Event Abstract Back to Event Functional Division Among Monkey Prefrontal Areas in Goal-Directed Behavior Keiji Tanaka1* 1 RIKEN Brain Science Institute, Japan The neural circuitries in the prefrontal cortex are thought to be critical for the flexible control of behavior in primates, but the mechanisms remain largely unknown. In the cases of sensory and motor cortices, the understanding of the functional mechanisms has advanced along with the elucidation of functional differentiation among areas. Because the prefrontal cortex is composed of multiple areas each with unique anatomical connections to other brain sites, we expected that the comparison of functional roles among the prefrontal sub-areas would help disentangle the processes of flexible behavioral control. Selection of actions based on recent experience of action-outcome contingency is one of the features of goal-directed behavior that distinguish it from habitual behavior. Therefore, we trained monkeys on a task in which the relation between visual cues, action types, and reward conditions changed regularly, such that the monkeys selected their actions based on anticipated reward conditions, and recorded neuronal activity from various subareas of the prefrontal cortex during the task performance. Neurons in the medial prefrontal cortex first showed activity, immediately after cue presentation and well before motor execution, which represented particular action-reward combinations. This activity might underlie goal-based action selection. Application of previously learned behavioral rules beyond simple stimulus-response or outcome-response associations is often necessary in goal-directed behavior, and the currently relevant rule is often not directly indicated by sensory cues. The Wisconsin Card Sorting Test (WCST) mimics such a situation. We have developed an animal version of WCST and trained monkeys with the task. The monkey selected one of the three test stimuli by matching it with the sample stimulus in color or in shape. The matching rule was constant within a block of trials, but changed between blocks without giving any notice to the monkey. There was no cue to indicate the currently relevant rule: the monkey had to determine the rule based on the reward given for correct responses. Lesioning of the principal sulcus region (PS), the orbitofrontal region (OFC) and the anterior cingulated sulcus region (ACCs) resulted in significant degradation of the overall performance of the monkeys. Further analyses of the monkeys’ performance in the task and in other probe tests showed that the reasons of the degradation were different. Only the PS lesion impaired maintenance of abstract rules in working memory; only the OFC lesion impaired rapid reward-based updating of rule value; and the ACCs lesion impaired active reference to the content of the rule working memory. These results show that the prefrontal subareas contribute to the flexible control of behavior through individually specific functions.