Correlations among minimal neurotoxicity, anticonvulsant activity, and displacing potencies in [3H]flunitrazepam binding of benzodiazepines.

Five clinically available 1,4-benzodiazepines (BDZs) (chlordiazepoxide, diazepam, oxazepam, nitrazepam, and clonazepam) and four investigational 1,4-BDZs (BDZ I, BDZ II, BDZ III, and BDZ IV) were tested in vivo for minimal neurotoxicity (TD50) and for ability to obtund seizures (ED50) induced by a battery of five well-standardized procedures (maximal electroshock, strychnine, pentylenetetrazol, bicuculline, and picrotoxin). In addition, these BDZs were also tested in vitro as inhibitors of [3H]flunitrazepam binding to BDZ receptors. The results with each of the six in vivo tests were compared with those with the in vitro receptor binding test, and the correlation coefficients (r) were calculated. There was a high correlation between the inhibitor constants (Ki) derived from BDZ binding studies and the TD50 values (r = 0.882) and the pentylenetetrazol ED50 values (r = 0.946). There was also good correlation between ED50 values of BDZs effective by the bicuculline and picrotoxin tests and their Ki values in the BDZ receptor binding studies (r = 0.868 and 0.892, respectively). However, BDZ I, BDZ II, and BDZ IV had Ki values of 1.830, 0.075, and 0.015 microM, respectively, but BDZ I was ineffective in nontoxic doses by the bicuculline and picrotoxin tests, BDZ II by the picrotoxin test, and BDZ IV by the bicuculline test. In contrast, there was no correlation between the BDZs' anticonvulsant potency, determined by either the maximal electroshock or strychnine test, and their inhibitory potency on [3H]flunitrazepam binding to receptor sites. The possible methodological and neurochemical bases for these differences are discussed. These studies indicate that BDZ binding studies can be used to screen BDZs for anticonvulsant activity (selectively for antipentylenetetrazol activity and less selectively for antibicuculline and antipicrotoxin activity). More importantly, they complement the conventional in vivo anticonvulsant threshold tests. Therefore, BDZ binding studies can be used in concert with conventional in vivo procedures for the pharmacological differentiation of candidate antiepileptic BDZs.