Episodic-like memory consolidation during slow-wave sleep

Abstract

Episodic memory as a form of declarative memory allows an individual to remember events from his or her own past; it is defined as the ability to replay in mind – with autonoetic consciousness – an event as it happened in a specific spatio-temporal context. Animal experiments on episodic memory, because they cannot address the subjectivity of autonoetic consciousness, focus on its major feature, i.e., the binding of an event into its spatio-temporal context. Two major paradigms have been used to investigate episodic memory in rodents: the ‘what-where-when’ (WW-When) and the ‘what-where-which’ (WW-Which) task. Both tasks measure exploration preferences for objects and locations to assess behavioral correlates of event-context binding underlying episodic memory. Sleep supports the consolidation of episodic memories and particularly the components dependent on hippocampal function. The systems consolidation model proposes that slow wave sleep (SWS) orchestrates different brain rhythms, namely the co-occurrence of neocortical slow oscillations (SOs), thalamic sleep spindles and hippocampal ripples. The latter accompany memory reactivation. In this thesis, I investigate in rats how sleep affects the consolidation of episodic memory at behavioral and electrophysiological level. In the first set of studies (1 and 2), I demonstrated that sleep is critical for preserving an integrated episodic representation over intermediate time intervals. The first study also adds novel evidence supporting a positive relationship between the amount of slow oscillatory EEG activity during SWS and the successful performance of WW-When task. Moreover, sleep not only supports the consolidation of episodic memory measured by this task, but is also critical when the broader context in which the episode took place is considered (WW-Which task). In the third and fourth studies, I focused on brain oscillatory dynamics in neocortex and hippocampus during sleep. First, I addressed the question whether sleep and its composing sleep stages, i.e., SWS, intermediate stage and REM sleep, occur as unitary phenomena affecting the whole brain in a congruent fashion. The results show that although SWS congruently arose in signals covering the activity of both regions, REM sleep often started substantially earlier in the hippocampus than in neocortex. This not only shows a region-specific regulation of REM sleep, but also might give a unique time window for memory-related synaptic plasticity. Finally, the dialogue between neocortex and hippocampus during SWS presents a loop-like interaction of oscillatory events, where SOs can trigger spindles, and spindles can regulate the occurrence of ripples, independently from the occurrence of SOs. In turn, ripples can contribute to the emergence of SOs independently from spindles. These results shed new light on the role of sleep for the consolidation of episodic memory by unraveling its electrophysiological underpinnings and the temporal dynamics between neocortical and hippocampal networks. Together, these findings pave the way for future studies exploring the mechanisms mediating the dissociation between hippocampal and neocortical networks and its relevance for episodic memory consolidation.