Effect of task novelty on reactivation of spike patterns in the rat prefrontal cortex.
1. NSMA, Univ Arizona, Tucson, AZ, USA
Studies in both hippocampus and cortex have shown that neural patterns induced by a repetitive experience are often replayed in the subsequent sleep period, a phenomenon referred to as “reactivation”. Reactivation is a central component of several influential theories of memory consolidation. Most reactivation studies, however, have employed highly overlearned tasks, and it is of interest to know whether task novelty affects reactivation strength. For example, consolidation of novel experiences might be facilitated by minimizing the time spent reactivating familiar experiences. To assess the effects of experience on reactivation, multiple cells in the medial prefrontal cortex (mPFC) were recorded while a rat ran to a sequence of locations for rewarding brain stimulation (Bower et al., 2005). Sequences alternated between “cued”, during which lights indicated the correct zone and “non-cued”, during which behavior was memory-guided. The rat learned the task well, typically completing 100 cycles of the 8-element sequence in a 50 min period. A total of 130 sessions were run (40-100 cells/session), of which 36 were analyzed for reactivation. The pattern of pairwise firing rate correlations during post-behavior sleep were more similar to those during the task than were correlations during sleep prior to the task (higher “explained variance”; Kudrimoti et al., 1999). Reactivation strength, assessed via explained variance, peaked in the first 3 days of experience with the sequence task and stayed relatively elevated for the first two weeks. Behavioral evidence suggests that, during this period, the rat was still learning the general task features. Novelty was also introduced throughout the experiment by switching the sequence after the rat had reached asymptotic performance (typically after 3-4 days). The rat’s performance showed a clear decrement after each switch, indicating that new learning was taking place; however, reactivation strength was unaffected by the number of days since a switch. In conclusion, during the learning of a new environment and/or the procedural aspects of a novel task, memory trace reactivation was higher in rat mPFC than after the task was well learned, but changing task contingencies (i.e., the sequence of reward zones) within an otherwise familiar task had no effect on reactivation. These findings are consistent with our previous finding that mPFC neurons are highly selective for subtle changes in behavior but are not sensitive to the sequential context (Euston et al., SfN, 2005).
Support Contributed By: MH046823
Key words: consolidation, memory, multielectrode, sleep