Effect of GABA agonists on the neurotoxicity and anticonvulsant activity of benzodiazepines

Synergistic combinations of anticonvulsant agents: what is the evidence from animal experiments?

PURPOSE

Combination therapy is often used in the treatment of seizures refractory to monotherapy. At the same time, the pharmacodynamic mechanisms that determine the combined efficacy of antiepileptic drugs (AEDs) are unknown, and this prevents a rational use of these drug combinations. We critically evaluate the existing evidence for pharmacodynamic synergism between AEDs from preclinical studies in animal models of epilepsy to identify useful combinations of mechanisms and to determine whether study outcome depends on the various research methods that are in use.

METHODS

Published articles were included if the studies were placebo-controlled, in vivo, or ex vivo animal studies investigating marketed or experimental AEDs. The animal models that were used in these studies, the primary molecular targets of the tested drugs, and the methods of interpretation were recorded. The potential association of these factors with the study outcome (synergism: yes or no) was assessed through logistic regression analysis.

RESULTS

In total, 107 studies were identified, in which 536 interaction experiments were conducted. In 54% of these experiments, the possibility of a pharmacokinetic interaction was not investigated. The majority of studies were conducted in the maximal electroshock model, and other established models were the pentylenetetrazole model, amygdala kindling, and the DBA/2 model. By far the most widely used method for interpretation of the results was evaluation of the effect of a threshold dose of one agent on the median effective dose (ED50) of another agent. Experiments relying on this method found synergism significantly more often compared with experiments relying on other methods (p<0.001). Furthermore, experiments including antagonists of the AMPA receptor were more likely to find synergism in comparison with all other experiments (p<0.001).

CONCLUSIONS

Intensive preclinical research into the effects of AED combinations has not led to an understanding of the pharmacodynamic properties of AED combinations. Specifically, the majority of the preclinical studies are not adequately designed to distinguish between additive, synergistic, and antagonistic interactions. Quantitative pharmacokinetic-pharmacodynamic studies of selectively acting AEDs in a battery of animal models are necessary for the development of truly synergistic drug combinations.

Attenuation of anticonvulsant effects of diazepam after chronic treatment with bicuculline

Changes in the GABAergic system after chronic treatment with bicuculline were examined in two strains of inbred rats, Fischer 344 (F344) and Lewis (LEW). Rats received an IP injection of either bicuculline (2 mg/kg) or vehicle once a day for 12 days. After this chronic treatment, the effects of diazepam (1 mg/kg, IP) and pentobarbital (20 mg/kg, IP) on bicuculline-induced convulsions were measured. Bicuculline was acutely infused into a tail vein at 0.0415 mg/min, and the infusion was terminated when rats showed seizure. Following the chronic bicuculline treatment, the anticonvulsant effect of diazepam, but not of pentobarbital, was significantly reduced as compared to its effect following chronic vehicle treatment in both strains. Both diazepam and pentobarbital showed a significant difference in anticonvulsant effects between strains (F344 > LEW). The hypnotic effects of muscimol, barbital, pentobarbital, and ethanol following chronic bicuculline treatment were examined. There was no significant difference in sleep time induced by these drugs between bicuculline- and vehicle-treated rats. These results suggest that the attenuation of diazepam's anticonvulsant effect after chronic bicuculline treatment may result from functional changes in benzodiazepine receptors and that the anticonvulsant effects of diazepam and pentobarbital may be influenced by genetic factors. Moreover, the hypnotic effects of several drugs tested are apparently not affected by chronic bicuculline treatment.

Do benzodiazepines induce sleep by a GABAergic mechanism?

In order to further characterize the possible role of GABA function in the sleep-inducing properties of benzodiazepines (BZs), we have administered the GABA agonist muscimol (0.05 and 0.1 mg/kg) and the GABA antagonist bicuculline (1.25 and 2.5 mg/kg) IP, alone and in combination with triazolam (0.8 mg/kg). There was no evidence of interaction of these compounds with triazolam vis a vis sleep. These data are consistent with an earlier report indicating a lack of interaction of muscimol with flurazepam, and suggest that non-GABAergic mechanisms may be involved in the hypnotic properties of benzodiazepines.

The role of technical, biological and pharmacological factors in the laboratory evaluation of anticonvulsant drugs. IV. Protective indices

Calculation of protective or therapeutic indices is widely used in primary and secondary screening for drugs with selective anticonvulsant activity. The protective index is the median minimal 'neurotoxic' dose, TD50, divided by median effective dose, ED50. TD50s are usually determined by tests, such as the rotarod test or the chimney test, for quantification of 'minimal neurological deficit', such as motor impairment, while median effective doses are commonly determined in the maximal electroshock seizure (MES) test or the s.c. pentylenetetrazol (PTZ) seizure test in mice or rats. For antiepileptic drug development, it has been proposed previously that only compounds with an estimated protective index of at least 5 should proceed to further evaluation. However, various technical, biological and pharmacological factors can influence anticonvulsant or 'neurotoxic' potencies and thereby protective indices. In order to reevaluate the value of protective indexes in the prediction of drugs with selective anticonvulsant action, protective indices were determined for various clinically used antiepileptic drugs in standardized seizure tests, i.e. MES and s.c. PTZ tests in mice and rats, as well as in seizure threshold tests. For most drugs, similar TD50s were determined in the rotarod and chimney test. When protective indices were calculated for the different seizure models, only few drugs reached an index of 5 (some not even reaching an index of 2) in the traditional MES or s.c. PTZ tests in mice and rats. In contrast, using anticonvulsant doses determined by seizure threshold tests, the 5 primary drugs against generalized tonic-clonic seizures, i.e., carbamazepine, phenytoin, phenobarbital, primidone and valproate, had indices of more than 5 in the MES threshold model, while drugs with efficacy against absence and myoclonic seizures, i.e., valproate, ethosuximide and the benzodiazepines, had protective indices of at least 5 in the i.v. PTZ seizure threshold model. The data substantiate that valuable information can be obtained by estimation of protective indices. However, in order to minimize the possibility that an interesting new anticonvulsant compound is overlooked during primary or secondary screening, a protective index of 2 should be considered sufficient in case of traditional MES or s.c. PTZ models with fixed seizure stimulus. Alternatively, seizure threshold models could be used for calculation of protective indices in order to avoid underestimation of anticonvulsant selectivity of test compounds.

The role of technical, biological and pharmacological factors in the laboratory evaluation of anticonvulsant drugs. III. Pentylenetetrazole seizure models

Although seizure models using systemic administration of the chemoconvulsant pentylenetetrazol (PTZ) for induction of generalized clonic seizures in rodents are widely employed to identify potential anticonvulsants, the important role of diverse technical, biological and pharmacological factors in interpretation of results obtained with these models is often not recognized. The aim of this study was to delineate factors other than sex, age, diet, climate, and circadian rhythms, which are generally known. For this purpose, experiments with 8 clinically established antiepileptic drugs were undertaken in the following PTZ models: (1) the threshold for different types of PTZ seizures, i.e., initial myoclonic twitch, generalized clonus with loss of righting reflexes, and tonic backward extension of forelimbs (forelimb tonus), in mice; (2) the traditional PTZ seizure test with s.c. injection of the CD97 for generalized clonic seizures in mice; and (3) the s.c. PTZ seizure test in rats. In rats, in addition to evaluating drug effects on generalized clonic seizures, a ranking system was used to determine drug effects on other seizure types. When drugs were dissolved in vehicles which themselves did not exert effects on seizure susceptibility, the most important factors which influenced drug potencies were: (1) bishaped dose-response curves, i.e., a decline in anticonvulsant dose-response at high doses of some drugs, leading to misinterpretations of drug efficacy if only a single high drug dosage is tested; (2) effects of route of PTZ administration (i.v. infusion vs. s.c. injection) on estimation of anticonvulsant potency; (3) species differences in drug metabolism; (4) differences in drug potencies calculated on the basis of administered doses compared to potency calculations based on 'active' drug concentrations in plasma; (5) qualitative and quantitative species differences in drug actions; (6) endpoints used for PTZ tests; (7) misleading predictions from PTZ seizure models. Analysis of anticonvulsant drug actions indicated that myoclonic or clonic seizures induced by i.v. or s.c. PTZ might be suitable for predicting efficacy against myoclonic petit mal seizures in humans, but certainly not to predict efficacy against absence seizures. Tonic seizures induced by PTZ were blocked by drugs, such as ethosuximide, which exert no effect on tonic seizures in humans. In order to reduce the variability among estimates of anticonvulsant activity in PTZ seizure models, the various factors delineated in this study should be rigidly controlled in experimental situations involving assay of anticonvulsant agents.

The role of technical, biological and pharmacological factors in the laboratory evaluation of anticonvulsant drugs. II. Maximal electroshock seizure models

Although seizure models using electrical stimulation for the induction of generalized tonic-clonic seizures in rodents are widely employed to identify potential anticonvulsants, the important role of various technical, biological and pharmacological factors in the interpretation of results obtained with these models is often not recognized. The aim of this study was to delineate factors other than sex, age, diet, climate and circadian rhythms, which are generally known. For this purpose, experiments with 8 clinically established antiepileptic drugs were undertaken in the following electroshock seizure models: (1) the maximal (tonic extensor) electroshock seizure threshold (MEST) in mice; (2) the traditional maximal electroshock seizure (MES) test with supra-threshold stimulation in mice; and (3) the MES test with suprathreshold stimulation in rats. When drugs were dissolved in vehicles which did not themselves exert effects on seizure susceptibility, the most important factors which influenced drug potencies were (1) marked differences between drugs and species in terms of peak drug effect, duration of action and the formation of active metabolites; (2) differences in drug potencies calculated on the basis of administered doses compared to potency calculations based on active drug concentrations; (3) the equipment used for seizure induction; (4) marked effects of current strength on results obtained in electroshock seizure models; (5) site of application of the electrical stimulus (transcorneal vs. transauricular). In order to reduce the variability among estimates of anticonvulsant activity, the various factors delineated in this study should be rigidly controlled in experimental situations involving assay of anticonvulsant agents.

Glycine potentiation of anticonvulsant drugs in pentylenetetrazol seizures in rats

This study evaluates the glycine potentiation of anticonvulsant drugs in subcutaneous pentylenetetrazol seizures in rats. Administered alone, glycine (30 or 40 mM/kg, PO) induced no anticonvulsant effect or neurological deficit. Coadministered with anticonvulsants, glycine significantly enhanced the anticonvulsant potency of diazepam and sodium valproate without affecting the neurological deficit induced by the anticonvulsants. Glycine did not significantly alter the anticonvulsant activity of ethosuximide or phenobarbital. These findings indicate a possible glycine-sensitive component in the mechanism of action of diazepam and sodium divalproate in subcutaneous pentylenetetrazol seizures. With the possible exception of sodium valproate, the present study provides little support for a glycine and gamma-aminobutyric acid (GABA) interaction as a mechanism of anticonvulsant activity in SC PTZ seizures. Further studies are required to determine the role of strychnine-sensitive and strychnine-insensitive glycine receptors in this experimental model of absence epilepsy.

The role of technical, biological and pharmacological factors in the laboratory evaluation of anticonvulsant drugs. I. The influence of administration vehicles

Although animal models, such as electroshock seizures, pentylenetetrazol (PTZ)-induced seizures and the rotorod test, are widely employed in the search for and evaluation of new anticonvulsant drugs, the important role of diverse technical, biological and pharmacological factors in the interpretation of results obtained with these models is often not recognized. In order to delineate factors other than strain, sex, age, diet, climate, and circadian rhythms, which are generally known, a series of studies was undertaken. In the experiments described here, the influence of administration vehicles and drug formulations on bioavailability, potency and time course of anticonvulsant drugs was studied in mice. Two standard anticonvulsant drugs, primidone and carbamazepine, with poor aqueous solubility were used for these experiments, because water insolubility is a common problem in the laboratory evaluation of anticonvulsant agents. Since vehicles, especially organic solvents or detergents, may exert effects of their own, sensitive electroshock and PTZ seizure threshold tests were used for the assessment of vehicle-related actions. Of various aqueous or lipophilic vehicles tested, only glycofurol increased seizure thresholds, when concentrations exceeding 10% were administered. However, even at a concentration of 30%, the solubilizer did not exert measurable effects in the maximal electroshock seizure (MES) test in mice, but markedly potentiated the effect of primidone. In contrast, polyethylene glycol 400 (PEG 400) up to a concentration of 30% did not affect electrical or chemical seizure thresholds nor did it alter the pharmacological potency of primidone. When primidone or carbamazepine were administered as a suspension in a Tween/water vehicle, their anticonvulsant effects were considerably lower compared to injections of the same doses as a solution.(ABSTRACT TRUNCATED AT 250 WORDS)

Potentiation by glycine of anticonvulsant drugs in maximal electroshock seizures in rats

This study evaluated the potentiation by glycine of anticonvulsant drugs in maximal electroshock seizures in rats. Administered alone, glycine (40 mmol/kg, p.o.) induced no anticonvulsant effect or neurotoxicity. Administered together with the anticonvulsants, glycine significantly enhanced the anticonvulsant potency of phenobarbital and carbamazepine. Glycine also potentiated the anticonvulsant actions of MK-801 and diazepam but did not improve the selectivity of the drugs, as effective doses were still associated with neurotoxicity. Glycine did not potentiate phenytoin or sodium divalproate. Administration together with glycine had no significant effect on the concentrations of phenobarbital or carbamazepine in the brain. Administration together with phenobarbital had no relevant effect on the concentration of glycine in the brain but administration of glycine and carbamazepine together resulted in an increased concentration of glycine in the hippocampus and brainstem. These findings indicate a possible glycine-sensitive component in the mechanism of action of phenobarbital, carbamazepine and diazepam in maximal electroshock seizures. Although the mechanism may not be mediated by a glycine-GABA interaction, the evidence does implicate a possible interaction between glycine and anticonvulsant drugs at NMDA receptors.

Effects of muscimol and flurazepam on the sleep EEG in the rat

In order to assess the possible role of GABA receptor function in the hypnotic property of benzodiazepines, we have examined the sleep EEG in rats given the GABA agonist muscimol, alone and in combination with flurazepam. Muscimol 0.05 and 0.1 mg/kg IP failed to alter sleep latency or total sleep time, and did not interact with the sleep-enhancing properties of flurazepam 20 mg/kg IP. These observations, in conjunction with a previous study of bicuculline, suggest that the hypnotic property of benzodiazepines may not be mediated by alteration of GABAergic activity.

Injection of benzodiazepines but not GABA or muscimol into pars reticulata substantia nigra suppresses pentylenetetrazol seizures

The substantia nigra pars reticulata (SNpr), a brain area rich in GABA and benzodiazepine receptors, is thought to be involved in the regulation of seizure activity. It has been shown to be a site of anticonvulsant action of substances that affect GABA transmission. The anti-pentylenetetrazol (PTZ) activities of intranigral of muscimol, a GABAA receptor agonist; two benzodiazepines, midazolam and flurazepam; and GABA were examined. Microinjection of a wide dose range of both GABA and muscimol into the SNpr failed to show anti-PTZ seizure activity. Intranigral injections of midazolam and flurazepam showed clear, dose-dependent anti-PTZ effects. Ro15-1788, a benzodiazepine receptor antagonist, reversed the anticonvulsant effects of midazolam when both were infused intranigrally. Intranigral infusion of muscimol or flurazepam protected rats from bicuculline-induced tonic seizures. The results suggest that the anti-PTZ effects of benzodiazepines in SNpr might not be mediated through GABAA receptors. Another possibility is that nigral neurons bearing GABAA receptors functionally linked to benzodiazepine sites may not be representative of the whole population of nigral neurons inhibited by GABA agonists. This could result in different patterns of inhibition of nigral efferent activity by GABAA agonists and benzodiazepines.

Binding pattern of [35S]t-butylbicyclophosphorothionate is not altered following electroconvulsive shock treatment in rats

Single or repeated electroconvulsive shock (ECS) treatment-induced changes in [35S]t-butylbicyclophosphorothionate [( 35S]TBPS) binding patterns in specific regions, i.e., cerebral cortex, cerebellum, hippocampus, and striatum of rat brain were investigated. Specific [35S]TBPS binding in these brain regions was not altered following a single or repeated administration of ECS, nor was the inhibition of [35S]TBPS binding to GABA affected. These observations tend to suggest that the picrotoxin-site on the GABA receptor complex may not be directly involved in electroconvulsive shock.

A comparison of the anticonvulsant effects of 1,4- and 1,5-benzodiazepines in the amygdala-kindled rat and their effects on motor function

Studies suggest that the 1,5-benzodiazepine clobazam possesses a favorable anticonvulsant profile due to its minimal neurotoxicity. The anticonvulsant and motor impairment effects of clobazam and 2 1,4-benzodiazepine, diazepam and clonazepam, were compared by dose-response analysis in amygdala-kindled rats and on 3 tests of motor function: gross motor impairment, a vertical screen test, and muscle tone. All drugs produced a significant, dose-dependent decrease in the duration of both behavioral and electrographic kindled seizure measures. Forelimb clonus suppression was the most sensitive measure of anticonvulsant drug effect. The order of potency for all effects was clonazepam greater than diazepam greater than clobazam. ED50s for the benzodiazepines' effects on motor impairment were compared to their ability to protect rats from forelimb clonus. Different spectrums of action for the various benzodiazepines were found depending on the comparison measure. Clonazepam had the most favorable ratio of potency for anticonvulsant vs. motor impairment activity when ataxia rating was the comparison measure. Diazepam had the most advantageous profile when the more sensitive screen test was used for comparison. Clobazam was not found to have a superior spectrum of action when compared across these measures. The results emphasize the importance of dose-response analyses and the consideration of behavioral measures used to assess beneficial and adverse effects of anticonvulsants.

Electrophysiology of benzodiazepine receptor ligands: multiple mechanisms and sites of action

Electrophysiology of BZR ligands has been reviewed from different points of view. A great effort was made to critically discuss the arguments for and against the temporarily leading hypothesis of the mechanism of action of BZR ligands, the GABA hypothesis. As has been discussed at length in the present article, an impressive body of electrophysiological and biochemical evidence suggests an enhancement of GABAergic inhibition in CNS as a mechanism of action of BZR agonists. Biochemical data even indicate a physical coupling between GABA recognition sites and BZR which, together with the effector site build-up by Cl- channels, form a supramolecular GABAA/BZR complex. By binding to a specific site on this complex, BZR agonists allosterically increase and BZR inverse agonists decrease the gating of GABA-linked Cl- channels, whereas BZR antagonists bind to the same site without an appreciable intrinsic activity and block the binding and action of both agonists as well as inverse agonists. While this model is supported by many electrophysiological experiments performed with BZR ligands in higher nanomolar and lower micromolar concentrations, it does not explain much controversial data from animal behavior and, more importantly, is not in line with electrophysiological effects obtained with low nanomolar BZ concentrations. The latter actions of BZR ligands in brain slices occur within a concentration range compatible with concentrations of BZ observed in CSF fluid, which would be expected to be found in the biophase (receptor level) during anxiolytic therapy in man. Enhanced K+ conductance seems to be a suitable candidate for this effect of BZR ligands. This direct action on neuronal membrane properties may underlie the many electrophysiological observations with extremely low systemic doses of BZR ligands in vivo which demonstrated a depressant effect on spontaneous neuronal firing in various CNS regions. Skeletomuscular spasticity and epilepsy are two neurological disorders, where both the enhanced GABAergic inhibition and increased K+ conductance may contribute to the therapeutic effect of BZR agonists, since electrophysiological and behavioral studies strongly support GABA-dependent as well as GABA-independent action of BZR ligands elicited by low to intermediate doses of BZ necessary to evoke anticonvulsant and muscle relaxant effects. Somewhat higher doses of BZR ligands, inducing sedation and sleep, lead perhaps to the only pharmacologically relevant CNS concentrations (ca. 1 microM) which might be due entirely to increased GABAergic inhibition.(ABSTRACT TRUNCATED AT 400 WORDS)

Temperature dependence and GABA modulation of [3H]triazolam binding in the rat brain

The hypnotic triazolam (TZ), a triazolobenzodiazepine displays a short physiological half life and has been used for the treatment of insomnia related to anxiety states. Our major objectives were the direct measurement of the temperature dependence and the gamma-aminobutyric acid (GABA) effect of [3H]TZ binding in the rat brain. Saturation studies showed a shift to lower affinity with increasing temperatures (Kd = 0.27 +/- 08 nM at 0 degree C; Kd = 1.96 +/- 0.85 nM at 37 degrees C) while the Bmax values remained unchanged (1220 +/- 176 fmoles/mg protein at 0 degree C and 1160 +/- 383 fmoles/mg protein at 37 degrees C). Saturation studies of [3H]TZ binding in the presence or absence of GABA (100 microM) showed a GABA-shift. At 0 degrees C the Kd values were (Kd = 0.24 +/- 0.03 nM/-GABA; Kd = 0.16 +/- 0.04/+GABA) and at 37 degrees C the Kd values were (Kd = 1.84 +/- 0.44 nM/-GABA; Kd = 0.95 +/- 0.29 nM/+GABA). In contrast to reported literature, our findings show that TZ interacts with benzodiazepine receptors with a temperature dependence and GABA-shift consistent with predicted behavior for benzodiazepine agonists.

Glycine potentiates diazepam anticonvulsant activity in electroshock seizures of rats: possible sites of interaction in the brainstem

The effect of orally or intracerebrally injected glycine on the anticonvulsant actions of intraperitoneal diazepam was examined using a tonic-clonic electroshock seizure model in the rat. Orally administered glycine (1.125 g/kg) potentiated the anticonvulsant effect of diazepam (DZP) to convert tonic-clonic electroshock seizures to less severe subthreshold clonic seizures. Oral glycine by itself had no effect on the tonic-clonic seizure response. Bilateral substantia nigra (SN) microinjections of glycine (125 micrograms/site) failed to potentiate intraperitoneal DZP when compared with the most appropriate control, animals treated with DZP and intranigral saline. It was not possible to determine whether bilateral glycine microinjections into the inferior olivary nucleus (IO) potentiated anticonvulsant effects of DZP since glycine alone converted all tonic-clonic seizures to the clonic response. Finally, bilateral glycine microinjection alone into the nucleus reticularis pontis oralis (PNO) produced an anticonvulsant effect when compared to untreated control responses but did not potentiate the anticonvulsant actions of DZP. Although these results may indicate that the glycinergic potentiation of DZP involves a direct pharmacodynamic interaction between these two compounds at specific brain sites, these sites have not yet been demonstrated conclusively.

Glycine potentiates the anticonvulsant action of diazepam and phenobarbital in kindled amygdaloid seizures of rats

The effect of glycine on the anticonvulsant activity of diazepam and phenobarbital in fully developed kindled amygdaloid seizures in rats was determined. Glycine alone had no significant effect on the seizure response, either when administered orally 1 hr prior to the seizure test or when given chronically in a 0.5 M solution as the source of water. Administration of glycine (10 mmol/kg, oral) together with diazepam produced a significant reduction in both cortical epileptiform afterdischarge and the severity of seizures at doses of diazepam which had no significant effect on the seizures when administered alone. Glycine potentiated the effects of phenobarbital on the cortical afterdischarge but not the severity of the seizures. The observed potentiation of the effects of diazepam and phenobarbital suggests a glycinergic mechanism in the anticonvulsant action of these drugs which may be mediated in part by the inhibitory gamma-aminobutyric acid (GABA) systems.

Correlation of the hypnotic potency of benzodiazepines with inhibition of voltage-dependent calcium uptake into mouse brain synaptosomes

Nine benzodiazepines were tested for their ability to inhibit 45Ca2+ uptake into mouse whole brain synaptosomes and for hypnotic activity as indicated by their ability to produce loss of the righting reflex. Eight of the benzodiazepines significantly inhibited fast-phase voltage-dependent 45Ca2+ uptake and five exhibited hypnotic activity. There was a direct correlation between the hypnotic potency of these five benzodiazepines and their ability to inhibit 45Ca2+ uptake. There does not appear to be a correlation between the anticonvulsant potency of the benzodiazepines and their potency for inhibiting 45Ca2+ uptake. These results support previous findings with other sedative/hypnotic drugs and suggest that inhibition of presynaptic calcium uptake may be linked with hypnotic but not anticonvulsant actions of benzodiazepines.

gamma-Aminobutyric acid modulation of benzodiazepine receptor binding in vitro does not predict the pharmacologic activity of all benzodiazepine receptor ligands

gamma-Aminobutyric acid (GABA) modulation of triazolam and nicotinamide binding to benzodiazepine (BDZ) receptors in vitro was compared with the neurotoxicity and anticonvulsant activity of these two drugs in vivo. GABA had no significant effect on the inhibitory potency of triazolam in [3H]flunitrazepam receptor binding, whereas GABA decreased the inhibitory potency of nicotinamide. When administered to mice, both triazolam and nicotinamide exhibited neurotoxicity by the rotorod test and anticonvulsant activity by the pentylenetetrazol seizure threshold test. This suggests that GABA modulation of the receptor binding of a BDZ ligand in vitro is not a reliable predictor of the pharmacologic activity of the ligand.