Discriminative-stimulus effects of midazolam in squirrel monkeys: comparison with other drugs and antagonism by Ro 15-1788.

Squirrel monkeys were trained to respond on one of two levers depending on whether midazolam (0.3 mg/kg) or saline had been injected. After i.v. injections of midazolam 10 consecutive responses on one lever either produced food or terminated a stimulus associated with electric shock, whereas after i.v. injections of saline 10 consecutive responses on the other lever either produced food or terminated the stimulus. The discriminative-stimulus effects of drugs were determined by administering cumulative doses i.v. during timeout periods that preceded sequential components of the experimental session. The benzodiazepines midazolam, chlordiazepoxide, diazepam and N-desmethyldiazepam, the cyclopyrrolone zopiclone and the triazolopyridazine CL 218,872 had qualitatively similar stimulus effects regardless of the type of consequence (food presentation or stimulus-shock termination) that maintained responding. Administration of each of these drugs resulted in greater than 90% of responses on the midazolam-associated lever at cumulative doses that did not severely suppress the overall rate of responding. The order of potency was: midazolam = diazepam greater than or equal to N-desmethyldiazepam greater than or equal to zopiclone greater than CL 218,872 greater than or equal to chlordiazepoxide. Administration of the 5-hydroxytryptamine antagonists cyproheptadine and cinanserin also resulted in greater than 90% of responses on the midazolam-associated lever in about half the subjects, although these effects were observed only with cumulative doses that markedly reduced the overall rate of responding. Administration of pentobarbital, barbital, clozapine, muscimol, buspirone, diphenhydramine, tripelennamine, caffeine and Ro 15-1788 did not result in substantial responding on the midazolam-associated lever at doses up to those that reduced or eliminated responding.(ABSTRACT TRUNCATED AT 250 WORDS)