5-HT Receptor Antagonists and Anxiety

European Neuropsychopharmacology 10 (2000) 77±95 www.elsevier.com/locate/euroneuro

Review

5-HT3 receptor antagonists and anxiety; a preclinical and clinical review

Berend Oliviera,b,c , Ineke van Wijngaarden d , Willem Soudijn d

aDepartment of Psychopharmacology, Faculty of Pharmacy, Rudolf Magnus Institute for Neurosciences, University of Utrecht, Utrecht, The Netherlands bPsychoGenics Inc., 4 Skyline Drive, Hawthorn, NY 10532, USA cDepartment of Psychiatry, Yale University School of Medicine, New Haven, CT, USA dLeiden/Amsterdam, Center for Drug Research, Leiden, The Netherlands

Received 25 May 1999; received in revised form 25 September 1999; accepted 19 October 1999

Abstract

The present paper reviews the evidence for anxiolytic activity of 5-HT3 receptor antagonists in animal models of anxiety and in clinical trials in humans. Compared to the established anxiolytics (benzodiazepine receptor agonists and, to a lesser extent, 5-HT1A receptor agonists) 5-HT3 receptor antagonists display a different anxiolytic pro®le. They are anxiolytic in a limited number of animal anxiety models. If active, they often are very potent and display bell-shaped dose response curves, whereas the ratio between therapeutic activity and side effects appears remarkably large. 5-HT3 receptor antagonists remain active after chronic dosing and no indications for tolerance, dependence or rebound effects were found, which seems to make these drugs an attractive alternative to the benzodiazepines. However, the large body of animal data indicating a complete lack of psychotropic activity of 5-HT3 receptor antagonists weakens the prediction of anxiolytic activity in these drugs. Human data are also controversial; some investigators have reported positive effects in anxiety disorders

(panic disorder, GAD), others did not. It can be concluded that 5-HT3 receptor antagonists do not represent a breakthrough in the treatment of various anxiety disorders, as initially suggested.  2000 Elsevier Science B.V. All rights reserved.

Keywords: 5-HT31A3 receptor; 5-HT receptor; 5-HT receptor antagonists; Benzodiazepines; Anxiety; Animal models; Con¯ict; Potentiated startle; Periaquaductal gray stimulation; Conditioned place preference; Defensive burying; Ultrasonic pup vocalization; Adult ultrasonic vocalization; Defeat- induced analgesia; Elevated plus-maze; Social interaction; Light±dark exploration; Stress-induced hyperthermia; Drug discrimination; Generalized anxiety disorder; Panic disorder

1. Introduction compounds like meprobamate, but these compounds, which were still considered `risky', were rapidly Traditionally, pharmacological research in the area of abandoned when the benzodiazepines were discovered and anxiety and stress treatment is very much in¯uenced by the introduced to the market in the late ®fties and early sixties. availability of anxiolytic drugs. Throughout recorded his- The benzodiazepines (BDZ) have been, and still are, the tory ethanol was, and is the standard drug for (self)treat- drugs of choice for the treatment of anxiety and stress over ment of feelings of discomfort, tension, anxiety and stress. the last three decades, because of their effectiveness and Mankind has sought purposefully for substances that relative safety. As with most other drugs, continued use modify these feelings and in the last century compounds revealed side effects, the most serious of them being such as bromide salts, paraldehyde and chloral hydrate tolerance and both psychological and physical dependence. were used in medical practice for such symptoms. This once again has led to the search for new and better In the early 1900s the barbiturates were introduced, anxiolytics without the side effects of the BDZ. Serotoner- which were by far the dominant anxiolytic agents through- gic agents seem to be promising in this regard. The out the ®rst half of the twentieth century. But considerable accidental ®nding of buspirone, a 5-HT1A receptor agonist concern arose about their safety, leading to the search for having anxiolytic effects (Goldberg and Finnerty, 1979), better alternatives. This search resulted in development of focused on the role of serotonin in anxiety processes,

leading to the development of various 5-HT1A -anxiolytics, 1.2. 5-HT3 receptor antagonists like ipsapirone, gepirone, ¯esinoxan and others (Yocca,

1990; Glennon et al., 1991a). Since the ®rst published 5-HT3 (M) receptor antagonist Halfway through the eighties, 5-HT3 receptors were bemesetron (MDL 72222) in 1983, many potent and identi®ed and this discovery was soon followed by the selective 5-HT3 receptor antagonists, belonging to different synthesis of 5-HT3 receptor antagonists. In 1987 the ®rst chemical classes, have been reported (for reviews see reports appeared describing the anxiolytic effects of the Gaster and King (1997), Gozlan (1997), King (1994) and

5-HT3 receptor antagonist ondansetron in different animal Kilpatrick et al. (1990). species and models (Costall et al., 1987a), followed by The best known compounds are tropisetron, granisetron numerous investigations on this and other 5-HT3 receptor and ondansetron (for reviews see Lee et al. (1993), Yarker antagonists, like tropisetron (ICS 205,930), bemesetron and McTavish (1994), Wilde and Markham (1996), respec- (MDL 72222), granisetron (BRL 43694) and others (Kil- tively). These compounds display nanomolar af®nities for patrick et al., 1990). Although several positive studies 5-HT3 receptors, whereas the af®nities for other transmitter were reported, negative studies also appeared quite fre- receptors tested are at least two orders of magnitude less. quently. The explanation for this perhaps can be found in the fact that the anxiolytic (and other) properties of 5-HT3 1.3. Structure of the 5-HT3 receptor receptor antagonists have some unique characteristics.

Firstly, anxiolytic effects reported occur at extremely low The 5-HT3 receptor, a member of the superfamily of doses (often in the nano or microgram/kg range). Second- ligand gated ion channels (Ortells and Lunt, 1995), is ly, often an inverted U-shaped dose-response curve is cation selective. Boess et al. (1995) demonstrated that the found and third, the extremely large safety margin between 5-HT3 receptor isolated from mouse neuroblastoma 3 rat effective dose and side effects (if any) is very remarkable. glioma hybrid cells (NG 108-15) is a pentameric homo-

Besides the effects of 5-HT3 receptor antagonists on oligomer that when modelled as a cylinder has a length of anxiety states (Barnes et al., 1992), these compounds are 11 nm, a diameter of 8 nm and a central cavity 3 nm in also claimed to be active in animal models of psychosis diameter. (Kilpatrick et al., 1990; Costall et al., 1991), memory and In analogy to other members of the superfamily the learning (Barnes et al., 1990b), and withdrawal/drug abuse subunits consist of a large extracellular NH2 terminal with (Costall et al., 1990a,b). a highly conserved 15 residues containing a Cys±Cys loop,

In the present contribution the available animal be- four putative transmembrane segments M14 ±M , a large havioural data on anxiolytic effects of 5-HT334 receptor cytoplasmic loop connecting M and M and a short antagonists and potential human anxiolytic effects are -COOH terminal (Maricq et al., 1991). In the large critically reviewed. Before dealing with the behavioural cytoplasmic loop there are two potential phosphorylation studies a short overview will be given of the history of the sites.

5-HT33 receptor, the available 5-HT receptor antagonists, Due to the spacing of polar and negatively charged their localization in the brain and some receptor neuro- amino acids in the M22 region the M helix seems a likely biology, including (electro)physiology and functional cor- candidate for the lining of the ion channel as was post- relates. ulated for the structurally closely related neuronal nicotinic acetylcholine receptor (see e.g. Leonard et al., 1988; Unwin, 1989).

1.1. 5-HT3 receptors The concept that the binding site(s) of 5-HT3 receptor agonists are situated at the extracellular N-terminal is

The name `5-HT3 receptor' was ®rst proposed in 1986 strengthened by the results published, using a receptor (Bradley et al., 1986), and immediately gained worldwide chimaera whereby the N-terminal of the 5-HT3 receptor acceptance. In fact this was a mere renaming of the subunit is exchanged for (parts of) the N-terminal of the `serotonin-M' receptor, as described by Gaddum and nicotinic a-7 receptor subunit. Functional expression of Picarelli (1957). The ®rst reference to this receptor how- one of the chimaeric constructs was induced by acetyl- ever, dates from 1953, when Rocha e Silva and coworkers choline but not by 5-HT and inhibited by nicotinic reported on the antagonistic activity of cocaine against antagonists. serotonin-induced contractions of guinea pig ileum. It was Cloning of the 5-HT3 receptor subunits from different not until the late seventies that studies by Fozard et al. animal species including man resulted in either a long

(1978, 1979) using metoclopramide and cocaine renewed (5-HT3L -A ) or a short (5-HT 3S -A ) splice variant. In the interest in what are now called 5-HT3 receptors. The short variant a string of 6 amino acids (GSDLLP) is subsequent discovery of extremely potent and selective deleted from the large internal loop between M34 and M 5-HT3 receptor antagonists ignited an explosion of re- (see Belelli et al., 1995; Miyaka et al., 1995 and references search on 5-HT3 receptor pharmacology. therein). The functional properties of the homo-oligomeric B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95 79

complexes of both variants expressed in different cell types 1.5. Functional correlates of central 5-HT3 receptors are comparable though not identical with those of the wild type 5-HT3 receptor. Central 5-HT3 receptors are involved in the release of various neurotransmitters in the brain (for review see Saito

et al., 1996). Activation of 5-HT3 receptors facilitates the 1.4. Localization of central 5-HT3 receptors release of 5-HT from guinea-pig hypothalamic, frontal cortical and hippocampal slices in vitro and rat hippocam-

Central 5-HT3 receptors have been labelled using vari- pus in vivo. This response is blocked by 5-HT3 receptor ous radioligands such as [33 H]-GR65630; [ H]-tropisetron; antagonists, providing support for the anxiolytic effect of [3 H]-Ly 278584; [ 3 H]-(S)-zacopride; [ 125 I]-(S)-zacopride these compounds.

(for review see Gozlan, 1997). Quantitative autoradio- Stimulation of 5-HT3 receptors decreases the release of graphic studies in several species, including man show that noradrenaline from rat hypothalamic slices and in vivo the highest concentration of 5-HT3 receptors is present in from rat hippocampus. This result is consistent with the four areas of the medulla oblongata: nucleus tractus antidepressant effect of 5-HT3 receptor antagonists in the solitarius, dorsal motor nucleus of the vagus nerve, nucleus learned helplessness paradigm. of the spinal tract of the trigeminal nerve and the area 5-HT3 receptor activation increases the release of dopa- postrema. Lower levels are detectable in cortical areas (the mine in vitro (rat striatal and nigral slices) and in vivo (rat piriform and entorhinal cortex), limbic areas (hippocam- nucleus accumbens). This effect is probably due to a direct pus, amygdala, septum), subcortical areas (nucleus accum- or indirect effect on the dopamine transporter or may bens, hypothalamus) and spinal cord (dorsal horn) (for involve cholecystokinine, which is released by activation review see Laporte et al., 1992). In situ hybridization of 5-HT3 receptors (Raiteri et al., 1993). experiments con®rm the presence of 5-HT3 receptor In in vitro experiments using cortical slices or human mRNA in cortical, limbic, subcortical areas and the spinal cortical synaptosomes, acetylcholine release is inhibited by cord (dorsal horn). In addition, mRNA is also detectable in stimulation of 5-HT3 receptors. In vivo, 5-HT has a dual striatum, olfactory tubercle and ventral horn of the spinal effect on the release of acetylcholine: inhibition in the cord (Tecott et al., 1993; Johnson and Heinemann, 1993). guinea pig cerebral cortex and stimulation in the rat No mRNA is found in the nucleus tractus solitarius, the hippocampus. dorsal motor nucleus of the vagus and the area postrema, areas with a high level of 5-HT3 receptors. Tecott et al. (1993) suggest that the lack of mRNA in these areas might be due to a presynaptic localization of 5-HT3 receptors on 2. Animal models of anxiety peripheral afferents to the dorsal vagal complex. This suggestion is in agreement with lesion studies (Tecott et There are basically two types of animal behaviour al., 1993) and with `in vitro' and `in vivo' release models used to detect anxiolytic drugs. Models can be experiments (vide infra). based on conditioned behaviour and involve responses

The localization of 5-HT3 receptors in the dorsal vagal controlled by operant conditioning procedures. The other complex is in good agreement with the antiemetic effect of type of models involves unconditioned behaviour which

5-HT3 receptor antagonists in cancer chemotherapy (for rely on natural behavioural reactions and do not require review see Sorbe, 1996). The presence of 5-HT3 receptors speci®c training. Often the latter models rely on species- in cortical and limbic region is supportive for a possible speci®c responses (e.g. social interaction, ultrasonic vocali- role of 5-HT3 receptors in `emotional' behaviour (this zation) and are sometimes referred to as `ethologically- review). based' models (Lister, 1990). Table 1 summarizes these

Table 1 Animal behaviour paradigms frequently used in the study of anxiolytic drug effects Conditioned models Unconditioned models Con¯ict procedures Light±Dark exploration Fear-potentiated startle Elevated Plus Maze Periaqueductal Gray Stimulation Defensive Burying Drug Discrimination Ultrasonic Vocalization Conditioned Tasted Aversion Social Interaction Conditioned Place Preference or Aversion Defeat-induced Analgesia Conditioned ultrasonic vocalization Stress-induced Hyperthermia 80 B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95 models currently in use to study the anxiolytic effects of some (ondansetron, granisetron, zacopride, tropisetron) but drugs. not all (bemesetron, DAU 6215) 5-HT3 receptor antago- In the following sections the effects of 5-HT3 receptor nists. In a modi®ed Geller-Seifter procedure, extinction of antagonists in conditioned and unconditioned anxiolytic con¯ict behaviour (Ketelaars et al., 1988), which is models will be described. Each model will be shortly sensitive for benzodiazepines and 5-HT1A receptor agon- described and its speci®c character outlined; when possible ists, tropisetron was also inactive (Ketelaars and Bruinvels, a short comparison with effects of benzodiazepine and 1991). In the four plate test (Mos et al., 1989) no

5-HT1A anxiolytics will be given. anxiolytic effects of ondansetron were found. In pigeons, Gleeson et al. (1989) studied ondansetron (0.001±1 mg/kg 2.1. Con¯ict models IM), tropisetron (0.001±0.3 mg/kg IM) and bemesetron (0.01±3 mg/kg IM). Ondansetron exhibited no anxiolytic In con¯ict models ongoing behaviour is suppressed due activity at any dose, but tropisetron showed an anxiolytic to the presentation of aversive consequences for the effect at two doses (0.003 and 0.01 mg/kg) and bemeset- behaviour. Punished consequences suppress either lever- ron at one dose (0.03 mg/kg). Since increases in punished pressing that is reinforced by food in hungry rats (Geller responding were variable within as well as across animals, and Seifter, 1960), key-pecking reinforced by food intake and often were not replicated when the effects of the in hungry pigeons (Barrett et al., 1986), water-licking in relevant doses were redetermined, the authors conclude thirsty rats (Vogel et al., 1971) or exploration of mice of a that tropisetron and bemesetron were at best only margi- new environment as in the four-plate test (Aron et al., nally active in the con¯ict test, which was con®rmed by 1971; Boissier et al., 1968). Signalled or unsignalled Colpaert and Koek (1991). Artaiz et al. (1995) studied the delivery of electric shock pulses may be used as aversive 5-HT3 receptor antagonists ondansetron, granisetron, consequence. The release of suppressed behaviour without tropisetron and VA 21B7 (3-[2-(49-piperonylpiperazinyl) in affecting the levels of unpunished responding following a Vogel test in rats. Ondansetron, granisetron and tropiset- pharmacological intervention is predictive of an `an- ron were all anxiolytic in a narrow dose range showing a xiolytic' effect. In rodents, pigeons and primates benzo- bell-shaped dose-response curve, whereas VA 21B7 was diazepines (BDZ) are consistently found to be effective in anxiolytic (both IP and PO) over the whole dose range these models (Barrett and Gleeson, 1991; Barrett, 1991). tested.

In rat con¯ict models non-benzodiazepine anxiolytics Overall, 5-HT3 receptor antagonists do not show up as such as buspirone have no or only weak anxiolytic effects consistent anxiolytics in con¯ict tests. If active, very (Barrett and Gleeson, 1991). Increases in punished re- narrow dose-ranges are general found. Compared to benzo- sponding in rats after 5-HT1A receptor agonists are more diazepines, 5-HT3 receptor antagonists can be considered likely seen with the Vogel-type procedure than with the as very weakly anxiolytic in these procedures, which are Geller con¯ict procedure (Barrett and Gleeson, 1991). In considered as very reliable anxiolytic paradigms. contrast, Gleeson et al. (1989) found anxiolytic activity not only of 5-HT1A receptor agonists in pigeons but to a lesser 2.2. Fear-potentiated startle response extent also of mixed 5-HT12 receptor agonists and 5-HT receptor antagonists (Barrett and Gleeson, 1991). A model consisting of a mixture of conditioned and

5-HT3 receptor antagonists have been studied over a unconditioned behaviour is the fear-potentiated startle wide range of doses. In rats, Jones et al. (1988) examined response (Davis, 1979, 1986). ondansetron in a dose range of 0.05±1.6 mg/kg IP in the The startle response of a rat to a loud tone can be Vogel water lick test, whereas Piper et al. (1988) examined augmented by prior pavlovian fear conditioning. During ondansetron (0.0005±5 mg/kg PO) and BRL 43694 the fear conditioning phase a light stimulus, serving as the (0.0005±50 mg/kg PO) both in food-reinforced and water- conditioned stimulus signals the presence of a shock reinforced con¯ict tests. Dunn et al. (1991) studied ondan- (unconditioned stimulus). During the startle response mea- setron (0.01±0.1 mg/kg IP), zacopride (0.1±1 mg/kg IP), surement, presentation of the light will augment the startle tropisetron (0.5±1 mg/kg IP) and bemesetron (MDL amplitude. Both the anxiolytic effect of benzodiazepines 72222) (10±20 mg/kg IP) in a modi®ed food±reinforced and the anxiogenic effect of yohimbine and DMCM can be con¯ict procedure. Higgins et al. (1991) studied ondanset- detected in this model (Davis, 1979, 1986; Davis et al., ron (10±1000 ng) and tropisetron (1±100 ng) in the 1979; Glenn and Green, 1989; Hijzen et al., 1995). The water-lick con¯ict test following microinjections in the partial 5-HT1A receptor agonists buspirone, gepirone amygdala. Borsini et al. (1993) studied several 5-HT3 (Davis et al., 1988; Kehne et al., 1988) and the full 5-HT1A receptor antagonists in a con¯ict test. Cervo and Samanin receptor agonists ¯esinoxan and 8-OH-DPAT (Joordens et (1995) studied ondansetron and tropisetron in punished al., 1996, 1998) are also effective in this model. However, responding in rats. None of these studies reported any Davis et al. (1988) found that the full agonist 8-OH-DPAT anxiolytic activity in con¯ict tests. However, Filip et al. and the partial agonist ipsapirone are not effective and the (1992) found anxiolytic activity in the Vogel-lick test for effect of buspirone could not be antagonised by 5-HT B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95 81 receptor antagonists or lesioning of the 5-HT system regard to potential anxiolytic effects. Using this paradigm

(Davis et al., 1988). In line with this, Joordens et al. Jenck et al. (1989a) have studied the effects of 5-HT3 (1998) also were not able to antagonize the anxiolytic receptor antagonists. Tropisetron, bemesetron and ondan- effects of 8-OH-DPAT and ¯esinoxan by 5-HT1A receptor setron had no effects, neither anti-aversive nor pro-aver- antagonists. The latter ®ndings suggest that 5-HT1A re- sive. The predictive validity of this model as an anxiolytic ceptor stimulation is not involved in the anxiolytic pro®le test is not established yet, but regarding the effects of the of 5-HT1A receptor agonists in this model. Glenn and benzodiazepines and the serotonergic drugs, the absense of Green (1989) showed, in a somewhat modi®ed procedure, any effect of 5-HT3 receptor antagonists points to their that ondansetron reduced the potentiated startle amplitude weak or absent psychotropic activities. in this model, suggesting anxiolytic activity comparable to the benzodiazepines. Nevins and Anthony (1994) studied 2.4. Defensive burying the effects of ondansetron, (R)-zacopride and granisetron in the fear-potentiated startle paradigm using two training Defensive burying, i.e. pushing and spraying of bedding protocols, one similar to those used in most laboratories to material with alternating thrusting movements of forepaws test for anxiolytic anxiety and one using a less intensive and shovelling movements of the snout directed towards a aversive stimulation. Their hypothesis was that 5-HT3 discrete source of unfamiliar, aversive or noxious stimuli, receptor antagonists are more effective under low aversive has been demonstrated in rats (Pinel and Treit, 1978), mice than under high aversive conditions. Under high aversive (Broekkamp et al., 1986), gerbils (Davis et al., 1982) and conditions, the control substances diazepam and buspirone ground squirrels (Heynen et al., 1989). The behaviour is effectively reduced the potentiated startle, but neither of interpreted as an unconditioned, species-speci®c response the 5-HT3 receptor antagonists did so, whereas under low toward certain olfactory (Jackson et al., 1984; Pinel et al., aversive conditions both diazepam/buspirone and the 5- 1981), tactile (Broekkamp et al., 1986) and visual stimuli

HT3 receptor antagonists were able to reduce it. Bill et al. (Pinel et al., 1980), which will elicit avoidance behaviour (1992) showed that the 5-HT3 receptor antagonist WAY under appropriate conditions. Together with ¯ight, freezing 100289 was active in the rat fear potentiated startle and certain forms of agonistic behaviour it constitutes the paradigm, showing a bell-shaped dose-response curve like defensive behavioural repertoire in these species. ondansetron in the Glenn and Green (1989) study. Nevins Defensive burying is also easily evoked as a conditioned and Anthony (1994) suggest that decreasing the shock response toward originally neutral stimuli with acquired intensity as a tool to manipulate anxiety, may either aversive properties, e.g. by one-trial experience with produce qualitatively different types of conditioned fear or electric shock through a wire-wrapped dowel (Hudson, alternatively may activate different brain mechanisms 1950; Pinel and Treit, 1978). Defensive burying, both as a involved in the generation of different anxiety-like states. conditioned and an unconditioned response, has been

In any case, these data suggest that 5-HT3 receptor proposed as a useful paradigm for screening of novel antagonists may display anxiolytic activity only under mild anxiolytic drugs (Treit et al., 1981; Treit, 1985; Broek- aversive conditions. kamp et al., 1986). Anxiolytics belonging to two different pharmacological

2.3. Periaquaductal gray stimulation classes (benzodiazepines, 5-HT1A receptor agonists) have been found to suppress defensive burying both in rats and Electrical stimulation of speci®c sites in the midbrain mice (De Boer et al., 1991), though the behavioural central gray (PAG) of rats induces aversive effects. Ani- selectivity of the effects of 5-HT1A receptor agonists has mals can readily be trained to switch off the current to been questioned in mice (Broekkamp et al., 1989). Broek- escape the aversive stimulation. kamp and Jenck (1989) and Njung'e and Handley (1991) This paradigm has been used by Graeff and coworkers did ®nd a selective blockade by a number of 5-HT uptake- (Graeff, 1984; Graeff et al., 1986) who reported anti- blocking drugs. Both groups also studied the effects of aversive effects of benzodiazepines, a ®nding replicated by ondansetron and tropisetron in a wide dose range. Both

Bovier et al. (1982). This suggests that aversive PAG 5-HT3 receptor antagonists were found ineffective. stimulation may be of interest for the study of mechanisms Njung'e and Handley (1991) found however, that tropiset- possibly underlying fear and anxiety. In contrast to what ron in the very high dose of 10 mg/kg (IP) facilitated the would be expected if this were the case, the 5-HT1A inhibition induced by zimelidine (10 mg/kg IP). It should receptor agonist 8-OH-DPAT had anxiogenic properties be noticed that various other 5-HT receptor antagonists had

(Jenck et al., 1989b), whereas the 5-HT2 receptor antago- a similar potentiating effect, thereby questioning the nist ketanserin had anxiolytic effects in this model, an speci®city of 5-HT3 receptors in this case. effect shared with 5-HT2C receptor agonists (Jenck et al., 1990), of which mCPP has been reported to be clinically 2.5. Ultrasonic pup vocalisation anxiogenic in sensitive patients. Thus, it can be stated that the observed effects in this model are equivocal with In the rat, removal of the pup from its nest, mother and 82 B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95 littermates elicits ultrasonic calling. These ultrasounds are have no effects whatsoever in the pup ultrasonic test. One in the frequency range of 30 to 50 kHz. The function of possibility is that 5-HT3 receptors in the CNS are not calling is probably to alert the mother and to stimulate her properly functioning at this age. to search and retrieve pups (Allin and Banks, 1972). Ultrasonic calling shows a characteristic developmental pattern. HardÊ et al. (1982) have shown that sounds increase 2.6. Vocalizations in aversive situations gradually to peak levels at about 9 to 11 days. Thereafter it gradually decreases and disappears around 16 to 18 days of Adult rats emit ultrasonic vocalizations (in approximate- age. ly the 20±28 kHz range) under various aversive conditions Several authors have reported that ultrasonic calling is such as, for example, in the presence of a predator sensitive to treatment with clinically proven anxiolytics as (Blanchard et al., 1991), an aggressive male opponent well as with putative anxiolytic compounds (Winslow and (Tornatzky and Miczek, 1995; Vivian and Miczek, 1993), Insel, 1991). Gardner and Budhram (1987) and many after exposure to a painful shock (Molewijk et al., 1995, others (cf. Olivier et al. 1998a for a review) have reported Van der Poel et al., 1989), a loud acoustic (Kaltwasser, that benzodiazepines reduce ultrasounds in rats, and simi- 1991; Miczek and Vivian, 1993) or tactile startle stimulus lar results have been found by Nastiti et al. (1991) in mice. (Barros and Miczek, 1996). Conversely, benzodiazepine inverse receptor agonists in- Interestingly, rats can also produce ultrasonic vocaliza- crease ultrasounds suggestive of anxiogenic effects. 5- tions in association with a prior aversive event without the

HT1A anxiolytics like buspirone, ¯esinoxan and ipsapirone actual physical presence of threat. This latter feature may reduce pup sounds too (Mos and Olivier, 1989; Winslow have face validity with regard to situational panic attacks, and Insel, 1991; Olivier et al., 1998b). These authors also where environmental stimuli may acquire aversive prop- established the fact that different 5-HT receptor agonists erties and become triggers for panic attacks (Molewijk et have differential effects on pup calls. Reuptake blockers of al., 1995). Ultrasonic vocalizations in adult rats may serotonin also reduce ultrasounds (Winslow and Insel, represent affective expressions by rats exposed to threaten- 1991; Mos and Olivier, 1989; Olivier et al., 1998b), but ing, startling or painful stimuli (cf. Miczek et al., 1995). noradrenaline reuptake inhibitors like desipramine increase Socially experienced rats emit ultrasonic vocalizations at a calling. Although this paradigm, like any other anxiety higher rate after being withdrawn from chronic treatment test, has its drawbacks and limitations, all authors so far with opoid drugs in a time- and dose-dependent manner have agreed upon the effects of benzodiazepines and 5- (Vivian and Miczek, 1991). Similarly, when rats are

HT1A anxiolytics as reducing ultrasounds. startled 24 h after the last of 10 diazepam injections, they Not many studies have been reported on the effects of emit high rates of ultrasonic vocalizations (Miczek and

5-HT3 receptor antagonists on rat pup ultrasounds. Mos et Vivian, 1993). These ultrasonic vocalizations during with- al. (1989), Mos and Olivier (1989) and Olivier et al. drawal from diazepam can be attenuated with the 5-HT1A (1998b), did not observe any signi®cant change in re- partial receptor agonist gepirone or with diazepam (Vivian sponse to treatment with ondansetron over a wide dose and Miczek, 1993). Withdrawal from cocaine, either via range (0.001±1.0 mg/kg IP) both in a stressful (188C) as long-term access via the drinking ¯uid or via intravenous well as in a less stressful (378C) condition. Kehne et al. self-administration in `binges', produces large increases in (1991) also reported the lack of effects on rat pup calls ultrasonic vocalizations in rats, mimicking the clinically after MDL 73,147EF (1.25±10 mg/kg). Olivier et al. observed `crash' phase (Barros and Miczek, 1996).

(1998b) tested the 5-HT3 receptor agonist phenylbiguanide Anxiolytic targets at 5-HT1A and benzodiazepine- (1±10 mg/kg IP) under high (188C) and low (378C) stress GABAA receptors decrease these ultrasounds, when rats conditions and found neither effects on ultrasounds, nor on are exposed to acoustic startle stimuli (Kaltwasser, 1991) the negative geotaxis, a measure for side effects. In a or to footshock, but not to tail shock (Cuomo et al., 1988; distress vocalization paradigm in guinea pig pups (sonic De Vrij et al., 1993; Sanchez, 1993), and also when they vocalizations), ondansetron (0.001±0.1 mg/kg IP), in are anticipating a confrontation with an aggressive oppo- contrast to benzodiazepines and 5-HT1A receptor agonists, nent (Tornatzky and Miczek, 1995). was not able to change the vocalizations (Molewijk et al., Under a conditioning procedure, when rats were ex- 1996). posed to the environment where they have received electric

No other publications have dealt with 5-HT3 receptor foot shock previously, they emitted fewer ultrasounds after ligands in such paradigms. Since in this test the results of treatment with 5-HT1A receptor agonists, but not with most other compounds are quite comparable between benzodiazepines, except for alprazolam (Molewijk et al., different laboratories, it must be assumed that minor 1995). This pattern of results suggests that this experimen- modi®cations of experimental procedures by other inves- tal procedure may serve as model for detecting anti-panic tigators will not lead to positive results with 5-HT3 activity. In this procedure, ondansetron (0.001±0.1 mg/kg receptor antagonists. IP) had no activity whatsoever. Schreiber et al. (1998)

It is as yet not clear why 5-HT3 receptor antagonists found that ondansetron (0.1 mg/kg IP) was not able to B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95 83 antagonize the decrease in ultrasonic vocalizations induced been developed, like the zero-maze (Shepherd and Rod- by paroxetine, a selective serotonin reuptake inhibitor. gers, 1990). In these paradigms the anxiolytic activity of benzo- 2.7. Defeat-induced analgesia in mice diazepines can reliably be demonstrated by the increase in percentage of open arm entries and time spent in the open

A characteristic feature of conspeci®c defence in mice is arms. Drugs stimulating the 5-HT1A receptor, such as the reduction in pain sensitivity that follows the experience 8-OH-DPAT, buspirone and ipsapirone are dif®cult to of defeat by a conspeci®c. Mice exposed to a resident in detect as anxiolytic or have either been reported to be his territory show two different forms of analgesia upon ineffective or even to show an opposite effect, i.e. decreas- defeat. In one, an opioid substrate is involved. The other is ing the percentage of entries or time spent in the open arms a non-opioid form, and behavioural and pharmacological (File et al., 1987; Pellow et al., 1987; Critchley and evidence indicates that this is a consequence of anxiety Handley, 1987; Moser, 1989; Griebel, 1995). This latter provoked by ecologically relevant aspects of the stimulus effect may be due to some anxiogenic effects of these situation (Rodgers and Randall, 1986; 1986b). Rodgers and drugs, but it may also represent a `false negative' in this coworkers (Rodgers and Randall, 1987, 1987a, 1988) have test. However, using ethologically derived measures in this shown that some benzodiazepines block the non-opioid paradigm (risk assessment and exploration) 5-HT1A re- short-lasting analgesic effects of defeat, i.e. mice show no ceptor agonists have anxiolytic potential (Griebel et al., longer the decrease in pain sensitivity. However, non- 1997). neuronal benzodiazepine agonists also reduced defeat-in- Similar results have been reported for 5-HT2 receptor duced analgesia, whereas for example chlordiazepoxide antagonists. Thus the 5-HT2 receptor antagonist ritanserin and midazolam were without effect. Thus the precise role has been reported to show both anxiolytic and anxiogenic of the benzodiazepine receptor is not yet de®nitively activity in this model (Pellow et al., 1987; Critchley and established. Handley, 1987).

5-HT1A receptor agonists like buspirone, ipsapirone and Costall et al. (1989a, 1989b) reported an anxiolytic gepirone also blocked the non-opioid defeat-induced anal- effect of the 5-HT3 receptor antagonists ondansetron, gesia, although bell-shaped dose-response curves were zacopride, tropisetron, bemesetron, granisetron and RS- observed for buspirone and ipsapirone (Rodgers and 42358-197 (Costall et al., 1993) as revealed by the Randall, 1988). The effects of benzodiazepines and 5- increase in time spent exploring the open arms of an

HT1A receptor agonists, although far from fully under- elevated plus-maze. Similar results were found by Dunn et stood, point to a role of anticipatory anxiety in the non- al. (1991) and Silvestre et al. (1996) for granisetron. File opioid defeat-induced or defensive analgesia (Rodgers, and Johnston (1989), Piper et al. (1988), Kshama et al. 1991). (1990) and Rodgers et al. (1995) failed to detect any effect

The ®rst report on 5-HT3 receptor antagonists (Rodgers of various 5-HT3 receptor antagonists in the same test. et al., 1990) showed that ondansetron suppressed the Griebel et al. (1997) tested (R,S)-zacopride, ondanset- defeat-induced analgesia. Similar, but less strong reactions ron, tropisetron and bemesetron in a rat elevated plus-maze were also observed for tropisetron, MDL 72222 and MDL test using spatiotemporal (open arm time/entry) and 73147 (Rodgers and Shepherd, 1990). It is therefore ethologically derived measures (risk-assessment/explora- suggested that 5-HT3 receptor antagonists might act as tion). Over quite a dose-range (0.01±1 mg/kg SC) no anxiolytics in this paradigm, although a peripheral mecha- effect on spatiotemporal parameters were found, whereas nism of action cannot be excluded. only zacopride and ondansetron reduced the number of aborted attempts to enter the open arms. It is unclear why

2.8. Elevated plus-maze different 5-HT3 receptor antagonists display differential effects. The elevated plus maze as used by Handley and Mithani Andrews and File (1993) showed that (R, S) Zacopride (1984) is a procedure that relies on exploratory behaviour had an anxiolytic effect in this model only in handling- by rats that are placed in an elevated maze, consisting of naive rats, but not in handling-habituated rats. (R, S) two (opposite) open and two walled alleys. The animals Zacopride is considered a partial receptor agonist, which, will explore the different alleys (total number of entries). according to Andrews and File (1993) may behave as a The open arms are more aversive to the rats than the weak agonist in conditions of low 5-HT functioning, but as closed ones, as revealed by a preference of the animals to a full antagonist under conditions of high 5-HT func- explore the closed alleys. Anxiolytic drugs will help to tioning. overcome the fear-induced inhibition of open alley explo- Rodgers et al. (1997b) administered ondansetron (0.1± ration. The ratio of open- versus total arm entries gives a 100 mg/kg) subchronically (21 days) to mice, and found sensitive measure of the anxiolytic effect of a drug and this no evidence for anxiolytic activity using an ethologically measure is not very sensitive for general effects on derived elevated-plus maze scoring system in mice, in exploration. Recently, modi®cations of the plus-maze have contrast to chlordiazepoxide (Rodgers et al., 1997a). 84 B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95

The lack of effect and the inconsistency of effects of 5-HT3 receptor antagonist S(2)-zacopride was devoid of new putative anxiolytic drugs such as 5-HT1A receptor activity (Barnes et al., 1990a). The ®nding that S(2)- agonists and 5-HT3 receptor antagonists is at variance with zacopride exerted 5-HT3 receptor agonistic activity in the the effects of classic anxiolytics in the elevated plus-maze ferret, whereas the R(1)-isomer behaved as an antagonist (Moser, 1989). (Middlefell et al., 1990) may (partly) explain the confus- ing effects of racemic zacopride. 2.9. Social interaction test 2.10. Light-dark exploration In this test, ®rst described by File (1980), either two rats or two mice unfamiliar to each other are placed in an This model was originally described by Crawley and arena. This arena may either be familiar or unfamiliar to Goodwin (1980) and is based on the natural aversion of the animals, and may be brightly lit or not. The most mice and rats for brightly lit places. In a two compartment aversive situation is obviously a brightly lit, unfamiliar box, one dark and one brightly lit, the total activity, the environment in which low levels of social interaction time spent in the light compartment and the crossings occur. The time spent by the animals in social interaction between the light and dark compartment provides in- is measured as well as locomotor activity. Benzodiazepines formation about the preference of the animal for the dark increase the time spent in social interaction. A similar compartment. As anxiolytics should reduce the natural effect has been reported for the 5-HT1A receptor agonists aversion to light, the essential feature of this model is that buspirone and ipsapirone (Gardner, 1985; Schuurman et anxiolytic drugs increase the number of crossings and/or al., 1987), though File (1985) could not con®rm the the time spent in the light compartment. The latter anxiolytic potential of buspirone in this model. parameter is generally considered to be the most relevant

The 5-HT3 receptor antagonist ondansetron increased one. social interaction (Jones et al., 1988; Piper et al., 1988; Benzodiazepines are reliably detected in this paradigm Dunn et al., 1991), a result which could not be con®rmed using mice (Crawley, 1981; Jones et al., 1988; Kilfoil et by File and Johnston (1989). Controversies exist over al., 1989; Costall et al., 1989c). Also the 5-HT1A receptor zacopride (active according to Dunn et al., 1991 and agonist buspirone has been reported to show an anxiolytic Costall et al., 1988a,b,c; not active according to File and pro®le in this test (Costall et al., 1988b; Carli et al., 1989;

Johnston, 1989) and granisetron (active according to Piper Kilfoil et al., 1989). The 5-HT3 receptor antagonists et al., 1988 and Kennett et al., 1990; inactive according to ondansetron, bemesetron, tropisetron, zacopride, itasetron File and Johnston, 1989). Dunn et al. (1991) found the (DAU 6215), WAY 100289, RS-42385-197 and WAY- anxiolytic activity of bemesetron in this paradigm at SEC-579 have been reported to produce similar effects as relatively high doses concomitantly producing sedation. the benzodiazepines (Costall et al., 1987a, 1993; Kilfoil et

RS-42358-197 appeared a very potent 5-HT3 receptor al., 1989; Onavi and Martin, 1989; Bill et al., 1992; antagonist in this test (Costall et al., 1993). Cutler (1990) Borsini et al., 1993; Middlefell et al., 1996a,b). Morinan reported anxiolytic effects of granisetron and tropisetron in (1989) did not ®nd anxiolytic activity of ondansetron and a social interaction test in gerbils. BRL 24924 in a rat black/white test box, but Sanchez

Because 5-HT3 receptor antagonists may display an- (1996) found ondansetron active in a similar set-up. Also, xiolytic effects, 5-HT3 receptor agonists may exert an- Kshama et al. (1990) using mice were not able to detect xiogenic effects. Mitchell et al. (1991) tested metach- anxiolytic activity in a number of 5-HT3 receptor antago- lorophenylbiguanide (mCPBG; intraventricularly, because nists. The anxiolytic activity of the two isomers of they suggest that the compound does not penetrate the zacopride does not parallel their 5-HT3 receptor blocking blood-brain barrier) in the social interaction test in rats. activity (Pinkus et al., 1990; Waeber et al., 1990). The mCPBG reduced social interaction under low light/famil- most active antagonist has been found to be less active iar conditions, but not under high light/unfamiliar con- than the weaker R(1)-isomer of zacopride, which is highly dition, which is considered the most anxiogenic condition. active (Barnes et al., 1990a; Young and Johnson, 1991).

Apparently, when anxiety levels are not high, a 5-HT3 On the other hand, Bill et al. (1991) could not ®nd receptor agonist may have anxiogenic effects. differences in the anxiolytic potential of R(1), S(2) and Thus some groups report an anxiolytic pro®le of the new racemic zacopride in the mouse light/dark box using drug types (5-HT1A receptor agonists and 5-HT 3 receptor female mice. Moreover, they found a good correlation antagonists) whereas others fail to ®nd such an effect. The between the minimally effective doses in the light/dark recent ®ndings with the optical isomers of zacopride add box and the minimum doses required to block the von further confusion to the understanding of possible an- Bezold±Jarisch re¯ex in mice. This may point to species xiolytic activity of 5-HT3 receptor antagonists. It was differences. Both isomers of zacopride possess long acting reported that the least active R(1)-isomer of zacopride anxiolytic-like effects (.16 h) (Young and Johnson, (de®ned biochemically: Pinkus et al., 1990; Waeber et al., 1991). 1990) increased social interaction, whereas the more active Costall and Naylor (1991) tested ondansetron for 14 B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95 85 days (10 mg/kg IP b.i.d.) and showed that it retained its 1990a,b). The same authors showed that ondansetron did anxiolytic activity in a white/dark box in mice and in a not affect the stress-induced hyperthermia, which was social interaction test in rats. Moreover, no evidence for con®rmed by Zethof et al. (1991, 1994, 1995). Borsini et tolerance or sensitization occurred. After stopping treat- al. (1993) could also ®nd no evidence for anxiolytic effects ment, the anxiolytic effects slowly waned until control of 5-HT3 receptor antagonists in the stress-induced hy- levels were reached after about 96 h. No rebound anxiety perthermia test. was observed. A similar pattern was observed with buspirone (1.0 mg/kg IP b.i.d.), but not with diazepam, which, upon cessation of treatment, led to anxiogenic effects. 2.12. Drug discrimination When diazepam (7 days treatment, 10 mg/kg IP b.i.d.) was withdrawn, a clear anxiogenic effect was noted when mice Drug discrimination has not been developed as a model were tested 8, 24, 48 and 96 h after withdrawal in the for anxiety and cannot simply be ®tted into a categoriza- black-white exploration test (Costall and Naylor, 1991). tion of animal models as homologous, physiological or This withdrawal anxiety could not be antagonized by empirical (Slangen, 1991). It is suggested that the drug- buspirone, but ondansetron (10 mg/kg IP b.i.d.) was able discrimination procedure in animals is the closest available not only to restore control levels but indeed gain the experimental model for assessing the subjective effects of anxiolytic level previously obtained by diazepam. A drugs. Moreover, drug discrimination studies may provide similar result was found in the social interaction test in additional information about receptor mechanisms in- rats. RS-42358-197 (Costall et al., 1993) appeared very volved in the action of diverse anxiolytic drugs. Drug effective in reducing withdrawal anxiety after stopping discrimination procedures can reveal very subtle differ- treatment with alcohol, nicotine, cocaine or diazepam. ences in the mechanism of action of different drugs, as e.g. WAY-SEC-579 was also able to antagonize the anxiolytic- evidenced in the 5-HT system (Stolerman et al., 1987). like pro®le after diazepam-withdrawal in the light/dark Drug discrimination seems remarkably sensitive for in vivo box (Middlefell et al., 1996a,b). Repeated co-administra- measurement of a selective drug effect, especially with tion of WAY-SEC-579 with diazepam prevented the de- regard to the mechanism of action or the involvement of velopment of diazepam withdrawal-induced anxiety (Mid- various neurochemical systems. In this sense it is worth- dlefell et al., 1996a,b). WAY-SEC-579 itself did not lead to while to consider whether it is feasible that manipulation withdrawal anxiety. However, when rats were made com- of a central 5-HT3 receptor leads to a discriminative pletely tolerant to the anxiolytic effects of oxazepam in a stimulus (cue) for an animal. The available evidence is modi®ed light-dark box (21 days, 5 mg/kg/day IP), very meager in contrast to the evidence on 5-HT1A ondansetron (0.1 mg/kg IP) did not have anxiolytic receptor agonists and benzodiazepines (Barrett and activity, whereas in partially tolerant rats (7 or 14 days Gleeson, 1991). Glennon et al. (1991b, 1992) describe that oxazepam), ondansetron was active. Apparently, ondanset- the 5-HT3 receptor agonists 2-methyl-5-HT (5 mg/kg) and ron displays different activities during stages of full metachlorophenylbiguanide (mCPBG) could be discrimi- tolerance or withdrawal (Nowakowska et al., 1998). nated from saline. The 2-methyl 5-HT stimulus (ED5052.6 The most unequivocal results with the new types of mg/kg) generalized to mCPBG (ED5051.7 mg/kg) and anxiolytic drugs are obtained in this model. The results was antagonized by zacopride and tropisetron, but never obtained with the isomers of zacopride, however, suggest completely by the quaternary salt of tropisetron, suggesting the involvement of another (subtype?) receptor, at least in that the cue was centrally mediated. Till now no reports of rats. 5-HT3 receptor antagonists as training stimuli have appeared, although Tricklebank (1989) suggests that tropiset-

2.11. Stress-induced hyperthermia ron has weak stimulus properties. 5-HT3 receptor antagonists have been used as tools in order to study whether The arousing effect of handling can induce hyperthermia substitution or antagonism of a certain drug cue occurs. in rats (Poole and Stephanson, 1977; Eikelboom, 1986). 5-HT3 receptor antagonists were neither able to antagonize Using this phenomenon of stress- or handling-induced 5-HT1A receptor mediated cues (Arnt, 1989; Nader et al., hyperthermia in mice, Borsini et al. (1989) suggested that 1989; Ybema et al., 1991), nor a mixed 5-HT1A/1B the higher rectal temperature of mice removed last from receptor mediated cue (eltoprazine; Ybema et al., 1992). their home cage compared to those removed ®rst from the The 5-HT3 receptor antagonists tropisetron and bemeset- same cage can be regarded as a form of anticipatory ron, but not zacopride were able to antagonize the ethanol anxiety. cue in pigeons (Grant and Barrett, 1991) and in rats (Grant This increase in rectal temperature was prevented in a and Colombo, 1993). Although structural/chemical fea- dose dependent way by prior treatment with benzodiaze- tures may underlie the differential effects of tropisetron pines and the 5-HT1A receptor agonists 8-OH-DPAT and and bemesetron on one side and zacopride on the other buspirone, and was not affected by non-anxiolytic com- side in antagonizing the ethanol cue in pigeons, it is also pounds (e.g. imipramine, haloperidol) (Lecci et al., possible that the 5-HT4 receptor agonistic activity of 86 B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95

zacopride (Bockaert et al., 1990) interferes with the 5-HT3 could be trained and the 5-HT precursor 5-HTP fully receptor antagonistic activity. substituted (dose-dependently) for mCPBG. The 5-HT3 Stefanski et al. (1996) trained rats on ethanol (1 g/kg) receptor antagonist zacopride only partially antagonized versus saline in a two lever operant conditioning paradigm. the cue and the rate-suppressant effects of mCPBG,

None of a very broad range of doses (0.001±10 mg/kg) of suggesting a partial 5-HT3 receptor mediated cue. Cocaine the two 5-HT3 receptor antagonists tropisetron or ondan- did only partially (50%) mimic the mCPBG-cue in substi- setron was able to antagonize the discriminative stimulus tution tests. Glennon et al. (1992) were able to train 4 out properties of ethanol. The 5-HT3 receptor agonist mCPBG, of 6 rats to discriminate 2-methyl 5-HT (5 mg/kg IP) from centrally or intraperitoneally administered, did not substi- saline (ED5052.6 mg/kg). The cue generalized to the tute for the ethanol cue. In an alternative to operant drug 5-HT350 receptor agonist mCPBG (ED 51.6 mg/kg), but discrimination, the cross-familiarization conditioned taste not to the 5-HT21/2 receptor agonist DOM or the 5-HT aversion (CTA) procedure, pre-exposure to centrally ad- receptor agonist 5-OMe-DMT. The stimulus of 2-methyl- ministered mCPBG was also not able to alter the ethanol- 5-HT could be almost completely antagonized by tropiset- induced CTA, whereas ethanol itself was (Bienkowski et ron (0.1±10 mg/kg), but not by a quaternary amine analog al., 1998). These data suggest that 5-HT3 receptors are not of tropisetron (even not at doses 10,000 times the dose of primarily involved in ethanol's stimulus properties. Stefan- tropisetron), strongly suggesting that 2-methyl-5-HT's cue ski et al. (1996) discuss several possible reasons why their is mediated via the 5-HT3 receptor and is centrally ®ndings deviate from those found by Grant and Barrett mediated. Further support (Young and Glennon, 1992) for

(1991) and Grant and Colombo (1993), e.g. dose, route of a 5-HT3 receptor mediated cue of 2-methyl-5-HT came administration, and training dose of ethanol. from its antagonism by (6) zacopride and its isomers in a

Similarly, 5-HT3 receptor antagonists were not able to stereo selective way consistent with their 5-HT3 receptor antagonize the discriminative stimulus properties of other binding pro®le (S(2).RS(6).R1). These data suggest major drugs of abuse, such as morphine (Joharchi et al., that the cue of mCPBG is serotonergic in nature and only

1993), cocaine (Paris and Cunningham, 1991; Lane et al., partially mediated by 5-HT3 receptors, whereas 2-methyl- 1992; Schechter, 1993; Johanson and Barrett, 1995; De La 5-HT seems to engender a full 5-HT3 receptor-mediated Garza et al., 1996) or d-amphetamine (Moser, 1992). cue.

Cocaine has af®nity for the 5-HT3 receptor (Kilpatrick et However, in an unpublished study, we (Slangen, 1991) al., 1987) and has structural similarity to some 5-HT33 tested two 5-HT receptor agonists, 2 methyl-5-HT and receptor antagonists, e.g. tropisetron and bemesetron 3,4-dichlorophenylbiguanide in a two lever drug-discrimi- (Richardson et al., 1985). In cocaine-trained (10 mg/kg) nation paradigm in rats. 2-methyl-5-HT, in a dose range of rats (Paris and Cunningham, 1991), tropisetron (2±24 0.1 to 4.0 mg/kg PO could not be discriminated from mg/kg) and bemesetron (2±16 mg/kg) were not able to vehicle in more than 80 training sessions in contrast to either substitute for or antagonize the cocaine-cue. This what was found by Glennon et al. (1991b, 1992; Young was somewhat unexpected because 5-HT3 receptor bloc- and Glennon, 1992). However, the same animals thereafter kade can interfere with the behavioural effects of several rapidly learned to discriminate chlordiazepoxide. Further- drugs of abuse like cocaine or amphetamine (Costall et al., more it was demonstrated that 2-methyl-5-HT, which very

1987b; Van der Hoek and Cooper, 1990). Apparently, weakly binds to 5-HT1A -receptors, does not substitute for stimulation of 5-HT3 receptors does not play an important ¯esinoxan in ¯esinoxan-trained animals. In a completely role in the discriminative stimulus and rewarding prop- parallel way, ondansetron (0.1 to 4.0 mg/kg) was also erties of stimulants (d-amphetamine, cocaine), although neither able to function as a discriminable cue, nor to they may do in the case of alcohol (Grant and Barrett, replace the chlordiazepoxide-cue.

1991). However, the 5-HT3 receptor agonist mCPBG The effects of the 5-HT3 receptor agonist 3,4-dich- showed partial substitution (36%) for the cocaine stimulus lorophenylbiguanide (3,4DCPBG) were rather complex. in rats trained to discriminate 10 mg/kg cocaine from The results suggest that, although 3,4DCPBG may be able saline. Moreover, pretreatment with a 5.6 mg/kg dose of to induce discriminable stimulus effects in rats, these mCPBG shifted the dose-response curve for lower doses of effects are obscured because food rewarded bar pressing is cocaine to the left, but not for higher doses (Koetzner et disrupted at higher doses (8 mg/kg). It may be suggested al., 1995). These data indicate only a very limited role for that other drug discrimination tests, not involving food

5-HT3 receptors in cocaine's discriminative stimulus, rewards, are more suited for training rats to discriminate which was further supported by a study where animals between 3,4DCPB and saline. were either trained on cocaine or on mCPBG (De La In conclusion: Activation or blocking of centrally lo-

Garza et al., 1996). mCPBG only very marginally (max. cated 5-HT3 receptors does not generate clear discrimina- 50%) substituted for cocaine, whereas coadministration of tive stimulus effects. Whether certain stimuli (e.g. alcohol) mCPBG and cocaine did not change generalization. Nei- can be blocked by 5-HT3 receptor antagonists (but not by ther ondansetron, nor bemesetron were able to antagonize others) is still a matter of controversy. The data sofar are cocaine's cue. In this study, also mCPBG (15 mg/kg) not pointing to clearly detectable (discriminable) stimulus B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95 87

effects of 5-HT3 receptor antagonists, in contrast to what no anti-aggressive effects, in contrast to 8-OH-DPAT, has been demonstrated for compounds acting at 5-HT1A , which had an aggression-enhancing effect. 5-HT1B , 5-HT 2 and GABA A -benzodiazepine receptors. When mice were placed in a so called communication box for 16 h, in which non-shocked mice (responders) 2.13. Miscellaneous were placed adjacent to mice receiving shocks (senders), both types of mice developed gastric ulcers. The responder When marmosets or cynomolgous monkeys are con- mice are thought to be psychologically stressed and fronted with a human observer, they retreat from the cage developed more severe gastric ulcers than the physically front to the rear and show characteristic postures indicating stressed animals. 5-HT3 receptor antagonists (ondansetron fear (Costall et al., 1988). Treatment with diazepam or (ID5050.1 mg/kg PO), tropisetron (ID 5050.2 mg/kg PO) ondansetron makes animals less fearful, i.e. they spend and MDL 72222 (ID5050.4 mg/kg PO) were able to more time in the front of the cage and show less fearlike prevent ulcer formation, whereas a peripherally acting behaviour (Jones et al., 1988). In subsequent experiments 5-HT3 receptor antagonist (M-840) was not (Nomura et other 5-HT3 receptor antagonists were also found effective al., 1994), suggesting the involvement of centrally located (Tyers et al., 1987; Piper et al., 1988; Borsini et al., 1993; 5-HT3 receptors. In a procedure in female mice where Costall et al., 1993). animals displayed cataleptic-like immobility after repeated Cutler (1991) studied the effects of buspirone, graniset- pinching at the scruff, ondansetron (0.1 and 1 mg/kg IP) ron and tropisetron (Cutler and Dixon, 1989) in an decreased the frequency and shortened the duration of approach±avoidance situation, where female mice in a pinch-induced catalepsy at non-sedating doses (Fundaro, neutral cage were confronted with an unfamiliar male. 1998). This immobility resembles behaviour seen in Both buspirone and granisetron (given for 5±10 days in natural situations under high stress or fear conditions the drinking water) decreased ¯ight of the females and (inescapable stress). enhanced active social investigation, which was taken as indicative of an anxiolytic action of the drugs. Such an 2.14. General discussion animal data effect was not found in male mice. Moreover no anti- aggressive effects of granisetron were found, which is in In the present review the effects of 5-HT3 receptor line with similiar ®ndings using ondansetron (Mos et al., antagonists in various putative animal paradigms of anxie- 1989; Sanchez et al., 1993) and zacopride (White et al., ty are described and these effects are compared with those

1991) in mice and ondansetron in rats (Mos and Olivier, of the established anxiolytic benzodiazepines and 5-HT1A 1990; Sanchez, 1993). Cutler and Piper (1990) found receptor agonists. Table 2 summarizes the overall conclu- enhanced social investigation after administration of sions drawn by us from the various studies performed and granisetron to male gerbils. When granisetron, BRL 46470 described. and diazepam were given to intruder mice in a resident- In practically all paradigms used, benzodiazepine re- intruder situation (Cutler and Piper, 1991), all three ceptor agonists show up as potent anxiolytic agents. 5- compounds exerted behavioural changes indicative of HT1A receptor agonists have a different pro®le, being very anxiolytic activity. Placing a hungry rat into an unfamiliar environment Table 2 with food, suppresses feeding behaviour compared to a Summary of anxiolytic effects of 5-HT3 receptor antagonists, benzo- familiar environment. Benzodiazepines (Cooper, 1980) and diazepine (BZ) receptor agonists and 5-HT1A receptor agonists in anxiety paradigmsa 5-HT1A receptor agonists (Dourish et al., 1986) reduce this hyponeophagia. Fletcher and Davis (1990) con®rmed these 5-HT31A BZ 5-HT effects for chlordiazepoxide, 8-OH-DPAT and buspirone. receptor receptor receptor Tropisetron (0.01±1 mg/kg) however had an opposite antagonists agonists agonists effect, although only at the lowest dose tested. Con¯ict Models 0 111 0/1 Shepherd and Rodgers (1990) used an elegant experi- Fear-potentiated startle 1111 ment to determine whether feeding behaviour in male mice Periaqueductal Gray stimulation 0 11 2 Conditioned Ultrasonic vocalization 0 0/1 111 was differentially affected under basal (free-feeding) or Vocalizations in Aversive situations 0 111 111 social con¯ict conditions. They reasoned that elicitation of Defensive Burying 0 11 11 feeding under con¯ict situations may point to a speci®c Ultrasonic Pupvocalisation 0 11 111 direct effect of the drug on mechanisms controlling food Defensive Analgesia 11111 intake. 8-OH-DPAT enhanced feeding both under basal Elevated Plus-maze 1/0 11 0/2 Social Interaction 1/0 11 1/0 and con¯ict conditions. The 5-HT3 receptor antagonist Light-Dark Exploration 11 11 1/0 ondansetron (1.0±2.0 mg/kg) enhanced feeding only under Stress-induced Hyperthermia 0 111 11 basal conditions. Rodgers and Shepherd conclude that Drug-discrimination (cue) 0 111 111 ondansetron does not show up as an anxiolytic agent in a ±5anxiogenic effect; 05no effect; 1/11/1115weak/moderate/ this con¯ict procedure. In the same set-up ondansetron had strong (anxiolytic) effects. 88 B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95 active in some paradigms, and not (or even anxiogenic) in highly unwanted; clinical proof of anxiolytic activity of a some others. drug not predicted by the animal models is rather frequent,

It is remarkable that the anxiolytic activity of 5-HT1A e.g. in the case of 5-HT 1A receptor agonists. The 5-HT 3 receptor agonists was not predicted from animal studies. receptor antagonists may be proof of false positives, at

Thus, buspirone, a 5-HT1A receptor agonist and dopamine least in some animal paradigms. This should be considered D2 -receptor antagonist, was tested clinically as an antipsy- a serious explanation for the results described in this chotic, but found inactive; concomitantly anxiolytic activi- review. ty was shown, leading to its development as an agent in Thus, some preclinical evidence for anxiolytic effects of

Generalized Anxiety Disorder (Goa and Ward, 1986). This 5-HT3 receptor antagonists is present, though not as ®nding triggered a new area of research and development massive as one would hope for. Especially the data on of 5-HT1A (partial) receptor agonists, of which several are chronic application look promising. The results obtained so in advanced stages of development (ipsapirone, gepirone, far, though derived from a limited number of laboratories ¯esinoxan). Although 5-HT was already earlier implied in and in need of independent replication, show evidence for the modulation of anxiety (Stein et al., 1973), the an- lack of tolerance and absence of withdrawal symptoms and xiolytic effects of 5-HT1A receptor agonists renewed the rebound anxiety, thereby sharply contrasting the effects of interest in the role of other serotonergic receptors in the benzodiazepines. Such effects would strongly favour anxiety. The synthesis of speci®c 5-HT3 receptor antago- the development of this class of drugs for treatment of nist and (besides other effects like anti-emesis) the sub- anxiety disorders, including generalized anxiety disorder, sequent ®nding of their highly potent anxiolytic effects in panic disorder, phobias and others. The ®nding that some animal anxiety tests, led to an enormous effort in ondansetron antagonizes the rebound anxiety after benzo- several pharmaceutical companies to synthesize and de- diazepine withdrawal (Costall et al., 1989a,b; Goudie and velop such speci®c agents. Leathley, 1990; Costall and Naylor, 1991) may be of great

Considering the activity of 5-HT3 receptor antagonists importance because of its implications for withdrawal in various anxiety tests (Table 2), it is clear that these anxiety upon cessation of drugs of abuse, like alcohol, agents deviate in their behavioural pro®le from both the cocaine or nicotine. benzodiazepines and the 5-HT1A -ligands. In several paradigms (con¯ict, periaquaductal gray stimulation, defensive burying, ultrasonic pupvocalization, stress-induced 3. Clinical studies hyperthermia and drug-discrimination) they are completely inactive, whereas in other models contradicting results are The ®nal proof for putative anxiolytic effects of a new described. Only in one test, the light/dark exploration in class of drugs is testing in humans. Several 5-HT3 receptor mice, 5-HT3 receptor antagonists are always consistently antagonists are presently in development or are marketed found. for various therapeutic areas (emesis, anxiety, cognitive

In some tests there are remarkable differences; in the dysfunctions, drug addiction and psychosis). Several 5-HT3 social interaction test in rats, some groups ®nd very potent receptor antagonists have been or are studied clinically as and very consistent effects of all 5-HT3 receptor antago- potential therapeutics in the treatment of anxiety disorders nists (Costall, Glaxo), whereas other groups (File) essen- (Roca et al., 1995). However, there is still a paucity of tially ®nd no effects. A similar situation holds for the fully published papers on the results of these studies. elevated plus-maze. It is unclear why such dramatic Most clinical work has been done on ondansetron. In differences are found, but it is conceivable that subtle (but human volunteers no sedation was observed, whereas maybe also less subtle) differences in methodology constipation emerged as the main signi®cant side-effect (strains, handling, housing etc, etc) can account for such (Millson and Preston, 1991; Hall and Ceuppens, 1991). A effects. Although positive reports in the elevated plus- similar pro®le was observed for granisetron (Leigh et al., maze, social interaction and light/dark exploration tests by 1991). far outnumber the negative reports, this may be partly due Lader (1991) summarized the preliminary ®ndings of to a publication bias; negative results in a test are mostly ondansetron in a multi center, double-blind, placebo-con- not published in the understanding that more investment is trolled, parallel group study with 58 general practitioners needed to optimalize the model in order to ®nd the and a total of 402 patients, suffering from generalized

(dramatic) effects of 5-HT3 receptor antagonists. Alter- anxiety disorder. Two doses of ondansetron (1 mg/kg t.i.d. natively, a number of animal paradigms may be too and 4 mg/kg t.i.d.) were compared to diazepam (2 mg/kg sensitive and pick up properties of drugs which are not t.i.d.) and placebo for a treatment period of 4 weeks, related to anxiety thus generating false positives. Although followed by a two-week withdrawal period. Notwithstand- speculative, the social interaction test and to a lesser ing a very high placebo response, some ef®cacy was extent, the elevated plus maze, could be such tests. The shown, the 1 mg/kg dose being slightly superior to the 4 predictive validity of animal models of anxiety has always mg/kg dose, which could be interpreted as being in line to be considered cautiously. Mostly, false negatives are with the bell-shaped dose-response curves emerging in B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95 89 animal models. Diazepam also had a limited ef®cacy in ef®cacy of ondansetron in panic disorder. In a multicentre this set-up. There was no rebound anxiety after ondanset- double-blind, placebo-controlled clinical trial ondansetron ron cessation, in contrast to diazepam, which con®rms the reduced frequency and intensity of panic attacks. In animal data and also suggests no dependence liability. another paper presented by Bell and De Veaugh-Geiss the Mathew and Wilson (1991) measured the effects of ef®cacy of ondansetron was also shown in social phobia diazepam (0.12 mg/kg), ondansetron (0.24 mg/kg) and (275 patients). Full papers of these studies however have saline on cerebral blood ¯ow (CBF) and anxiety in GAD- not been published yet. In a small study, 14 patients with patients. Diazepam, but not ondansetron or saline reduced panic disorder or social phobia, were challenged with CBF and anxiety. pentagastrin, a CCK-receptor agonist, which induces anxi- Schweizer and Rickels (1991) are less optimistic about ety and panic in such patients (MCann et al., 1997). the anxiolytic properties of 5-HT3 receptor antagonists in Pentagastrin induced anxiety, physical symptoms of panic humans. They shortly describe two ondansetron (daily attacks, pulse, plasma ACTH and cortisol, which could not dosage range 8±48 mg) and one zacopride (daily dose be blocked by ondansetron, suggesting that pentagastrin's range 0.04 mgto400mg) studies in GAD, but report no effects (behavioural and physiological) are not 5-HT3 clear-cut signi®cant treatment effects different from receptor mediated. The ef®cacy of ondansetron as adjunc- placebo. Negative results were also reported in a review tive medication in the discontinuation of chronic benzo- article by Wilde and Markham (1996). In a randomized diazepine treatment (alprazolam or lorazepam) was investi- double blind trial 0.24 mg/kg ondansetron did not reduce gated in 108 patients (Romach et al., 1998). 97 patients anxiety in GAD patients whereas 0.12 mg/kg diazepam completed a randomized double-blind discontinuation did. In the review by Greenshaw and Silverstone (1997) in treatment program during which they received either a preliminary study, it was shown that the anxiolytic effect ondansetron (2 mg twice daily) or placebo and ¯exibly of ondansetron in GAD patients was similar to that of tapered their benzodiazepine over a 6-week period. There diazepam (57±59% vs. placebo 45%) when the Hamilton were no signi®cant effects of ondansetron on any outcome Anxiety Rating scores were used. Using the Montgomery parameter and the authors (Romach et al., 1998) concluded AsbergÊ Depression Rating Scale the anxiolytic effect is a that ondansetron had no signi®cant effects on severity of little higher than that of diazepam (54±59%) vs. 51% of withdrawal symptoms or levels of anxiety. placebo. In a small study, 10 OCD patients which were on a Lecrubier et al. (1993) observed in a double-blind, stable treatment with the SSRI ¯uvoxamine (200±300 mg placebo-controlled evaluation of 0.5, 5, or 25 mg/day of daily) were treated with ondansetron (8 mg/day) or tropisetron in 91 GAD patients a dose-dependent ef®cacy placebo in a double-blind randomized controlled design for after 7 days of treatment (Hopkins Symptom Checklist two weeks (Smeraldi et al., 1992). Ondansetron was not total score, Global Impression Scale). However using the effective in further changing the antiobsessional effects of Hamilton Anxiety Scale statistical signi®cance was not ¯uvoxamine. reached. Freeman et al. (1997) did a randomized, double- Of course, much more studies over an extensive dose blinded, placebo-controlled study in Generalized anxiety range are needed to decisively conclude whether 5-HT3 disorder patients which was part of a parallel multicenter receptor antagonists are the `anxiolytics' for the coming study with eight participating groups. Ondansetron (0.25 decades, but the data so far are far from promising. mg b.i.d., 0.5 mg b.i.d.), diazepam (5 mg b.i.d.) or placebo were given after a placebo run-in phase for 1±2 weeks. The study lasted eight weeks. Only 1 mg daily ondansetron signi®cantly decreased anxiety as measured by the HAM.A Acknowledgements and CGI severity scales. GAD was diagnosed in this study using DSM-III-R diagnostic (18) criteria, whereas in We thank Marijke Mulder for secretarial assistance. DSM-IV six core symptoms are used. Since ondansetron was found earlier ineffective in GAD, and suggested to be effective (absence of placebo group) in panic disorder References (Schneier et al., 1996; Ballenger et al., in press), the Freeman et al. (1997) study could be confounded by Allin, J.T., Banks, E.M., 1972. Functional aspects of ultrasound pro- mixing diagnostic criteria for GAD and panic disorder. duction by infant albino rats (Rattus norvegious). Anim. Behav. 20, Freeman et al. (1997) question that if ondansetron treats 175±185. panic disorder but not GAD, there must be pharmaco- Andrews, N., File, S.E., 1993. Handling history of rats modi®es be- logical evidence and pathophysiological implications for havioural effects of drugs in the elevated plus-maze test of anxiety. the 5-HT receptor in panic disorder. Eur. J. Pharmacol. 235, 109±112. 3 Arnt, J., 1989. Characterization of the discriminative stimulus properties During a meeting of the American College of Neuro- induced by 5-HT12 and 5-HT agonists in rats. Pharmacol. Toxicol. 64, psychopharmacology (Puerto Rico, 1994) a paper was 165±172. presented by Metz, Evoniuk and De Veaugh-Geiss on the Aron, C., Simon, P., Larousse, C., Boissier, J.R., 1971. Evaluation of a 90 B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95

rapid technique for detecting minor tranquilizers. Neuropharmacology Borsini, F., Brambilla, A., Cesana, R., Donetti, A., 1993. The effect of

10, 459±469. DAU 6215, a novel 5-HT3 antagonist, in animal models of anxiety. Artaiz, I., Romero, G., Zazpe, A., Monge, A., Caldero,Â J.M., Roca, J., Pharmacol. Res. 27, 151±163. Lasheras, B., DelRio, J., 1995. The pharmacology of VA 21B7: an Borsini, F., Lecci, A., Volterra, G., Meli, A., 1989. A model to measure atypical 5-HT3 receptor antagonist with anxiolytic-like properties in anticipatory anxiety in mice? Psychopharmacology 98, 207±211. animal models. Psychopharmacology 117, 137±148. Bovier, P., Broekkamp, C.L.E., Lloyd, K.G., 1982. Enhancing GABAer- Ballenger, J.C., DeVeaugh-Geiss, J., Evoniuk, G. et al. Multicenter trial of gic transmission reverses the aversive state in rats induced by electrical a 5-HT antagonist, ondansetron, in panic disorder (in press). 3 stimulation of the periaqueductal gray region. Brain Res. 248, 313± Barnes, J.M., Barnes, N.M., Cooper, S.J., 1992. Behavioural pharma- 320. cology of 5-HT receptor ligands. Neurosci. Biobehav. Rev. 16, 3 Bradley, P.B., Engel, G., Feniuk, W., Fozard, J.R., Humphrey, P.A., 107±113. Middlemiss, D.N., Mylecharane, E.J., Richardson, B.P., Saxena, P.R., Barnes, J.M., Barnes, N.M., Costall, B., Domeney, A.M., Johnson, D.N., 1986. Proposals for the classi®cation and nomenclature of functional Kelly, M.E., Munson, H.R., Naylor, R.J., Young, R., 1990a. The receptors for 5-HT. Neuropharmacology 25, 563±576. differential activities of R(1)- and S(2)-zacopride as 5-HT receptor 3 Broekkamp, C.L., Jenck, F., 1989. The relationship between various antagonists. Pharmacol. Biochem. Behav. 37, 717±727. animal models of anxiety, fear-related psychiatric symptoms and Barnes, J.M., Costall, B., Coughlan, J., Domeney, A.M., Gerrard, P.A., Kelly, M.E., Naylor, R.J., Onaivi, E.S., Tomkins, D.M., Tyers, M.B., response to serotonergic drugs. In: Bevan, P., Cools, A.R., Archer, T. (Eds.), Behavioural Pharmacology of 5-HT, Lawrence Erlbaum, 1990b. The effects of ondansetron, a 5-HT3 receptor antagonist, on cognition in rodents and primates. Pharmacol. Biochem. Behav. 35, Hillsdale, New Jersey, pp. 321±335. 955±962. Broekkamp, C.L., Rijk, H.W., Joly-Gelouin, D., Lloyd, K.L., 1986. Major Barrett, J.E., 1991. Animal behavioral models in the analysis and tranquillizers can be distinguished from minor tranquillizers on the understanding of anxiolytic drugs at serotonin receptors. In: Olivier, basis of effects on marble burying and swim-induced grooming in B., Mos, J., Slangen, J.L. (Eds.), Animal models in psychopharmacol- mice. Eur. J. Pharmacol. 126, 223±229. ogy, Birkhauser,È Basel, pp. 37±52. Broekkamp, C.L.E., Berendsen, H.H.G., Jenck, F., Van Delft, A.M.L., 1989. Animal models for anxiety and response to serotonergic drugs. Barrett, J.E., Gleeson, S., 1991. Anxiolytic effects of 5-HT1A agonists, Psychopathology 22 (suppl 1), 2±12. 5-HT3 antagonists and benzodiazepines: con¯ict and drug discrimina- Carli, M., Prontera, C., Samanin, R., 1989. Evidence that central 5- tion studies. In: Rodgers, R.J., Cooper, S.J. (Eds.), 5-HT1A agonists, hydroxy-tryptaminergic neurones are involved in the anxiolytic activi- 5-HT3 antagonists and benzodiazepines; their comparative behavioural pharmacology, Wiley, Chichester, pp. 59±105. ty of buspirone. Br. J. Pharmacol. 96, 829±836.

Barrett, J.E., Witkins, J.M., Mansbach, R.S., Skolnick, P., Weissman, Cervo, L., Samanin, R., 1995. 5-HT1A receptor full and partial agonists

B.A., 1986. Behavioral studies with anxiolytic drugs. III. Antipunish- and 5-HT2C (but not 5-HT 3 receptor antagonists increase rates of ment actions of buspirone in the pigeon do not involve benzodiazepine punished responding in rats. Pharmacol. Biochem. Behav. 52, 671± receptor mechanisms. J. Pharmacol. Exp. Ther. 238, 1009±1013. 676. Barros, H.M.T., Miczek, K.A., 1996. Withdrawal from oral cocaine in Colpaert, F.C., Koek, W., 1991. 5-HT receptor ligands and con¯ict rats: ultrasonic vocalizations and tactile startle. Psychopharmacology behavior. In: Second Polish-Swedish Symposium on Neurotransmitters 125, 379±384. and neuropharmacology, Zakopane, p. 5. Belelli, D., Balcarek, J.M., Hope, A.G., Peters, J.A., Lambert, J.J., Cooper, S.J., 1980. Benzodiazepines as appetite enhancing compounds. Blackburn, T.P., 1995. Cloning and functional expression of a human Appetite 1, 7±19.

5-hydroxytryptamine Type 3 AS receptor subunit. Mol. Pharmacol. 48, Costall, B., Domeney, A., Kelly, M.E., Tomkins, D.M., Naylor, R.J., 1054±1062. Wong, E.H.F., Smith, W.L., Whiting, R.L., Egler, R.M., 1993. The

Bienkowski, P., Iwinska, K., Koros, E., Panocka, I., Piasecki, J., Kostow- effects of the 5-HT3 receptor antagonist, RS-42358-197, in animal

ski, W., 1998. Prior repeated exposure to a 5-HT3 receptor agonist does models of anxiety. Eur. J. Pharmacol. 234, 91±99. not alter the ethanol-induced conditioned taste aversion in rats. Costall, B., Domeney, A.M., Gerrard, P.A., Kelly, M.E., Naylor, R.J., Pharmacol. Biochem. Behav. 59, 975±980. 1988a. Zacopride: anxiolytic pro®le in rodent and primate models of Bill, D.J., Fletcher, A., Glenn, B.D., Knight, M., 1991. Behavioural anxiety. J. Pharm. Pharmacol. 40, 302±305.

effects of WAY 100289, a novel 5-HT3 receptor antagonist, in two Costall, B., Domeney, A.M., Gerrard, P.A., Naylor, R.J., 1988b. A models of anxiety. In: Serotonin: 5-hydroxytryptamine-CNS receptors primate model for the assessment of anxiolytic drug action. Br. J. and brain function, Abstract, p. 113. Pharmacol. 92, 881±894. Bill, D.J., Fletcher, A., Glenn, B.D., Knight, M., 1992. Behavioural Costall, B., Domeney, A.M., Hendrie, C.A., Kelly, M.E., Naylor, R.J.,

studies on WAY 100289, a novel 5-HT3 receptor antagonist in two Tyers, M.B., 1987a. The anxiolytic activity of GR 38032F in the animal models of anxiety. Eur. J. Pharmacol. 218, 327±334. mouse and marmoset. Br. J. Pharmacol. 90, 257P. Blanchard, R.J., Blanchard, D.C., Agullana, R., Weiss, S.M., 1991. Costall, B., Domeney, A.M., Kelly, M.E., Naylor, R.J., 1991. Pharmaco- Twenty-two kHz alarm cries to presentation of a predator, by logical models in the development of antipsychotic drugs-new strate- laboratory rats living in visible burrow systems. Physiol. Behav. 50, gies. In: Olivier, B., Mos, J., Slangen, J.L. (Eds.), Animal models in 967±972. psychopharmacology, Birkhauser,È Basel, pp. 253±263. Blandina, P., Goldfarb, J., Craddock-Royal, B., Green, J.P., 1989. Release Costall, B., Domeney, A.M., Naylor, R.J., Tyers, M.B., 1987b. Effects of

of endogenous dopamine by stimulation of 5-hydroxytryptamine 3 the 5-HT3 antagonist, GR 38032F, on raised dopaminergic activity in receptors in rat striatum. J. Pharmacol. Exp. Ther. 251, 803±809. the mesolimbic system of the rat and marmoset brain. Br. J. Phar- Bockaert, J., Sevven, M., Dumuis, A., 1990. Pharmacological characteri- macol. 92, 881±894.

zation of 5-hydroxytryptamine 4 (5-HT4 ) receptors positively coupled Costall, B., Jones, B.J., Kelly, M.E., Naylor, R.J., Oakley, N.R., Onaivi, to adenylate cyclase in adult guinea pig hippocampal membranes: E.S., Tyers, M.B., 1989a. The effects of ondansetron (GR 38032F) in effect of substituted benzamide derivatives. Mol. Pharmacol. 37, 408± rats and mice treated subchronically with diazepam. Pharmacol. 411. Biochem. Behav. 34, 769±778. Boess, F.G., Beroukhim, R., Martin, I.L., 1995. Ultrastructure of the Costall, B., Jones, B.J., Kelly, M.E., Naylor, R.J., Onaivi, E.S., Tyers,

5-hydroxytryptamine3 receptor. J. Neurochemistry 64, 1401±1405. M.B., 1990a. Sites of action of ondansetron to inhibit withdrawal from Boissier, J.R., Simon, P., Aron, C., 1968. A new method for rapid drugs of abuse. Pharmacol. Biochem. Behav. 36, 97±104. screening of minor tranquilizers in mice. Eur. J. Pharmacol. 4, 145± Costall, B., Jones, B.J., Kelly, M.E., Naylor, R.J., Onaivi, E.S., Tyers, 151. M.B., 1990b. Ondansetron inhibits a behavioural consequence of B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95 91

withdrawing from drugs of abuse. Pharmacol. Biochem. Behav. 36, Dourish, C.T., Hutson, P.H., Kennett, G.A., Curzon, G., 1986. 8-OH- 339±344. DPAT induced hyperphagia: its neural basis and possible therapeutic Costall, B., Kelly, M.E., Tomkins, D.M., Tyers, M.B., 1989b. Pro®le of relevance. Appetite (Suppl) 7, 127±140.

action of diazepam and 5-HT3 receptor antagonists in the elevated Dunn, R.W., Carlezon, W.A., Corbett, R., 1991. Preclinical anxiolytic X-maze. J. Psychopharmacol. 3, 10P. versus antipsychotic pro®les of the 5-HT3 antagonists ondansetron, Costall, B., Jones, B.J., Kelly, M.E., Naylor, R.J., Tomkins, D.M., 1989c. zacopride, 3a-tropanyl-1H-indole-3-carboxylic acid ester, and 1aH, 3a, Exploration of mice in a black and white box: validation as a model of 5aH-tropan-3-yl-3,5-dichlorobenzoate. Drug Devl. Res. 23, 289±300. anxiety. Pharmacol. Biochem. Behav. 32, 777±785. Eikelboom, R., 1986. Learned anticipatory rise in body temperature due Costall, B., Naylor, R.J., 1991. Anxiolytic effects of 5-HT antagonists in 3 to handling. Physiol. Behav. 37, 649±653. animals. In: Rodgers, R.J., Cooper, S.J. (Eds.), 5-HT agonists, 1A File, S.E., 1980. The use of social interaction as a method for detecting 5-HT antagonists and benzodiazepines: their comparative behavioural 3 anxiolytic activity of chlordiazepoxide-like drugs. J. Neurosci. Meth. pharmacology, Wiley, Chichester, pp. 133±157. 2, 219±238. Crawley, J.N., 1981. Neuropharmacological speci®city of a simple animal File, S.E., 1985. Animal models for predicting clinical ef®cacy of model for the behavioural actions of benzodiazepines. Pharmacol. anxiolytic drugs. Neuropsychobiology 13, 55±62. Biochem. Behav. 15, 695±699. File, S.E., Johnston, A.L., 1989. Lack of effect of 5-HT receptor Crawley, J.N., Goodwin, F.K., 1980. Preliminary report of a simple 3 animal behaviour model for the anxiolytic effects of benzodiazepines. antagonists in the social interaction and elevated plus-maze tests of Pharmacol. Biochem. Behav. 13, 167±170. anxiety in the rat. Psychopharmacology 99, 248±251. Critchley, M.A.E., Handley, S.L., 1987. Effects in the X-maze anxiety File, S.E., Johnston, A.L., Pellow, S., 1987. Effects of compounds acting at CNS 5-hydroxytryptamine systems on anxiety in the rat. Brit. J. model of agents acting at 5-HT12 and 5-HT receptors. Psycho- pharmacology 93, 502±506. Pharmacol. 90, 265P. Cuomo, V., Ciagiano, R., Salvia, M.A., 1988. Ultrasonic vocalization in Filip, M., Baran, L., Siwanowicz, J., Chojnacka-Wojcik,ÂÂ E., Przegalõnski, response to unavoidable aversive stimuli in rats: effects of benzo- E., 1992. The anxiolytic-like effects of 5-hydroxytryptamine33 (5-HT ) diazepines. Life Sci. 43, 485±491. receptor antagonists. Pol. J. Pharmacol. Pharm. 44, 261±269. Fletcher, P.J., Davis, M., 1990. Effects of 8-OH-DPAT, buspirone and Cutler, M.G., 1990. Behavioural effects in gerbils of the 5-HT3 receptor antagonists, BRL 43694 and ICS 205-930, under circumstances of ICS 205-930 on feeding in a novel environment: comparisons with high and low light intensity. Neuropharmacology 29, 515±520. chlordiazepoxide and FG 7142. Psychopharmacology 102, 301±308. Cutler, M.G., 1991. An ethological study of the effects of buspirone and Fozard, J.R., Mobarek Ali, A.T.M., 1978. Blockade of neuronal trypt-

the 5-HT3 antagonist, BRL43694 (Granisetron) on behaviour during amine receptors by metoclopramide. Br. J. Pharmacol. 49, 109±112. social interactions in female and male mice. Neuropharmacology 30, Fozard, J.R., Mobarek Ali, A.T.M., Newgrosh, G., 1979. Blockade of 299±306. serotonin receptors on autonomic neurones by (-) cocaine and some

Cutler, M.G., Dixon, A.K., 1989. Effects of the 5-HT3 antagonist, ICS related compounds. Eur. J. Pharmacol. 59, 195±210. 205-930, on behaviour of mice during social encounters. Br. J. Freeman, A.M., Westphal, J.R., Norris, G.T., Roggero, B.A., Webb, P.B., Pharmacol. 96, 12P. Freeman, K.L., Rush, J.A., Hearne, E.M., Evoniuk, G., 1997. Ef®cacy

Cutler, M.G., Piper, D.C., 1990. Chronic administration of the 5-HT3 of ondansetron in the treatment of generalized anxiety disorder. receptor antagonist BRL 43694; effects on re¯ex epilepsy and social Depression Anxiety 5, 140±141. behaviour of the Mongolian gerbil. Psychopharmacology 101, 244± Fundaro, A., 1998. Pinch-induced catalepsy in mice: a usefull model to 249. investigate antidepressant or anxiolytic drugs. Progr. Neuro-Psycho-

Cutler, M.G., Piper, D.C., 1991. Effects of diazepam and the 5-HT3 pharmacol. Biol. Psychiat. 22, 147±158. antagonists granisetron and BRL 46470, on the behaviour of intruder Gaddum, J.H., Picarelli, Z.P., 1957. Two kinds of tryptamine receptor. Br. mice. Br. J. Pharmacol. 104, 254P. J. Pharmacol. 12, 323±328. Davis, M., 1979. Effect on conditioned fear as measured with the Gardner, C.R., 1985. Pharmacological studies on the role of serotonin in potentiated startle paradigm. Psychopharmacology 62, 1±7. animal models of anxiety. In: Green, A.R. (Ed.), Neuropharmacology Davis, M., 1986. Pharmacological and anatomical analysis of fear of serotonin, Oxford University Press, Oxford, pp. 281±325. conditioning using the fear-potentiated startle paradigm. Behav. Gardner, C.R., Budhram, P., 1987. Effects of agents which interact with Neurosci. 100, 814±824. central benzodiazepine binding sites on stress-induced ultrasounds in Davis, M., Cassella, J.V., Kehne, J.H., 1988. Serotonin does not mediate rat pups. Eur. J. Pharmacol. 134, 275±283.

anxiolytic effect of buspirone in the fear-potentiated startle paradigm: Gaster, L.M., King, F.D., 1997. Serotonin 5-HT34 and 5-HT receptor comparison with 8-OH-DPAT and ipsapirone. Psychopharmacology antagonists. Med. Res. Rev. 17, 163±214. 94, 14±20. Geller, I., Seifter, J., 1960. The effects of meprobamate, barbiturates, Davis, M., Redmond, Jr. D.E., Baraban, J.M., 1979. Noradrenergic d-amphetamine and promazine on experimentally induced con¯ict in agonists and antagonists: effects on conditioned fear as measured by the rat. Psychopharmacologia 1, 482±492. the potentiated startle paradigm. Psychopharmacology 65, 111±118. Gleeson, S., Ahlers, S.T., Mansbach, R.S., Foust, J.M., Barrett, J.E., Davis, S.F., Moore, S.A., Cowen, C.L., Thurston, D.K., Maggio, J.C., 1989. Behavioral studies with anxiolytic drugs.VI. Effects on punished 1982. Defensive burying in the Mongolian gerbil (Meriones un- responding of drugs interacting with serotonin receptor subtypes. J. guiculatus) as a function of size and shape of the test chamber. Anim. Pharmacol. Exp. Ther. 250, 809±817. Learn. Behav. 10, 516±520. Glenn, B., Green, S., 1989. Anxiolytic pro®le of GR 38032F in the De Boer, S.F., Van der Gugten, J., Slangen, J.L., 1991. Behavioural and potentiated startle paradigm. Behav. Pharmacol. 1, 91±94. hormonal indices of anxiolytic and anxiogenic drug action in the shock Glennon, R.A., Westkaemper, R.B., Bartyzel, P., 1991a. Medicinal prod defensive burying/avoidance paradigm. In: Olivier, B., Mos, J., chemistry of serotonergic agents. In: Peroutka, S.J. (Ed.), Serotonin Slangen, J.L. (Eds.), Animal Models in Psychopharmacology, Birkh- receptor subtypes: basic and clinical aspects, Wiley-Liss, New York, auser,È Basel, pp. 81±96. pp. 19±64.

De La Garza, R., Callahan, P.M., Cunningham, K.A., 1996. Detailed Glennon, R.A., Young, R., Dukat, M., 1991b. 5-HT3 serotonin agonists as

investigations of 5-HT3 compounds in a drug discrimination model. discriminative stimuli. Soc. Neurosci. Abstr. 17 (1), 601.

Pharmacol. Biochem. Behav. 54, 533±540. Glennon, R.A., Young, R., Dukat, M., 1992. 5-HT3 agonist 2-methyl De Vrij, J., Benz, U., Schreiber, R., Traber, J., 1993. Shock-induced serotonin as a training drug in drug discrimination studies. Pharmacol. ultrasonic vocalization in young adult rats: a model for testing putative Biochem. Behav. 41, 361±364. anti-anxiety drugs. Eur. J. Pharmacol. 249, 331±339. Goa, K.L., Ward, A., 1986. Buspirone. A preliminary review of its 92 B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95

pharmacological properties and therapeutic ef®cacy as an anxiolytic. in the serotonergic control of periaqueductal gray induced aversion in Drugs 32, 114±129. rats. Psychopharmacology 100, 372±376. Goldberg, H.L., Finnerty, R.J., 1979. The comparative ef®cacy of Johanson, C.E., Barrett, J.E., 1995. The dricriminative stimulus effects of buspirone and diazepam in the treatment of anxiety. Am. J. Psychiatry. cocaine in pigeons. J. Pharmacol. Exp. Ther. 267, 1±8.

136, 1184±1187. Joharchi, N., Sellers, E.M., Higgins, G.A., 1993. Effects of 5-HT3

Goudie, A.J., Leathley, M.J., 1990. Effects of the 5-HT3 antagonist receptor antagonists on the discriminative stimulus properties of GR38032F (ondansetron) on benzodiazepine withdrawal in rats. Eur. J. morphine in rats. Psychopharmacology 112, 111±115.

Pharmacol. 185, 179±186. Johnson, D.S., Heinemann, S.F., 1993. Rat 5-HT3 receptor expression in Gozlan, H., 1997. Serotonin receptors and their ligands. In: Olivier, B., the central and peripheral nervous system-an in situ hybridization

Van Wijngaarden, I., Soudijn, W. (Eds.), 5-HT3 Receptors, Phar- study. Soc. Neurosci. Abstr. 19, 632. macochemistry Library 27, Elsevier Science Publishers, Amsterdam, Jones, B.J., Costall, B., Domeney, A.M., Kelly, M.E., Naylor, R.J., pp. 221±257. Oakley, N.R., Tyers, M.B., 1988. The potential anxiolytic activity of

Graeff, F.G., 1984. The anti-aversive action of minor tranquilizers. Tr. GR38032F, a 5-HT3 -receptor antagonist. Br. J. Pharmacol. 93, 985± Pharmacol. Sci. 5, 230±233. 993. Graeff, F.G., Brandao, M.L., Audi, E.A., Schuetz, M.T., 1986. Modula- Joordens, R.J.E., Hijzen, T.H., Peeters, B.W.M.N., Olivier, B., 1996. tion of the brain aversive system by GABA-ergic and serotonergic Fear-potentiated startle response is remarkably similar in two lab- mechanisms. Behav. Brain Res. 21, 65±72. oratories. Psychopharmacology 126, 104±109.

Grant, K.A., Barrett, J.E., 1991. Blockade of the discriminative stimulus Joordens, R.J.E., Hijzen, T.H., Olivier, B., 1998. The effects of 5-HT1A

effects of ethanol with 5-HT3 receptor antagonists. Psychopharmacolo- receptor agonists, 5-HT1A receptor antagonists and their interaction in gy 104, 451±456. the fear-potentiated startle response paradigm. Psychopharmacology

Grant, K.A., Colombo, G., 1993. The ability of 5-HT3 antagonists to 139, 383±390. block ethanol discrimination: effect of route of administration. Alcohol Kaltwasser, M.T., 1991. Acoustic startle-induced ultrasonic vocalizations Clin. Exp. Res. 17, 497±503. in the rat: a novel animal model of anxiety? Behav. Brain Res. 3, Greenshaw, A.J., Silverstone, P.H., 1997. The non-emetic uses of 133±137.

serotonin 5-HT3 receptor antagonists: Clinical pharmacology and Kehne, J.H., Cassella, J.V., Davis, M., 1988. Anxiolytic effects of therapeutic applications. Drugs 53, 20±39. buspirone and gepirone in the fear-potentiated startle paradigm. Griebel, G., 1995. 5-hydroxytryptamine-interacting drugs in animal Psychopharmacology 94, 8±13. models of anxiety disorders: more than 30 years of research. Pharmac. Kehne, J.H., McCloskey, T.C., Baron, B.M., Chi, E.M., Harrison, B.L., Ther. 65, 319±395. Whitten, J.P., Palfreyman, M.G., 1991. NMDA receptor complex Griebel, G., Rodger, R.J., Perrault, G., Sanger, D.J., 1997. Risk assess- antagonists have potential anxiolytic effects as measured with sepa- ment behaviour: evaluation of utility in the study of 5-HT related ration-induced ultrasonic vocalizations. Eur. J. Pharmacol. 193, 283± drugs in the rat elevated plus-maze test. Pharmacol. Biochem. Behav. 292. 57, 817±827. Kennett, G.A., Blackburn, T.P., Harlow, U.K., 1990. Anxiolytic-like

Hall, S.T., Ceuppens, P.R., 1991. A study to evaluate the effects of actions of BRL 46470-a novel 5-HT3 antagonist. J. Psychopharmacol. ondansetron on psychomotor performance after repeated oral dosing in 4, 291. healthy subjects. Eur. J. Pharmacol. 104, 86±90. Ketelaars, C.E.J., Bollen, E.L., Rigter, H., Bruinvels, J., 1988. GABA-B Handley, S.L., Mithani, S., 1984. Effects of alpha-adrenoceptor agonists receptor activation and con¯ict behaviour. Life Sci. 42, 933±942. and antagonists in a maze-exploration model of `fear'-motivated Ketelaars, C.E.J., Bruinvels, B., 1991. Extinction of con¯ict behaviour. behaviour. Naunyn-Schmiedeberg's. Arch. Pharmacol. 327, 1±5. In: In: Olivier, B., Mos, J., Slangen, J.L. (Eds.), Animal Models in Hard,Ê E., Engel, J., Musi, B., 1982. The ontogeny of defensive reactions Psychopharmacology, Birkhauser,È Basel, pp. 103±106. in the rat: in¯uence of the monoamine transmission systems. Scand. J. Kilfoil, T., Michel, A., Montgomery, D., Whiting, R.L., 1989. Effects of Psychol. Suppl 1, 90±96. anxiolytic and anxiogenic drugs on exploratory activity in a simple Heynen, A.J., Sainsbury, R.S., Montoya, C.P., 1989. Cross-species model of anxiety in mice. Neuropharmacology 28, 901±905.

responses in the defensive burying paradigm: a comparison between Kilpatrick, G.J., Bunce, K.T., Tyers, M.B., 1990. 5-HT3 receptors. Med. Long-Evans rats (Rattus norvegicus). Richardson's ground squirrels Res. Rev. 10, 441±475. (Spermophilus richardsonii), and Thirteen-lined ground squirrels Kilpatrick, G.J., Jones, B.J., Tyers, M.B., 1987. Identi®cation and

(Catellus tridecemlineatus). J. Comp. Psychol. 103, 184±190. distribution of 5-HT3 receptors in rat brain using radioligand binding. Higgins, G.A., Jones, B.J., Oakley, N.R., Tyers, M.B., 1991. Evidence Nature 330, 746±748.

that the amygdala is involved in the disinhibitory effects of 5-HT3 King, F.D., 1994. 5-hydroxytryptamine-3-receptor antagonists. In: King, receptor antagonists. Psychopharmacology 104, 545±551. F.D., Jones, B.J., Sanger, G.J. (Eds), CRC: Boca Raton FLA, pp. 1-44. Hijzen, T.H., Houtzager, S.W., Joordens, R.J.E., Olivier, B., Slangen, J.L., Koetzner, K., Mastropaolo, J., Deutsch, S.I., 1995. Facilitating roleof 1995. Predictive validity of the potentiated startle response as a m-chlorophenyl-biguanide in a cocaine discrimination. Pharmacol. behavioral model for anxiolytic drugs. Psychopharmacology 118, 150± Biochem. Behav. 52, 815±818. 154. Kshama, D., Hrishikeshavan, H.J., Shanbhogue, R., Manonyedi, U.S., Hudson, B.B., 1950. One-trial learning in the domestic rat. Genetic 1990. Modulation of baseline behavior in rats by putative serotonergic Psychol. Monogr. 41, 99±145. agents in three ethoexperimental paradigms. Behav. Neural. Biol. 54, Jackson, R.L., Garbin, C.P., Hollingsworth, E.M., 1984. Defensive 234±253. burying of aversive ¯uids in rats: the possible role of odor. Learn Lader, M.H., 1991. Ondansetron in the treatment of anxiety. In: Satellite Motivation 15, 85±105. Symp 5th World Congress Biol. Psychiatry, In: The role of ondanset-

Jenck, F., Broekkamp, C.L.E., Van Delft, A.M.L., 1989a. Effects of ron, a novel 5-HT3 antagonist, in the treatment of psychiatric serotonin receptor antagonists on PAG stimulation induced aversion: disorders, pp. 17±19.

different contributions of 5-HT12 , 5-HT and 5-HT 3 receptors. Psycho- Lane, J.D., Pickering, C., Hooper, M., Fagan, K., Tyers, M.B., Emmett- pharmacology 97, 489±495. Oglesby, M.B., 1992. Failure of ondansetron to block discriminative or Jenck, F., Broekkamp, C.L.E.,Van Delft, A.M.L., 1989b. Opposite control reinforcing stimulus effects of cocaine in the rat. Drug Alcohol

mediated by 5-HT1A and non 5-HT 1A (5-HT 1B or 5-HT 1C ) receptors Depend. 30, 151±162. on periaqueductal gray aversion. Eur. J. Pharmacol. 161, 219±221. Laporte, A.M., Kidd, E.J., Verge,Â D., Gozlan, H., Hamon, M., 1992.

Jenck, F., Broekkamp, C.L.E., Van Delft, A.M.L., 1990. 5-HT1C receptors Autoradiographic mapping of central 5-HT3 receptors. In: Hamon, M B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95 93

(Ed.), Central and peripheral 5-HT3 receptors, Academic Press, male rats as a behavioural paradigm for screening anti-panic drugs. London, pp. 157±187. Psychopharmacology 117, 32±40.

Lecci, A., Borsini, F., Mancinelli, A., D'Aranno, V., Stasi, M.A., Volterra, Morinan, A., 1989. Effects of the 5-HT3 receptor antagonists. GR 38032F G., Meli, A., 1990a. Effect of serotonergic drugs on stress-induced and BRL 24924 on anxety in socially isolated rats. Br. J. Pharmacol. hyperthermia (SIH) in mice. J. Neural. Transm. 2, 219±230. 97, 457P. Lecci, A., Borsini, F., Volterra, G., Meli, A., 1990b. Pharmacological Mos, J., Olivier, B., 1989. Ultrasonic vocalizations by rat pups as an validation of a novel animal model of anticipatory anxiety in mice. animal model for anxiolytic activity: effects of serotonergic drugs. In: Psychopharmacology 101, 255±261. Bevan, P., Cools, A.R., Archer, T. (Eds.), Behavioural Pharmacology Lecrubier, Y., Puech, A.J., Azcona, A., Bailey, P.E., Lataste, X., 1993. A of 5-HT, Lawrence Erlbaum, Hillsdale, New Jersey, pp. 361±366. randomized double-blind placebo controlled study of tropisetron in the Mos, J., Olivier, B., 1990. Behavioural and neuropharmacological aspects treatment of outpatients with generalized anxiety disorder. Psycho- of maternal aggression in rodents. Aggr. Behav. 16, 145±163. pharmacology 112, 129±133. Mos, J., Van der Heyden, J., Olivier, B., 1989. Behavioural effects of

Lee, C.R., Plosker, G.L., McTavish, D., 1993. Tropisetron. A review of 5-HT3 antagonists in animal models for aggression, anxiety and its pharma-codynamic and pharmacokinetic properties and therapeutic psychosis. In: Bevan, P., Cools, A.R., Archer, T. (Eds.), Behavioural potential as an antiemetic. Drugs 46, 925±943. Pharmacology of 5-HT, Lawrence Erlbaum, New Jersey, pp. 389±395. Leigh, T.J., Link, C.G.G., Fell, G.L., 1991. Effects of granisetron and Moser, P.C., 1989. An evaluation of the elevated plus-maze test using the lorazepam, alone and in combination, on psychometric performance. novel anxiolytic buspirone. Psychopharmacology 99, 48±53.

Biol. Psychiatry 29, 602S. Moser, P.C., 1992. The effects of 5-HT3 receptor antagonists on the Leonard, R.J., Labarca, C.G., Charnet, P., Davidson, N., Lester, H.A., discriminative stimulus properties of amphetamine. Eur. J. Pharmacol.

1988. Evidence that the M2 membrane-spanning region lines the ion 212, 271±276. channel pore of the nicotinic receptor. Science 242, 1578±1581. Nader, M.A., Hoffman, S., Gleeson, S., Barrett, J.E., 1989. Further Lister, R.G., 1990. Ethologically-based models of anxiety disorders. characterisation of the discriminative stimulus effects of buspirone Pharmac. Ther. 46, 321±340. using monoamine agonists and antagonists in the pigeon. Behav. Maricq, A.V., Peterson, A.S., Brake, A.J., Meyers, R.M., Julius, D., 1991. Pharmacol. 1, 57±67.

Primary structure and functional expression of the 5-HT3 receptor, a Nastiti, K., Benton, D., Brain, P.F., 1991. The effects of compounds serotonin-gated ion channel. Science 254, 432±437. acting at the benzodiazepine receptor complex on the ultrasonic calling Mathew, R.J., Wilson, W.H., 1991. Evaluation of the effects of diazepam of mouse pups. Behav. Pharmacol. 2, 121±128.

and an experimental anti-anxiety drug on regional cerebral blood ¯ow. Nevins, M.E., Anthony, E.W., 1994. Antagonists at the serotonin-3 Psychiatry. Res.; Neuro imaging 40, 125±134. receptor can reduce the fear-potentiated startle response in the rat: MCann, U.D., Morgan, C.M., Geraci, M., Slate, S.O., Murphy, D.L., Post, evidence for different types of anxiolytic activity? J. Pharmacol. Exp.

R.M., 1997. Effects of the 5-HT3 antagonist, ondansetron, on the Ther. 268, 248±268. behavioural and physiological effects of pentagastrin in patients with Njung'e, K., Handley, S.L., 1991. Effects of 5-HT uptake inhibitors, panic disorder and social phobia. Neuropsychopharmacology 17, 360± agonists and antagonists on the burying of harmless objects by mice; a 369. putative test for anxiolytic agents. Br. J. Pharmacol. 104, 105±112. Miczek, K.A., Weerts, E.M., Vivian, J.A., Barros, H.M.T., 1995. Aggres- Nomura, K., Maeda, N., Yoshino, T., Yamaguchi, I., 1994. A mechanism

sion, anxiety and vocalizations in animals: GABAA3 and 5-HT an- of 5-HT receptor mediation is involved etiologically in the psycho- xiolytics. Psychopharmacology 121, 38±56. logical stress lesion in the stomach of the mouse. J. Pharmacol. Exp. Miczek, K.A., Vivian, J.A., 1993. Automatic quanti®cation of withdrawal Ther. 271, 100±106. from 5-day diazepam in rats: ultrasonic distress vocalizations and Nowakowska, E., Chodera, A., Kus, K., 1998. An anxiolytic-like effect of hyperre¯exia to acoustic startle stimuli. Psychopharmacology 110, ondansetron disappears in oxazepam-tolerant rats. Pharmacol. Bio- 379±382. chem. Behav. 59, 935±938. Middlefell, V.C., Bill, D.J., Brammer, N.T., Coleman, J., Fletcher, A., Olivier, B., Molewijk, H.E., Van Oorschot, R., Van der Heyden, J.A.M., Hallett, I., Rhodes, K.F., Wainwright, T.L., Ward, T.J., 1996a. WAY- Ronken, E., Mos, J., 1998a. Rat pup ultrasonic vocalization: effects of

SEC 579: a novel 5-HT3 receptor antagonist. CNS Drug Res. 2, benzodiazepine receptor ligands. Eur. J. Pharmacol. 358, 117±128. 269±293. Olivier, B., Molewijk, H.E., Van Oorschot, R., Ronken, E., Mos, J.,

Middlefell, V.C., Price, T.L., Thompson, I.A., Spokes, R., 1990. 5-HT3 Miczek, K.A., 1998b. Utrasonic vocalizations in rat pups: effects of agonist-effects of zacopride in the ferret. Brit. J. Pharmac. 101, 560. serotonergic ligands. Neurosci. Biobehav. Rev. 23, 217±227. Middlefell, V.C., Bill, D.J., Brammer, N.T., Coleman, J., Fletcher, A., Onavi, E.S., Martin, B.R., 1989. Neuropharmacological and physiological Hallett, I., Rhodes, K.F., Wainwright, T.L., Ward, T.J., 1996b. WAY- validatiopn of a computer-controlled two-compartment black and

SEC-579: a novel 5-HT3 receptor antagonist. CNS Drug Res. 2, white test box for the assessment of anxiety. Prog. Neuro-psycho- 269±293. pharmacol. Biol. Psychiat. 13, 963±976. Millson, D.S., Preston, G.C., 1991. The clinical psychopharmacology of Ortells, M.I., Lunt, G., 1995. Evolutionary history of the ligand-gated ondansetron. In: Satellite Symp 5th World Congress Biol. Psychiatry, ion-channel superfamily of receptors. Tr. Neurosci. 18, 121±127.

The role of ondansetron, a novel 5-HT3 antagonist, in the treatment of Paris, J.M., Cunningham, K.A., 1991. Serotonin 5-HT3 antagonists do not psychiatric disorders, pp. 13±15. alter the discriminative stimulus properties of cocaine. Psycho-

Mitchell, E.A., Pratt, J.A., Kilpatrick, G.J., 1991. Effects of the 5-HT3 pharmacology 104, 475±478. receptor agonist mCPBG in the social interaction test of anxiety. Br. J. Pellow, S., Johnston, A.L., File, S.E., 1987. Selective agonists and Pharmacol. 104, 252P. antagonists for 5-hydroxytryptamine receptor subtypes, and interac- Miyaka, A., Mochizuki, S., Takemoto, Y., Akuzawa, S., 1995. Molecular tions with yohimbine and FG7142 using the elevated plus-maze test in

cloning of human 5-hydroxytryptamine3 receptor: heterogeneity in the rat. J. Pharm. Pharmacol. 39, 917±928. distribution and function among species. Mol. Pharmacol. 48, 407± Pinel, J.P.J., Gorzalka, B.B., Ladak, F., 1981. Cadaverine and putrescine 416. initiate the burial of dead conspeci®cs by rats. Physiol. Behav. 27, Molewijk, H.E., Hartog, K., Van der Poel, A.M., Mos, J., Olivier, B., 819±824. 1996. Reduction of guinea pig pup isolation calls by anxiolytic and Pinel, J.P.J., Treit, D., 1978. Burying as a defensive response in the rat. J. antidepressant drugs. Psychopharmacology 128, 31±38. Comp. Physiol. Psychol. 92, 708±712. Molewijk, H.E., Van der Poel, A.M., Mos, J., Van der Heyden, J.A.M., Pinel, J.P.J., Treit, D., Wilkie, D.M., 1980. Stimulus control of defensive Olivier, B., 1995. Conditioned ultrasonic distress vocalizations in adult burying in the rat. Learn Motivation 11, 150±163. 94 B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95

Pinkus, L.M., Sarbin, N.S., Gordon, J.C., Munson, H.R., 1990. Antagon- Sanchez, C., 1996. 5-HT1A receptors play an important role in modulation 3 ism of [ H]zacopride binding to 5-HT3 recognition sites by its (R) and of behavior of rats in a two-compartment black and white box. (S) enantiomers. Eur. J. Pharmacol. 179, 231. Pharmacol. 7, 788±797. Piper, D., Upton, N., Thomas, D., Nicholas, J., 1988. The effects of the Schechter, M.D., 1993. Cocaine discrimination is attenuated by isradipine

5-HT3 receptor antagonists BRL 43694 and GR38032F in animal and CGS 10746B. Pharmacol. Biochem. Behav. 44, 661±664. behavioural models of anxiety. Br. J. Pharmacol. 94, 314P. Schneier, F.R., Garfunkel, R., Kennedy, B., Campea, R., Fallon, B., Poole, S., Stephanson, J.D., 1977. Core temperature: some shortcomings Marshall, R., O'Donnell, L., Hogan, T., Liebowitz, M., 1996. Ondan- of rectal temperature measurements. Physiol. Behav. 18, 203±205. setron in the treatment of panic disorder. Anxiety 2, 199±202.

Raiteri, M., Paudice, P., Vallebuona, F., 1993. Inhibition by 5-HT3 Schreiber, R., Melon, C., De Vrij, J., 1998. The role of 5-HT receptor receptor antagonists of release of cholecystokinin-like immunoreactivi- subtypes in the anxiolytic effects of selective serotonin reuptake ty from the frontal cortex of freely moving rats. Naunyn- inhibitors in the rat ultrasonic vocalization test. Psychopharmacology Schmiedeberg's. Arch. Pharmacol. 347, 111±114. 135, 383±391. Richardson, B.P., Engel, G., Donatsch, P., Stadler, P.A., 1985. Identi®ca- Schuurman, T., Spencer, D.G., Traber, J., 1987. Behavioral effects of the

tion of serotonin M-receptor subtypes and their speci®c blockade by a 5-HT1A -receptor ligand ipsapirone (TVX Q7821): a comparison with new class of drugs. Nature 316, 126±131. 8-OH-DPAT and diazepam. Psychopharmacology 89, S54.

Roca, J., Artaiz, I., DelRio, J., 1995. 5-HT3 receptor antagonists in Schweizer, E., Rickels, K., 1991. Serotonergic anxiolytics: a review of

development as anxiolytics. Exp. Opin. Invest. Drugs 4, 333±342. their clinical ef®cacy. In: Rodgers, R.J., Cooper, S.J. (Eds.), 5-HT1A

Rocha e Silva, M., Velle, J.R., Picarelli, Z.P., 1953. A pharmacological Agonists, 5-HT3 Antagonists and Benzodiazepines: Their Comparative analysis of the mode of action of serotonin (5-HT) upon the guinea- Behavioural Pharmacology, John Wiley & Sons Ltd, Chichester, pp. pig ileum. Br. J. Pharmacol. 8, 378±388. 365±376. Rodgers, R.E., Cutler, M.G., Jackson, J.E., 1997a. Behavioural effects in Shepherd, J.K., Rodgers, R.J., 1990. 8-OH-DPAT speci®cally enhances mice of subchronic chlordiazepoxide, maprotiline and ¯uvoxamine. II. feeding behaviour in mice: evidence from behavioural competition. The elevated plus-maze. Pharmacol. Biochem. Behav. (in press). Psychopharmacology 101, 408±413.

Rodgers, R.J., 1991. Effects of benzodiazepine and 5-HT receptor ligands Silvestre, J.S., Fernandez,Â A.G., Palacios, J.M., 1996. Effects of 5-HT4 on aggression and defence in animals. In: Rodgers, R.J., Cooper, S.J. receptor antagonists on rat behaviour in the elevated plus-maze test.

(Eds), 5-HT1A agonists, 5-HT 3 antagonists and benzodiazepines: their Eur. J. Pharmacol. 305, 219±222. comparative behavioural pharmacology, pp. 195-231. Slangen, J.L., 1991. Drug discrimination and animal models. In: Olivier,

Rodgers, R.J., Cole, J.C., Tredwell, J.M., 1995. Pro®le of action of 5-HT3 B., Mos, J., Slangen, J.L. (Eds.), Animal models in Psychopharmacol- receptor antagonists, ondansetron and WAY100289 in the elevated ogy, Birkhauser,È Basel, pp. 359±373.

plus-maze test of anxiety on mice. Psychopharmacology 177, 306± Smeraldi, E., Mundo, E., Erzegovesi, S., 1992. 5-HT3 receptor and 312. antiobsessional effect. Human Psychopharmacol. 7, 291±292.

Rodgers, R.J., Cutler, M.G., Jackson, J.E., 1997b. Behavioural effects in Sorbe, B., 1996. 5-HT3 antagonists as antiemetic agents in cancer mice of subchronic buspirone, ondansetron and tianeptine. II The chemotherapy. Exp. Opin. Invest. Drugs 5, 389±407. elevated Plus-maze. Pharmacol. Biochem. Behav. 56, 295±303. Stefanski, R., Bienkowski, P., Kostowski, W., 1996. Studies on the role of

Rodgers, R.J., Randall, J.I., 1986. Acute Non-opioid analgesia in defeated 5-HT3 receptors in the mediation of the ethanol interoceptive cue. Eur. male mice. Physiol. Behav. 36, 947±950. J. Pharmacol. 309, 141±147. Rodgers, R.J., Randall, J.I., 1986b. Resident's Scent: a critical factor in Stein, L., Wise, C.D., Berger, B.D., 1973. Antianxiety action of benzo- acute analgesic reaction to defeat experience in male mice. Physiol. diazepines: decrease in activity of serotonergic neurones in the Behav. 37, 317±322. punishment system. In: Garattini, S., Mussini, E., Randell, L.O. (Eds.), Rodgers, R.J., Randall, J.I., 1987. Are the analgesic effects of social Benzodiazepines, Raven Press, New York, pp. 229±326. defeat mediated by benzodiazepine receptors? Physiol. Behav. 41, Stolerman, I.P., Stephenson, J.D., Rauch, J.R., 1987. Classi®cation of 279±289. opioid and 5-hydroxytryptamine receptors by means of driscriminative Rodgers, R.J., Randall, J.I., 1987a. Benzodiazepine ligands, nociception drug effects. Neuropharmacology 26, 867±875. and `defeat' analgesia in male mice. Psychopharmacology 91, 305± Tecott, L.H., Maricq, A.V., Julius, D., 1993. Nervous system distribution

315. of the serotonin 5-HT3 receptor mRNA. Proc. Natl. Acad. Sci. USA Rodgers, R.J., Randall, J.I., 1988. Blockade of non-opioid analgesia in 90, 1430±1434. intruder mice by selective neuronal and non-neuronal benzodiazepine Tornatzky, W., Miczek, K.A., 1995. Alcohol, anxiolytics and social stress recognition site ligands. Psychopharmacology 96, 45±54. in rats. Psychopharmacology 121, 135±144. Rodgers, R.J., Shepherd, J.K., 1990. Evidence for a peripheral component Treit, D., 1985. Animal models for the study of anti-anxiety agents: a

in the inhibitory actions of 5-HT3 antagonists on defeat analgesia. Br. review. Neurosci. Biobehav. Rev. 9, 203±222. J. Pharmacol. 100, 415P. Treit, D., Pinel, J.P.J., Fibiger, H.C., 1981. Conditioned defensive Rodgers, R.J., Shepherd, J.K., Randall, J.I., 1990. Highly potent inhib- burying: a new paradigm for the study of anxiolytic agents. Pharmacol.

itory effects of 5-HT3 receptor antagonist, GR38032F, on non-opiod Biochem. Behav. 15, 619±626. defeat analgesia in male mice. Neuropharmacology 29, 17±23. Tricklebank, M.D., 1989. Behavioural correlates of the activation of 5-HT Romach, M.K., Kaplan, H.L., Busto, U.E., Somer, G., Sellers, E.M., receptors. In: Mylecharane, E.J., Angus, J.A., De la Lande, I.S.,

1998. A controlled trial of ondansetron, a 5-HT3 antagonist in Humphrey, P.P.A. (Eds.), Serotonin: Actions, receptors, pathophysiol- benzodiazepine discontinuation. J. Clin. Psychopharmacol. 18, 121± ogy, MacMillan, London, e, pp. 87±94. 131. Tyers, M.B., Costall, B., Domeney, A.M., Jones, B.J., Kelly, M.E.,

Saito, H., Matsumoto, M., Togashi, H., Yoshioka, M., 1996. Functional Naylor, R.J., Oackley, N.R., 1987. The anxiolytic activities of 5-HT3 interaction between serotonin and other neuronal systems: Focus on in receptor antagonists in laboratory animals. Neurosci. Lett., Suppl. 29, vivo microdialysis studies. Jap. J. Pharmacol. 70, 203±225. S68. Sanchez, C., 1993. Effects of serotonergic drugs on footshock-induced Unwin, N., 1989. The structure of ion channels in membranes of excitable ultrasonic vocalization in adult male rats. Behav. Pharmacol. 4, 269± cells. Neuron. 3, 665±676. 277. Van der Hoek, G.A., Cooper, S.J., 1990. Evidence that ondansetron, a

Sanchez, C., Arnt, J., Hyttel, J., Moltzen, E.K., 1993. The role of selective 5-HT3 antagonist, reduce cocaine's psychomotor stimulant serotonergic mechanisms in inhibition of isolation-induced aggression effects in the rat. Psychopharmacology 101, S59. in male mice. Psychopharmacology 110, 53±59. Van der Poel, A.M., Noach, E.J., Miczek, K.A., 1989. Temporal B. Olivier et al. / European Neuropsychopharmacology 10 (2000) 77 ±95 95

patterning of ultrasonic distress calls in the adult rat: effects of stimulus properties of the serotonergic compound eltoprazine. J. morphine and benzodiazepines. Psychopharmacology 97, 147±148. Pharmacol. Exp. Therap. 260, 1045±1051. Vivian, J.A., Miczek, K.A., 1991. Ultrasounds during morphine with- Ybema, C., Van den Broek, I., Slangen, J.L., Olivier, B., Mos, J., 1991. drawal in rats. Psychopharmacology 104, 187±193. Further characterization of the discriminative stimulus induced by Vivian, J.A., Miczek, K.A., 1993. Diazepam and gepirone selectively ¯esinoxan. In: Olivier, B., Mos, J., Slangen, J.L. (Eds.), Animal attenuate either 20±32 kHz or 32±64 kHz ultrasonic vocalizations Models in Psychopharmacology, Birkhauser,È Basel, pp. 387±393. during aggressive encounters. Psychopharmacology 112, 66±73. Yocca, F.D., 1990. Novel anxiolytic agents: actions on speci®c subtypes Vogel, J.R., Beer, B., Clody, D.E., 1971. A simple and reliable con¯ict of central 5-HT receptors. In: Meldrum, B.S., Williams, M. (Eds.), procedure for testing anti-anxiety agents. Psychopharmacologia 21, Current and future trends in anticonvulsant, anxiety and stroke therapy, 1±7. Wiley-Liss, New York, pp. 145±167. Waeber, C., Pinkus, L.M., Palacios, J.M., 1990. The (S)-isomer of Young, R., Glennon, R.A., 1992. Zacopride and its optical isomers 3 [ H]zacopride labels 5-HT3 receptors with high af®nity in rat brain. produce stereoselective antagonism of a 2-methyl serotonin dis- Eur. J. Pharmacol. 181, 283±287. criminative stimulus. Eur. J. Pharmacol. 212, 117±119. White, S.M., Kucharik, R.F., Moyer, J.A., 1991. Effects of serotonergic Young, R., Johnson, D.N., 1991. Anxiolytic-like activity of R(1)- and agents on isolation-induced aggression. Pharmacol. Biochem. Behav. S(2)-zacopride in mice. Eur. J. Pharmacol. 201, 151±155. 39, 729±736. Zethof, T.J.J., Van der Heyden, J.A.M., Olivier, B., 1991. A new animal Wilde, M.I., Markham, A., 1996. Ondansetron. A review of its pharma- model for anticipatory anxiety. In: Olivier, B., Manceaux, A. (Eds.), cology and preliminary clinical ®ndings in novel applications. Drugs Destructive drives and impulse control. Preclinical considerations, 52, 773±794. Solvay, Weesp, pp. 34±40. Winslow, J., Insel, T.R., 1991. The infant rat separation paradigm: a Zethof, T.J.J., Van der Heyden, J.A.M., Tolboom, J.T.B.M., Olivier, B., novel test for novel anxiolytics. Tr. Pharmacol. Sci. 12, 402±404. 1994. Stress-induced hyperthermia in mice: a methodological study. Yarker, Y.E., McTavish, D., 1994. Granisetron. An update of its therapeu- Physiol. Behav. 55, 109±115. tic use in nausea and vomiting induced by antineoplastic therapy. Zethof, T.J.J., Van der Heyden, J.A.M., Tolboom, J.T.B.M., Olivier, B., Drugs 48, 761±793. 1995. Stress-induced hyperthermia as a putative anxiety model. Eur. J. Ybema, C., Slangen, J.L., Olivier, B., Mos, J., 1992. Discriminative Pharmacol. 294, 125±135.