Rationale: Effects of compounds on punished responding have been predictive of anxiolytic efficacy in humans. The use of mice in these tests has been limited, but the utility of this species in drug discovery and for neurobiological inquiry would benefit from a rapid, reliable method.
Objectives: The present experiments were designed to validate a new procedure in mice.
Methods: Male, NIH Swiss mice were food deprived and placed in an experimental chamber with two nose-poke holes. Every nose poke (FR1) produced a 20 mg food pellet. On the following day, a drug vehicle was administered and the mice were again exposed to the FR1 schedule. On day 3, a compound was given and the mice were run under a mixed FR1 (food), FR1 (food+shock) schedule in alternating, unsignalled periods of 4 and 10 min for three cycles. In the 10-min periods, nose-pokes produced both food plus brief electrification of the grid floor (0.5 mA for 100 ms). Effects of compounds on food intake were also evaluated in separate groups of mice.
Results: The introduction of shock substantially decreased responding during the 10-min punishment periods without significantly affecting responding during the non-punishment periods. The clinically effective anxiolytic agents chlordiazepoxide, pentobarbital, and bretazenil, but not buspirone, produced dose-dependent increases in suppressed responding, whereas d-amphetamine, chlorpromazine, and morphine were not effective. Chlordiazepoxide and bretazenil increased food consumption.
Conclusions: The present method enables rapid and reliable evaluation of potential anxiolytic agents in mice with minimal training. Increases in food intake are not necessary for anxiolytic-like effects under these conditions.