LEARNING OF REWARD LOCATIONS UNDER PATH INTEGRATION AND CUE-BASED NAVIGATION

Background: The hippocampus is considered a hub for episodic memory and spatial memory. It has been established that a combination of both environmental cues and self-motion information is used to form a cognitive map of the environment in the hippocampus. Furthermore, recently acquired memories are reactivated during replay events in NREM sleep, a phenomenon believed to contribute to memory consolidation. Yet, it is still unclear how replay events contribute to the formation of cognitive maps and to the integration of information such as rewards.
Methods and results: Mice were trained to run head-fixed on a treadmill apparatus equipped with a 2-meter-long belt enriched with visual-tactile cues. The belt presented a cue-enriched zone and a cue-impoverished zone promoting cue-based and path-integration navigation strategies, respectively. Mice learned to navigate to a pair of reward locations, with each reward location falling within one of the two zones. Within a session, mice first ran for a familiar pair of reward locations and then experienced a new pair of reward locations for a few trials. Then a period of rest/sleep was induced by increasing the resistance of the treadmill belt rotation, following which mice ran again for the same pair of novel reward locations. Hippocampal neural activity was recorded using silicon probes. Place cells and ripple events were analyzed. Place cells exhibited various patterns of remapping upon the change in reward locations and following the period of sleep.