Datensatz

Zusammenfassung

The orbitofrontal cortex (OFC) is necessary for using learned representations of outcomes to guide behavior, i.e. model-based cognition. For example, lesions of the OFC have long been known to impair changes in conditioned responding following reinforcer devaluation. These findings, among others, have fueled the hypothesis that the OFC contributes to the use of mental representations that link cues to the incentive properties of outcomes and support inferential thinking (which is necessary for reinforcer devaluation effects). However, it is unclear whether the OFC is necessary for creating such representations or only for their use in guiding subsequent decision making. To answer this question, we conducted a simple experiment in which we used high-potency chemogenetics to inactivate OFC in rats only during the first, rather than in the final, phase of an outcome-specific Pavlovian reinforcer devaluation task. Rats were conditioned to associate two cues with two different rewards, during which OFC was inactivated in the experimental group. Subsequently, one reward was devalued by pairing with LiCl, and, finally, responding to the two cues was assessed in an extinction test. During conditioning, groups showed robust learning to both cues, although the experimental group responded less in the final two sessions. Both groups also similarly reduced consumption of the outcome paired with illness in taste aversion training. However, in the probe test, rats in the experimental group responded equally to cues predicting the devalued versus the non-devalued outcome. Importantly, they suppressed responding to both cues, indicating a loss in response specificity rather than an abolishment of devaluation effects. One interpretation is that, without OFC, rats were still able to form a model, but one without outcome-specific information. Therefore, transient and reversible OFC inactivation during conditioning affected the rats’ ability to use the learned information to support selective response changes in the probe test. This result implicates OFC in the formation of appropriate model-based and outcome-specific representations during cue-outcome learning.