The Search for Abstract Value: Encoding of Specific Outcome Features in Lateral OFC and Adjusted Value in Medial OFC Revealed by Pavlovian Unblocking

Abstract

The orbitofrontal cortex (OFC) has been broadly implicated in the ability to use the current value of expected outcomes to guide behavior. While value correlates have been prominently reported in lateral OFC (lOFC), they are more often associated with more medial areas. In this body of work, we tested the hypothesis that lOFC single-unit responses to cues predictive of reward size are a feature-based rather than value-based signal, in contrast to medial OFC (mOFC). My thesis followed a prior study that used a novel Pavlovian unblocking procedure to assess excitatory cue learning driven by amount shifts or flavor shifts in expected reward outcomes. This study showed that lOFC neurons respond to a predictive cue signaling a "valueless" change in outcome features. However, many lOFC neurons also fired to a cue that signaled more reward. In the first experiment, we recorded lOFC neurons in a variant of this prior task in which rats learned about cues that signaled either more (upshift), less (downshift), or the same (blocked) amount of reward. Neurons predominantly acquired responses specifically to one of these three cues. These results show that lOFC neurons fire to valued cues in a way that is more consistent with signaling of the predicted outcome's features than with signaling of a general, abstract or cached value that is independent of the outcome. In the second experiment, we recorded in mOFC using the same design to test whether such value signals might be found there. mOFC neurons acquired responses to size-predictive cues; however, again, these responses did not signal value across cues. However, counter to lOFC, most putative glutamatergic associative neurons in mOFC developed a stronger response to the downshift cue and few neurons attended to the cue predicting no change in reward. These results suggest a model in which medial areas of OFC signal information (emotional reactions, error signals, or strategy shifting) elicited by a change - particularly reduction - in expected reward value. This would be consistent with a special role for mOFC in representing current value to specifically support devaluation-sensitive changes in behavior, perhaps with input from representations in lOFC.