Standard Article

You have free access to this content

Sleep and Memory

  1. SJ Aton,
  2. J Seibt,
  3. MG Frank

Published Online: 15 SEP 2009

DOI: 10.1002/9780470015902.a0021395

eLS

eLS

How to Cite

Aton, S., Seibt, J. and Frank, M. 2009. Sleep and Memory. eLS. .

Author Information

  1. University of Pennsylvania, Philadelphia, Pennsylvania, USA

Publication History

  1. Published Online: 15 SEP 2009

This is not the most recent version of the article. View current version (15 AUG 2014)

Introduction

  1. Top of page
  2. Introduction
  3. Synaptic and Systems Mechanisms of Memory Consolidation
  4. Sleep and Memory Consolidation
  5. Sleep and Synaptic Consolidation
  6. Sleep and Systems Consolidation
  7. Summary, Caveats and Future Strategies
  8. Does Sleep Facilitate Consolidation through Permissive or Instructive Mechanisms?
  9. Conclusions
  10. References
  11. Further Reading

In the last decade, there has been a revival of interest in the theory that sleep is essential for memory formation. This has been fuelled primarily by behavioural studies in humans which show beneficial effects of sleep on memory consolidation (reviewed in Stickgold, 2005). However, the neurobiological processes underlying these effects remain unclear (Frank and Benington, 2006). In this article, we address the possible role of sleep in neurophysiological processes involved in memory consolidation at the synaptic and brain systems levels. We first briefly summarize key steps involved in these two levels of consolidation, then examine current evidence that suggests that sleep influences these processes. We conclude with a discussion of remaining key questions regarding sleep's role in consolidation, and suggest future strategies to better understand the underlying mechanisms.

Synaptic and Systems Mechanisms of Memory Consolidation

  1. Top of page
  2. Introduction
  3. Synaptic and Systems Mechanisms of Memory Consolidation
  4. Sleep and Memory Consolidation
  5. Sleep and Synaptic Consolidation
  6. Sleep and Systems Consolidation
  7. Summary, Caveats and Future Strategies
  8. Does Sleep Facilitate Consolidation through Permissive or Instructive Mechanisms?
  9. Conclusions
  10. References
  11. Further Reading

Memory formation can be broken down into several discrete steps. Relevant sensory information must first be attended to and registered as a memory trace – a process referred to as acquisition. Following acquisition, memories either undergo a process of consolidation (enhancement and/or stabilization for long-term storage) or are forgotten. Forgetting can be thought of as a process that acts in opposition to consolidation, and can occur either through interference (where prior or subsequent memory formation actively prevents consolidation of information) or decay (the passive disappearance of unconsolidated memories over time) (Wixted, 2004). Memory is often broadly categorized as either explicit or implicit. The main difference between the two memory types is that consolidation of explicit memory (e.g. episodic and semantic memory) is dependent on medial temporal lobe structures such as the hippocampus, whereas consolidation of implicit memory (e.g. priming and procedural/skill memory) is not. However, consolidation of both implicit and explicit memory seems to rely on two general mechanistic processes, referred to as synaptic and systems consolidation (Dudai, 2004). See also Learning and Memory

‘Synaptic consolidation’ (Figure 1a and d) refers to rapid, activity-dependent changes in synaptic strength within specific neural circuits following memory acquisition, which likely represents a cellular scaffold for subsequent memory consolidation. There are three defining features of this process. The first is that it occurs within the first hours after acquisition (Figure 1a). This can be seen, for example, in the decreasing sensitivity of motor skill learning to behavioural interference over the first few hours after acquisition (Brashers-Krug et al., 1996; reviewed in Dudai, 2004). A second feature of synaptic consolidation is that it requires activation of N-methyl-d-aspartate receptor (NMDAR)- and kinase-dependent intracellular signalling cascades, and synthesis of new messenger ribonucleic acid (mRNA) and proteins (Figure 1d). For example, during the first hours after fear conditioning training in rodents, consolidation is sensitive to pharmacological disruption of NMDAR-dependent kinase pathways and de novo mRNA and protein synthesis in brain areas activated by learning (Abel and Lattal, 2001). These cellular mechanisms are also required for long-term potentiation (LTP) and long-term depression (LTD) – well-studied processes of synaptic strengthening and weakening, respectively, which occur in vitro and in vivo. Importantly, the third defining feature of synaptic consolidation is that it involves rapid changes in synaptic strength within specific neural circuits. Examples of this include findings that specific types of learning are followed by the same postsynaptic changes (increased postsynaptic density area, phosphorylation of key target proteins and enhancement of synaptic responses) that occur during LTP (Horn, 2004; Whitlock et al., 2006). See also Long-term Depression and Depotentiation, Long-term Potentiation, NMDA Receptors, Protein Phosphorylation and Long-term Synaptic Plasticity, and Protein Synthesis and Long-term Synaptic Plasticity

thumbnail image

Figure 1. Synaptic and systems levels of memory consolidation. In the first hours following acquisition, a memory trace is represented in specific neural circuits (in this example, within the hippocampus, highlighted in grey, (a)). The memory trace is then gradually integrated into a broader neuronal network, including areas not involved in acquisition (in this example, in the neocortex, (b)), and distributed over time throughout the neocortex (c). Synaptic consolidation (schematized in (d)) involves activation of NMDAR- and kinase-dependent signalling cascades which promote de novo mRNA and protein synthesis. NMDAR activation increases calcium (Ca2+) influx, whereas neurotransmitter-mediated activation of G protein-coupled receptors stimulate cyclic adenosine monophosphate (cAMP) production. These intracellular signals in turn promote the phosphorylation-dependent activation of kinases (e.g. PKA, CaMKII and extracellular signal-regulated kinase (ERK)) (Wang et al., 2006). These kinases activate cAMP response element-binding (CREB) protein-dependent immediate early gene (IEG) expression (e.g. zif268, arc/arg3.1, c-fos) (Abel and Lattal, 2001). Products of these transcripts are then translated, together with additional effectors of synaptic potentiation or depression (e.g. glutamate receptor subunits, and CaMKII). Systems consolidation (schematized in (e)) requires subsequent activation of CaMKII and synthetic pathways over a period of days to months (Wang et al., 2006), as memory traces expand through more broadly distributed brain areas.

‘Systems consolidation’ (Figure 1b, c and e) refers to more protracted, activity-dependent expansion of memory traces from their original source to more widespread cortical areas. Two key defining features of systems consolidation are the gradual integration of memories through more distributed brain areas over a period of days to months (Figure 1b and c), and dependence on NMDAR-dependent kinase pathways and protein synthetic machinery during this process (Figure 1e). Evidence for the expansion and redistribution of memory traces comes from longitudinal brain-imaging studies, which demonstrate that long-term consolidation is associated with more distributed brain activation during recall (Frankland et al., 2004; Takashima et al., 2006). Additional, behavioural evidence for this gradual redistribution is seen in patients with damage to medial temporal lobe structures. These individuals frequently experience explicit memory loss for recent events, whereas more remote memories remain accessible (Dudai, 2004). Evidence that NMDARs, kinases and protein syntheses are required for this process have come from animal studies of learning where pharmacological and genetic blockade of these pathways days or even months after learning impairs subsequent recall (Wang et al., 2006). See also Amnesia

Sleep and Memory Consolidation

  1. Top of page
  2. Introduction
  3. Synaptic and Systems Mechanisms of Memory Consolidation
  4. Sleep and Memory Consolidation
  5. Sleep and Synaptic Consolidation
  6. Sleep and Systems Consolidation
  7. Summary, Caveats and Future Strategies
  8. Does Sleep Facilitate Consolidation through Permissive or Instructive Mechanisms?
  9. Conclusions
  10. References
  11. Further Reading

Despite decades of debate (Ellenbogen et al., 2006), consensus in the field is gradually building that sleep promotes memory consolidation (Stickgold, 2005). The critical question, which we address here, is how consolidation could be achieved during sleep. A simple explanation is that sleep protects memory traces from interference associated with wakefulness, by reducing sensory input to the brain (Wixted, 2004). However, recent studies specifically addressing this issue report improvements in memory associated with sleep which cannot be explained entirely by a reduction in interference (Mednick et al., 2002; Gais et al., 2006; Korman et al., 2007). For example, Mednick et al. found improvements in subjects’ performance on a visual perception task when a 1-h nap was allowed between training sessions. However, no improvement was seen in control subjects who were kept awake, but blindfolded, over the same time period – thus simply limiting visual interference (without sleep) was insufficient for consolidation (Mednick et al., 2002). Therefore, it is likely that sleep modulates memory via more direct effects on synaptic and systems consolidation.

Sleep and Synaptic Consolidation

  1. Top of page
  2. Introduction
  3. Synaptic and Systems Mechanisms of Memory Consolidation
  4. Sleep and Memory Consolidation
  5. Sleep and Synaptic Consolidation
  6. Sleep and Systems Consolidation
  7. Summary, Caveats and Future Strategies
  8. Does Sleep Facilitate Consolidation through Permissive or Instructive Mechanisms?
  9. Conclusions
  10. References
  11. Further Reading

Time-dependency

As discussed earlier, a key feature of synaptic consolidation is that it occurs within the first hours following acquisition. Thus if sleep plays a role in synaptic consolidation, one would expect that sleep immediately following acquisition would be critical for memory. There are several examples in the behavioural literature where sleep is indeed required during the first hours following acquisition in order for optimal consolidation to occur. One recent study found that memory for new vocabulary words was stabilized when subjects slept within 3 h of learning them, but impaired when sleep was delayed for several hours (Gais et al., 2006). A second example is contextual fear conditioning in rodents, which is impaired when animals are sleep deprived over the first 5 h following acquisition training, but not when they are deprived afterward (Graves et al., 2003). A third is visual imprinting in chicks, which requires sleep in the first few hours following exposure to imprinting stimuli, but is not facilitated by sleep after this critical window (Jackson et al., 2008).

Dependence on NMDA receptors, kinases and synthetic pathways

A second feature of synaptic consolidation is dependence on NMDAR- and kinase-dependent intracellular signalling cascades, and synthesis of new mRNA and proteins. There are two lines of evidence that these cellular mechanisms may be specifically activated during post-acquisition sleep (as schematized in Figure 2c). First are reports of activation of these pathways which is correlated with the requirement for sleep, and occurs in the first hours of memory consolidation. For example, sleep-dependent consolidation of two-way active avoidance learning in rats is correlated with hippocampal IEG expression and protein phosphorylation in the first few hours following training (Saha and Datta, 2005; Ulloor and Datta, 2005). Increased hippocampal expression of the LTP-related gene zif268 has been reported during rapid eye movement (REM) sleep in the first 1–2 h following exposure of rodents to novel, enriched environments (Ribeiro et al., 1999). And finally, sleep-dependent consolidation of chick imprinting occurs at a time when there are dramatic increases in IEG expression (including Fos and arc/arg3.1, both of which are involved in synaptic plasticity), and calcium-calmodulin-dependent protein kinase II (CaMKII) activation in neural circuits activated by training (Horn, 2004; Bock et al., 2005).

The second set of findings demonstrates that interfering with these mechanisms blocks improvements in consolidation which are associated with sleep. For example, one recent study found that subjects’ sleep-dependent improvements on a visual task could be blocked by administering an NMDAR antagonist over the first 6 h of sleep (Gais et al., 2008). In the example of contextual fear conditioning cited earlier, consolidation is disrupted by administration of protein synthesis and protein kinase A (PKA) inhibitors during the first 1–4 h following acquisition (the same timeframe when sleep is required) (Bourtchouladze et al., 1998; Schafe et al., 1999).

thumbnail image

Figure 2. Potential roles of sleep in synaptic and systems consolidation. (a) Representative changes in release of acetylcholine (ACh) and noradrenaline (NA) in the cortex (relative to waking) during REM and NREM sleep. Representative EEG traces for the two sleep states show cortical activity similar to that of waking during REM sleep, and high-amplitude EEG oscillations – including sleep spindles and slow-wave activity (SWA) – during NREM sleep. (b) Sequence of events showing the transfer over time of the memory trace from hippocampus to neocortical distributed network (see Figure 1a–c for detailed description). (c)–(e) Potential cellular mechanisms underlying sleep facilitation of consolidation. Functions demonstrated to occur specifically during sleep are shown in colour; mechanisms correlated with sleep (i.e. known to be active at the time when sleep is required for consolidation) are indicated with a question mark. Mechanisms that have not been studied with respect to sleep are shown in gray. (c) Mechanisms implicated in sleep facilitation of synaptic consolidation are schematized in a neuron within the circuits activated by prior learning (in this example, a neuron in the hippocampus). NMDAR activation during sleep may be necessary for some early forms of consolidation (Gais et al., 2008), whereas PKA, CaMKII and CREB activation and protein synthesis are correlated with the time-dependent requirement for sleep in other forms of learning (Bourtchouladze et al., 1998; Horn, 2004; Saha and Datta, 2005). Increased expression of the IEG zif268 occurs during REM sleep following some forms of learning (Ribeiro et al., 1999), whereas arc/arg3.1 and c-fos expression are correlated in time with the requirement for sleep following other learning paradigms (Horn, 2004; Bock et al., 2005). (d) The same signalling pathways may be activated in additional neural circuits during subsequent systems consolidation (schematized as a neuron in the neocortex which is synaptically connected with the original hippocampal neuron). Changes in neurotransmission or activity during sleep (e.g. synchronized bursting activity produced by SWA and spindles) could promote intercellular communication between these brain areas (represented by red arrows). During subsequent stages of systems consolidation (e), these same sleep-dependent mechanisms may promote further expansion of memory traces throughout the neocortex (indicated by red arrows).

Changes in synaptic strength

The third feature of synaptic consolidation is that it leads to rapid changes in synaptic strength in circuits activated by prior learning. Two lines of evidence indicate that this process is promoted by sleep. Ocular dominance plasticity (ODP), a canonical form of in vivo synaptic remodelling occurring in the visual cortex after a brief monocular visual experience, is consolidated by a few hours of subsequent sleep (but not a similar period of wakefulness) (Frank et al., 2001) and is dependent on postsynaptic cortical activity during sleep (Jha et al., 2005; Frank et al., 2006). Because this type of plasticity involves both synaptic weakening (of deprived eye inputs) and synaptic strengthening (of nondeprived eye inputs) within the visual cortex (Mioche and Singer, 1989), it is possible that either, or both, of these processes occurs during the sleep-dependent ODP consolidation. However, evidence for specific strengthening of synapses during sleep is suggested from studies of chick imprinting. In cerebral areas implicated in this form of learning, increases in both postsynaptic density size and glutamate receptor numbers occur at the time during which sleep is required for consolidation (Horn, 2004).

Some of these changes in synaptic strength may be mediated by thalamocortical and neocortical electroencephalogram (EEG) oscillations that occur during non-REM (NREM) sleep (Figure 2a). For example, neuronal firing patterns similar to sleep spindle (12–15 Hz) oscillations during NREM sleep have been shown to potentiate postsynaptic responses in cortical target neurons in vivo and in vitro (Steriade and Timofeev, 2003; Rosanova and Ulrich, 2005). Conversely, slow-wave EEG oscillations (0.5–4 Hz) may promote synaptic weakening (Tononi and Cirelli, 2003; Czarnecki et al., 2007). Whether or not NREM oscillations normally promote synaptic consolidation is unknown, but spindles and slow-wave oscillations are upregulated in cortical circuits involved in learning new tasks during the first hours of subsequent sleep. This upregulation is correlated with subsequent behavioural measures of memory consolidation (Huber et al., 2004; Nishida and Walker, 2007). Moreover, artificially augmenting these oscillations during the first hours of NREM sleep improves hippocampus-dependent declarative memory consolidation during sleep (Marshall et al., 2006). See also Neural Activity and the Development of Brain Circuits, and Sleep

Sleep and Systems Consolidation

  1. Top of page
  2. Introduction
  3. Synaptic and Systems Mechanisms of Memory Consolidation
  4. Sleep and Memory Consolidation
  5. Sleep and Synaptic Consolidation
  6. Sleep and Systems Consolidation
  7. Summary, Caveats and Future Strategies
  8. Does Sleep Facilitate Consolidation through Permissive or Instructive Mechanisms?
  9. Conclusions
  10. References
  11. Further Reading

Time-dependent expansion of memory traces

Systems consolidation involves gradual integration of memory traces through more spatially distributed brain areas over a period of days to months. A hallmark of this process is a corresponding expansion of activation patterns in hippocampal and neocortical structures during recall. Several findings indicate that sleep may facilitate these events (Figure 2b). Two recent brain-imaging studies assessed the effect of a night of post-acquisition sleep or sleep deprivation on explicit memory recall days to months later. Both report that sleep (but not sleep deprivation) aided subsequent recall (2 days or 6 months later, respectively) and led to greater overall activation and/or functional connectivity between hippocampal and neocortical brain areas during recall (Gais et al., 2007; Sterpenich et al., 2007). Similar effects of post-acquisition sleep on an implicit motor skill task have also been reported (Fischer et al., 2005).

Findings from animal studies further suggest that sleep promotes the transfer of memory traces between different brain areas. During sleep, hippocampal and neocortical neurons in rodents and primates ‘replay’ patterns of firing triggered during prior waking (e.g. during maze running). To what extent this reflects actual systems consolidation is not clear, but coordinated replay has been reported in the neocortex and hippocampus (Ji and Wilson, 2007) and between various sensory and motor cortical areas following a sequential reaching task (Hoffman and McNaughton, 2002). In addition, correlational analyses suggest that slow-wave and spindle oscillations can synchronize neural activity across various brain regions (e.g. between neocortex and hippocampus) during NREM sleep (Isomura et al., 2006; Hoffman et al., 2007).

Dependence on kinases and synthetic pathways

A second key feature of systems consolidation is that it requires continued activation of NMDARs, kinase pathways and protein synthesis machinery over a period of days to months (Wang et al., 2006). The possibility that sleep may promote such events (Figure 2d and e) has been minimally explored, but is suggested by several findings. For example, whole-brain measurements show that sleep is associated with an upregulation of many genes important for synaptic plasticity and memory consolidation (Mackiewicz et al., 2007). Many of these genes are likely to be translated during sleep, as sleep promotes cerebral protein synthesis (Ramm and Smith, 1990). More direct evidence for an expansion of a memory trace during sleep via these signalling cascades comes from a study of IEG expression during sleep. In vivo LTP induction in the hippocampus during waking leads to expanding waves of zif268 expression, that gradually propagate throughout the neocortex during subsequent REM sleep episodes (Ribeiro and Nicolelis, 2004). This broadening expression pattern could reflect the activity-dependent consolidation of synaptic changes associated with LTP induction across the neural network during sleep.

Summary, Caveats and Future Strategies

  1. Top of page
  2. Introduction
  3. Synaptic and Systems Mechanisms of Memory Consolidation
  4. Sleep and Memory Consolidation
  5. Sleep and Synaptic Consolidation
  6. Sleep and Systems Consolidation
  7. Summary, Caveats and Future Strategies
  8. Does Sleep Facilitate Consolidation through Permissive or Instructive Mechanisms?
  9. Conclusions
  10. References
  11. Further Reading

In summary, current findings in humans and animals support the theory that sleep promotes memory consolidation. The vast majority of behavioural studies show that sleep promotes explicit and implicit memory consolidation. Investigations using brain imaging in human subjects, and electrophysiological and molecular measures in animals, have revealed an intriguing pattern of neural and intracellular events consistent with this theory. More specifically, sleep immediately following acquisition appears to promote the activation of intracellular cascades necessary for synaptic consolidation and can lead to additional changes in synaptic strength. Human and animal studies further suggest that during sleep, memory traces expand from circuits initially engaged by acquisition (e.g. the hippocampus) to become more distributed across cortical areas over time (Figure 2b, d and e).

Although these findings are impressive when taken together, several caveats regarding their interpretation merit discussion. First, as described earlier, some behavioural and neurophysiological findings are consistent with the theory that sleep promotes memory consolidation. However, it does not always follow that sleep is required for these processes to occur. Without additional experiments, in many cases one cannot rule out that similar consolidation processes might also occur during waking. Second, sleep is associated with intriguing and suggestive changes in neuronal activity, gene transcription and translation; however, in most studies these events are not directly shown to be necessary for memory consolidation. Third, there are few animal models where sleep and waking behaviour, and synaptic and network events, can be studied simultaneously in vivo. Currently, findings at various levels (molecular, neuronal, systems and behavioural) have been collected from different model organisms, leading to a morass of disconnected results. A fourth and final caveat is that sleep – like waking – affects physiology at many levels, and should not be considered a homogeneous brain state. In addition to diverse, large-scale changes in brain activity and consciousness, REM and NREM sleep are associated with distinct profiles of hormone and neurotransmitter release (Figure 2a) and massive changes in transcription and translation across the brain. This diversity makes the isolation of putative causal factors exceedingly difficult. An important future strategy to overcome this issue will be to focus on defining which aspects of sleep are not only necessary, but also sufficient, for consolidation.

Does Sleep Facilitate Consolidation through Permissive or Instructive Mechanisms?

  1. Top of page
  2. Introduction
  3. Synaptic and Systems Mechanisms of Memory Consolidation
  4. Sleep and Memory Consolidation
  5. Sleep and Synaptic Consolidation
  6. Sleep and Systems Consolidation
  7. Summary, Caveats and Future Strategies
  8. Does Sleep Facilitate Consolidation through Permissive or Instructive Mechanisms?
  9. Conclusions
  10. References
  11. Further Reading

A fruitful approach may be to consider whether sleep facilitates memory consolidation through permissive or instructive mechanisms. Instructive mechanisms are generally defined as those that both directly mediate a process, and also provide information about how it should proceed. Thus we define instructive mechanisms of consolidation as those that are sufficient to mediate the potentiation or depression of specific synapses within neural circuits engaged by prior learning. In contrast, permissive mechanisms are defined as those that can facilitate consolidation, but do not inform how it progresses. For example, permissive mechanisms could be required for changes in synaptic strength, but would not determine which specific synapses are strengthened or weakened. Sleep may facilitate memory consolidation through either permissive or instructive mechanisms, or through a combination of permissive and instructive mechanisms, as described in the next sections.

Potential permissive mechanisms

Sleep is associated with large-scale changes in neurotransmission, transcription, translation and neural activity across the brain, which could promote synaptic plasticity without informing how specific synaptic connections should change. For example, expression of genes required for macromolecular biosynthesis and transport, and overall protein synthesis, are preferentially upregulated in the cortex during sleep (Ramm and Smith, 1990; Mackiewicz et al., 2007); these changes likely play a supportive role in synaptic plasticity and memory consolidation (Figure 2c–e). However, an upregulation of synthesis and ‘housekeeping’ activities alone would be insufficient for informing specific synaptic changes. Sleep also leads to widespread changes in neurotransmitter release throughout the cortex (Figure 2a). These changes include increases in acetylcholine release (and decreases in serotonin and noradrenaline release) relative to wakefulness during REM sleep, and decreases in acetylcholine, noradrenaline and serotonin release during NREM sleep (Gottesmann, 1999). Specific changes in the neurotransmitter milieu during sleep may facilitate synaptic plasticity and memory consolidation (Benington and Frank, 2003), but because they appear to be uniform throughout structures like the cortex, they are unlikely to modify specific synapses activated by prior learning. For example, the decrease in overall cholinergic transmission during NREM has long been hypothesized to mediate transfer of memory traces from hippocampus to the neocortex (Hasselmo, 1999). In support of this idea, artificially augmenting cholinergic neurotransmission (by administering a cholinesterase inhibitor) during post-acquisition NREM sleep interferes with some forms of declarative memory consolidation (Gais and Born, 2004). Finally, a recent hypothesis suggests that NREM slow-wave activity downregulates synaptic strength across the entire neocortex (Tononi and Cirelli, 2003). This homeostatic downscaling is hypothesized to offset Hebbian synaptic strengthening that occurs during waking, and presumably, memory formation. Because it is proposed that all cortical synapses are scaled back during sleep – without corresponding instructions for specific changes at specific synapses – this hypothesis is consistent with a permissive mechanism.

Potential instructive mechanisms

As discussed earlier, some features of sleeping brain activity depend on prior experience during wakefulness, and have the potential to specifically activate (and modify) circuits and synapses engaged by prior learning. One potentially instructive mechanism for sleep-dependent consolidation is ‘replay’ of neuronal activity present during wakefulness. Replay during sleep is clearly associated with prior waking experience (Hoffman and McNaughton, 2002; Ji and Wilson, 2007), involves reactivation of specific synaptic connections involved in performance of the learned task, and has unique features that may promote synaptic modifications. For example, replay events during NREM sleep occurs at a faster rate than that observed during actual task performance (Ji and Wilson, 2007). This time compression would increase the frequency of activation for a given set of synapses, and thus may increase their likelihood for potentiation (Benington and Frank, 2003). A second use-dependent feature of NREM sleep, which may represent an instructive mechanism of consolidation, are localized changes in sleep spindles and slow-wave activity in the cortex. Recent findings indicate that the prevalence of these oscillations in specific neural circuits is associated with their activation during prior learning (Huber et al., 2004; Nishida and Walker, 2007), and that their ‘up’ states activate specific synaptic connections between neurons (Luczak et al., 2007).

Conclusions

  1. Top of page
  2. Introduction
  3. Synaptic and Systems Mechanisms of Memory Consolidation
  4. Sleep and Memory Consolidation
  5. Sleep and Synaptic Consolidation
  6. Sleep and Systems Consolidation
  7. Summary, Caveats and Future Strategies
  8. Does Sleep Facilitate Consolidation through Permissive or Instructive Mechanisms?
  9. Conclusions
  10. References
  11. Further Reading

Despite a wealth of behavioural, molecular and electrophysiological findings, a definitive explanation of how sleep influences memory continues to elude scientists. Several important clues, however, have emerged over the past few years. Rather than a simple, passive reduction of internal brain activity and external sensory input, sleep is a time of massive central nervous system activation. Some of these sleep-associated events can influence synaptic plasticity within the hippocampus and neocortex, and in some studies these are correlated with synaptic or systems-level memory consolidation. Because sleep is not a uniform brain state, the precise mechanisms underlying memory consolidation have proven difficult to isolate. However, a first step in addressing this problem will be to determine if putative causal mechanisms are permissive or instructive, as that information will guide more focused future investigations. One major goal for future investigations will be to develop animal models where molecular and neurophysiological events can be studied in parallel with behavioural outcomes. A second important strategy will be to develop simple, repeatable behavioural assays that will clearly identify which aspects of consolidation are influenced by specific aspects of sleep.

References

  1. Top of page
  2. Introduction
  3. Synaptic and Systems Mechanisms of Memory Consolidation
  4. Sleep and Memory Consolidation
  5. Sleep and Synaptic Consolidation
  6. Sleep and Systems Consolidation
  7. Summary, Caveats and Future Strategies
  8. Does Sleep Facilitate Consolidation through Permissive or Instructive Mechanisms?
  9. Conclusions
  10. References
  11. Further Reading