The reward factor in the control of action

Abstract

Our brains process ubiquitous reward information on a daily basis to learn, choose, prepare and execute goal-directed behaviour. Although there has been a proliferation of valuable research examining reward-related neural mechanisms, exactly how the expectation of reward is integrated into decision-making (i.e., choosing one option over another) and motor processes still largely remains enigmatic. The central tenet of the current research was to further contribute to our understanding of the cognitive and neurochemical mechanisms involved in visuospatial processing of reward-oriented behaviour in rats using three variants of an asymmetrical reward paradigm (ARP). More specifically, investigations of the specific effects of both central D1-like and D2-like dopamine (DA) receptors and the N-methyl-D-aspartate (NMDA) receptor were explored after the systemic administration of prototypical agonists and antagonists respectively. Using variants of the ARP in three separate studies (comprising a total of three behavioural and 13 pharmacological experiments), the effects of reward expectation on reaction time distributions of visuospatial discrimination were assessed. To elaborate, Study I explored the implementation of an automatic stimulus-reward reversal in a lever-press version of the ARP. Study II examined different receptor contributions to reward-oriented behaviour using a nose-poke analogue of the ARP. Finally, Study III aimed to further tease apart the different receptor contributions to goal-directed behaviour by imposing greater visuospatial complexity in the form of distraction in a modified version of the ARP nose-poke task. Overall, the findings revealed that rats showed a reward-dependent spatial preference for the expected larger magnitude of reward by consistently responding faster in a direction associated with this outcome. Furthermore, although there were unique characteristic profiles of each of the five different pharmacological treatments administered across each of the three studies, no form of receptor manipulation entirely extinguished the reward factor. A further additional ambit of Study III was to demonstrate that the behavioural changes associated with reward expectation may be underpinned by two different parallel cognitive mechanisms: perceptual sensitivity, on the one hand, and response bias, on the other. The current data provide evidence in favour of response bias, thus lending further support that neural activity is initiated in advance of sensory information processing - at least in the case of well-learned visuospatial processing of different reward magnitudes. Taken altogether, with regards to DA, the most parsimonious conclusion to arise from the current research is that this neuromodulator appears to exert an influence on the learning of reward associations rather than on reward per se. With regards to glutamate, the intriguing findings revealed that blocking the action of this neurotransmitter, specifically by antagonising NMDA receptors, facilitated ARP task performance in small-rewarded trials in the absence of distraction, and is suggested to pertain to a counteractive neural mechanism that operates in opposition to response bias. In closing summary, this research demonstrates a novel way of evaluating the effect of reward on decision-making behaviour in rodents and invites the future collaboration of neurophysiology and psychopharmacology approaches in one and the same animal to further extend the current investigations. The general theoretical and practical implications, as well as the strengths and limitations of this research are discussed.