Neural Substrates of Reward Processing and Emotion

National Institute Of Mental Health

Investigators

Linked publications & trials

Abstract

Lesion studies suggest dissociable functions of medial prefrontal cortex (MFC) and orbitofrontal cortex (OFC), with MFC being essential for social cognition and OFC being essential for value-based decision making. One puzzling finding is that animals with bilateral damage to the OFC made by aspiration, which directly removes the cortex, exhibit a different set of behavioral impairments than animals with excitotoxic lesions of the same region. Specifically, animals with bilateral aspiration removals of OFC are impaired on tests of cognitive flexibility and emotion regulation, whereas those with bilateral excitotoxic lesions of OFC are not (Rudebeck et al., Nat. Neurosci., 2013). This discrepancy is attributed to the inadvertent disruption of fibers of passage by aspiration lesions but not by excitotoxic lesions. The underlying question is: Which fibers of passage are responsible for impairments in cognitive flexibility and emotion regulation when compromised? One candidate is the cholinergic fibers originating in the nucleus basalis magnocellularis (NBM) and passing nearby or through OFC on their way to other frontal cortex regions (Kitt et al., Brain Res. 1987). To investigate this possibility, we gave three animals unilateral aspiration lesions of the OFC, and then we compared cholinergic innervation of the anterior cingulate cortex (ACC), a part of the MFC, between hemispheres with and without OFC. Our dependent measure was AChE expression. The assessment revealed diminished AChE expression in the ACC of hemispheres with OFC lesions, particularly in layers I and III/V. This disparity was more pronounced in the rostral ACC and less so in the more caudal regions. No difference was observed in cholinergic staining in the striatum, which receives cholinergic innervation from local interneurons. These findings suggest that aspiration lesions of the OFC disrupt cholinergic fibers of passage, potentially explaining the significant impairments in cognitive flexibility and emotion regulation observed in comparison to excitotoxic lesions. Further research is necessary to demonstrate a causal relationship between the interruption of cholinergic innervation in the frontal cortex and its effects on behavior. As indicated above, the MFC, together with the amygdala and cortex near the temporal-parietal junction is implicated in social cognition. This network of brain regions is specialized for processing social information. For example, neurons in the MFC encode not only reward value but also features of conspecifics and of oneself. Our previous research has demonstrated that MFC damage blunts learning of prosocial preferences. It has been difficult, however, to disentangle MFC contributions to social attention, social preference, and social reward, as measured by vicarious reinforcement (i.e., the willingness to give reward to others). In addition, MFC, including the ACC, contributes to nonsocial reward processing. To further investigate the role of the MFC in nonsocial reward processing, which would serve as a comparison with our social (vicarious reinforcement) measures, we assessed the ACCs causal contribution to the expression of reward expectancy via an autonomic measure. Three animals received bilateral neurotoxic lesions of ACC. Together with four unoperated controls (CON), they were presented with a Pavlovian procedure whereby visual stimuli on a monitor screen predicted reward delivery, independent of the animals actions. In each trial, animals saw two blue circles slowly move towards each other for 15 seconds until they touched, disappeared, and fluid reward was delivered. A 15-s intertrial interval ensued, during which the screen was blank. Heart rate (HR) was acquired with a plethysmography sensor and was extracted off-line. We found a group difference in HR in anticipation of reward. Whereas CON animals exhibited a gradual HR increase while the circles moved closer together, animals with ACC lesions did not. This implies that the autonomic output that reflects reward expectancy was lost as a result of the ACC lesion. In contrast, we observed a significant HR change after reward delivery in all animals, controls and animals with ACC lesions alike, indicating that the autonomic response to receipt of reward was intact after ACC lesions. Taken together, the results support an essential role for the ACC specifically in generating autonomic responses to reward expectancy. Thus, the ACC may contribute to positive affect by generating arousal in anticipation of positive emotional events.