Astra A ward Lecture

Plitrr.riruco/ogy & To.yico/ogj 1992, 7 I 40 I 415.

The Psychopharmacology of 5-HT3Receptors* Brenda Costall and Robert J. Naylor Postgraduate Studies i n Pharmacology. The School of Pharmacy, University of Bradford. Bradford, West Yorkshire, BD7 IDP, United Kingdom (Received August 25, 1992; Accepted August 27, 1992) Ahs/r.rrc/: 5-HT, receptors have an exclusive neuronal location and evidence is presented of their involvement in behaviour. 5-HT, receptor antagonists such as ondansetron, tropisetron and zacopride have provided the critical pharmacological tools to reveal a potent and efficacious ability to regulate disturbed behaviour. Thus the 5-HT, receptor antagonists will restore to normal rodent and primate behaviour disturbed by increasing limbic dopamine function, aversive situations, cognitive impairments and drug abuse. The remarkable feature of their action is a failure to modify normal behaviour. This unique pharmacological signature has ensured a wide interest in the potential role of the 5-HT3 receptor antagonists in the treatment of schizophrenia, anxiety, age related memory impairment and the problems of withdrawal from drugs of abuse. The preclinical data and preliminary clinical observations are presented.

Ahhrc4rr/ion.v: 5-HT = 5-hydroxytryptaniine. BPRS = British Psychiatric Rating Scale, DSM-III =Diagnostic and Statistical Manual 111, HAM-A score = Hamilton-A, 5-HTP= 5-hydroxytryptophan, 5-HIAA = 5-hydroxyindole acetic acid. AAMI =Age Associated Memory Impairment.


Thc authority of any scientific study is limited by the available methodologies and investigative tools. The period of 5-HT research between 1950 and 1980 established the presence of 5-HT in many tissues of many species and, in animals, the ability of 5-HT to modify cardiovascular function, gastrointestinal motility and secretion, temperature, hormonal status, motor behaviours and other functions. Such findings were initially obtained on the basis of the administration of 5-HT and its precursor 5-hydroxytryptophan, or the inhibition of 5-HT synthesis and lesions of the central 5-HT pathways. Such manipulations frequently caused a gross disruption or stimulation of 5-HT function and a selective effect could not be achieved before the use of discrete pharmacological probes. Initially this was limited t o thc use of agents that had varying degrees of selectivity for the 5-HT, and 5-HT2receptors, e.g. methysergide, methiothcpin and cyproheptadinc. It was the introduction of kctanserin and subsequently other compounds with highly selective effects on the 5-HTz systems that allowed a discrimination of drug action on the 5-HT, and 5-HT2systems (see Bradley et rd. 1986). It was the use of such tools that revealed a 5-HT,/5HT, receptor involvement in the 5-HT induced behavioural syndrome of reciprocal forepaw treading, hindlimb abduction, lateral head weaving, straub tail and hyperactivity in the rodent (Grahame-Smith 1971; Jacobs 1976; Deakin & Grccn 1978; review by Heal " t crl. 1992). Yet, with other tests. for example anxiety, whilst depletion of brain 5-HT released punished responding in animal tests for anxiety (see rcview by Iversen 1980), the evidence was conflicting

* Lecture presented at the XX Nordic Congress of Physiology & Pharniacology, August 16-1 9. 1992. Copenhagen.

that 5-HT, /5-HT2 receptor ligands could modify aversive behaviour (see review by Deakin et rrl. 1992). Such evidence was based on the methodologies and tools available and, using such compounds, the evidence indicating that 5-HT was involved in aversive behaviour was confusing. The situation has improved but remains complex, with the recent identification of drug action o n the 5-HT3 receptors moderating anxiety related behaviours in novel animal tests. Furthermore, the breadth of action of the 5-HT3 receptor antagonists in behavioural models of cognition, schizophrenia and drug addiction attests to a much wider spectrum of activities.

Identification of the 5 - H T , receptor. An interest in 5-HT and the cardiovascular system, together with a measure of serendipity, provided the initial basis lor Fozard and co-workers to identify a novel 5-HT receptor. It was found that the ability of 5-HT to depolarise the sympathetic nerve terminals in the rabbit heart causing a release of noradrenaline was blocked by (-)cocaine but not by methysergide (Fozard & Mwaluko 1975 & 1976; Fozard & Mobarok-Ali 1978). The antagonism afforded by (-)cocaine at the non-5-HT, /5-HTz site was competitive and surmountable, and at concentrations too low for interference from its anaesthetic actions. Procainamide was also an effective 5-HT antagonist at concentrations below those causing local anaesthesia and the results from this compound prompted the use of the cogener metoclopramide (see Fozard 1989). Metoclopramide was also revealed as a potent and surmountable 5-HT antagonist (Fozard & Mobarok-Ali 1978). Based on the tropane nucleus of cocaine, the substituted benzamide structure of metoclopramide or the indole nucleus of 5-HT, other compounds with a greater selectivity and specificity of action for the neuronal



5-HT receptor were successfully synthesised and tested, e.g. MDL72222, tropisetron, ondansetron, zacopride (Fozard 1984; Richardson et d. 1985; Brittain et NI. 1987; Smith et (/I. 1988). The use of such antagonists, and of 2-methyl-5-HT a s a reasonably selective neuronal 5-HT receptor agonist, allowed a characterisation of the receptor as mediating a depolarisation response on peripheral autonomic, afferent and enteric neurones (see reviews by Wallis 1989; Gershon ct a/. 1989); the receptor so defined was designated the S H T , receptor (Bradley et NI. 1986). A S H T , receptor antagonist action on the afferent and enteric neurones was the basis for the clinical investigation of ondansetron, MDL72222, tropisetron and other agents in migraine, gastrokinesis, the carcinoid syndrome and the nausea and vomiting induced by chemo- and radiotherapy (see reviews by Fozard 1989; Costall & Naylor 1990). The subsequent development of radiolabelled S H T , receptor antagonists allowed the direct demonstration of 5HT, receptors on the vagus nerve, enteric elements and within the brain (Kilpatrick et trl. 1987 & 1989; Barnes et oI. 1988; Waeber et 01. 1988; Pinkus 1’1 rrl. 1989; Gordon i’t crl. 1990; Champaneria et uf. 1992). Concomitant electrophysiological studies revealed that, from amongst the various 5-HT receptors, only the 5-HT3 receptor is representative of the ligand gated ion channels. Thus 5-HT3 receptor activation ‘opens’ the 5-HT3 channel. increasing its conductance for monovalent cations. Due to the input of sodium ions, a depolarisation occurs causing excitation of the neurone (Lambert et a/. 1989). The importance of the 5-HT3 receptors is their exclusive neuronal location and, for the purposes of the present review. the consequences of drug interaction at the S H T , reccptor on behaviour. The rffict of 5- HT, receptor arilrrgonists on norninl behhnviout:

Probably the first test to be developed for the evaluation of 5-HT3 receptor blockade in man was based on the ability of an intradermal injection of 5-HT to induce a flare response which is caused by an activation of an axon reflex vasodilatation. The response was antagonised by the intravenous injection of MDL72222 (0.3 mg/ kg) (Orwin & Fozard 1986). Fozard was a subject of the study and reported as a personal obscrvation that ”substantial blockade of peripheral 5-HT, receptors does not result in overt (behavioural) pharmacological activity” (see Fozard 1989).This observation has been confirmed in tens of thousands of patients receiving ondansetron as an antiemetic agent during chemotherapy and to prevent post-operative nausea and vomiting. Such observations in man simply reflect the earliest findings of the effects of ondansetron, tropisetron and other 5HT, receptor antagonists in animals. They were without overt effect to modify normal behaviour (Costall et al. 1987 & 1990b). This contrasts the actions of the 5-HT3 receptor antagonists with other types of psychopharmacological agents which can induce sedation, motor impairment or disorders of cognition as dose-ranges are fully explored.


Tlie Cffict of 5 - H T 3 receptor antrrgonists in ~ i n i m ~rnorlels rl of disturbed behaviour. In 1983 the Neuropharmacology Research group at the University of Bradford commenced a collaborative research project with Glaxo Group Research to investigate the potential psychopharmacological activity of a series ofS-HT, receptor antagonists structured around the indole nucleus of 5-HT. The 5-HT3 receptor antagonist properties of ondansetron (GR38032F) were detected on the basis of the in vitro tests identified above: ondansetron was shown to bc a highly potent S H T , receptor antagonist lacking any significant affinity for other neurotransmitter receptors (see Butler et t i / . 1988). Also, ondansetron was shown to lack affinity for other 5-HT receptor subtypes, a profile of action shared by the other 5-HT, receptor antagonists with the exception of tropisetron. The latter agent, at exceptionally high micromolar concentrations, has a 5-HT4 receptor antagonist effect (Dumuis o! al. 1988).

The ej’ect o j S - H T , receptor ~ntrrgoriiststo rmdifj bruin clopumine,funrtion. The first behavioural experiments which used ondansetron focused on its potential to modify brain dopamine function. Dopamine is found in highest concentration in the striatal, limbic and tuberoinfundibular systems; the ascending 5-HT system from the raphe nuclei innervates all forebrain structures. It was a straightforward procedure to show that ondansetron did not effect the tuberoinfundibular system t o modify prolactin release or the release of other hormones. It was also a simple procedure to show that ondansetron did not cause sedation or the motor impairments caused by a reduced striatal dopamine function (Costall et al. 1987). This is characteristically revealed by dopamine receptor antagonists as a motor immobility or catalepsy in rodents or a Parkinsonian-like rigidity and akinesia in primates. The 5HT, receptor antagonists are dramatically different from the neuroleptics in this respect and this has been consistentlyconfirmed in man: there has been no induction of extrapyramidal side effects. Also, in the traditional animal models to detect a reduced dopamine function, the abilityofa drugto inhibit the stereotyped motor behaviours induced by dopamine agonists such as amphetamine has been correlated with neuroleptic or antipsychotic potential (Costall & Naylor 1980). Ondansetron and other 5-HT3 receptor antagonists !‘ail to block stereotyped responding. Therefore it can be safely concluded that the 5-HT, receptor antagonists have to known ability to comprise striatal or hypothalamic/pituitary dopamine function. It should also be noted that if the ability of the 5-HT1 receptor antagonists to antagonise brain dopamine function had been measured only using the classical tests for neuroleptic action, then their ability to modify mesolimbic dopaminc activity would have passed unnoticed. It has been known for many years that the injection ofdopamine and dopamine agonists into the nucleus accumbens of the rat can increase locomotor activity to provide a model of limbic dopamine overactivity (Costall & Naylor 1976). which can detect drugs which will antagonise limbic dopamine ac-



tivity. Such actions would indicate an antischizophrenic profile. Thus the mesolimbic dopamine hyperactivity model detected the actions of the atypical agent clozapine, which has a useful role to play in the modern treatment of schizophrenia (Costall & Naylor 1976). The model IS also useful since the nucleus accumbens receives a 5-HT innervation and manipulations of the 5-HT system in the accumbens can influence locomotor responding (Jackson et crl. 1975; Pijnenburg et al. 1975; Makanjuola et NI. 1980; Costall et NI. 1979). A 5-HT3 receptor involvement in the 5-HT regulation of the limbic dopamine system was investigated using an elevated dopamine function. Limbic dopamine activity was increased by three procedures; firstly by the injection of single doses of amphetamine directly into the nucleus accumbens, secondly by the constant infusion of dopamine directly into the nucleus accumbens over a 13 day period and thirdly, by discrete activation of the midbrain dopamine cells using the localised micro-infusion of the neurokinin agonist Di-MeC7, causing a release of endogenous dopamine in limbic and other forebrain structures (Hagan et d.1987). In the first experiments the effects of amphetamine were enhanced by the intra-accumbens injections of the 5-HT3 receptor agonist 2-methyl-5-HT and this enhancement was antagonised by a local or peripheral injection of ondansetron. The peripheral administration of ondansetron and other 5-


HT3 receptor antagonists also antagonised the behavioural hyperactivity caused by the infusion of dopamine into the nucleus accumbens (fig. I ) , and was also successful in inhibiting both the hyperactivity response and the increased turnover of limbic dopamine following the Di-Me-C7 midbrain injection (Hagan rt ul. 1987 & 1990). The direction of the behavioural change would indicate that activation of 5HT3receptors will enhance dopamine function, and it is interesting that dopamine release in the nucleus accumbens is reported to be increased following activation of the 5-HT, receptors (Jiang ef al. 1990; Chen ot d.1991). The ability of ondansetron to antagonise a raised limbic dopamine function should be placed within the following perspective. Ondansetron is as effective as the neuroleptic agents with the additional advantage that the raised locomotor activity is simply returned to control levels. Unlike the use of neuroleptic agents, activity is not depressed to levels below control values. This profile of action is also obtained using tropisetron, zacopride and granisetron and their common feature of action with ondansetron is a 5-HT3receptor blockade. Following withdrawal from a neurolepticidopamine infusion treatment, rodents demonstrate a rebound hyperactivity response. This does not occur following withdrawal from an ondansetronidopaminc infusion regimen and indeed, coadministration of ondansetron with a neuroleptic treatment

Fig. I . Influence of a classical neuroleptic, fluphenazine. and 5-HT3receptor antagonists on the hyperactivity caused by dopamine infused slowly and persistently into the rat nucleus accumbens (25 pg/24 hr. 0.48 plihr) for 13 days. Locomotor activity was measured in individual boxes equipped with photocell units: the data was plotted as locomotor activity in countsi5 min. for each day of lest and the area under the locomotor activity curve was computed (shown as AUC (area under the curve). represented in arbitrary units). n = 5- 10, S.E.M.son original data were less C indicates animals given intra-accumbens dopamine alone. Significant reductions in the dopamine hyperactivity are indicated as than 12.7'%~ *P

Astra Award Lecture. The psychopharmacology of 5-HT3 receptors.

5-HT3 receptors have an exclusive neuronal location and evidence is presented of their involvement in behaviour. 5-HT3 receptor antagonists such as on...
3MB Sizes 0 Downloads 0 Views