- antagonists block these effects.[92] Hence, the 5-HT3
creased levels of burst firing in mesocorticolimbic receptor is a site of action for ethanol in the
dopaminergic neurons[84] and, later, through recipro- brain.[93,94]
cal feedback inhibition of dopaminergic activity.[85] The rewarding effects of alcohol, and those of
We were the first to examine the efficacy of other abused substances, are mediated by mesocor-
ritanserin for the treatment of alcoholism.[86] In a ticolimbic dopamine pathways.[6,95-98] 5-HT3 recep-
rigorous multicentre clinical trial, 423 alcoholics tors are densely distributed in the terminals of
received 12 weeks of ritanserin (2.5 mg/day or 5 mg/ mesocorticolimbic dopamine-containing neurons,
day) or placebo as adjunctive medication to cogni- where they regulate dopamine release in these brain
tive behavioural therapy. Significant reductions in regions.[99-101] Thus, the rewarding effects of alcohol

 2004 Adis Data Information BV. All rights reserved. CNS Drugs 2004; 18 (15)
1110 Johnson

can be modulated by interactions between dopamine beneficial treatment for alcoholism, rigorous doub-
and 5-HT3 receptors in the midbrain and cor- le-blind clinical studies were needed to establish its
tex.[6,96,97,102,103] Demonstration that 5-HT3 receptor efficacy.
blockade reduces dopamine activity and, therefore,
Clinical trial evidence supports the efficacy of
the rewarding effects of alcohol and other abused
ondansetron for treating alcoholism. In a 6-week
drugs is derived from at least three different animal
double-blind trial of non-severely alcohol-depen-
paradigms. Firstly, 5-HT3 receptor antagonists at-
dent males (n = 71), Sellers and colleagues[117]
tenuate hyperlocomotion in the rat induced by dop-
showed that the 0.5-mg/day dose but not the 4-mg/
amine or ethanol injection into the nucleus accum-
day dose of ondansetron was associated with a non-
bens.[104] Secondly, 5-HT3 receptor antagonists sup-
significant trend (p = 0.06) toward a reduction in
press DiMe-C7 (a neurokinin)-induced hyper-
locomotion, an effect also diminished by the dop- alcohol consumption. Secondary data analysis to
amine antagonist fluphenazine.[105,106] Thirdly, exclude those who consumed more than ten drinks
5-HT3 receptor antagonists reduce alcohol con- per drinking day (n = 11) yielded a significant
sumption in a variety of animal models and across difference (p = 0.001) between the ondansetron 0.5
different species (see Meert,[64] Barnes and mg/day and placebo groups. Excluding this group
Sharp,[102] Costall et al.,[107] Hodge et al.,[108] Fadda can be justified by the fact that such individuals
et al.,[109] Rodd-Henricks et al.,[110] McBride and would typically not be enrolled in contemporary
Li,[111] and Tomkins et al.,[112] compared with clinical trials (and are more likely to have received
Beardsley et al.[113]). Essentially, the results of inpatient treatment) because their persistent level of
preclinical studies show that ondansetron is a prom- intoxication may have reduced the validity of their
ising medication for treating alcoholism. self-reported drinking. Nevertheless, as with other
preliminary studies, there are some limitations that
Human laboratory studies also have mainly sup-
should be taken into consideration when evaluating
ported a role for 5-HT3 antagonists in the treatment
these results. Firstly, ondansetron was provided for a
of alcoholism. In two independent studies, we were
relatively short treatment period (6 weeks rather
the first to show that ondansetron pretreatment at-
than the typical 12 weeks). Since the therapeutic
tenuates low-dose alcohol-induced positive subjec-
effects of ondansetron appear to take several weeks
tive effects, including the desire to drink.[103,114] That
is, ondansetron reduced the rewarding effects of to emerge, there may have been insufficient time to
alcohol associated with its abuse liability. Swift and demonstrate the full magnitude of the medication’s
colleagues,[115] using much higher alcohol and effects. Secondly, the relatively low subject num-
ondansetron doses, also observed that ondansetron bers would have decreased statistical power; hence,
pretreatment, compared with placebo, decreased al- the treatment differences observed in this small co-
cohol preference; however, a mixture of both stimu- hort may have been even more striking with a larger
lant and sedative interactions between ondansetron study population. Thirdly, the study cohort was lim-
and alcohol was also reported. In contrast, Doty and ited to mostly white males, thereby limiting the
coworkers[116] did not find an effect of ondansetron generalisability of the results. Despite these limita-
on alcohol-induced mood. These investigators may tions, these results were supportive of a treatment
have failed to detect the effects of ondansetron be- effect for ondansetron that required testing in a
cause administration of ondansetron via the intrave- large-scale clinical trial. Further, the study of Sellers
nous rather than the oral route, together with use of a and colleagues[117] indicated the need for future
group rather than an individual experimental setting, studies to determine whether ondansetron had an
may have reduced the sensitivity of their mood inverted U-shaped dose-response curve. Prospective
assessments. Thus, although three human laboratory empirical characterisation of what type of alcoholic
studies had suggested that ondansetron would be a responds best to ondansetron treatment was also

 2004 Adis Data Information BV. All rights reserved. CNS Drugs 2004; 18 (15)
Alcoholism and the Serotonergic System 1111

needed, given that psychosocial factors in the cohort problems compared with their late-onset alcoholic
had important effects on treatment outcome. counterparts.
In a recent large-scale (n = 321), randomised, In sum, our results show that ondansetron is
double-blind, placebo-controlled, 12-week clinical efficacious in treating early- but not late-onset al-
trial, we tested the efficacy of ondansetron (1, 4 and coholics, as exemplified by improved drinking out-
16 µg/kg twice daily) in alcohol-dependent men and comes and decreased alcohol craving.
women receiving weekly cognitive behavioural The differential effects of specific medication
therapy.[118] The first treatment week consisted of treatment for a particular subtype of alcoholic are
single-blind placebo followed by 11 weeks of doub- intriguing. Ondansetron improves the drinking out-
le-blind treatment. Thus, double-blind treatment comes of early-onset alcoholics but has no effect on
was not provided until study week 2. The cohort was late-onset alcoholics.[118] In contrast, SSRIs improve
about equally divided between early- and late-onset the drinking outcomes of late-onset[55] but not early-
alcoholics. The results of this study showed that onset[49] alcoholics. We have proposed an hypo-
early-onset alcoholics who received ondansetron (1, thesis, as yet untested, that these treatment differ-
4 and 16 µg/kg twice daily), compared with those ences may be related to variations in the interaction
who were administered placebo, had fewer drinks between allelic variants of the serotonin transporter
per day (1.89, 1.56 and 1.87 vs 3.30; p = 0.03, p = (SERT) and alcohol drinking in subtypes of alcohol-
0.01 and p = 0.02, respectively) and fewer drinks/ ic. Genotypic variation of the SERT 5′regulatory
drinking day (4.75, 4.28 and 5.18 vs 6.90; p = 0.03, p promoter region (5′-HTTLPR) on chromosome
= 0.004 and p = 0.03, respectively). Ondansetron (4 17p12 consists of three types,[121-123] the long (LL),
µg/kg twice daily) was superior to placebo at in- short (SS) and heterozygous (SL) forms. While nor-
creasing the percentage of days abstinent (70.10 vs mal individuals with the LL genotype experience
50.20; p = 0.02) and total days abstinent per study facilitated serotonin uptake,[124,125] the converse ap-
week (6.74 vs 5.92; p = 0.03). Among early-onset pears to be the case among chronic heavy drinkers.
alcoholics, there also was a significant difference in We have hypothesised that because (i) early-onset
the mean log carbohydrate-deficient transferrin ratio alcoholics may have greater likelihood of possess-
ing the LL variant of the SERT transporter (see
(a biochemical marker of heavy alcohol consump-
Johnson[126] for a review); (ii) those with the LL
tion) between those who received ondansetron (1
form of 5′-HTTLPR are more vulnerable to chronic
and 4 µg/kg twice daily) and those who received
alcohol-induced reductions in SERT density;[127]
placebo (–0.17 and –0.19 vs 0.12; p = 0.03 and p =
and (iii) decreased somatodendritic SERT density
0.01, respectively).[118] Of the 321 subjects in this
and function in early-onset alcoholics with the LL
study, we identified a cohort (n = 253) who had at variant of 5′-HTTLPR can be associated with re-
least 1 week of randomised medication (i.e. had duced serotonin firing rates (i.e. by increasing ex-
completed at least study week 2). In this group, we tracellular serotonin levels and correspondingly in-
examined whether ondansetron also had a differ- creasing autoreceptor-mediated inhibition), it is rea-
ential effect on the craving dimension derived by sonable to predict that chronically drinking early-
factor analysis of seven visual analogue scales. We onset alcoholics would have reduced trans-synaptic
found that ondansetron (4 µg/kg twice daily), com- serotonin neurotransmission and up-regulated
pared with placebo, was associated with a signif- 5-HT3 receptors. The therapeutic effects of ondanse-
icant decrease in the craving dimension.[119] More tron may, therefore, be associated with blockade and
recently, Kranzler and colleagues[120] have also normalisation of these up-regulated 5-HT3 recep-
shown that early-onset alcoholics treated with tors.[126,128] My colleagues and I recognise that alco-
ondansetron (4 µg/kg twice daily) had significantly holism is a multifactorial and complex disease and
better drinking outcomes and fewer alcohol-related that it is likely that other genotypes (even those not

 2004 Adis Data Information BV. All rights reserved. CNS Drugs 2004; 18 (15)
1112 Johnson

in the serotonin system) and environmental circum- nale is supported by four streams of scientific evi-
stances also may contribute to ondansetron respon- dence. Firstly, early-onset alcoholics may have ab-
siveness among early-onset alcoholics. However, normalities at both the opioid and serotonergic sys-
what we have proposed is a theoretical starting point tems;[57] thus, treating each of these abnormalities
from which to launch future research. with a selective agent can be expected to be asso-
ciated with a superior improvement in drinking out-
6. Other 5-HT Receptor Subtypes comes compared with treatment with either agent
5-HT4 receptor antagonists may play a role in alone.
alcohol-induced brain reward mechanisms.[129] It is, Secondly, evidence from preclinical research
therefore, of interest that Panocka and col- shows that there might be synergistic interactions
leagues[130] have shown that subcutaneous injection between the 5-HT3 and endogenous opioids in the
of the 5-HT4 antagonist GR113808 (1, 3 or 10 mg/ mesocorticolimbic dopamine reward system.[57]
kg) significantly reduced volitional ethanol intake. Binding studies demonstrate that 5-HT3 receptors
Additional animal studies are, however, needed to are highly concentrated within the mesocor-
establish this result. To date, no human studies on ticolimbic system, including the nucleus accum-
the effects of 5-HT4 antagonists on alcohol con- bens,[99,131,132] which receives its serotonergic inner-
sumption have been conducted. Relatively little is vation from the dorsal raphe nucleus.[133] The re-
known about the effects of medications that affect warding effects of alcohol through enhancement of
other serotonin receptor subtypes and drinking be- dopamine release in the nucleus accumbens are me-
haviour. diated through activation of 5-HT3 receptors.[134] It
is also known that alcohol stimulates the release of
7. Combining Serotonergic Agents with β-endorphins and enkephalins,[135] which also pro-
Other Types of Medication duce dopamine release through activation of mu and
Recently, there has been intense scientific and delta opioid receptors. Importantly, morphine can
clinical interest in combining therapeutic agents for increase ethanol intake,[136-138] and morphine-in-
the treatment of alcohol dependence. This is based duced increases in extracellular dopamine levels and
upon the hypothesis that multiple neurochemical place preference are attenuated by 5-HT3 antagon-
pathways may be dysregulated as either ‘state’ or ists.[139-142] Ondansetron also attenuates morphine-
‘trait’ effects of the drinking behaviour, and com- or TAN-67 (a selective delta opioid agonist)-in-
bining effective medications working at different duced enhancement of alcohol place preference, an
neurotransmitters may produce a synergistic or at effect that appears greater at the mu than at the delta
least an added therapeutic response. Although this opioid receptors.[143] Alcohol-induced increases in
hypothesis is intriguing, knowledge about how the dopamine levels can also be attenuated by both mu
various neurotransmitters interact in the living brain and delta opioid receptor antagonists.[144] Taken to-
of alcohol-dependent individuals and how this may gether, these data suggest that 5-HT3 receptors may
differ under different stages of the addiction is not themselves mediate alcohol reward via activation of
well known. Hence, at present, the practical option the endogenous opioid system. Additionally, the
is to combine medications with some demonstrated augmentation of this effect may be due to the fact
effectiveness in the clinical setting and determine that the opioid and 5-HT3 systems also show a
the treatment response. distribution that, combined, results in more exten-
Of the medications reviewed, perhaps the most sive dopaminergic modulation throughout the
intriguing rationale for a combination has been de- mesocorticolimbic system, rather than more selec-
veloped for the addition of naltrexone (a mu opiate tively at the centrally located nucleus accumbens.
antagonist) to the 5-HT3 antagonist ondansetron in For example, while ondansetron has additional anti-
the treatment of early-onset alcoholism. This ratio- dopaminergic actions in more rostral structures such

 2004 Adis Data Information BV. All rights reserved. CNS Drugs 2004; 18 (15)
Alcoholism and the Serotonergic System 1113

as the frontal cortex, dopamine release may also be have been in the moderate range.[118] There are,
effectively modulated at the more caudally located however, two important caveats to this interpreta-
ventral tegmental area through endogenous opioids. tion. Firstly, the independent effects of ondansetron
In effect, the net result of combining ondansetron versus the combination of ondansetron and naltrex-
and naltrexone would be to diminish dopamine one were not tested in the same experiment. There-
function across the mesocorticolimbic system axis, fore, even though these studies were conducted by
thereby resulting in a synergistic effect on attenuat- the same experimental group, we cannot control for
ing alcohol-induced reward. Indeed, the result of a the possibility that nonspecific factors might ac-
recent animal study demonstrates that the combina- count for these differential treatment effects. We are
tion of ondansetron and naltrexone might be more not aware of any studies that have examined the
effective than either alone at reducing ethanol in- effects of naltrexone in treating early-onset al-
take.[145] coholics. Secondly, although our data do not offer
Thirdly, the clinical efficacy of naltrexone is direct proof that the ondansetron plus naltrexone
dependent upon high levels of compliance. Com- combination will have an additive or synergistic
pliance is reduced, particularly in the early days of effect, these results would suggest that the med-
treatment, by the ability of naltrexone to induce ication combination is likely to be superior to either
nausea.[146] Ondansetron has both anti-nausea and alone. A large-scale study testing the independent
anti-emetic properties.[147,148] Therefore, it would be effects of ondansetron and naltrexone is now needed
reasonable to expect that the addition of ondansetron to build upon these promising preliminary results.
would improve the tolerability of naltrexone.
Mechanistically, there are, of course, other po-
Fourthly, apart from its actions on the mesocor- tential combinations worthy of scientific explora-
ticolimbic system, naltrexone has other nonspecific tion. For instance, animal studies suggest that while
effects, such as those that reduce general consum- naltrexone may reduce alcohol consumption in non-
matory behaviours. The impact of these nonspecific stressed conditions, fluoxetine administration is
naltrexone-induced reductions in general consum- associated with blockade of stress-induced reinstate-
matory behaviours may play a greater role in de- ment.[151] Although the current preclinical evidence
creasing alcohol consumption under conditions of does not suggest that a combination of naltrexone
increased compliance. Hence, the combination of and fluoxetine would be superior to either alone,[152]
ondansetron and naltrexone might be associated it is possible that the provision of these medications
with both specific and nonspecific neurobiological in combination may be particularly useful in
changes that work together to reduce alcohol con- humans, since fluoxetine may assist with the initial
sumption. As a preliminary exploration of this con- period of post-cessation craving and anxiety, there-
cept, my group compared the drinking outcomes of by allowing more time for naltrexone to establish its
individuals who received ondansetron plus naltrex- drinking-reductive effects. Obviously, the combina-
one (n = 10) with those who received placebo (n = tion of an SSRI and naltrexone might be an interest-
10). Despite the small sample size, the medication ing area of investigation for the treatment of al-
combination was superior to placebo at reducing the
coholics with concurrent comorbid major depres-
self-reported drinking outcomes of drinks per day,
sion.
drinks per drinking day, and percentage of days
abstinent,[149] an effect that was corroborated by In sum, studying the effectiveness of combina-
differentially greater decreases for the biochemical tions of serotonergic agents with other types of
marker among those who took the medication com- medication, both alone and together, is a growth
bination.[118,150] Importantly, the treatment effect area in the pharmacotherapy field because it affords
sizes for the medication combination were large; in us the potential to improve our understanding of the
contrast, effect sizes reported for ondansetron alone biological underpinnings of alcoholism as a disease,

 2004 Adis Data Information BV. All rights reserved. CNS Drugs 2004; 18 (15)
1114 Johnson

and may provide promising new treatment strate- infancy. Nevertheless, there appears to be promise
gies. in combining naltrexone with ondansetron for the
treatment of early-onset alcoholism.
8. Conclusions
Acknowledgements
Preclinical studies have contributed greatly to our
understanding of the neurochemical pathways asso- This work was supported by grants #AA10522-05 and
ciated with the development and maintenance of AA10522-0551 from the National Institute on Alcohol Abuse
and Alcoholism. I would also like to thank the following
alcohol-seeking behaviour. These studies have dem- colleagues affiliated with my group (alphabetical order) – Drs
onstrated the important role of serotonin pathways, N. Ait-Daoud, M. Devous, J. Hensler, M. Javors, R. Lamb,
particularly as they relate to dopaminergic function, and J. Roache – for their comments in developing this hypo-
which mediates alcohol-induced reward associated thesis for the differential effectiveness of specific sero-
with its abuse liability. Naturally, this has led to the tonergic agents in treating alcoholic subtypes. I also am
grateful to Mr Robert Cormier BA for his skilled assistance in
study of serotonergic agents as treatments for alco- the preparation of this manuscript. The author has no poten-
holism. tial conflicts of interest directly related to the contents of this
Serotonergic agents remain an intense area of review.
pharmacological study as treatments for alcoholism.
SSRIs do not appear to be effective treatment for a References
heterogeneous alcoholic group. However, they may 1. Audet MA, Descarries L, Doucet G. Quantified regional and
be useful as treatment for late-onset alcoholics, or laminar distribution of the serotonin innervation in the anterior
half of adult rat cerebral cortex. J Chem Neuroanat 1989; 2:
alcoholism complicated by comorbid major depres- 29-44
sion. Buspirone, a 5-HT1A partial agonist, does not 2. Dahlstrom A, Fuxe K. Evidence for the existence of monoam-
appear to be an effective treatment for alcoholics ine-containing neurons in the central nervous system: I. Dem-
onstration of monoamines in the cell bodies of brain stem
without comorbid disease. Buspirone may, how- neurons. Acta Physiol Scand 1964; 62 Suppl. 232: 1-55
ever, have some utility for treating alcoholics with 3. Molliver ME. Serotonergic neuronal systems: what their ana-
tomic organization tells us about function. J Clin Psycho-
comorbid anxiety disorder. The 5-HT2 antagonist pharmacol 1987; 7: 3S-23S
ritanserin at pharmacologically relevant clinical 4. Herve D, Pickel VM, Joh TH, et al. Serotonin axon terminals in
doses does not appear to be an effective treatment the ventral tegmental area of the rat: fine structure and synaptic
input to dopaminergic neurons. Brain Res 1987; 435: 71-83
for alcoholism. Ondansetron, a 5-HT3 antagonist, is 5. Nedergaard S, Hopkins C, Greenfield SA. Do nigro-striatal
an efficacious and promising medication for the neurones possess a discrete dendritic modulatory mechanism?
treatment of early-onset alcoholism. Electrophysiological evidence from the actions of amphetam-
ine in brain slices. Exp Brain Res 1988; 69: 444-8
The differential treatment effect of SSRIs and 6. Hemby SE, Johnson BA, Dworkin SI. Neurobiological basis of
ondansetron among various subtypes of alcoholic is drug reinforcement. In: Johnson BA, Roache JD, editors. Drug
addiction and its treatment: nexus of neuroscience and behav-
intriguing. Future research is needed to understand ior. Philadelphia: Lippincott-Raven, 1997: 137-69
more clearly the molecular genetic differences and 7. Bockaert J, Sebben M, Dumuis A. Pharmacological characteri-
the interactions of these differences with the envi- zation of 5-hydroxytryptamine4 (5-HT4) receptors positively
coupled to adenylate cyclase in adult guinea pig hippocampal
ronment that typify a particular alcoholic subtype. membranes: effect of substituted benzamide derivatives. Mol
Such an understanding could enable us to make Pharmacol 1990; 37: 408-11
comfortable predictions as to which alcoholic sub- 8. Ford AP, Baxter GS, Eglen RM, et al. 5-Hydroxytryptamine
stimulates cyclic AMP formation in the tunica muscularis
type might respond best to a particular serotonergic mucosae of the rat oesophagus via 5-HT4 receptors. Eur J
agent, which could then be provided. Relatively Pharmacol 1992; 211: 117-20
little is known about the effects that medications 9. Monferini E, Gaetani P, Rodriguez y Baena R, et al. Pharmaco-
logical characterization of the 5-hydroxytryptamine receptor
acting on serotonin receptor subtypes other than coupled to adenylyl cyclase stimulation in human brain. Life
those described have on drinking behaviour. Sci 1993; 52: PL61-5
10. Grossman CJ, Kilpatrick GJ, Bunce KT. Development of a
Research into adding other types of medication to radioligand binding assay for 5-HT4 receptors in guinea-pig
enhance the efficacy of serotonergic agents is in its and rat brain. Br J Pharmacol 1993; 109: 618-24

 2004 Adis Data Information BV. All rights reserved. CNS Drugs 2004; 18 (15)
Alcoholism and the Serotonergic System 1115

11. Roychowdhury S, Haas H, Anderson EG. 5-HT1A and 5-HT4 31. Stellar JR, Stellar E. The neurobiology of motivation and re-
receptor colocalization on hippocampal pyramidal cells. ward. New York: Springer-Verlag, 1985
Neuropharmacology 1994; 33: 551-7 32. Barreto Medeiros JM, Cabral Filho JE, De Souza SL, et al.
12. Waeber C, Sebben M, Grossman C, et al. [3H]-GR113808 Early malnourished rats are not affected by anorexia induced
labels 5-HT4 receptors in the human and guinea-pig brain. by a selective serotonin reuptake inhibitor in adult life. Nutr
Neuroreport 1993; 4: 1239-42 Neurosci 2002; 5: 211-4
13. Waeber C, Sebben M, Nieoullon A, et al. Regional distribution 33. Wurtman JJ, Wurtman RJ. Fenfluramine and fluoxetine spare
and ontogeny of 5-HT4 binding sites in rodent brain. Neuro- protein consumption while suppressing caloric intake by rats.
pharmacology 1994; 33: 527-41 Science 1977; 198: 1178-80
14. Andrade R, Chaput Y. 5-Hydroxytryptamine4-like receptors 34. Wurtman JJ, Wurtman RJ. Drugs that enhance central sero-
mediate the slow excitatory response to serotonin in the rat toninergic transmission diminish elective carbohydrate con-
hippocampus. J Pharmacol Exp Ther 1991; 257: 930-7 sumption by rats. Life Sci 1979; 24: 895-903
15. Chaput Y, Araneda RC, Andrade R. Pharmacological and func- 35. Li ET, Anderson GH. 5-Hydroxytryptamine control of meal to
tional analysis of a novel serotonin receptor in the rat hippo- meal composition chosen by rats. Fed Proc 1983; 42: 542-8
campus. Eur J Pharmacol 1990; 182: 441-56 36. Smith GP. The physiology of the meal. In: Silverstone T, editor.
16. Steward LJ, Brown DC, Stokes PR, et al. Antagonism of the (S)- Drugs and appetite. London: Academic Press, 1982
zacopride-induced increase in dopamine release from rat 37. Fantino M. Role of sensory input in the control of food intake. J
striatal slices by the 5-HT receptor antagonist SDZ 205-557 Auton Nerv Syst 1984; 10: 347-58
[abstract]. Third International Union of Pharmacology Satel- 38. Wise RA, Raptis L. Effects of pre-feeding on food-approach
lite Meeting on Serotonin; 1994 Jul 30-Aug 3; Chicago: 84 latency and food consumption speed in food deprived rats.
17. Steward LJ, Barnes NM. The 5-HT4 receptor agonists Physiol Behav 1985; 35: 961-3
renzapride and (S)-zacopride stimulate dopamine release from 39. Haraguchi M, Samson HH, Tolliver GA. Reduction in oral
rat striatal slices [abstract]. Br J Pharmacol 1994; 111 Suppl.: ethanol self-administration in the rat by the 5-HT uptake
155P blocker fluoxetine. Pharmacol Biochem Behav 1990; 35:
18. Myers RD, Veale WL. Alcohol preference in the rat: reduction 259-62
following depletion of brain serotonin. Science 1968; 160: 40. Murphy JM, Waller MB, Gatto GJ, et al. Effects of fluoxetine
1469-71 on the intragastric self-administration of ethanol in the alcohol
19. Nachman M, Lester D, Le Magnen J. Alcohol aversion in the preferring P line of rats. Alcohol 1988; 5: 283-6
rat: behavioral assessment of noxious drug effects. Science 41. Naranjo CA, Sellers EM, Roach CA, et al. Zimelidine-induced
1970; 168: 1244-6 variations in alcohol intake by nondepressed heavy drinkers.
20. Daoust M, Chretien P, Moore N, et al. Isolation and striatal Clin Pharmacol Ther 1984; 35: 374-81
(3H) serotonin uptake: role in the voluntary intake of ethanol 42. Naranjo CA, Sellers EM, Sullivan JT, et al. The serotonin
by rats. Pharmacol Biochem Behav 1985; 22: 205-8 uptake inhibitor citalopram attenuates ethanol intake. Clin
Pharmacol Ther 1987; 41: 266-74
21. Geller I. Effects of para-chlorophenylalanine and 5-hydroxy-
tryptophan on alcohol intake in the rat. Pharmacol Biochem 43. Naranjo CA, Sellers EM. Serotonin uptake inhibitors attenuate
Behav 1973; 1: 361-5 ethanol intake in problem drinkers. Recent Dev Alcohol 1989;
7: 255-66
22. Gill K, Amit Z, Koe BK. Treatment with sertraline, a new
serotonin uptake inhibitor, reduces voluntary ethanol con- 44. Naranjo CA, Kadlec KE, Sanhueza P, et al. Fluoxetine differen-
sumption in rats. Alcohol 1988; 5: 349-54 tially alters alcohol intake and other consummatory behaviors
in problem drinkers. Clin Pharmacol Ther 1990; 47: 490-8
23. Gill K, Filion Y, Amit Z. A further examination of the effects of
45. Naranjo CA, Poulos CX, Bremner KE, et al. Citalopram de-
sertraline on voluntary ethanol consumption. Alcohol 1988; 5:
creases desirability, liking, and consumption of alcohol in
355-8
alcohol-dependent drinkers. Clin Pharmacol Ther 1992; 51:
24. Zabik JE, Binkerd K, Roache JD. Serotonin and ethanol aver- 729-39
sion in the rat. In: Naranjo CA, Sellers EM, editors. Research
46. Gorelick DA, Paredes A. Effect of fluoxetine on alcohol con-
advances in new psychopharmacological treatments for alco-
sumption in male alcoholics. Alcohol Clin Exp Res 1992; 16:
holism: proceedings of the symposium; 1984 Oct 4-5; Toronto.
261-5
Amsterdam: Excerpta Medica, 1985
47. Naranjo CA, Bremner KE, Lanctot KL. Effects of citalopram
25. Blundell JE, Latham CJ. Behavioural pharmacology of feeding. and a brief psycho-social intervention on alcohol intake, de-
In: Silverstone T, editor. Drugs and appetite. London: Aca- pendence and problems. Addiction 1995; 90: 87-99
demic Press, 1982
48. Kabel DI, Petty F. A placebo-controlled, double-blind study of
26. Blundell JE. Serotonin and appetite. Neuropharmacology 1984; fluoxetine in severe alcohol dependence: adjunctive pharma-
23: 1537-51 cotherapy during and after inpatient treatment. Alcohol Clin
27. Gill K, Amit Z. Serotonin uptake blockers and voluntary alcohol Exp Res 1996; 20: 780-4
consumption: a review of recent studies. Recent Dev Alcohol 49. Kranzler HR, Burleson JA, Korner P, et al. Placebo-controlled
1989; 7: 225-48 trial of fluoxetine as an adjunct to relapse prevention in al-
28. Gottfries CG. Influence of depression and antidepressants on coholics. Am J Psychiatry 1995; 152: 391-7
weight. Acta Psychiatr Scand Suppl 1981; 290: 353-6 50. Buydens-Branchey L, Branchey MH, Noumair D. Age of alco-
29. Simpson RJ, Lawton DJ, Watt MH, et al. Effect of zimelidine, a holism onset: I. Relationship to psychopathology. Arch Gen
new antidepressant, on appetite and body weight. Br J Clin Psychiatry 1989; 46: 225-30
Pharmacol 1981; 11: 96-8 51. Linnoila M, Virkkunen M. Biologic correlates of suicidal risk
30. Leander JD. Fluoxetine suppresses palatability-induced inges- and aggressive behavioral traits. J Clin Psychopharmacol
tion. Psychopharmacology 1987; 91: 285-7 1992; 12 (2 Suppl.): 19S-20S

 2004 Adis Data Information BV. All rights reserved. CNS Drugs 2004; 18 (15)
1116 Johnson

52. Linnoila M, De Jong J, Virkkunen M. Family history of alcohol- 70. Blier P, Ward NM. Is there a role for 5-HT1A agonists in the
ism in violent offenders and impulsive fire setters. Arch Gen treatment of depression? Biol Psychiatry 2003; 53: 193-203
Psychiatry 1989; 46: 613-6 71. Malcolm R, Anton RF, Randall CL, et al. A placebo-controlled
53. Fils-Aime ML, Eckardt MJ, George DT, et al. Early-onset trial of buspirone in anxious inpatient alcoholics. Alcohol Clin
alcoholics have lower cerebrospinal fluid 5-hydroxy- Exp Res 1992; 16: 1007-13
indoleacetic acid levels than late-onset alcoholics. Arch Gen 72. Kranzler HR, Burleson JA, Del Boca FK, et al. Buspirone
Psychiatry 1996; 53: 211-6 treatment of anxious alcoholics: a placebo-controlled trial.
54. Kranzler HR, Burleson JA, Brown J, et al. Fluoxetine treatment Arch Gen Psychiatry 1994; 51: 720-31
seems to reduce the beneficial effects of cognitive-behavioral 73. Bruno F. Buspirone in the treatment of alcoholic patients.
therapy in type B alcoholics. Alcohol Clin Exp Res 1996; 20: Psychopathology 1989; 22 Suppl. 1: 49-59
1534-41 74. Malec TS, Malec EA, Dongier M. Efficacy of buspirone in
55. Pettinati HM, Volpicelli JR, Kranzler HR, et al. Sertraline alcohol dependence: a review. Alcohol Clin Exp Res 1996; 20:
treatment for alcohol dependence: interactive effects of med- 853-8
ication and alcoholic subtype. Alcohol Clin Exp Res 2000; 24: 75. George DT, Rawlings R, Eckardt MJ, et al. Buspirone treatment
1041-9 of alcoholism: age of onset, and cerebrospinal fluid 5-hydroxy-
56. Johnson BA, Cloninger CR, Roache JD, et al. Age of onset as a indolacetic acid and homovanillic acid concentrations, but not
discriminator between alcoholic subtypes in a treatment-seek- medication treatment, predict return to drinking. Alcohol Clin
ing outpatient population. Am J Addict 2000; 9: 17-27 Exp Res 1999; 23: 272-8
57. Johnson BA, Ait-Daoud N. Neuropharmacological treatments 76. Meert TF, Awouters F, Niemegeers CJ, et al. Ritanserin reduces
for alcoholism: scientific basis and clinical findings. Psycho- abuse of alcohol, cocaine, and fentanyl in rats. Pharmacop-
pharmacology 2000; 149: 327-44 sychiatry 1991; 24: 159-63
58. Cornelius JR, Salloum IM, Ehler JG, et al. Fluoxetine in 77. Myers RD, Lankford M, Bjork A. Selective reduction by the
depressed alcoholics: a double-blind, placebo-controlled trial. 5-HT antagonist amperozide of alcohol preference induced in
Arch Gen Psychiatry 1997; 54: 700-5 rats by systemic cyanamide. Pharmacol Biochem Behav 1992;
43: 661-7
59. Mason BJ, Kocsis JH, Ritvo EC, et al. A double-blind, placebo-
controlled trial of desipramine for primary alcohol dependence 78. Svensson L, Fahlke C, Hard E, et al. Involvement of the
stratified on the presence or absence of major depression. serotonergic system in ethanol intake in the rat. Alcohol 1993;
JAMA 1996; 275: 761-7 10: 219-24
60. McGrath PJ, Nunes EV, Stewart JW, et al. Imipramine treat- 79. Myers RD, Lankford MF. Suppression of alcohol preference in
ment of alcoholics with primary depression: a placebo-control- high alcohol drinking rats: efficacy of amperozide versus
led clinical trial. Arch Gen Psychiatry 1996; 53: 232-40 naltrexone. Neuropsychopharmacology 1996; 14: 139-49
80. Myers RD, Lankford M. Action of the 5-HT2A antagonist
61. Pettinati HM, Volpicelli JR, Luck G, et al. Double-blind
amperozide on alcohol-induced poikilothermia in rats.
clinical trial of sertraline treatment for alcohol dependence. J
Pharmacol Biochem Behav 1998; 59: 91-5
Clin Psychopharmacol 2001; 21: 143-53
81. Biggs TA, Myers RD. Naltrexone and amperozide modify choc-
62. Collins DM, Myers RD. Buspirone attenuates volitional alcohol olate and saccharin drinking in high alcohol-preferring P rats.
intake in the chronically drinking monkey. Alcohol 1987; 4: Pharmacol Biochem Behav 1998; 60: 407-13
49-56
82. Overstreet DH, McArthur RA, Rezvani AH, et al. Selective
63. Privette TH, Hornsby RL, Myers RD. Buspirone alters alcohol inhibition of alcohol intake in diverse alcohol-preferring rat
drinking induced in rats by tetrahydropapaveroline injected strains by the 5-HT2A antagonists amperozide and FG 5974.
into brain monoaminergic pathways. Alcohol 1988; 5: 147-52 Alcohol Clin Exp Res 1997; 21: 1448-54
64. Meert TF. Effects of various serotonergic agents on alcohol 83. Lankford MF, Bjork AK, Myers RD. Differential efficacy of
intake and alcohol preference in Wistar rats selected at two serotonergic drugs FG5974, FG5893, and amperozide in re-
different levels of alcohol preference. Alcohol Alcohol 1993; ducing alcohol drinking in P rats. Alcohol 1996; 13: 399-404
28: 157-70 84. Ugedo L, Grenhoff J, Svensson TH. Ritanserin, a 5-HT2 recep-
65. Wilson AW, Costall B, Neill JC. Manipulation of operant re- tor antagonist, activates midbrain dopamine neurons by block-
sponding for an ethanol-paired conditioned stimulus in the rat ing serotonergic inhibition. Psychopharmacology 1989; 98:
by pharmacological alteration of the serotonergic system. J 45-50
Psychopharmacol 2000; 14: 340-6 85. Awouters F, Niemegeers CJ, Megens AA, et al. The pharmaco-
66. Rezvani AH, Overstreet DH, Janowsky DS. Genetic serotonin logical profile of ritanserin, a very specific central serotonin-
deficiency and alcohol preference in the fawn hooded rats. S2 antagonist. Drug Dev Res 1988; 15: 61-73
Alcohol Alcohol 1990; 25: 573-5 86. Johnson BA, Jasinski DR, Galloway GP, et al. Ritanserin in the
67. Gongwer MA, Murphy JM, McBride WJ, et al. Regional brain treatment of alcohol dependence: a multi-center clinical trial.
contents of serotonin, dopamine and their metabolites in the Ritanserin Study Group. Psychopharmacology 1996; 128:
selectively bred high- and low-alcohol drinking lines of rats. 206-15
Alcohol 1989; 6: 317-20 87. Wiesbeck GA, Weijers HG, Chick J, et al. Ritanserin in relapse
68. McBride WJ, Bodart B, Lumeng L, et al. Association between prevention in abstinent alcoholics: results from a placebo-
low contents of dopamine and serotonin in the nucleus accum- controlled double-blind international multicenter trial. Ritan-
bens and high alcohol preference. Alcohol Clin Exp Res 1995; serin in Alcoholism Work Group. Alcohol Clin Exp Res 1999;
19: 1420-2 23: 230-5
69. Korpi ER, Paivarinta P, Abi-Dargham A, et al. Binding of 88. LeMarquand D, Pihl RO, Benkelfat C. Serotonin and alcohol
serotonergic ligands to brain membranes of alcohol-preferring intake, abuse, and dependence: clinical evidence. Biol Psychi-
AA and alcohol-avoiding ANA rats. Alcohol 1992; 9: 369-74 atry 1994; 36: 326-37

 2004 Adis Data Information BV. All rights reserved. CNS Drugs 2004; 18 (15)
Alcoholism and the Serotonergic System 1117

89. Lovinger DM, White G. Ethanol potentiation of 5-hydroxy- sumption in alcohol-preferring rats. Alcohol Alcohol 1991; 26:
tryptamine3 receptor-mediated ion current in neuroblastoma 107-10
cells and isolated adult mammalian neurons. Mol Pharmacol 110. Rodd-Henricks ZA, McKinzie DL, Li T-K, et al. Intracranial
1991; 40: 263-70 self-administration of ethanol into the posterior VTA of Wistar
90. Zhou Q, Lovinger DM. Pharmacologic characteristics of poten- rats is mediated by 5-HT3 receptors [abstract]. Alcohol Clin
tiation of 5-HT3 receptors by alcohols and diethyl ether in Exp Res 1999; 23 Suppl. 5: 49A
NCB-20 neuroblastoma cells. J Pharmacol Exp Ther 1996; 111. McBride WJ, Li TK. Animal models of alcoholism: neurobiolo-
278: 732-40 gy of high alcohol-drinking behavior in rodents. Crit Rev
91. Lovinger DM, Zhou Q. Alcohols potentiate ion current medi- Neurobiol 1998; 12: 339-69
ated by recombinant 5-HT3RA receptors expressed in a mam-
112. Tomkins DM, Le AD, Sellers EM. Effect of the 5-HT3 ant-
malian cell line. Neuropharmacology 1994; 33: 1567-72
agonist ondansetron on voluntary ethanol intake in rats and
92. Lovinger DM. Inhibition of 5-HT3 receptor-mediated ion cur- mice maintained on a limited access procedure. Psycho-
rent by divalent metal cations in NCB-20 neuroblastoma cells. pharmacology 1995; 117: 479-85
J Neurophysiol 1991; 66: 1329-37
113. Beardsley PM, Lopez OT, Gullikson G, et al. Serotonin 5-HT3
93. Lovinger DM. Ethanol potentiates ion current mediated by
antagonists fail to affect ethanol self-administration of rats.
5-HT3 receptors on neuroblastoma cells and isolated neurons.
Alcohol 1994; 11: 389-95
Alcohol Alcohol Suppl 1991; 1: 181-5
94. Lovinger DM. 5-HT3 receptors and the neural actions of alco- 114. Johnson BA, Campling GM, Griffiths P, et al. Attenuation of
hols: an increasingly exciting topic. Neurochem Int 1999; 35: some alcohol-induced mood changes and the desire to drink by
125-30 5-HT3 receptor blockade: a preliminary study in healthy male
volunteers. Psychopharmacology 1993; 112: 142-4
95. Di Chiara G, Imperato A. Drugs abused by humans preferential-
ly increase synaptic dopamine concentrations in the 115. Swift RM, Davidson D, Whelihan W, et al. Ondansetron alters
mesolimbic system of freely moving rats. Proc Natl Acad Sci human alcohol intoxication. Biol Psychiatry 1996; 40: 514-21
U S A 1988; 85: 5274-8 116. Doty P, Zacny JP, de Wit H. Effects of ondansetron pretreat-
96. Koob GF. Neural mechanisms of drug reinforcement. Ann N Y ment on acute responses to ethanol in social drinkers. Behav
Acad Sci 1992; 654: 171-91 Pharmacol 1994; 5: 461-9
97. Wise RA, Bozarth MA. A psychomotor stimulant theory of 117. Sellers EM, Toneatto T, Romach MK, et al. Clinical efficacy of
addiction. Psychol Rev 1987; 94: 469-92 the 5-HT3 antagonist ondansetron in alcohol abuse and de-
98. Bloom FE, Morales M. The central 5-HT3 receptor in CNS pendence. Alcohol Clin Exp Res 1994; 18: 879-85
disorders. Neurochem Res 1998; 23: 653-9 118. Johnson BA, Roache JD, Javors MA, et al. Ondansetron for
99. Kilpatrick GJ, Jones BJ, Tyers MB. Identification and distribu- reduction of drinking among biologically predisposed alcohol-
tion of 5-HT3 receptors in rat brain using radioligand binding. ic patients: a randomized controlled trial. JAMA 2000; 284:
Nature 1987; 330: 746-8 963-71
100. Kilpatrick GJ, Hagan RM, Gale JD. 5-HT3 and 5-HT4 receptors 119. Johnson BA, Roache JD, Ait-Daoud N, et al. Ondansetron
in terminal regions of the mesolimbic system. Behav Brain Res reduces the craving of biologically predisposed alcoholics.
1996; 73: 11-3 Psychopharmacology 2002; 160: 408-13
101. Oxford AW, Bell JA, Kilpatrick GJ, et al. Ondansetron and 120. Kranzler HR, Pierucci-Lagha A, Feinn R, et al. Effects of
related 5-HT3 antagonists: recent advances. Prog Med Chem ondansetron in early- versus late-onset alcoholics: a prospec-
1992; 29: 239-70 tive, open-label study. Alcohol Clin Exp Res 2003; 27: 1150-5
102. Barnes NM, Sharp T. A review of central 5-HT receptors and 121. Heils A, Teufel A, Petri S, et al. Allelic variation of human
their function. Neuropharmacology 1999; 38: 1083-152 serotonin transporter gene expression. J Neurochem 1996; 66:
103. Johnson BA, Cowen PJ. Alcohol-induced reinforcement: dop- 2621-4
amine and 5-HT3 receptor interactions in animals and humans.
122. Heils A, Mossner R, Lesch KP. The human serotonin transport-
Drug Dev Res 1993; 30: 153-69
er gene polymorphism: basic research and clinical implica-
104. Bradbury AJ, Costall B, Domeney AM, et al. Laterality of tions. J Neural Transm 1997; 104: 1005-14
dopamine function and neuroleptic action in the amygdala in
the rat. Neuropharmacology 1985; 24: 1163-70 123. Lesch KP, Meyer J, Glatz K, et al. The 5-HT transporter gene-
linked polymorphic region (5-HTTLPR) in evolutionary per-
105. Hagan RM, Jones BJ, Jordan CC, et al. Effect of 5-HT3
spective. Alternative biallelic variation in rhesus monkeys:
receptor antagonists on responses to selective activation of
rapid communication. J Neural Transm 1997; 104: 1259-66
mesolimbic dopaminergic pathways in the rat. Br J Pharmacol
1990; 99: 227-32 124. Greenberg BD, Tolliver TJ, Huang SJ, et al. Genetic variation
106. Eison AS, Iversen SD, Sandberg BE, et al. Substance P analog, in the serotonin transporter promoter region affects serotonin
DiMe-C7: evidence for stability in rat brain and prolonged uptake in human blood platelets. Am J Med Genet 1999; 88:
central actions. Science 1982; 215: 188-90 83-7
107. Costall B, Domeney AM, Naylor RJ, et al. Effects of the 5-HT3 125. Lesch KP, Bengel D, Heils A, et al. Association of anxiety-
receptor antagonist, GR38032F, on raised dopaminergic ac- related traits with a polymorphism in the serotonin transporter
tivity in the mesolimbic system of the rat and marmoset brain. gene regulatory region. Science 1996; 274: 1527-31
Br J Pharmacol 1987; 92: 881-94 126. Johnson BA. Serotonergic agents and alcoholism treatment:
108. Hodge CW, Samson HH, Lewis RS, et al. Specific decreases in rebirth of the subtype concept: an hypothesis. Alcohol Clin
ethanol- but not water-reinforced responding produced by the Exp Res 2000; 24: 1597-601
5-HT3 antagonist ICS 205-930. Alcohol 1993; 10: 191-6 127. Heinz A, Jones DW, Mazzanti C, et al. A relationship between
109. Fadda F, Garau B, Marchei F, et al. MDL 72222, a selective serotonin transporter genotype and in vivo protein expression
5-HT3 receptor antagonist, suppresses voluntary ethanol con- and alcohol neurotoxicity. Biol Psychiatry 2000; 47: 643-9

 2004 Adis Data Information BV. All rights reserved. CNS Drugs 2004; 18 (15)
1118 Johnson

128. Enoch M-A, Schuckit MA, Johnson BA, et al. Genetics of amine release and function in the rat. Eur J Pharmacol 1993;
alcoholism using intermediate phenotypes. Alcohol Clin Exp 230: 63-8
Res 2003; 27: 169-76
143. Matsuzawa S, Suzuki T, Misawa M, et al. Roles of 5-HT3 and
129. Bisaga A, Sikora J, Kostowski W. The effect of drugs interact- opioid receptors in the ethanol-induced place preference in rats
ing with serotonergic 5HT3 and 5HT4 receptors on morphine exposed to conditioned fear stress. Life Sci 1999; 64: PL241-9
place conditioning. Pol J Pharmacol 1993; 45: 513-9
144. Widdowson PS, Holman RB. Ethanol-induced increase in en-
130. Panocka I, Ciccocioppo R, Polidori C, et al. The 5-HT4 receptor dogenous dopamine release may involve endogenous opiates. J
antagonist, GR113808, reduces ethanol intake in alcohol-pre- Neurochem 1992; 59: 157-63
ferring rats. Pharmacol Biochem Behav 1995; 52: 255-9
145. Le AD, Sellers EM. Interaction between opiate and 5-HT3
131. Barnes JM, Barnes NM, Champaneria S, et al. Characterisation
receptor antagonists in the regulation of alcohol intake. Alco-
and autoradiographic localisation of 5-HT3 receptor recogni-
tion sites identified with [3H]-(S)-zacopride in the forebrain of hol Alcohol Suppl 1994; 2: 545-9
the rat. Neuropharmacology 1990; 29: 1037-45 146. Volpicelli JR, Rhines KC, Rhines JS, et al. Naltrexone and
132. Perry DC. Autoradiography of [3H]quipazine in rodent brain. alcohol dependence: role of subject compliance. Arch Gen
Eur J Pharmacol 1990; 187: 75-85 Psychiatry 1997; 54: 737-42

133. Yoshimoto K, McBride WJ. Regulation of nucleus accumbens 147. Gale JD. Serotonergic mediation of vomiting. J Pediatr Gas-
dopamine release by the dorsal raphe nucleus in the rat. troenterol Nutr 1995; 21 Suppl. 1: S22-8
Neurochem Res 1992; 17: 401-7 148. Wilde MI, Markham A. Ondansetron: a review of its pharmacol-
134. Campbell AD, McBride WJ. Serotonin-3 receptor and ethanol- ogy and preliminary clinical findings in novel applications.
stimulated dopamine release in the nucleus accumbens. Drugs 1996; 52: 773-94
Pharmacol Biochem Behav 1995; 51: 835-42
149. Johnson BA, Ait-Daoud N, Prihoda TJ. Combining ondansetron
135. Gianoulakis C. The effect of ethanol on the biosynthesis and and naltrexone effectively treats biologically predisposed al-
regulation of opioid peptides. Experientia 1989; 45: 428-35 coholics: from hypotheses to preliminary clinical evidence.
136. Reid LD, Hunter GA. Morphine and naloxone modulate intake Alcohol Clin Exp Res 2000; 24: 737-42
of ethanol. Alcohol 1984; 1: 33-7 150. Ait-Daoud N, Johnson BA, Javors M, et al. Combining ondan-
137. Hubbell CL, Czirr SA, Reid LD. Persistence and specificity of setron and naltrexone treats biological alcoholics: corrobora-
small doses of morphine on intake of alcoholic beverages. tion of self-reported drinking by serum carbohydrate deficient
Alcohol 1987; 4: 149-56 transferrin, a biomarker. Alcohol Clin Exp Res 2001; 25:
847-9
138. Hubbell CL, Czirr SA, Hunter GA, et al. Consumption of
ethanol solution is potentiated by morphine and attenuated by 151. Le AD, Poulos CX, Harding S, et al. Effects of naltrexone and
naloxone persistently across repeated daily administrations. fluoxetine on alcohol self-administration and reinstatement of
Alcohol 1986; 3: 39-54 alcohol seeking induced by priming injections of alcohol and
139. Carboni E, Acquas E, Leone P, et al. 5-HT3 receptor antagon- exposure to stress. Neuropsychopharmacology 1999; 21:
ists block morphine- and nicotine-induced place-preference 435-44
conditioning. Eur J Pharmacol 1988; 151: 159-60 152. Gardell LR, Whalen CA, Chattophadyay S, et al. Combination
140. Carboni E, Acquas E, Frau R, et al. Differential inhibitory of naltrexone and fluoxetine on rats’ propensity to take alco-
effects of a 5-HT3 antagonist on drug-induced stimulation of holic beverage. Alcohol Clin Exp Res 1997; 21: 1435-9
dopamine release. Eur J Pharmacol 1989; 164: 515-9
141. Imperato A, Angelucci L. 5-HT3 receptors control dopamine
release in the nucleus accumbens of freely moving rats. Correspondence and offprints: Dr Bankole A. Johnson, De-
Neurosci Lett 1989; 101: 214-7 partment of Psychiatric Medicine, University of Virginia,
142. Pei Q, Zetterstrom T, Leslie RA, et al. 5-HT3 receptor antagon- PO Box 800623, Charlottesville, VA 22908-0623, USA.
ists inhibit morphine-induced stimulation of mesolimbic dop- E-mail: [email protected]

 2004 Adis Data Information BV. All rights reserved. CNS Drugs 2004; 18 (15)