Differential contributions of infralimbic prefrontal cortex and nucleus accumbens during reward-based learning and extinction.

Using environmental cues for the prediction of future events is essential for survival. Such cue-outcome associations are thought to depend on mesolimbic circuitry involving the nucleus accumbens (NAc) and prefrontal cortex (PFC). Several studies have identified roles for both NAc and PFC in the expression of stable goal-directed behaviors, but much remains unknown about their roles during learning of such behaviors. To further address this question, we used in vivo oxygen amperometry, a proxy for blood oxygen level-dependent (BOLD) signal measurement in human functional magnetic resonance imaging, in rats performing a cued lever-pressing task requiring discrimination between a rewarded and nonrewarded cue. Simultaneous oxygen recordings were obtained from infralimbic PFC (IFC) and NAc throughout both acquisition and extinction of this task. Activation of NAc was specifically observed following rewarded cue onset during the entire acquisition phase and also during the first days of extinction. In contrast, IFC activated only during the earliest periods of acquisition and extinction, more specifically to the nonrewarded cue. Thus, in vivo oxygen amperometry permits a novel, stable form of longitudinal analysis of brain activity in behaving animals, allowing dissociation of the roles of different brain regions over time during learning of reward-driven instrumental action. The present results offer a unique temporal perspective on how NAc may promote actions directed toward anticipated positive outcome throughout learning, while IFC might suppress actions that no longer result in reward, but only during critical periods of learning.