The orbitofrontal and anterior cingulate cortex are inextricably linked with processing emotional information, generating appropriate autonomic responses and guiding behavior. Both areas accomplish these functions via reciprocal connections with the amygdala and brain structures related to visceral function. Recent neurophysiological reports have shown that the activity of single neurons in both areas is related to the emotional or rewarding nature of different decisions. What is less clear, however, is how the reward related activity of single neurons in the orbitofrontal cortex and the anterior cingulate cortex drives bodily sensations during behavior and how this activity, both neural and autonomic, may be altered following damage to the amygdala. This project probes the contribution of the amygdala to orbitofrontal and anterior cingulate cortex function and to autonomic responses based on reward expectation. Specifically, single-cell recording in orbitofrontal cortex and the anterior cingulate cortex have been combined with autonomic measures (pupil diameter, heart rate, and skin conductance) while subjects performed two kinds of reward-guided tasks. In one task, subjects choose between two stimuli that they know will yield different amounts of reward;in the other they learn how much reward a novel stimulus will yield. We hypothesize that, in intact brains, the activity of orbitofrontal and anterior cingulate cortex neurons and autonomic outputs will reflect reward magnitude, but following amygdala lesions these signals will be attenuated or absent. This project is in its early stages, but so far we have confirmed previous reports that both the activity of single orbitofrontal cortex neurons and autonomic responses reflect the amount of reward the subjects can predict for different decisions. We have also found that lesions of the amygdala block the autonomic arousal that occurs before large rewards, as reflected in pupillary diameter. These results indicate a role for the amygdala in mediating the emotions related to reward expectation. Our future neurophysiological analysis should reveal whether the amygdala relays this information to the orbitofrontal and anterior cingulate cortex, as well. Another important initial finding from this project is that while amygdala lesions abolish the relationship between autonomic signals and reward magnitude, the subjects task performance is unaltered by the lesion. This uncoupling of autonomic output and behavior suggests that distinct parts of orbitofrontal cortex (or distinct neurons within the same parts) play different roles: those controlling autonomic output cease to reflect anticipated reward magnitude after amygdala lesions, whereas cells correlated with the subjects decisions continue to encode this information. We are now testing this prediction experimentally. This project represents an important new paradigm for studying the interactions of the amygdala with the frontal cortex, one that should yield important insights into the functional localization and mechanisms of emotion, reward expectation, and valuation. In an fMRI study of the role of the amygdala in reward and emotion, we are collaborating with Hadj-Bouziane and others in the Laboratory of Brain and Cognition, NIMH. As is well known, when subjects view emotional facial expressions (compared to unemotional faces), there is a greater increase in regional cerebral blood flow in areas of visual cortex, such as the inferior temporal cortex. We have tested the hypothesis that amygdala lesions would eliminate the emotional feedback to the visual cortex that causes this effect. Face-selective regions were identified in the inferior temporal cortex in both control subjects and subjects with amygdala lesions. In controls, faces with emotional expressions produced enhanced responses in these regions relative to neutral faces, as expected. In subjects with amygdala lesions, these effects were disrupted. In the hemisphere with the most complete amygdala lesion, face-selective patches in inferior temporal cortex were unaffected by facial expressions. Overall, our results support the idea that the amygdala is the source of emotion-related signals seen in the visual cortex, a finding that yields important predictions that can be tested with the neurophysiological paradigm described above.