In the accompanying paper we describe how MRK-409 unexpectedly produced sedation in man at relatively low levels of GABA(A) receptor occupancy (∼10%). Since it was not clear whether this sedation was mediated via the α2/α3 or α1 GABA(A) subtype(s), we characterized the properties of TPA023B, a high-affinity imidazotriazine which, like MRK-409, has partial agonist efficacy at the α2 and α3 subtype but is an antagonist at the α1 subtype, at which MRK-409 has weak partial agonism. TPA023B gave dose- and time-dependent occupancy of rat brain GABA(A) receptors as measured using an in vivo [(3)H]flumazenil binding assay, with 50% occupancy corresponding to a respective dose and plasma drug concentration of 0.09 mg/kg and 19 ng/mL, the latter of which was similar to that observed in mice (25 ng/mL) and comparable to values obtained in baboon and man using [(11)C]flumazenil PET (10 and 5.8 ng/mL, respectively). TPA023B was anxiolytic in rodent and primate (squirrel monkey) models of anxiety (elevated plus maze, fear-potentiated startle, conditioned suppression of drinking, conditioned emotional response) yet had no significant effects in rodent or primate assays of ataxia and/or myorelaxation (rotarod, chain-pulling, lever pressing), up to doses (10 mg/kg) corresponding to occupancy of greater than 99%. In man, TPA023B was well tolerated at a dose (1.5 mg) that produced occupancy of >50%, suggesting that the sedation previously seen with MRK-409 is due to the partial agonist efficacy of that compound at the α1 subtype, and highlighting the importance of antagonist efficacy at this particular GABA(A) receptor population for avoiding sedation in man.

A radioiodinated ligand that binds to muscarinic acetylcholine receptors was shown to distribute in the brain by a receptor-mediated process. With single-photon-emission imaging techniques, radioactivity was detected in the cerebrum but not in the cerebellum, whereas with a flow-limited radiotracer, radioactivity was detected in cerebrum and cerebellum. Single-photon-emission computed tomography showed good definition of the caudate putamen and cortex in man.

Summary: Radioiodinated R-and S-Quinuclidinyl deriv atives of RS-benzilate (R-and S-1 2 5IQNB) have been syn thesized for quantitative evaluation of muscarinic acetyl choline receptor binding in vivo. Two sets of experiments were performed in rats. The first involved determining the metabolite-corrected blood concentration and tissue distribution of tracer R-IQNB (active enantiomer) and S IQNB (inactive enantiomer) in brain 1 min to 26 h after intravenous injection. The second involved the measure ment of brain tissue washout over a 2-min period after loading the brain by an intracarotid artery injection of the ligands. Various pharmacokinetic models were tested, which included transport across the blood-brain barrier (BBB), nonspecific binding, low-affinity binding, and high-affinity binding. Our analysis demonstrated that the assumptions of rapid equilibrium across the BBB and rapid nonspecific binding are incorrect and result in erro neous estimates of the forward rate constant for binding at the high-affinity receptor sites (k3). The estimated val ues for influx across the BBB (K1), the steady-state ac cumulation rate in cerebrum (K), and the dissociation rate constant at the high-affinity site (k4) of R-IQNB were in dependent of the specific compartmental model used to analyze these data (K1 = 0.23 mllmin/g, K = 0.13 mil min/g, and k4 = 0.0019 min -I for caudate). In contrast, the estimated values of k3 and the efflux rate constant (k 2 ) 781varied over a 10-fold range between different compart mental models (k3 = 2.3-22 min -I and k2 = 1.6-16 min-I in caudate), but their ratios were constant (k31k2 = 1.4). Our analysis demonstrates that the estimates of k3 (and derived values such as the binding potential) are model dependent, that the rate of R-IQNB accumulation in ce rebrum depends on transport across the BBB as well as the rate of binding, and that uptake in cerebrum is essen tially irreversible during the first 360 min after intrave nous administration. Graphical analysis was consistent with compartmental analysis of the data and indicated that steady-state uptake of R-IQNB in cerebrum is estab lished within 1-5 min after intravenous injection. We pro pose a new approach to the analysis of R-IQNB time activity data that yields reliable quantitative estimates of k3, k4, and the nonspecific binding equilibrium constant (Keq) by either compartmental or graphical analysis. The approach is based on determining the free unbound frac tion of radiolabeled ligand in blood and an estimate of K1• The analysis can be applied to time-activity data obtained in a clinical setting using either positron emission tomog raphy or single photon emission computed tomography and has general application for other highly lipophilic ligands. Key Words: Blood-brain barrier-Muscarinic acetylcholine receptor binding-Quinuclidinyl benzilate.