Mediotemporal cross-frequency coupling during sleep and its relevance for memory consolidation

点击翻译按钮获取中文摘要

Previous research suggests that EEG cross-frequency coupling is an important mechanism supporting the representation of information in working and long-term memory. According to an influential model multiple items are represented in working memory by gamma cycles superposed on theta oscillations and this oscillatory representation is transferred to long-term memory. This neural code would be reflected by a modulation of gamma amplitudes by theta phases and by phase-phase coupling between gamma and theta oscillations. On the other hand, two stage models of memory formation postulate that long-term memories are reactivated and consolidated during sleep involving an information transfer from the hippocampus to neocortical sites. Based on the implications of both models, a reactivation of cross-frequency coupling patterns, particularly in mediotemporal brain regions, should occur during sleep, which resemble encoding-related patterns and support memory consolidation. However, direct evidence for this prediction is yet lacking. Therefore, intracranial night sleep and nap recordings will be obtained from epilepsy patients undergoing presurgical evaluation with implanted depth electrodes. Different associative long-term memory tasks will be conducted with encoding phases before night sleep, as well as before and briefly after the naps, and retrieval phases after night sleep, as well as after the naps. Phase amplitude and phase-phase coupling will be evaluated with regard to different sleep stages and recording sites (e.g. hippocampal, rhinal, temporo-lateral etc.) and the interrelation with sleep graphoelements related to memory consolidation, such as slow waves, sleep spindles and ripples will be investigated. Furthermore, individual behavioral consolidation measures and neuropsychological scores will be correlated to the crosscoupling measures and the occurrence of sleep graphoelements. The capability of the different intracranial EEG measures to predict memory performance, in particular crosscoupling measures versus graphoelements, will be evaluated and compared by multiple regression analyses. These data will elucidate, which patterns of mediotemporal crossfrequency coupling occur during sleep and whether these coupling patterns are relevant for memory consolidation.