An assembly-line model of sleep, dreaming and memory enhancement: investigating the coordinated roles of REM and nonREM sleep in human memory consolidation

快速眼动（REM）和非快速眼动（NREM）睡眠都有助于改善记忆，某些振荡事件，如睡眠纺锤波（11-16赫兹脑电图活动的爆发），是这种活动的可靠生物标志物。关于做梦是否在记忆中起作用，特别是新记忆的梦境回放是否也构成这样一种生物标记，证据更加复杂。我们最近重复了NREM和REM睡眠以一种互补的方式影响记忆的发现，更重要的是，我们首次表明NREM睡眠纺锤波、REM睡眠的时间和做梦的长度和内容都是相互关联的。这些相关性在第三个睡眠周期中尤为强烈，这表明这是一个特别有利于记忆可塑性的时间，这种现象以前只在啮齿动物身上得到证实。作为新发现，这些NREM-REM-dreaming相关性需要被复制，扩展到记忆改善的其他睡眠测量中，并使用允许得出因果结论的工具进行评估。拟议的研究旨在通过仔细复制和使用经过验证的、有针对性的记忆再激活（TMR）方法在睡眠中增强记忆来实现这些目标。使用TMR，音调在任务期间呈现，然后在非快速眼动睡眠期间再次呈现；这些声音可能会重新激活在任务学习过程中形成的神经回路。我们将在非快速眼动睡眠中呈现与任务相关的音调，并测量非快速眼动、快速眼动和做梦的测量是如何改变的，从而增强学习能力。在一个典型的实验中，我们的任务将在虚拟现实（VR）环境中完成，并将涉及导航迷宫。我们将评估，在非快速眼动睡眠中，纺锤体活动是否增加；在快速眼动睡眠中，如果眼球运动发生变化；在梦境报告中，如果VR任务回放出现。如果睡眠周期3是记忆处理的特权窗口，那么在周期3播放TMR声音的受试者的任务改善和睡眠/梦的改变应该是最大的（与周期2或4相比）。

Both rapid eye movement (REM) and nonREM (NREM) sleep actively contribute to memory improvement and certain oscillatory events, such as sleep spindles (bursts of 11-16 Hz EEG activity), are reliable biomarkers of this activity. Evidence is more mixed as to whether dreaming plays a role in memory and, especially, whether dream replays' of new memories also constitute such a biomarker. We recently replicated findings that NREM and REM sleep influence memory in a complementary fashion and, more importantly, showed for the first time that NREM sleep spindles, amount of REM sleep and dreaming length and content are all inter-correlated measures. These correlations were preferentially strong for the third sleep cycle, suggesting that it is a time particularly conducive to memory plasticity, a phenomenon previously shown only for rodents. As new discoveries, these NREM-REM-dreaming correlations need to be replicated, extended to other sleep measures of memory improvement, and evaluated with tools that permit drawing conclusions about cause and effect. The proposed research aims to achieve these objectives by careful replication and by enhancing memory during sleep using a validated, targeted memory reactivation (TMR), method. With TMR, tones are presented during a task and then again during NREM sleep; the sounds presumably reactivate neural circuits that were formed during task learning. We will enhance learning by presenting task-associated tones in NREM sleep and measuring how NREM, REM and dreaming measures are altered. In a typical experiment, our task will be completed in a virtual reality (VR) setting and will involve navigating a maze. We will assess, in NREM sleep, if spindle activity increases; in REM sleep, if eye movements change; and, in dream reports, if VR task replays appear. If sleep cycle 3 is a privileged window for memory processing, then task improvement and sleep/dream changes should be greatest for subjects to whom TMR sounds are played during cycle 3 (vs. during cycles 2 or 4). The research will be of interest to persons studying sleep and memory, dreaming or TMR and should provide insights into how sleep controls memory storage. It will have a particular relevance for showing how timing and stimulating sleep episodes can optimize memory and thus enhance learning with greater precision. Findings may also help explain if and how dreaming is implicated in memory, i.e., by clarifying its relationship to sleep spindles and whether it replays elements of new learning. Thus, the work may lead to a more profound understanding of both memory and sleep and support a growing outreach to Canadians to improve their cognitive functioning by getting more sleep. This type of research may lead to the development of sleep-based methods for enhancing, optimizing and rehabilitating memory and in protecting it in the face of age-related or illness-related decline.