In vivo bidirectional modulatory effect of benzodiazepine receptor ligands on GABAergic transmission evaluated by positron emission tomography in non-human primates

The central type benzodiazepine receptor (BDZr), an allosteric modulatory site of the GABAA receptor-anion channel, has been shown in vitro to respond to drugs with positive efficacy (agonists), zero efficacy (competitive antagonists) and drugs with negative efficacy (inverse agonists). However, this general concept of the function of BDZr drugs has rarely been assessed in intact living brain. We report here in on a non-invasive in vivo assessment of the intrinsic efficacies of BDZr drugs in the brain of non-human primates. We have performed an in vivo simultaneous determination of fractional BDZr occupancy and the resulting pharmacological efficacies of the full agonist diazepam, the partial agonist bretazenil, the antagonist flumazenil (Ro15-1788), the partial inverse agonist Ro15-4513 and the full inverse agonist methyl beta-carboline-3-carboxylate (beta-CCM). Positron emission tomography (PET) was used to estimate fractional BDZr occupancy measured as the in vivo displacement in the brain of the positron emitter radioligand, [11C]flumazenil. Simultaneously, the proconvulsant or anticonvulsant efficacies of the BDZr drugs were measured as their abilities to facilitate or counteract the central effects of an infusion of pentylenetetrazol, a non-competitive GABA antagonist acting on the picrotoxin site of the receptor complex. This was measured using electroencephalographic recording (EEG). Our results show that, in vivo, the fractional receptor occupancy by a given drug is perfectly correlated with its resulting graded pharmacological effects, as predicted from the competitive drug receptor interaction theory. Furthermore, the slope of the relationship between fractional receptor occupancies and the resulting pharmacological effects (an index of intrinsic efficacy) strictly depends on the BDZr ligand considered. Diazepam displayed a strong positive intrinsic efficacy, and, in contrast, beta-CCM a marked negative one. Between these two extremes, the partially active drugs bretazenil and Ro15-4513, which required a large fractional receptor occupancy to produce significant anti- or proconvulsant effects, respectively, displayed only a weak intrinsic efficacy. Flumazenil did not produce any significant pharmacological effect. We observed that the in vivo intrinsic efficacies of diazepam, flumazenil and beta-CCM correlate with their intrinsic efficacies as measured by their modulatory effects on the GABA-dependent membrane chloride conductance in vitro. Thus, the intrinsic efficacies measured using PET and EEG are likely to reflect the different in vivo abilities of BDZr drugs to induce or stabilize the GABAA-benzodiazepine chloride channel in a given conformation.(ABSTRACT TRUNCATED AT 400 WORDS)