Attenuation of the bidirectional effects of chlordiazepoxide and FG 7142 on conditioned response suppression and associated cardiovascular reactivity by loss of cortical cholinergic inputs

RATIONALE: Basal forebrain cortical cholinergic projections have been hypothesized to mediate the enhanced cardiovascular defensive response initiated by the putative anxiogenic benzodiazepine receptor (BZR) partial inverse agonist FG 7142 (FG). The present study was designed to test the broader hypothesis that the integrity of this cholinergic projection is required for the mediation of the bidirectional modulatory effects of BZR agonists and inverse agonists on anxiety and associated cardiovascular reactivity. OBJECTIVES: The interactions between the effects of 192 IgG-saporin-induced lesions of basal forebrain corticopetal cholinergic neurons and of the BZR agonist chlordiazepoxide (CDP) and FG on the performance of rats tested in a conditioned suppression paradigm and on associated cardiovascular reactivity were assessed. METHODS: Lesioned and control animals were equipped with a telemetric device to record heart rate, trained in an operant lever task, and then tested for suppression of responding during presentation of a conditioned stimulus (CS) and a general contextual cue that was previously associated with shock. FG, CDP (8 mg/kg) and vehicle were administered IP in separate extinction sessions. RESULTS: In control animals, operant responding was suppressed during presentation of the CS and contextual cue. Administration of FG enhanced this suppression, while CDP attenuated it. Lesions attenuated overall response suppression as well as the modulatory effects of BZR ligands on responding during presentation of the contextual stimulus. Likewise, lesions attenuated the cardioacceleratory response to the contextual stimulus and the ability of the BZR ligands to modulate this response. CONCLUSIONS: The behavioral and autonomic responses to anxiety-related stimuli, as well as the modulatory effects of BZR ligands, are mediated in part via cortical cholinergic inputs.     

Effects of benzodiazepine receptor ligands on the performance of an operant delayed matching to position task in rats: opposite effects of FG 7142 and lorazepam

The effects of a series of benzodiazepine (BZ) receptor ligands, ranging from a full agonist through to partial inverse agonists, were examined on short term working memory in the rat. The behavioural paradigm used was a discrete trial, operant delayed matching to position task, as originally described by Dunnett (1985), with delays of 0, 5, 15 and 30 s. The benzodiazepine receptor (BZR) full agonist lorazepam (0.25, 0.375 and 0.5 mg/kg) dose and delay dependently impaired matching accuracy. Lorazepam also increased the latency to respond and decreased the number of nose pokes made into the food tray during the delays. In contrast, the BZR partial agonist ZK 95 962 (1, 3, 10 mg/kg) did not affect matching accuracy, but did increase the speed of responding. The BZR antagonist ZK 93 426 (1.25, 5, 25 mg/kg) had no effects in this paradigm. The BZR weak partial inverse agonists Ro 15-4513 (0.1, 1 and 10 mg/kg) and ZK 90 886 (1, 3 and 10 mg/kg) did not affect accuracy of performance. However, both of these drugs increased the latency to respond and decreased nose poke responses. These motoric effects were particularly strong following 10 mg/kg Ro 15-4513. This shows that the effects of drugs on the accuracy of responding and on the speed of responding can be dissociated. The BZR partial inverse agonist FG 7142 had effects on matching accuracy that were dependent upon dose. The lowest dose of FG 7142 (1 mg/kg) significantly improved accuracy, whereas the highest dose (10 mg/kg) impaired accuracy. This impairment induced by FG 7142 (10 mg/kg) was accompanied by an increase in the latency to respond and a decrease in the number of nose pokes. Taken together, these results show that the accuracy of delayed matching performance can be modulated in opposite ways by the BZR full agonist lorazepam and a low dose of the BZR partial inverse agonist, FG 7142.     

FG 7142 selectively decreases nonpunished responding,but has no anxiogenic effects on time allocation in a conflict schedule

Previous work (Thomas et al. 1990) showed that an anxiolytic benzodiazepine increased the time allocated to responding in a conflict situation (where responses were both food-reinforced and shock-punished) versus a nonpunishment situation. The present experiment tested whether a benzodiazepine-receptor inverse agonist (FG 7142, 1–30 mg/kg) would have the opposite effect (i.e., decrease time spent responding in a punishment situation). Chain pulls determined whether a rat's lever presses were reinforced on 1) a lean variable-interval schedule, or 2) a richer variable-interval schedule in which responding also produced shock intermittently. FG 7142 dose-dependently decreased nonpunished lever responding, but did not affect punished responding. The drug nonselectively decreased chain pulling (the schedule-switching response). Like chlordiazepoxide, FG 7142 increased the time spent in the punishment component, showing that not all effects of benzodiazepine-receptor agonists and inverse agonists are opposite. These results are inconsistent with expectations that anxiogenic actions of FG 7142 should 1) decrease punished responding; 2) increase the rate of responses that terminate the punishment condition; and 3) decrease time spent in the punishment component. Rather, nonsuppressed responding seems most sensitive to decreases by FG 7142.     

Evidence that the amygdala is involved in benzodiazepine and serotonergic effects on punished responding but not on discrimination

Interactions between the benzodiazepines (BZs) chlordiazepoxide (CDP) and midazolam (MDZ), the BZ antagonist R0 15-1788, the inverse BZ receptor agonists CGS 8216 and FG 7142, -aminobutyrate (GABA), serotonin (5-HT), the 5-HT₂ antagonist methysergide and the putative 5-HT agonist 8-hydroxy-2-(di-n-propylamino) tetralin (8-OH-DPAT) were investigated using peripheral and intra-amygdaloid treatments. A multiple schedule consisting of rewarded, nonrewarded (Time out: TO) and conflict periods was used to compare in parallel effects on successive discrimination between rewarded and nonrewarded periods and punished responding. The three components were presented in both a fixed order (Experiment 1) and a random order (Experiments 2 and 3).Intra-amygdaloid treatments with GABA and the BZs selectively increased rates of punished responding. CDP given systemically, on the other hand, increased both TO and conflict rates, suggesting an additional impairment of discrimination, which was more marked in the random than the fixed order condition. R0 15-1788, CGS 8216 and FG 7142 given by both routes counteracted the anti-conflict effects of CDP given centrally or systemically. However increases in TO rates induced by IP CDP were antagonized only by IP treatments with these compounds. The two inverse agonists, but not R0 15-1788, also counteracted increases in punished responding which were found after intra-amygdaloid GABA infusions. In Experiments 2 and 3 where baseline rates of pressing in Conflict periods were sufficiently high to detect decreases, CGS 8216 and FG 7142 reduced responding below control level, suggesting a specific anxiogenic activity. Evidence for effects of R0 15-1788 by itself was inconclusive. 5-HT injected into the amygdala also reduced punished responding below control level, whereas methysergide increased it with both central and peripheral treatment. Effects of 8-OH-DPAT varied according to route of administration. With IP treatment Conflict rates were increased, but after amygdaloid infusion both TO and Conflict rates were marginally reduced below control level, with a more consistent depression of punished responding.These results provide evidence that effects of BZs on punished responding are mediated by a GABAergic system which includes the lateral/basolateral amygdala, but which does not participate in BZ-induced disruption of discrimination. They also indicate that the antagonistic effects of CGS 8216 and FG 7142 involve a decrease in GABA transmission, and that these compounds may also be anxiogenic. Finally, the results suggest that 5-HT utilizes the same system for regulating resistance to punishment, but plays no significant part in reward-nonreward successive discrimination, which is impaired after systemic BZs.This work was supported by the award of the MRC project Grant No. 8313090N to SG     

Behavioral studies with the beta-carboline FG 7142 combined with related drugs in monkeys

The beta-carboline FG 7142 was studied alone and in combination with Ro 15-1788, CGS 8216 and lorazepam in squirrel monkeys trained to respond under a fixed-interval (FI) schedule of food presentation. FG 7142 (0.3-5.6 mg/kg i.v.) produced dose-related decreases in the rate of FI responding, effects opposite to those of moderate doses of lorazepam (0.03-0.3 mg/kg i.v.). Pretreatment with low doses of Ro 15-1788 (0.1 and 0.3 mg/kg i.v.) shifted the dose-response curve of FG 7142 progressively to the right indicating pharmacological antagonism at benzodiazepine recognition sites. In comparison, pretreatment with the pyrazoloquinoline CGS 8216 (0.1-1.0 mg/kg i.v.), which alone decreased responding, did not alter the effects of FG 7142 in a systematic manner. Combinations of behaviorally active doses of FG 7142 and lorazepam had primarily additive effects: the opposing actions of one drug tended to cancel the other's effect on responding. These results show that the reduction in behavior by FG 7142 is modified predictably by Ro 15-1788 but not by CGS 8216, and behaviorally active doses of both FG 7142 and lorazepam may be needed for their mutual antagonism.     

Effects of diazepam,FG 7142, and RO 15-1788 on schedule-induced polydipsia and the temporal control of behavior

Although benzodiazepine agonists and inverse agonists have opposite effects on drinking elicited by water deprivation, there is much less information about the effects of these drugs on nonhomeostatic drinking. In this experiment the effects of diazepam (0.3–5.0 mg/kg), a benzodiazepine receptor agonist, and FG 7142 (1.0–9.0 mg/kg), an inverse agonist, were determined on drinking elicited by a FT-60 schedule of food delivery (SIP). Both diazepam and FG 7142 dose-dependently reduced SIP, measured as either licking or volume consumed. In addition, diazepam reduced panel pressing for food, decreased locomotor activity, and changed the time course of each behavior. In contrast, FG 7142 reduced schedule-induced drinking without significantly altering other behaviors. The antagonist RO 15-1788, when given in combination with these drugs, only partially restored the reductions in licking produced by diazepam, but was much more effective in reversing the effects of FG 7142 at doses of the antagonist that failed by themselves to affect responding. The opposite pattern of effects was seen on the volume of water consumed. These effects are discussed in terms of the behavioral and pharmacological specificity of these drugs.     

beta-Carboline FG 7142-reduced aggression in male rats: reversed by the benzodiazepine receptor antagonist, Ro15-1788

The beta-carboline FG 7142 decreases conspecific aggression in male hooded rats. The purpose of this study was to examine the effects of pretreatment with Ro15-1788 or chlordiazepoxide (CDP) in this paradigm. The six groups (n = 8) were saline, FG 7142 (5 mg/kg, immediate, IP), CDP (5 mg/kg, -10 min, IP), CDP (5 mg/kg, -10 min) plus FG 7142 (5 mg/kg, immediate), Ro15-1788 (10 mg/kg, -10 min, IP), and Ro15-1788 (10 mg/kg, -10 min) plus FG 7142 (5 mg/kg, immediate). Following injection of the more aggressive member of a pair of isolation-housed rats, the pair was observed in a living cage over four 6-min trials interpolated over a 40 min session. In the first 20 min after the injection FG 7142 decreased aggression, decreased the pinning of the other animal, and increased avoiding behavior. These effects were the opposite of those seen in the Ro15-1788-injected rats and Ro15-1788 pretreatment reversed the effects of FG 7142. CDP alone caused prolonged aggressive behavior but as a pretreatment only partially reversed the effects of FG 7142.     

Reversal of alcohol-induced amnesia by the benzodiazepine inverse agonist Ro 15-4513

We investigated the effects of benzodiazepine inverse agonists on ethanol-induced amnesia using a passive avoidance task. Pretraining treatment of mice with ethanol significantly impaired the passive avoidance response: there was a significant reduction in the % retention and step-down latency. The benzodiazepine inverse agonists, Ro 15-4513 and beta-CCM, significantly increased the % retention and prolonged the step-down latencies in mice treated with ethanol, but FG 7142 did not. The anti-amnesic effects of Ro 15-4513 were completely antagonized by co-administration of Ro 15-1788, a benzodiazepine antagonist. These results suggest that the anti-amnesic effect of Ro 15-4513 on alcohol-induced amnesia is mediated by benzodiazepine receptors.     

Inverse agonist properties of the THBC discriminative stimulus: asymmetrical generalization with FG 7142.

Male rats were trained to discriminate the stimulus properties of the beta-carbolines 1,2,3,4-tetrahydro-beta-carboline (THBC) (15.0 mg/kg) or FG 7142 (5.0 mg/kg) from vehicle in a two-lever, food-motivated operant task. Consistent with the serotonergic properties of THBC, administration of the 5HT1B agonists TFMPP and mCPP to THBC-trained rats resulted in THBC-appropriate responding. Norharmane, a beta-carboline metabolite of THBC, also mimicked the THBC discriminative stimulus. In contrast, the benzodiazepine receptor partial inverse agonist FG 7142, the anxiogenic/convulsant pentylenetetrazole (PTZ), two physiological stressors and the alpha 2 adrenergic antagonists yohimbine and idazoxan failed to produce THBC-appropriate responding. In the FG 7142-trained rats, THBC and norharmane dose-dependently mimicked the FG 7142 discriminative stimulus. This generalization was not based upon the serotonergic properties of THBC and norharmane since administration of the serotonin agonist mCPP to FG 7142-trained rats failed to produce FG 7142-appropriate responding. The ability of THBC to substitute for the FG 7142 discriminative stimulus was antagonized by the benzodiazepine receptor mixed agonist/antagonist CGS 9896 and the benzodiazepine receptor antagonist RO 15-1788, indicating that THBC produces an inverse agonist stimulus in FG 7142-trained rats. These results suggest that THBC produces a discriminative stimulus which consists of both serotonergic and inverse agonist components.     

Effects of prenatal exposure to benzodiazepine-related drugs on early development and adult social behaviour in Swiss mice--III. Inverse agonists.

1. The present experiment examined, using a battery of tests, the effects of in utero exposure to the benzodiazepine inverse agonists DMCM and FG 7142 upon the early development and adult behaviour of Swiss Mice. 2. Early development was respectively retarded and augmented by treatments with the higher and lower doses of DMCM. FG 7142 suppressed later development. DMCM retarded righting reflex on certain days. FG 7142 had a significant, biphasic effect on eye opening. 3. Adult social behaviour was examined using the resident-intruder paradigm, by determining the time spent in broad behavioural categories. Male offspring of dams treated with DMCM showed increased threat. FG 7142 had no significant effects.     

Differential pharmacological reactivity of aversion induced by stimulation of periaqueductal gray or mesencephalic locomotor region

Rats were trained to switch-off aversive electrical brain stimulations applied to the periaqueductal gray (PAG) or mesencephalic locomotor region (MLR) by pressing a bar (switch-off behavior). We investigated the effects of IP injections of the benzodiazepine (BZ) receptor inverse agonist FG 7142 (2.5, 5, 10 mg/kg) or BZ receptor agonist chlordiazepoxide (CDP: 5 mg/kg) on the switch-off latency, i.e., the time elapsed between the onset of the stimulation and its offset by a press of the bar. It was found that FG 7142 decreased, whereas CDP increased the mean switch-off latency for electrical stimulation of the PAG, which is interpreted as a potentiating effect of FG 7142 and a reducing effect of CDP on the electrically induced aversive state. By contrast, neither FG 7142 nor CDP were found to affect the mean switch-off latency for MLR stimulations. These results suggest a difference in the pharmacological sensitivity to BZ receptor ligands between aversive states elicited by electrical stimulation of the PAG or MLR.     

Discriminative stimulus control by the anxiogenic beta-carboline FG 7142: generalization to a physiological stressor

Drug discrimination was employed to investigate the similarities between FG 7142-induced anxiogenesis and the stress produced by exposure to either a novel environment or to footshock. Eight rats were trained to discriminate between the stimulus properties of the beta-carboline FG 7142 (5.0 mg/kg) and its vehicle in a two-lever, food motivated operant task. Once trained, decreasing doses of FG 7142 produced fewer FG 7142-appropriate responses and the dose-response relationship yielded an ED50 of 1.45 mg/kg. Rats were subsequently subjected to two physiological/environmental stressors, footshock and novelty, and then tested in the discriminative paradigm. Exposure to novelty resulted in partial FG 7142-appropriate responding, whereas footshock sessions produced responding predominately on the FG 7142-appropriate lever. This is the first report of stimulus control by FG 7142 and it is likely that the interoceptive cue state produced by this compound is anxiogenic in nature, as reported to occur in man. The anxiogenic nature of the FG 7142 discriminative stimulus is supported by the generalization of FG 7142 to the state produced following stressful environmental manipulation.     

Antagonism of ethanol-reinforced behavior by the benzodiazepine inverse agonists Ro15-4513 and FG 7142: relation to sucrose reinforcement

The partial inverse benzodiazepine agonist Ro15-4513 has been shown to antagonize many of ethanol's actions, including the reduction of behavior reinforced with ethanol presentation. The studies reported here compared the effects of the Ro compound on sucrose reinforcement alone and concurrently available with ethanol reinforcement. Also, a second inverse agonist, FG 7142, was tested. The result indicated that ethanol reinforcement was more sensitive to the inverse agonists compared to sucrose reinforcement. This was seen as a graded effect upon ethanol responding at doses which failed to have any effect upon sucrose-reinforced behavior. The Ro compound was approximately three times more potent than the FG compound in suppressing ethanol-reinforced responding. Possible explanations for the greater sensitivity of ethanol reinforcement compared to sucrose reinforcement was discussed in terms of ethanol's potential actions at the benzodiazepine-GABA receptor complex.     

Chronic administration of beta-carboline-3-carboxylic acid methylamide by continuous intraventricular infusion increases GABAergic function.

The repeated, intraperitoneal administration of the benzodiazepine receptor inverse agonist, FG 7142 (beta-carboline-3-carboxylic acid methylamide), leads to pharmacological kindling and an associated decrease in GABA-stimulated influx of 36Cl- into cortical membrane preparations. The chronic administration of benzodiazepine agonists results in the development of tolerance and also results in a decrease in GABA-stimulated uptake of 36Cl-. The present study was designed to evaluate further the paradoxical reports that both chronic treatment with benzodiazepine receptor agonists and inverse agonists results in a decreased ability of GABA to stimulate uptake of 36Cl- into cortical membrane preparations. The effects of continuous administration of FG 7142 on GABA-stimulated uptake of 36Cl-, the threshold for bicuculline-induced seizures and the proconvulsant actions of acute administration FG 7142 were evaluated. The continuous administration of FG 7142 resulted in an increased capacity of GABA to stimulate the uptake of 36Cl- into cortical membrane preparations and a significant increase in the seizure threshold for bicuculline following the acute administration of FG 7142. These data, therefore, indicate that changes in GABAergic function following chronic administration of GF 7142 are dependent on the regimen of administration of drug. The results also suggest that the GABA receptor homeostatically responds to continuous occupation by inverse agonists by an upregulation of its functional response to GABA.     

Continuous exposure to FG 7142: behavioural sensitisation is not accompanied by changes in benzodiazepine/GABA receptor coupling

Chronic intermittent high-dose treatment with N-methyl-β-carboline-3-carboxamide (FG 7142) leads to kindling accompanied by reduction in γ-aminobutyric acid (GABA) receptor function, whereas chronic continuous administration may result in behavioural effects in the opposite direction from those of acute FG 7142. In the present study, we have investigated the effects of continuous administration of low doses of FG 7142 on the response to an acute challenge dose of FG 7142 in an ethologically based model of anxiety. Rats treated continuously for 14 days with FG 7142 delivered by osmotic minipump at a rate of 1.2-1.5 mg/kg/day showed sensitisation to the anxiogenic effects of a challenge dose of FG 7142 (6 mg/kg), as measured in the elevated plus-maze. This was not accompanied by any change in benzodiazepine/GABA receptor coupling, as assessed by the 'GABA shift'. These results indicate that continuous low-dose treatment with FG 7142 can elicit sensitisation to the behavioural effects of FG 7142, but that this is unlikely to be mediated by changes in benzodiazepine/GABA receptor coupling.     

Methylxanthine discrimination in the rat: possible benzodiazepine and adenosine mechanisms

Rats were trained to discriminate either caffeine or theophylline from saline using a two-lever discrimination paradigm. Since methylxanthines have been found to interfere with agonist binding at both adenosine and benzodiazepine (BDZ) receptors, chlordiazepoxide (CDP) and L-PIA (an adenosine analog) were tested for generalization to and blockade of both xanthine cues. Neither L-PIA nor CDP generalized to either xanthine cue, although both produced dose-related decreases in response rate. CDP, but not L-PIA, produced dose-related decreases in drug-lever responses when combined with training doses of caffeine or theophylline. Response rates indicated a complex interaction between the xanthines and both L-PIA and CDP. When combined with the caffeine training dose, pentobarbital also produced a dose-dependent decrease in response rate but not in drug lever choices. Finally, papaverine generalized to the caffeine cue in a dose-dependent fashion. In a second experiment, rats trained to discriminate CDP from saline showed no generalization in L-PIA tests. CDP-appropriate responding was not significantly affected when the CDP training dose was combined with caffeine. These data indicate that: (a) methylxanthine interactions with L-PIA and CDP on response rate likely involve blockade of adenosine mechanisms; (b) the xanthine cue does not appear to depend on interactions with adenosine receptors; and (c) the xanthine cue may involve effects on cyclic AMP activity and/or interaction with the BDZ/GABA receptor complex.     

Effects of chlordiazepoxide training dose on the mixed agonist-antagonist properties of benzodiazepine receptor antagonist Ro 15-1788, in a drug discrimination procedure

In experiment 1, rats (n=12) were trained to discriminate the benzodiazepine (BDZ) compound chlordiazepoxide (CDP, 20 mg/kg, IP) from saline in a two-lever food-reinforced procedure, and subsequently were tested for stimulus control with different doses of CDP, Ro 15-1788 (a proposed BDZ receptor antagonist) and Ro 15-1788 plus 20 mg/kg CDP. Ro 15-1788 (0.63–40 mg/kg) dose-dependently antagonized CDP, and induced predominantly saline appropriate responding when administered alone. Thereafter, the same rats were retrained by progressively decreasing the training dose, to discriminate 2.5 mg/kg CDP from saline, and were tested again with the same compounds. Ro 15-1788 (0.16–40 mg/kg) now failed to antagonize CDP (2.5 mg/kg) and increased the percentage of drug-appropriate responding in a dose-related manner when administered alone. In experiment 2, separate groups of rats (n=10) were similarly trained to discriminate either 15 or 3 mg/kg CDP from saline. Tests with CDP, Ro 15-1788 and Ro 15-1788 plus CDP (either 15 or 3 mg/kg) yielded similar results to experiment 1, suggesting that the training dose effects on generalization and antagonism of Ro 15-1788 were not affected by the manner in which the lower CDP dose acquired drug stimulus control. It is concluded that mixed agonist-antagonist properties are apparent after-variations of the BDZ training dose in a drug discrimination procedure.     

The effects of the beta-carboline FG 7142, on intracranial self-stimulation in the rat

The effects of FG 7142 were examined, alone and in combination with chlordiazepoxide, on self-stimulation of the mid-lateral hypothalamus. Rewarding stimuli were delivered according to a 10-sec variable-interval schedule of reinforcement. FG 7142 (1-20 mg/kg) produced a dose-related depression in responding, and chlordiazepoxide (5 mg/kg) enhanced it. When these two drugs were given together, response rates did not differ significantly from control rates.     

The effects of FG 7142 and RO 15-1788 on the release of punished responding produced by chlordiazepoxide and ethanol in the rat

Previous results in our laboratory have shown that both chlordiazepoxide and ethanol will release punished responding in a rat operant conflict test using incremental shock. In the present study, a benzodiazepine antagonist and a benzodiazepine inverse agonist were used to explore the neurochemical basis for this behavioral action. N-methyl--carboline-3-carboxamide (FG 7142) at high doses (20 and 40 mg/kg) produced suppression of both punished and unpunished responding, and reversed the release of punished responding produced by both chlordiazepoxide and ethanol, but only at doses that produced an effect on its own. FG 7142 thus acted to oppose the actions of both ethanol and benzodiazepines but in an additive, not interactive, manner. In contrast, RO 15-1788 produced no changes when injected by itself in doses as high as 12 mg/kg and reversed chlordiazepoxide-induced but not ethanol-induced release of punished responding. RO 15-1788 also reversed the decrease in punished responding produced by FG 7142. Results suggest that ethanol does not interact directly with the benzodiazepine binding sites on the GABA/benzodiazepine ionophore complex to produce its anxiolytic action.     

Electroencephalographic investigations in rabbits of drugs acting at GABA-benzodiazepine-barbiturate/picrotoxin receptors complex

This paper describes the EEG profiles, observed in rabbits, of drugs which affect GABA synaptic activity at GBB complex. Drugs which enhance GABA synaptic activity induce sedation associated with EEG synchronization. However, muscimol, THIP, GHB and baclofen induce signs of CNS stimulation (light tremors of the forelimbs, chewing, light nystagmus and hyperpnea) associated with EEG spikes. Signs of light stimulation (chewing and jerks of the head) also occur after BDZs and barbiturates, and are associated with the presence of 12-24 and 20-25 Hz waves, respectively. Drugs which reduce GABA synaptic activity (bicuculline, inverse BDZ agonists, PTZ, picrotoxin and Ro 5-3663) induce three dose-dependent stages of EEG changes: trains of slow waves, trains of spike-and-wave complexes and paroxysmal activity in the rostral encephalic structures without apparent changes of the electrical activity in the spinal cord. The first two stages are associated with a behavioral state of alert and the third stage with tonico-clonic convulsions. Among the inverse BDZ agonists, DMCM and beta-CCM elicit all three stages, whereas FG 7142 and beta-CCE induce only the first two and CGS 8216 only the first. The BDZ antagonists Ro 15-1788 and Ro 15-3505 (0.2-30 mg/kg IV) do not significantly affect the EEG pattern. However, they selectively inhibit the effects of diazepam and of the inverse BDZ agonists. In both cases, the inhibition is observed with doses as low as 0.2 mg/kg IV and leads to an EEG desynchronization. The possible involvement of the modifications of GABA synaptic activity in the etiology of both petit mal and grand mal epilepsies is discussed.