Abstract
The biochemical targets for antidepressants are relatively well established, but we lack a clear understanding of how actions at these proteins translate to clinical benefits. This study used a novel rodent assay to investigate how different antidepressant drugs act to modify affective biases that have been implicated in depression. In this bowl-digging task, rats encounter two equal value learning experiences on separate days (one during an affective manipulation and the other during control conditions). This induces an affective bias that is quantified using a preference test in which both digging substrates are presented together and the individual rats' choices recorded. The assay can be used to measure affective biases associated with learning (when the treatment is given at the time of the experience) or examine the modification of previously acquired biases (when the treatment is administered before the preference test). The rapid-onset antidepressant ketamine, but not the delayed-onset antidepressant, venlafaxine, attenuated the previously acquired FG7142-induced negative bias following systemic administration. Venlafaxine but not ketamine induced a positive bias when administered before learning. We then used local drug infusions and excitotoxic lesions to localize the effects of ketamine to the medial prefrontal cortex and venlafaxine to the amygdala. Using a modified protocol we also showed that positive and negative biases amplified further when the numbers of substrate-reinforcer associations are increased. We propose that this pattern of results could explain the delayed onset of action of venlafaxine and the rapid onset of action but lack of long-term efficacy seen with ketamine.