Effects of non-rapid eye movement sleep on the cortical synaptic expression of GluA1-containing AMPA receptors.

Converging electrophysiological, molecular and ultrastructural evidence supports the hypothesis that sleep promotes a net decrease in excitatory synaptic strength, counteracting the net synaptic potentiation caused by ongoing learning during waking. However, several outstanding questions about sleep-dependent synaptic weakening remain. Here, we address some of these questions by using two established molecular markers of synaptic strength, the levels of the AMPA (alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid) receptors containing the GluA1 subunit and the phosphorylation of GluA1 at serine 845 (p-GluA1(845)). We previously found that, in the rat cortex and hippocampus, these markers are lower after 6-8 h of sleep than after the same time spent awake. Here, we measure GluA1 and p-GluA1(845) levels in synaptosomes of mouse cortex after 5 h of either sleep, sleep deprivation, recovery sleep after sleep deprivation or selective REM sleep deprivation (32 C57BL/B6 adult mice, 16 females). We find that relative to after sleep deprivation, these synaptic markers are lower after sleep independent of whether the mice were allowed to enter REM sleep. Moreover, 5 h of recovery sleep following acute sleep deprivation is enough to renormalize their expression. Thus, the renormalization of GluA1 and p-GluA1(845) expression crucially relies on NREM sleep and can occur in a few hours of sleep after acute sleep deprivation.

Behavioral and cortical arousal from sleep, muscimol-induced coma, and anesthesia by direct optogenetic stimulation of cortical neurons.

The cerebral cortex is widely considered part of the neural substrate of consciousness, but direct causal evidence is missing. Here, we tested in mice whether optogenetic activation of cortical neurons in posterior parietal cortex (PtA) or medial prefrontal cortex (mPFC) is sufficient for arousal from three behavioral states characterized by progressively deeper unresponsiveness: sleep, a coma-like state induced by muscimol injection in the midbrain, and deep sevoflurane-dexmedetomidine anesthesia. We find that cortical stimulation always awakens the mice from both NREM sleep and REM sleep, with PtA requiring weaker/shorter light pulses than mPFC. Moreover, in most cases light pulses produce both cortical activation (decrease in low frequencies) and behavioral arousal (recovery of the righting reflex) from brainstem coma, as well as cortical activation from anesthesia. These findings provide evidence that direct activation of cortical neurons is sufficient for behavioral and/or cortical arousal from sleep, brainstem coma, and anesthesia.

Consciousness and sleep.

Sleep is a universal, essential biological process. It is also an invaluable window on consciousness. It tells us that consciousness can be lost but also that it can be regained, in all its richness, when we are disconnected from the environment and unable to reflect. By considering the neurophysiological differences between dreaming and dreamless sleep, we can learn about the substrate of consciousness and understand why it vanishes. We also learn that the ongoing state of the substrate of consciousness determines the way each experience feels regardless of how it is triggered-endogenously or exogenously. Dreaming consciousness is also a window on sleep and its functions. Dreams tell us that the sleeping brain is remarkably lively, recombining intrinsic activation patterns from a vast repertoire, freed from the requirements of ongoing behavior and cognitive control.

Cortical Tonic Inhibition Gates the Expression of Spike-and-Wave Discharges Associated with Absence Epilepsy.

OBJECTIVE: Absence seizures result from aberrant thalamocortical processing that confers synchronous, bilateral spike-and-wave discharges (SWDs) and behavioral arrest. Previous work has demonstrated that SWDs can result from enhanced thalamic tonic inhibition, consistent with the mechanism of first-line antiabsence drugs that target thalamic low-voltage-activated calcium channels. However, nearly half of patients with absence epilepsy are unresponsive to first-line medications. In this study we evaluated the role of cortical tonic inhibition and its manipulation on absence seizure expression. METHODS: We used video-electroencephalogram (EEG) monitoring to show that mice with a γ-aminobutyric acid type A (GABAA) receptor mutation (γ2R43Q) display absence seizures. Voltage-clamp recordings in brain slices from wild type and γ2R43Q mice were used to evaluate the amount of tonic inhibition and its selective pharmacological modulation. Finally, we determined whether modulating tonic inhibition controls seizure expression. RESULTS: γ2R43Q mice completely lack tonic inhibition in principal neurons of both layer 2/3 cortex and ventrobasal thalamus. Blocking cortical tonic inhibition in wild type mice is sufficient to elicit SWDs. Tonic inhibition in slices from γ2R43Q mice could be rescued in a dose-dependent fashion by the synthetic neurosteroid ganaxolone. Low-dose ganaxolone suppressed seizures in γ2R43Q mice. CONCLUSIONS: Our data suggest that reduced cortical tonic inhibition promotes absence seizures and that normal function can be restored via selective pharmacological rescue. These results, together with previous findings, suggest that deviations of tonic inhibition either above or below an optimal set point can contribute to absence epilepsy. Returning the thalamocortical system to this set point may provide a novel treatment for refractory absence epilepsy.

Sleep/wake changes in perturbational complexity in rats and mice.

In humans, the level of consciousness is assessed by quantifying the spatiotemporal complexity of cortical responses using Perturbational Complexity Index (PCI) and related PCIst (st, state transitions). Here we validate PCIst in freely moving rats and mice by showing that it is lower in NREM sleep and slow wave anesthesia than in wake or REM sleep, as in humans. We then show that (1) low PCIst is associated with the occurrence of an OFF period of neuronal silence; (2) stimulation of deep, but not superficial, cortical layers leads to reliable PCIst changes across sleep/wake and anesthesia; (3) consistent PCIst changes are independent of which single area is being stimulated or recorded, except for recordings in mouse prefrontal cortex. These experiments show that PCIst can reliably measure vigilance states in unresponsive animals and support the hypothesis that it is low when an OFF period disrupts causal interactions in cortical networks.

Ultrastructural effects of sleep and wake on the parallel fiber synapses of the cerebellum.

Multiple evidence in rodents shows that the strength of excitatory synapses in the cerebral cortex and hippocampus is greater after wake than after sleep. The widespread synaptic weakening afforded by sleep is believed to keep the cost of synaptic activity under control, promote memory consolidation, and prevent synaptic saturation, thus preserving the brain's ability to learn day after day. The cerebellum is highly plastic and the Purkinje cells, the sole output neurons of the cerebellar cortex, are endowed with a staggering number of excitatory parallel fiber synapses. However, whether these synapses are affected by sleep and wake is unknown. Here, we used serial block face scanning electron microscopy to obtain the full 3D reconstruction of more than 7000 spines and their parallel fiber synapses in the mouse posterior vermis. This analysis was done in mice whose cortical and hippocampal synapses were previously measured, revealing that average synaptic size was lower after sleep compared to wake with no major changes in synapse number. Here, instead, we find that while the average size of parallel fiber synapses does not change, the number of branched synapses is reduced in half after sleep compared to after wake, corresponding to ~16% of all spines after wake and ~8% after sleep. Branched synapses are harbored by two or more spines sharing the same neck and, as also shown here, are almost always contacted by different parallel fibers. These findings suggest that during wake, coincidences of firing over parallel fibers may translate into the formation of synapses converging on the same branched spine, which may be especially effective in driving Purkinje cells to fire. By contrast, sleep may promote the off-line pruning of branched synapses that were formed due to spurious coincidences.

Increase in NREM sleep slow waves following injections of sodium oxybate in the mouse cerebral cortex and the role of somatostatin-positive interneurons.

The systemic administration of sodium oxybate (SXB), the sodium salt of gamma-hydroxybutyric acid, promotes slow wave activity (SWA, 0.5-4 Hz EEG power) and increases non-rapid eye movement (NREM) sleep. These effects are mediated by the widely expressed GABAb receptors, and thus, the brain areas targeted by SXB remain unclear. Because slow waves are mainly a cortical phenomenon, we tested here whether systemic SXB promotes SWA by acting directly on the cortex. Moreover, because somatostatin (SOM) + cortical interneurons play a key role in SWA generation, we also assessed their contribution to the effects of SXB. In adult SOM-Cre mice, the injection of SXB in left secondary motor cortex increased SWA during NREM sleep in the first 30 min post-injection (11 mice: either sex). SWA, the amplitude and frequency of the slow waves, and the frequency of the OFF periods increased ipsilaterally and contralaterally to the SXB injection in frontal and parietal cortex. All these changes disappeared when the intracortical injection of SXB was preceded by the chemogenetic inhibition of the SOM+ cells. Thus, SXB may promote the slow waves of NREM sleep, at least in part, by acting directly on the cortex, and this effect involves GABAergic SOM+ interneurons. Our working hypothesis is that SXB potentiates the ability of these cells to inhibit all other cortical cell types via a GABAb mechanism, thus promoting the transition from ON to OFF periods during NREM sleep.

Multiple traces and altered signal-to-noise in systems consolidation: Evidence from the 7T fMRI Natural Scenes Dataset.

The brain mechanisms of memory consolidation remain elusive. Here, we examine blood-oxygen-level-dependent (BOLD) correlates of image recognition through the scope of multiple influential systems consolidation theories. We utilize the longitudinal Natural Scenes Dataset, a 7-Tesla functional magnetic resonance imaging human study in which ∼135,000 trials of image recognition were conducted over the span of a year among eight subjects. We find that early- and late-stage image recognition associates with both medial temporal lobe (MTL) and visual cortex when evaluating regional activations and a multivariate classifier. Supporting multiple-trace theory (MTT), parts of the MTL activation time course show remarkable fit to a 20-y-old MTT time-dynamical model predicting early trace intensity increases and slight subsequent interference (R2 > 0.90). These findings contrast a simplistic, yet common, view that memory traces are transferred from MTL to cortex. Next, we test the hypothesis that the MTL trace signature of memory consolidation should also reflect synaptic "desaturation," as evidenced by an increased signal-to-noise ratio. We find that the magnitude of relative BOLD enhancement among surviving memories is positively linked to the rate of removal (i.e., forgetting) of competing traces. Moreover, an image-feature and time interaction of MTL and visual cortex functional connectivity suggests that consolidation mechanisms improve the specificity of a distributed trace. These neurobiological effects do not replicate on a shorter timescale (within a session), implicating a prolonged, offline process. While recognition can potentially involve cognitive processes outside of memory retrieval (e.g., re-encoding), our work largely favors MTT and desaturation as perhaps complementary consolidative memory mechanisms.

Sleep and wake in a model of the thalamocortical system with Martinotti cells.

The mechanisms leading to the alternation between active (UP) and silent (DOWN) states during sleep slow waves (SWs) remain poorly understood. Previous models have explained the transition to the DOWN state by a progressive failure of excitation because of the build-up of adaptation currents or synaptic depression. However, these models are at odds with recent studies suggesting a role for presynaptic inhibition by Martinotti cells (MaCs) in generating SWs. Here, we update a classical large-scale model of sleep SWs to include MaCs and propose a different mechanism for the generation of SWs. In the wake mode, the network exhibits irregular and selective activity with low firing rates (FRs). Following an increase in the strength of background inputs and a modulation of synaptic strength and potassium leak potential mimicking the reduced effect of acetylcholine during sleep, the network enters a sleep-like regime in which local increases of network activity trigger bursts of MaC activity, resulting in strong disfacilitation of the local network via presynaptic GABAB1a -type inhibition. This model replicates findings on slow wave activity (SWA) during sleep that challenge previous models, including low and skewed FRs that are comparable between the wake and sleep modes, higher synchrony of transitions to DOWN states than to UP states, the possibility of triggering SWs by optogenetic stimulation of MaCs, and the local dependence of SWA on synaptic strength. Overall, this work points to a role for presynaptic inhibition by MaCs in the generation of DOWN states during sleep.

Identification of ultrastructural signatures of sleep and wake in the fly brain.

The cellular consequences of sleep loss are poorly characterized. In the pyramidal neurons of mouse frontal cortex, we found that mitochondria and secondary lysosomes occupy a larger proportion of the cytoplasm after chronic sleep restriction compared to sleep, consistent with increased cellular burden due to extended wake. For each morphological parameter, the within-animal variance was high, suggesting that the effects of sleep and sleep loss vary greatly among neurons. However, the analysis was based on 4-5 mice/group and a single section/cell. Here, we applied serial block-face scanning electron microscopy to identify signatures of sleep and sleep loss in the Drosophila brain. Stacks of images were acquired and used to obtain full 3D reconstructions of the cytoplasm and nucleus of 263 Kenyon cells from adult flies collected after a night of sleep (S) or after 11 h (SD11) or 35 h (SD35) of sleep deprivation (9 flies/group). Relative to S flies, SD35 flies showed increased density of dark clusters of chromatin and Golgi apparata and a trend increase in the percent of cell volume occupied by mitochondria, consistent with increased need for energy and protein supply during extended wake. Logistic regression models could assign each neuron to the correct experimental group with good accuracy, but in each cell, nuclear and cytoplasmic changes were poorly correlated, and within-fly variance was substantial in all experimental groups. Together, these results support the presence of ultrastructural signatures of sleep and sleep loss but underscore the complexity of their effects at the single-cell level.

Measuring Stimulus-Evoked Neurophysiological Differentiation in Distinct Populations of Neurons in Mouse Visual Cortex.

Despite significant progress in understanding neural coding, it remains unclear how the coordinated activity of large populations of neurons relates to what an observer actually perceives. Since neurophysiological differences must underlie differences among percepts, differentiation analysis-quantifying distinct patterns of neurophysiological activity-has been proposed as an "inside-out" approach that addresses this question. This methodology contrasts with "outside-in" approaches such as feature tuning and decoding analyses, which are defined in terms of extrinsic experimental variables. Here, we used two-photon calcium imaging in mice of both sexes to systematically survey stimulus-evoked neurophysiological differentiation (ND) in excitatory neuronal populations in layers (L)2/3, L4, and L5 across five visual cortical areas (primary, lateromedial, anterolateral, posteromedial, and anteromedial) in response to naturalistic and phase-scrambled movie stimuli. We find that unscrambled stimuli evoke greater ND than scrambled stimuli specifically in L2/3 of the anterolateral and anteromedial areas, and that this effect is modulated by arousal state and locomotion. By contrast, decoding performance was far above chance and did not vary substantially across areas and layers. Differentiation also differed within the unscrambled stimulus set, suggesting that differentiation analysis may be used to probe the ethological relevance of individual stimuli.

The why and how of sleep-dependent synaptic down-selection.

Sleep requires that we disconnect from the environment, losing the ability to promptly respond to stimuli. There must be at least one essential function that justifies why we take this risk every day, and that function must depend on the brain being offline. We have proposed that this function is to renormalize synaptic weights after learning has led to a net increase in synaptic strength in many brain circuits. Without this renormalization, synaptic activity would become energetically too expensive and saturation would prevent new learning. There is converging evidence from molecular, electrophysiological, and ultrastructural experiments showing a net increase in synaptic strength after the major wake phase, and a net decline after sleep. The evidence also suggests that sleep-dependent renormalization is a smart process of synaptic down-selection, comprehensive and yet specific, which could explain the many beneficial effects of sleep on cognition. Recently, a key molecular mechanism that allows broad synaptic weakening during sleep was identified. Other mechanisms still being investigated should eventually explain how sleep can weaken most synapses but afford protection to some, including those directly activated by learning. That synaptic down-selection takes place during sleep is by now established; why it should take place during sleep has a plausible explanation; how it happens is still work in progress.

Extended Visual Sequence Learning Leaves a Local Trace in the Spontaneous EEG.

We have previously demonstrated that, in rested subjects, extensive practice in a motor learning task increased both electroencephalographic (EEG) theta power in the areas involved in learning and improved the error rate in a motor test that shared similarities with the task. A nap normalized both EEG and performance changes. We now ascertain whether extensive visual declarative learning produces results similar to motor learning. Thus, during the morning, we recorded high-density EEG in well rested young healthy subjects that learned the order of different visual sequence task (VSEQ) for three one-hour blocks. Afterward, a group of subjects took a nap and another rested quietly. Between each VSEQ block, we recorded spontaneous EEG (sEEG) at rest and assessed performance in a motor test and a visual working memory test that shares similarities with VSEQ. We found that after the third block, VSEQ induced local theta power increases in the sEEG over a right temporo-parietal area that was engaged during the task. This local theta increase was preceded by increases in alpha and beta power over the same area and was paralleled by performance decline in the visual working memory test. Only after the nap, VSEQ learning rate improved and performance in the visual working memory test was restored, together with partial normalization of the local sEEG changes. These results suggest that intensive learning, like motor learning, produces local theta power increases, possibly reflecting local neuronal fatigue. Sleep may be necessary to resolve neuronal fatigue and its effects on learning and performance.

Effects of Severe Sleep Disruption on the Synaptic Ultrastructure of Young Mice.

There is molecular, electrophysiological, and ultrastructural evidence that a net increase in synaptic strength occurs in many brain circuits during spontaneous wake (SW) or short sleep deprivation, reflecting ongoing learning. Sleep leads instead to a broad but selective weakening of many forebrain synapses, thus preventing synaptic saturation and decreasing the energy cost of synaptic activity. Whether synaptic potentiation can persist or further increase after long sleep deprivation is unknown. Whether synaptic renormalization can occur during chronic sleep restriction (CSR) is also unknown. Here, we addressed these questions by measuring an established ultrastructural measure of synaptic strength, the axon-spine interface (ASI), in the primary motor cortex (M1) of (1) one-month-old adolescent mice CSR using a paradigm that decreases NREM and REM sleep by two/thirds; (2) in two-week-old mouse pups sleep deprived for 15 h, or allowed afterward to recover for 16 h. Both groups were compared with mice of the same age that were asleep or awake for a few hours (both sexes). The ASI size of CSR mice (n = 3) was comparable to that measured after SW or short sleep deprivation and larger than after sleep (n = 4/group). In pups, the ASI size increased after short sleep loss (n = 3) relative to sleep (n = 4), fell below sleep levels after long sleep deprivation (n = 4), and remained low after recovery (n = 3). Long sleep deprived pups also lost some weight. These results suggest that (1) severe sleep restriction is incompatible with synaptic renormalization; (2) very young mice cannot maintain high synaptic strength during prolonged wake.

Net decrease in spine-surface GluA1-containing AMPA receptors after post-learning sleep in the adult mouse cortex.

The mechanisms by which sleep benefits learning and memory remain unclear. Sleep may further strengthen the synapses potentiated by learning or promote broad synaptic weakening while protecting the newly potentiated synapses. We tested these ideas by combining a motor task whose consolidation is sleep-dependent, a marker of synaptic AMPA receptor plasticity, and repeated two-photon imaging to track hundreds of spines in vivo with single spine resolution. In mouse motor cortex, sleep leads to an overall net decrease in spine-surface GluA1-containing AMPA receptors, both before and after learning. Molecular changes in single spines during post-learning sleep are correlated with changes in performance after sleep. The spines in which learning leads to the largest increase in GluA1 expression have a relative advantage after post-learning sleep compared to sleep deprivation, because sleep weakens all remaining spines. These results are obtained in adult mice, showing that sleep-dependent synaptic down-selection also benefits the mature brain.

Neural fatigue due to intensive learning is reversed by a nap but not by quiet waking.

Do brain circuits become fatigued due to intensive neural activity or plasticity? Is sleep necessary for recovery? Well-rested subjects trained extensively in a visuo-motor rotation learning task (ROT) or a visuo-motor task without rotation learning (MOT), followed by sleep or quiet wake. High-density electroencephalography showed that ROT training led to broad increases in EEG power over a frontal cluster of electrodes, with peaks in the theta (mean ± SE: 24% ± 6%, p = 0.0013) and beta ranges (10% ± 3%, p = 0.01). These traces persisted in the spontaneous EEG (sEEG) between sessions (theta: 42% ± 8%, p = 0.0001; beta: 35% ± 7%, p = 0.002) and were accompanied by increased errors in a motor test with kinematic characteristics and neural substrates similar to ROT (81.8% ± 0.8% vs. 68.2% ± 2.3%; two-tailed paired t-test: p = 0.00001; Cohen's d = 1.58), as well as by score increases of subjective task-specific fatigue (4.00 ± 0.39 vs. 5.36 ± 0.39; p = 0.0007; Cohen's d = 0.60). Intensive practice with MOT did not affect theta sEEG or the motor test. A nap, but not quiet wake, induced a local sEEG decrease of theta power by 33% (SE: 8%, p = 0.02), renormalized test performance (70.9% ± 2.9% vs 79.1% ± 2.7%, p = 0.018, Cohen's d = 0.85), and improved learning ability in ROT (adaptation rate: 71.2 ± 1.2 vs. 73.4 ± 0.9, p = 0.024; Cohen's d = 0.60). Thus, sleep is necessary to restore plasticity-induced fatigue and performance.

Prior Practice Affects Movement-Related Beta Modulation and Quiet Wake Restores It to Baseline.

Beta oscillations (13.5-25 Hz) over the sensorimotor areas are characterized by a power decrease during movement execution (event-related desynchronization, ERD) and a sharp rebound after the movement end (event-related synchronization, ERS). In previous studies, we demonstrated that movement-related beta modulation depth (peak ERS-ERD) during reaching increases within 1-h practice. This increase may represent plasticity processes within the sensorimotor network. If so, beta modulation during a reaching test should be affected by previous learning activity that engages the sensorimotor system but not by learning involving other systems. We thus recorded high-density EEG activity in a group of healthy subjects performing three 45-min blocks of motor adaptation task to a visually rotated display (ROT) and in another performing three blocks of visual sequence-learning (VSEQ). Each block of either ROT or VSEQ was followed by a simple reaching test (mov) without rotation. We found that beta modulation depth increased with practice across mov tests. However, such an increase was greater in the group performing ROT over both the left and frontal areas previously involved in ROT. Importantly, beta modulation values returned to baseline values after a 90-min of either nap or quiet wake. These results show that previous practice leaves a trace in movement-related beta modulation and therefore such increases are cumulative. Furthermore, as sleep is not necessary to bring beta modulation values to baseline, they could reflect local increases of neuronal activity and decrease of energy and supplies.

Effects of sleep and waking on the synaptic ultrastructure.

We summarize here several studies performed in our laboratory, mainly using serial block-face scanning electron microscopy (SBEM), to assess how sleep, spontaneous waking and short sleep deprivation affect the size and number of synapses in the cerebral cortex and hippocampus. With SBEM, we reconstructed thousands of cortical and hippocampal excitatory, axospinous synapses and compared the distribution of their size after several hours of sleep relative to several hours of waking. Because stronger synapses are on average also bigger, the goal was to test a prediction of the synaptic homeostasis hypothesis, according to which overall synaptic strength increases during waking, owing to ongoing learning, and needs to be renormalized during sleep, to avoid saturation and to benefit memory consolidation and integration. Consistent with this hypothesis, we found that the size of the axon-spine interface (ASI), a morphological measure of synaptic strength, was on average smaller after sleep, but with interesting differences between primary cortex and the CA1 region of the hippocampus. In two-week-old mouse pups, the decline in ASI size after sleep was larger, and affected more cortical synapses, compared with one-month-old adolescent mice, suggesting that synaptic renormalization during sleep may be especially important during early development. This work is still in progress and other brain areas need to be tested after sleep, acute sleep loss and chronic sleep restriction. Still, the current results show that a few hours of sleep or waking lead to significant changes in synaptic morphology that can be linked to changes in synaptic efficacy. This article is part of the Theo Murphy meeting issue 'Memory reactivation: replaying events past, present and future'.

The evolving view of replay and its functions in wake and sleep.

The term hippocampal replay originally referred to the temporally compressed reinstantiation, during rest, of sequential neural activity observed during prior active wake. Since its description in the 1990s, hippocampal replay has often been viewed as the key mechanism by which a memory trace is repeatedly rehearsed at high speeds during sleep and gradually transferred to neocortical circuits. However, the methods used to measure the occurrence of replay remain debated, and it is now clear that the underlying neural events are considerably more complicated than the traditional narratives had suggested. "Replay-like" activity happens during wake, can play out in reverse order, may represent trajectories never taken by the animal, and may have additional functions beyond memory consolidation, from learning values and solving the problem of credit assignment to decision-making and planning. Still, we know little about the role of replay in cognition, and to what extent it differs between wake and sleep. This may soon change, however, because decades-long efforts to explain replay in terms of reinforcement learning (RL) have started to yield testable predictions and possible explanations for a diverse set of observations. Here, we (1) survey the diverse features of replay, focusing especially on the latest findings; (2) discuss recent attempts at unifying disparate experimental results and putatively different cognitive functions under the banner of RL; (3) discuss methodological issues and theoretical biases that impede progress or may warrant a partial revaluation of the current literature, and finally; (4) highlight areas of considerable uncertainty and promising avenues of inquiry.

Sleep and synaptic down-selection.

The synaptic homeostasis hypothesis (SHY) proposes that sleep is an essential process needed by the brain to maintain the total amount of synaptic strength under control. SHY predicts that by the end of a waking day the synaptic connections of many neural circuits undergo a net increase in synaptic strength due to ongoing learning, which is mainly mediated by synaptic potentiation. Stronger synapses require more energy and supplies and are prone to saturation, creating the need for synaptic renormalization. Such renormalization should mainly occur during sleep, when the brain is disconnected from the environment and neural circuits can be broadly reactivated off-line to undergo a systematic but specific synaptic down-selection. In short, according to SHY sleep is the price to pay for waking plasticity, to avoid runaway potentiation, decreased signal-to-noise ratio, and impaired learning due to saturation. In this review, we briefly discuss the rationale of the hypothesis and recent supportive ultrastructural evidence obtained in our laboratory. We then examine recent studies by other groups showing the causal role of cortical slow waves and hippocampal sharp waves/ripples in sleep-dependent down-selection of neural activity and synaptic strength. Finally, we discuss some of the molecular mechanisms that could mediate synaptic weakening during sleep.