Effects of chlordiazepoxide and flumazenil on preference for punished and unpunished response alternatives in rats.

Male food-restricted hooded rats were trained to respond on a two-component multiple schedule. Reinforcement density was several times higher in one component than in the other. However, responses were intermittently punished with shock in the richer reinforcement component (conflict situation). Shock intensities were adjusted to produce mild and strong suppression of responding in two separate phases. Half of the rats controlled which component was operating (Preference group) and half did not (Yoked group). The effect of chlordiazepoxide (CDZ; 0, 1, 3, and 10 mg/kg; IP) was measured on component preference and response rate. Chlordiazepoxide increased both time spent in the conflict situation and response rate in that component. This is the first study employing a schedule that permitted these two behavioral indices to be measured independently in a conflict paradigm. Response rates were also increased in the unpunished response alternative, but to a lesser degree than in the conflict situation. The effects of CDZ were at least partially mediated by the benzodiazepine receptor because CDZ's effects were diminished by flumazenil (10 mg/kg; IP), a benzodiazepine antagonist.