Behavioral and biochemical characterization of benzodiazepine receptor partial agonists in pigeons.

The ability of benzodiazepine receptor partial agonists to exhibit full efficacy in preclinical anxiolytic tests, in conjunction with initial clinical results, has suggested the possibility of a reduced clinical side-effect profile compared to benzodiazepine receptor full agonists like diazepam. Because punished behavior of pigeons has been useful in detecting effects of novel anxiolytic drugs, effects of imidazobenzodiazepine and beta-carboline benzodiazepine receptor partial agonists and some related compounds were evaluated in this species. The abilities of these compounds to substitute for the discriminative stimulus effects of the full agonists midazolam also was determined. Intrinsic efficacy was assessed by the degree to which gamma-aminobutyric acid increased ligand potency to displace [(3)H]Ro15-1788 (flumazinil) from membranes of pigeon cerebrum, and ranged from full agonist-like efficacy (Ro 19-5470; 7-(3-cyclopropyl-1,2,4-oxodiazol-5-yl)-5,6-dihydro-5-methyl-4H- imidazo[1,5a]-thieno[3,2-f]diazin-4-one) to minimal gamma-aminobutyric acid potentiations close to that of the antagonist flumazenil (abecarnil and Ro 41-7812; 7-chloro-4,5-dihydro-3-(3-hydroxy-1-propynyl)-5-methyl-6H-imidazo[1,5-a] -[1,4 ]benzodiazepine-6-one). Punished responding was increased markedly by midazolam and by all partial agonists, except Ro 41-7812 and Ro 42-8773 (7-chloro-3-[3-(cyclopropylmethoxy)-1-propynyl]-4,5-dihyro-5 -methyl-6H-imidaz o[1,5-a][1,4]benzodiazepine-6-one), at doses that did not affect nonpunished responding. In contrast to the full substitution generally observed in mammals, all of the partial agonists produced incomplete substitution (40-70%) in the midazolam drug discrimination procedure in pigeons. A positive relationship was observed between the degree of substitution and intrinsic efficacy. The benzodiazepine antagonists, flumazenil and ZK 93,426 (ethyl-5-isopropoxy-4-methoxymethyl-beta-carboline-3-carboxylate), neither increased punished responding nor substituted for midazolam. The results of the present study suggest that benzodiazepine receptor partial agonists and related compounds may provide full anxiolytic activity at doses that do not fully reproduce the subjective effect profile of full agonists.