Effects of cognitive judgement bias and acute antidepressant treatment on sensitivity to feedback and cognitive flexibility in the rat version of the probabilistic reversal-learning test.

Depressive disorders are often associated with cognitive biases. In this study, we investigated, in an animal model, whether cognitive judgement bias, measured as a stable and enduring behavioural trait, could modulate the effects of antidepressant drugs on other cognitive processes associated with depression. For this purpose, we identified rats displaying 'pessimistic' and 'optimistic' traits in a series of ambiguous-cue interpretation (ACI) tests. Subsequently, in the preclinical version of the probabilistic reversal-learning (PRL) test, allowing multiple reversals within a test session, we compared the effects of acute administration of 5 different antidepressant (AD) drugs (agomelatine, escitalopram, clomipramine, mirtazapine and venlafaxine) on cognitive flexibility and sensitivity to positive/negative feedback in optimistic and pessimistic animals. We report that, following acute treatment with agomelatine, the proportion of lose-shift behaviours in the PRL test was significantly reduced in pessimistic animals compared to optimists. We also demonstrate that acute treatment with another antidepressant drug, mirtazapine, significantly increased the sensitivity of rats to positive feedback, as indexed by the increased proportion of win-stay behaviour following probabilistic reward. This effect was independent of cognitive bias and was associated with a reduced number of reversals made by the animals. Three other tested drugs had no significant effects on behavioural measures assessed in our study.