Assessment of the anxiolytic and amnesic effects of three benzodiazepines, diazepam, alprazolam and triazolam, by conflict and non-matching to sample tests in mice.

The anti-conflict and amnesic effects of three benzodiazepines (diazepam, alprazolam and triazolam) po-administered in mice were assessed by a modified Geller Seifter conflict test and a non-matching to sample test using a 4-arm maze with 3 selectable arms, respectively. Diazepam (10 mg/kg), alprazolam (1-10 mg/kg) and triazolam (1 and 3 mg/kg) significantly increased the lever-press in the alarm period (punished responding) under the conflict test. Under the non-matching to sample test, the correct response (CR), but not non-reward response (NR), decreased depending on the delay time. Diazepam (3 mg/kg) significantly decreased the CR under 0-sec delay condition, but not that under 30-sec delay condition. Comparatively higher doses of the benzodiazepines, diazepam (10 mg/kg), alprazolam (1 and 3 mg/kg) and triazolam (0.3 and 1 mg/kg), significantly decreased the CR under both 0- and 30-sec delay conditions. However, no significant change in the NR was produced by any dose of the three benzodiazepines. These results suggest that the decrease in CR during the non-matching to sample test is caused by the impairment of cognitive function rather than working and reference memories, and that such effects appear at doses equal to and/or less than the effective doses for anti-conflict effect. However, the assumption that triazolam has the highest risk of induction of amnesia, as compared to the advantage of sleep induction or relief of anxiety in the benzodiazepine group, was not supported by the present results.