Reconsolidation of appetitive odor discrimination requires protein synthesis only when reactivation includes prediction error.

Reconsolidation theory has supported the notion that active memory is vulnerable to the effects of an amnesic agent. An extension of reconsolidation theory posits that active memory, while necessary for creating vulnerability in memory, is not sufficient. Prediction error (i.e., when expectation is inconsistent with reality) may be a key factor in the destabilization of memory. The present study examined the role of prediction error in appetitive memory reconsolidation. Rats learned to dig in cups of scented sand to retrieve buried sweet cereal rewards. Forty-eight hours following acquisition, a single reactivation trial was given during which a prediction error or no prediction error was included. The prediction error consisted of a single extinction trial, while the no prediction error condition consisted of an additional reinforced trial. Cycloheximide (CHX; 1 mg/kg) or vehicle (VEH: distilled water; 1 mg/kg) was administered intraperitoneally immediately following reactivation. One week following reactivation, rats received 2 nonreinforced test trials. Results showed longer latencies to dig for rats that received CHX following a prediction error (CHX/PE) compared to rats that received VEH (VEH/PE) or did not receive a prediction error (CHX/NoPE or VEH/NoPE). These results add to a growing literature demonstrating that protein synthesis is necessary in appetitive memory reconsolidation only when reactivation includes a prediction error. (PsycINFO Database Record