Search

Increasing evidence suggests that the metabolism of glutathione, an endogenous redox regulator, is abnormal in schizophrenia. Patients show a deficit in glutathione levels in the cerebrospinal fluid and prefrontal cortex and a reduction in gene expression of the glutathione synthesizing enzymes. We investigated whether such glutathione deficit altered synaptic transmission and plasticity in slices of rat hippocampus, with particular emphasis on NMDA receptor function. An approximately 40% decrease in brain glutathione levels was induced by s.c. administration of L-buthionine-(S,R)-sulfoximine, an inhibitor of glutathione synthesis. Such glutathione deficit did not affect the basal synaptic transmission, but produced several NMDA receptor-dependent and -independent effects. Glutathione deficit caused an increase in excitability of CA1 pyramidal cells. The paired-pulse facilitation was diminished in glutathione-depleted slices, in a manner that was independent of NMDA receptor activity. This suggests that lowering glutathione levels altered presynaptic mechanisms involved in neurotransmitter release. NMDA receptor-dependent long-term potentiation induced by high-frequency stimulation was impaired in glutathione-depleted slices. Pharmacologically isolated NMDA receptor-mediated field excitatory postsynaptic potentials were significantly smaller in L-buthionine-(S,R)-sulfoximine-treated than in control slices. Hypofunction of NMDA receptors under glutathione deficit was explained at least in part by an excessive oxidation of the extracellular redox-sensitive sites of the NMDA receptors. These results indicate that a glutathione deficit, like that observed in schizophrenics, alters short- and long-term synaptic plasticity and affects NMDA receptor function. Thus, glutathione deficit could be one causal factor for the hypofunction of NMDA receptors in schizophrenia.

Cognitive impairment is a prominent feature of schizophrenia. Currently there is no well-accepted explanation of the aetiology of this disorder, but recent evidence indicates that dysfunction of the habenula may be involved. We therefore examined whether habenula lesions in Sprague-Dawley rats cause behavioural changes resembling those of schizophrenia. Rats received either habenula lesions, a sham operation or a small lesion of the overlying dorsal hippocampus as a check that effects observed were not due to incidental damage to this structure. As there are alterations of social behaviour, sensorimotor gating and cognition in schizophrenia, we examined comparable behaviours. Social interaction time was measured during a 5-min encounter with a novel juvenile conspecific. Prepulse inhibition of an acoustic startle response, as an index of sensorimotor gating, was measured with prepulses of various amplitudes, and spatial cognitive performance was assessed in the Morris water maze task. Histological analysis showed that habenula lesions substantially damaged both medial and lateral habenula bilaterally while largely sparing neighbouring structures. Assay of choline acetyltransferase (ChAT) in the interpeduncular nucleus terminal region of the habenulo-interpeduncular tract, showed marked reduction (by 80%) in habenula-lesioned animals. Habenula-lesioned rats, but not the control group with small dorsal hippocampus lesions, showed marked impairment of Morris maze performance compared to the sham-operated control group. Social interaction time and prepulse inhibition were not significantly altered in either lesion group. The results are consistent with a role of the habenula in cognition, and with the view that pathology of the habenula may contribute to the cognitive impairments of schizophrenia.

The non-competitive NMDA receptor antagonist phencyclidine (PCP) is known to produce a discriminative stimulus in rats. The first aim of the present study was to investigate which NMDA receptor subtype(s) is involved in this effect of PCP. Rats were trained to discriminate PCP (2 mg/kg; i.p.) from saline in a two lever operant task. The NMDA channel blocker, (+)MK-801 (0.1 mg/kg; i.p.) and the competitive NMDA receptor antagonist SDZ 220-581 (3 mg/kg; i.p.) produced 76% of PCP-lever selection (ED50=0.045 and 2 mg/kg, respectively), whereas their respective inactive enantiomers (-)MK-801 (0.025-0.1 mg/kg) and SDZ 221-653 (2-5 mg/kg) induced less than 30% of PCP-appropriate responding. Another competitive NMDA antagonist, SDZ EAB-515 (30 mg/kg; i.p.), induced 63% of PCP-lever responding (ED50=23.48 mg/kg). The selective antagonist of NMDA receptors containing the NR1A/NR2B-subunits Ro 25-6981 (20 mg/kg; i.p.) resulted in a complete substitution (more than 80% of PCP-lever selection) for PCP (ED50=8.59 mg/kg). In contrast, the NR1A/NR2A NMDA receptor-preferring antagonist NVP-AAM077 (2-10 mg/kg; i.p.) failed to produce PCP-like discriminative stimuli. At high doses SDZ 220-581 (ED50=2.44), NVP-AAM077 (ED50=8.33) and SDZ EAB-515 (ED50=25.81) decreased the performance of the rats in this operant task. The ability of these NMDA receptor antagonists to disrupt the prepulse inhibition (PPI) of the startle response and to alter locomotor activity was also studied. PCP (0.5-2 mg/kg; s.c.), SDZ 220-581 (0.5-5 mg/kg; s.c.), SDZ EAB-515 (1-30 mg/kg; i.p.) and Ro 25-6981 (5-20 mg/kg; i.p.) disrupted PPI and at high doses produced hyperlocomotion. In contrast, NVP-AAM077 (5-20 mg/kg; i.p.) did not disrupt PPI and reduced locomotor activity. In conclusion, it appears that the NMDA receptor containing the NR2B, rather than the NR2A subunit, may play a major role in the PCP-like discriminative stimulus. In addition, sensory motor gating disturbances associated with NMDA antagonists do not seem to result from a blockade of NR1/NR2A-containing NMDA receptors.

The dopamine D(3) receptor agonist PD 128907 decreased body temperature in the rat. The selective dopamine D(3) and D(4) receptor antagonists, A-437203 and L-745,870, respectively, did not prevent this effect. In contrast, PD 128907-induced hypothermia was antagonized by SCH 23390, a selective D(1) receptor antagonist, and by L-741,626, a selective D(2) receptor antagonist. Moreover, the selective D(2) receptor agonist trihydroxy-N-n-propylnoraporphine (TNPA) elicited a robust hypothermia which was prevented by pretreatment with L-741,626 but not by A-437203. In agreement with previous data obtained in D(3) knock-out mice, present results suggest that D(2) rather than D(3) receptors mediate dopamine receptor agonist-induced hypothermia in rats. In addition, it appears that both D(1) and D(2) receptors may be involved in a cooperative manner.

Glutamatergic neurotransmission has been strongly implicated in the pathophysiology of affective disorders, such as major depression and anxiety. Of all glutamate receptors, the role of group III metabotropic glutamate receptors (mGluR4, mGluR6, mGluR7, mGluR8) in such disorders is the least investigated because of the lack of specific pharmacological tools. To this end, we examined the behavioural profiles of mice with a targeted deletion of the gene for mGluR7 (mGluR7-/-) in animal models of depression and anxiety. mGluR7-/- mice were compared with wild-type (mGluR7+/+) littermates and showed substantially less behavioural immobility in both the forced swim test and the tail suspension test. Both behavioural paradigms are widely used to predict antidepressant-like activity. Further, mGluR7-/- mice displayed anxiolytic activity in four different behavioural tests, i.e. the light-dark box, the elevated plus maze, the staircase test, and the stress-induced hyperthermia test, while their cognitive performance was normal in the passive avoidance paradigm. Analysis of locomotor activity in a novel environment demonstrated that mGluR7-/- mice were slightly more active in the initial minutes following placement in the chamber only. Together, these data suggest that mGluR7 may play a pivotal role in mechanisms that regulate behavioural responses to aversive states. Therefore, drugs acting at mGluR7 may provide novel treatments for psychiatric disorders such as depression and anxiety.

Recently, selective and systemically active antagonists for the metabotropic glutamate 5 receptor (mGlu(5)) were discovered, and the most potent derivative was found to be MPEP (2-methyl-6-(phenylethynyl)pyridine). Given the high expression of mGlu(5) receptors in limbic forebrain regions, it was decided to evaluate the anxiolytic potential of MPEP. After an acute oral administration, MPEP attenuated the anxiety-dependent variable in a variety of well established anxiety test paradigms. In rats, MPEP (10, 30, and 100 mg/kg) increased punished responses in the Geller-Seifter test, but none of these effects reached statistical significance. MPEP significantly increased the ratio (open/total arm entries; 0.1, 1, and 10 mg/kg), the number of open arm entries (0.1, 1, and 10 mg/kg), as well as time spent on open arm (0.1 and 1 mg/kg) in the elevated plus maze test. Furthermore, MPEP (0.3 and 1 mg/kg) significantly increased the time spent in social contact in the social exploration test. In mice, MPEP attenuated stress-induced hyperthermia (15 and 30 mg/kg) and decreased the number of buried marbles in the marble burying test (7.5 and 30 mg/kg). Finally, MPEP (0.01, 0.1, 1, 10, and 100 mg/kg) was tested on spontaneous locomotor activity in mice, and only a dose of 100 mg/kg significantly reduced vertical activity; no effect was seen on horizontal activity. MPEP (7.5, 15, and 30 mg/kg) was ineffective on d-amphetamine-induced (2.5 mg/kg) locomotor activity in mice and prepulse inhibition in rats (1, 3, or 10 mg/kg). Thus, these findings indicate that MPEP exhibits anxiolytic-like effects and low risks for sedation and psychotomimetic side-effects in rodents.

Noncompetitive N-methyl-D-aspartate (NMDA) receptor antagonists such as phencyclidine are psychotomimetics and disrupt prepulse inhibition (PPI), a measure of sensorimotor gating that is deficient in schizophrenia. Systemically administered competitive NMDA receptor antagonists do not disrupt PPI in rats, leading to speculation that these compounds might have use as neuroprotective agents without the risk of psychotomimetic side effects. The effects on sensorimotor gating and locomotor activity of competitive NMDA receptor antagonists that either penetrate (SDZ 220-581 and SDZ EAB-515) or poorly penetrate [SDZ EAA-494 (D-CPPene)] the blood-brain barrier were compared. Rats were treated with either SDZ 220-581 (0, 2.5, or 5.0 mg/kg) or SDZ EAB-515 (0, 3.0, 10.0, or 30.0 mg/kg) and tested for PPI and locomotor activity. Different rats were tested for PPI after either systemic (0, 0.5, 1.0, or 5.0 mg/kg) or intra-amygdala (0 or 1.0 microg/microl) administration of D-CPPene. Finally, rats were pretreated with clozapine (0 or 5.0 mg/kg) or haloperidol (0 or 0.1 mg/kg), together with SDZ 220-581 (0 or 2.5 mg/kg), and tested. SDZ 220-581 and SDZ EAB-515 decreased PPI without affecting startle magnitude. Reduced PPI was noted after central but not systemic administration of D-CPPene. The gating deficits produced by SDZ 220-581 were blocked by clozapine or haloperidol. Movement pattern analysis indicated that locomotor activity was increased by SDZ 220-581 and SDZ EAB-515 in a phencyclidine-like manner. These results indicate that competitive NMDA receptor antagonists, if they gain sufficient access to the brain, produce a behavioral profile that resembles that of the psychotomimetic noncompetitive antagonists.

A selection of biphenyl-analogues of 2-amino-7-phosphonoheptanoic acid (AP7), N-methyl-D-aspartate (NMDA) receptor antagonists with high affinity in vivo efficacy. The lead compound SDZ EAB 515 was found to inhibit L-phenylalanine uptake by the large neutral amino acid carrier in vitro and in vivo; active transport may thus confer a good bioavailability to this class of compounds. CNS effects were demonstrated by significant changes in 2-deoxyglucose-uptake in various brain regions at doses from 1 to 10 mg/kg i.p. With the most active agent, SDZ 220-581, full protection against maximal electroshock seizures (MES) was obtained at oral doses of 10 mg/kg in rats and in mice. The compound had a fast onset (< or = 1 hr) and a long duration (> or = 24 hr) of action. Motor-debilitating effects (impairment of rotarod performance) occurred at doses about 10 times higher than those required for protection against MES. Neuroprotective activity was demonstrated by the ability of the compounds to reduce the extent of quinolinic acid-induced striatal lesions in rats, in the dose range of 3-15 mg/kg (i.p.) or 10-50 mg/kg (p.o.). In the middle cerebral artery occlusion (MCAO) model of focal cerebral ischemia in rats, the test compounds reduced the infarct size by 40-50% when given i.v. before or by 20-30% when given i.v. 1 hr after MCAO. SDZ 220-581 provided 20-30% protection at > or = 2 x 10 mg/kg p.o. This compound also showed analgesic activity at low oral doses in a model of neuropathic pain, although higher doses were required in model of mechanical inflammatory hyperalgesia. Unexpectedly, SDZ 220-581 at low s.c. doses counteracted the antiparkinsonian effects of L-DOPA in MPTP-treated marmosets. (Sub)chronic administration of SDZ 220-581 did not reduce its ability to protect against quinolinic acid neurotoxicity, and no upregulation of NMDA receptors was detected using a [3H]CGP-39653 binding assay. In conclusion, from a series of biphenyl-AP7-derivatives, SDZ 220-581 is clearly the most active compound in vivo. Its pharmacological profile with a good, long-lasting oral activity might open up novel therapeutic applications for competitive NMDA receptor antagonists.

SDZ EAA 494 (D-CPPene) was characterized as a competitive NMDA antagonist, having a pA2 value against NMDA depolarizations in frog spinal cord and rat neocortex of 6.7-6.8 and a pKi of 7.5 in a [3H]CGP39653 binding assay, with no action on other receptors or amine reuptake. The compound was orally active in rodent maximal electroshock models with an ED50 of around 16 mg/kg, was protective in rats even 24 hours after oral application and had an oral therapeutic index of around 8. Muscle relaxation, ataxia, flattened body posture and reduced acquisition of a passive avoidance task, suggesting potential effects on memory formation, occurred at supra-anticonvulsant doses in rodents, with PCP-like stimulatory effects produced only by high i.p. doses or constant i.v. infusions. This favourable profile is discussed in relation to the negative outcome of a recent trial of the compound in patients with intractable epilepsy. The conclusion is drawn that standard models for screening new anticonvulsants are inappropriate to seeking drugs active in patients with a protracted convulsive history. The anti-ischaemic action of SDZ EAA 494 encourages further testing in brain trauma, in which the anticonvulsant action of the compound may be an added benefit.

D- and L-enantiomers of the competitive NMDA antagonists CPP and CPP-ene, as well as the non-competitive NMDA antagonist MK-801, inhibited spontaneous activity occurring in slices of rat cerebral cortex exposed to Mg2(+)-free medium. D-CPP-ene (SDZ EAA 494) was the most active competitive antagonist with a threshold concentration of 10 nM and an ED50 of 39 nM. The inhibitory effects of all competitive antagonists were reversible, whereas reversibility following MK-801 (ED50 = 33 nM) was incomplete and slow. D-CPP-ene was also the most potent competitive antagonist against NMDA-evoked depolarizations, having an apparent pA2 value of 6.8; its action was specific to the NMDA type of excitatory amino acid receptor.

[3H]ICS 205-930 labelled 5-HT3 recognition sites in membranes prepared from murine neuroblastoma N1E-115 cells. Binding was rapid, reversible, saturable and stereoselective to an apparently homogeneous population of sites. Kinetic studies revealed that agonists and antagonists produced a monophasic dissociation reaction of [3H]ICS 205-930 from its recognition sites. The dissociation rate constant of the radioligand was similar whether the dissociation was induced by an agonist or an antagonist. Competition studies carried out with agonists and antagonists also suggested the presence of a homogeneous population of [3H]ICS 205-930 recognition sites. Competition curves were best fit for a 1 site model. [3H]ICS 205-930 binding sites displayed the pharmacological profile of a 5-HT3 receptor. The interactions of agonists and antagonists with [3H]ICS 205-930 recognition sites were apparently competitive in nature, as demonstrated in kinetic and equilibrium experiments. In saturation experiments carried out with [3H]ICS 205-930 in the presence and the absence of unlabelled agonists and antagonists, apparent Bmax values were not reduced whereas apparent Kd values were increased in the presence of competing ligands. There was a good agreement between apparent pKB values calculated for the competing ligands in saturation experiments and pKd values calculated from competition experiments. The present data demonstrate that [3H]ICS 205-930 labels a homogeneous population of sites at which agonists and antagonists interact competitively.

Parallel electrophysiological and anatomical tracing studies on the projections from the hippocampal formation to the ventral striatum in the cat, were carried out with the aim of defining the nature and organization of the motor interfaces of the limbic cortex. In this context the main pathway was found to be that from the subiculum to the N. Accumbens; electrophysiological characteristics of this pathway are reported in the present study. This well organized pathway is primarily of an excitatory nature and consists of slow conducting fibers (1-2 ms-1). Three types of response patterns of single Accumbens units to subicular simulation were found: units which responded with a burst of action potentials (the majority); units which responded with iterative bursting; units responding primarily with an inhibition of firing. Often the primary excitatory response was followed by a decrease in firing rate. Analysis of Evoked Potentials indicate that the subicular inputs induce a monopolar positive field within the N. Accumbens; in combination with the evidence obtained from current source density (CSD) analysis these data suggest the following activation model: subicular inputs evoke depolarization of synapses lying peripherally in the dendrites of stellate neurons provoking peripheral sinks which form a concentric shell around sources at and in the neighbourhood of the cell bodies. In addition to the subiculum-Accumbens pathway evidence was also obtained for an excitatory input from the Entorhinal cortex to the N. Accumbens. Furthermore, an excitatory pathway from the prepyriform cortex to the olfactory tubercle was also electrophysiologically identified.(ABSTRACT TRUNCATED AT 250 WORDS)

1. The binding characteristics of [3H]ICS 205-930, a potent and selective 5-hydroxytryptamine 5-HT3 receptor antagonist, were investigated in membranes prepared from murine neuroblastoma-glioma NG 108-15 cells. 2. [3H]ICS 205-930 bound rapidly, reversibly and stereoselectively to a homogeneous population of high affinity recognition sites: Bmax = 58 +/- 3 fmol/mg protein, pKD = 9.01 +/- 0.08 (n = 11). Non linear regression and Scatchard analysis of saturation data suggested the existence of a single class of [3H]ICS 205-930 recognition sites on NG 108-15 cells. The binding was rapid, stable and reversible. The affinity of [3H]ICS 205-930 determined in kinetic studies was in agreement with that obtained under equilibrium conditions. 3. Competition studies performed with a variety of agonists and antagonists also suggested the presence of a homogeneous population of [3H]ICS 205-930 recognition sites. All competition curves were steep and monophasic and were best fit by a 1 receptor site model. [3H]ICS 205-930 binding sites displayed the pharmacological profile of a 5-HT3 receptor. Potent 5-HT3 receptor antagonists showed nanomolar affinities for [3H]ICS 205-930 binding sites with the following rank order of potency: SDZ 206-830 greater than ICS 205-930 greater than SDZ 206-792 greater than BRL 43694 greater than quipazine greater than BRL 24924 greater than SDZ 210-204 greater than MDL 72222 greater than SDZ 210-205. Metoclopramide, mCP and mianserin showed submicromolar affinity.(ABSTRACT TRUNCATED AT 250 WORDS)

[3H]ICS 205-930 recognition sites were analyzed in membranes prepared from murine neuroblastoma N1E-115 cells. [3H]ICS 205-930 bound rapidly, reversibly, and stereoselectively to a homogeneous population of high affinity recognition sites: Bmax = 40 +/- 5 fmol/mg of protein, pKD = 9.20 +/- 0.05 (n = 11). Nonlinear regression and Scatchard analysis of saturation data suggested the existence of a single class of [3H]ICS 205-930 recognition sites on N1E-115 cells. The affinity of [3H]ICS 205-930 determined in kinetic studies was in agreement with that obtained under equilibrium conditions. Competition studies carried out with a large variety of agonists and antagonists also suggested the presence of a homogeneous population of [3H]ICS 205-930 recognition sites. [3H]ICS 205-930-binding sites displayed the pharmacological profile of a 5-HT3 receptor. Potent 5-HT3 receptor antagonists showed nM affinities for [3H]ICS 205-930-binding sites with the following rank order of potency: SDZ 206-830 greater than SDZ 206-792 greater than ICS 205-930 greater than BRL 43694 greater than quipazine greater than BRL 24924 greater than MDL 72222 greater than GR 38032F. Methiothepine, mCPP, and metoclopramide showed sub-microM affinity. The rank order of potency of agonists was: 5-HT greater than phenylbiguanide = 2-methyl-5-HT much greater than 5-methoxytryptamine = 5-carboxamidotryptamine. All antagonist competition curves were steep (pseudo-Hill coefficients not lower than 1), monophasic, and best fit for a one-site model; 5-HT and 2-methyl-5-HT produced pseudo-Hill coefficients of 1.2-1.4. Drugs acting at 5-HT1, 5-HT2, alpha- and beta-adrenergic, dopaminergic, and histaminergic receptors (methysergide, ketanserin, propranolol, phentolamine, sulpiride, SCH 23390, cimetidine) were essentially inactive at 10 mumol/liter. The binding of [3H]ICS 205-930 was not affected by guanine and adenine nucleotides (GTP, GppNHp, and ATP) at 1 mmol/liter. These nucleotides did not affect the binding of agonists, suggesting that 5-HT3 recognition sites are not coupled to G-proteins. The interactions of agonists and antagonists with [3H]ICS 205-930 recognition sites were competitive in nature, as demonstrated by saturation experiments carried out with [3H]ICS 205-930 in the presence and the absence of unlabeled compounds: apparent Bmax values were not reduced, whereas apparent KD values were increased in the presence of competing ligands.(ABSTRACT TRUNCATED AT 400 WORDS)

The aim of this study was to investigate the pharmacological characteristics of the 5-hydroxytryptamine-(5-HT)-induced electrical response in cultured neuroblastoma N1E-115 cells of the mouse. In these cells 5-HT induces a transient membrane depolarization, which is associated with a transient inward current, that has been recorded in voltage clamp experiments on whole cells. The peak amplitude of the inward current depends on the concentration of 5-HT applied. Maximum peak inward current was evoked by 10 microM 5-HT and half maximum effect by 2 microM. Responses to 5-HT were blocked by nanomolar concentrations of selective 5-HT3-receptor antagonists, whereas the selective agonist 2-methyl-5-HT mimicked the membrane depolarization induced by 5-HT. A number of agonists and antagonists, which are known to act on 5-HT1-like, 5-HT2, dopaminergic and adrenergic receptors failed to affect the response to 5-HT in neuroblastoma cells. Observed antagonistic effects of SCH 23390 [(R)-(+)-8-chloro-2,3,4,5-tetrahydro-3-methyl-5-phenyl-1H-3-benzazepi n-7-ol hemimaleate] and haloperidol are discussed. The inhibitory effect of the 5-HT3 receptor antagonist, ICS 205-930 [(3 alpha-tropanyl)-1H-indole-3-carboxylic acid ester] has been demonstrated. When cells were exposed to 0.1 nM ICS 205-930 the maximum evoked response was reduced by about 50%, but a surmountable shift of the concentration-response curve of 5-HT was not observed. The kinetics of the 5-HT-induced inward current remained unchanged in the presence of ICS 205-930. Recovery from the block by ICS 205-930 was very slow.(ABSTRACT TRUNCATED AT 250 WORDS)

Serotonin (5HT) and dopamine (DA) induce, in neuroblastoma N1E-115 cells, a transient membrane depolarization associated with an inward current. The half-maximum response is obtained with 2 microM 5HT or 200 microM DA. The maximum response to 5HT is 2-3 times that to DA. The selective 5HT3 receptor antagonists ICS 205-930 and MDL 72222 at nanomolar concentrations block both the 5HT- and the DA-induced response. High concentrations (10 microM) of 5HT2 receptor antagonists are without effect. It is concluded that, in N1E-115 cells, 5HT and DA activate a single population of 5HT3 receptors.

1. Ionic currents mediated by serotonin 5-HT3 receptors were studied in the mouse neuroblastoma cell line N1E-115, using suction pipettes for intracellular perfusion and voltage clamp recording. The dependence of the kinetics of the membrane current on serotonin concentration was investigated. 2. At a holding potential of -70 mV application of 5-HT (5-hydroxytryptamine creatinine sulphate) causes a transient inward current. The i-V curve of the peak amplitude is linear between -80 and 60 mV. The reversal potential is 20 +/- 4 mV (mean +/- S.D.). The kinetics of the transient ionic current are independent of the holding potential. 3. In the presence of 5-HT the membrane current decays to a small steady-state level with a single-exponential time course. The time constant of decay decreases with increasing concentration of the agonist, to a minimum value of 6.5 +/- 1.5 s for concentrations of 5-HT greater than or equal to 3 microM. 4. When the agonist is rapidly removed, single-exponential decay of the ionic current is observed. The time constant of this decay in the absence of 5-HT amounts to 6.9 +/- 1.5 s and is independent of the membrane potential and of the concentration of 5-HT used. 5. In the presence of low concentrations of 5-HT the peak amplitude of the inward current evoked with a high concentration of agonist is gradually reduced. The onset of this desensitization follows the same time course as the decay of the membrane current. In the range from 0.7 to 1.5 microM-5-HT both kinetic processes show the same steep concentration dependence. 6. Recovery from desensitization, measured at variable intervals after removal of the agonist, can be fitted by a single-exponential function with a time constant of 18 +/- 4 s. 7. The results show that the kinetic properties of the 5-HT3 receptor-mediated ionic current can only be described by a complex, co-operative model.