GABA‐A Selective Positive Modulators: Anxiolytic Efficacy versus Side Effects

Benzodiazepines (BZ) have been the classic treatment for acute anxiety, but they are also known to produce several unwanted side effects. Recent GABA A selective agonists have been shown to have high margins between efficacy in anxiety models and disruption in locomotor behavior. However, some of the compounds appeared to be sedative near anxiolytic doses in clinical studies. To determine the safety margin between efficacy and side effects in rats, we tested classic BZ compared to selective GABA A compounds in a modified Geller‐Seifter conflict model (GS) and the differential reinforcement of low rates (DRL15) schedule, a task requiring attention and behavioral inhibition. The classic nonselective BZ, diazepam (ip) and alprazolam (po) significantly disrupted DRL accuracy at 2 mg/kg and 10 mg/kg respectively while the corresponding minimum anxiolytic doses were 0.3 mg/kg and 1 mg/kg. L838417 (po), a potent α2‐GABA selective positive modulator, decreased DRL accuracy as low as 3 mg/kg while anxiolysis occurred at 0.1 mg/kg. Another positive modulator, selective for α2 and α3 GABAA receptors, NS2710 was efficacious in GS at 0.3 mg/kg. However it impaired DRL accuracy at 1 mg/kg. Alpidem, an anxiolytic selective for α1 and α3 subunits, disrupted accuracy at 3 mg/kg and was efficacious in GS at 1 mg/kg. The estimated margins for diazepam, alprazolam, L838417, NS2710 and alpidem are 7x, 10x, 30x, 3x and 3x respectively. All GABA A compounds tested increased the percent of intertime responses that were less than 15 sec and disrupted accuracy at doses that were near the anxiolytic doses. The results suggest that the DRL schedule may be a useful tool in determining the inclination of BZ to cause side effects in man.