Investigating the effects of orexin receptor 1 antagonism on decision making under uncertainty.

Orexins/hypocretins are neuropeptides produced by several thousand neurons in the lateral hypothalamus. They project widely through the central nervous system where they release orexins which bind to two regionally selective G-protein coupled receptors: OX1R and OX2R. Orexins are well known as regulators of the sleep/wake cycle, however, recent investigations into orexinergic modulation of feeding and drug-seeking behaviour suggest they also play a role in reward processing and decision making. In the present study, we investigated the effects of OX1R antagonism on goal-directed decision making using an operant probabilistic reversal learning (PRL) task. 44 male C57/BL6 mice were dosed daily with an OX1R selective antagonist (1-SORA-51, 45 mg/kg) or vehicle (20 % w/v TPGS) while performing a PRL task consisting of 5 sessions on 5 consecutive days of probabilistic discrimination learning, followed by 5 sessions of reversal learning, both on and off drug, in a crossover design. Behaviours were then analysed within a reinforcement learning framework. Mice treated with 1-SORA-51 show a significant decrease in learning both initial and reversed reward contingencies, mediated largely through learning from positive outcomes. 1-SORA-51 also increased exploratory behaviours, both during learning and after reward contingencies had been learned. The findings suggest that OX1R signalling plays multiple roles in decision making in both learning and reward processing, largely by impacting the positive reward domain. As such, OX1R antagonists may be of therapeutic interest for improving abnormal reward processing and explore-exploit behaviours, such as the heightened sensitivity to drug cues and reduced responses to natural rewards, or heightened delay discounting as observed in people with substance use disorders.