Abstract
The past two decades have seen increased interest in the function of dopamine (DA) by virtue of its role in generating Reward-Prediction Errors (RPEs) in the service of Reinforcement Learning. From the perspective of systems neuroscience, most of this research has focused on the mesostriatal and mesoaccumbens DA pathways, as technical limitations have prevented a detailed examination of the role of RPEs in the medial frontal cortex (MFC). The recent development of DA-sensitive fluorescent sensors and fiber photometry now provides relatively selective measures of DA responses, with high temporal resolution. Here the technique was used to compare MFC DA responses to appetitive, aversive, and neutral events, along with violations in learned contingencies that should theoretically generate robust RPEs in male rats. Aversive and rewarding outcomes evoked DA responses of comparable magnitude, although responses to tones paired with footshock evoked larger responses than those paired with food. Conditions used successfully to create RPEs in striatum, such as omitting outcomes, failed to provide clear evidence of signed RPE-like DA responses in the MFC. Furthermore, DA responses were similar in magnitude regardless of whether outcomes were expected or unexpected/swapped. The most parsimonious conclusion from these experiments would be that while the MFC DA is capable of conveying weak RPEs, it mainly tracks the physiological arousal created by the affective properties of appetitive or aversive stimuli.