week 7 - sleep and memory Flashcards
what is sleep?
A reversible state of sensory disconnection from the environment, often with reduced mobility
what are good (non-eeg) indicators of sleep?
Reduced movement
Howveer, not all mammals. E.g whales sleep one hemisphere at a time so they can keep moving all the time
What is the role of the Cerebral Spinal Cord in sleep
Diffusion of CSF in the extracellular space clears waste products
The rate of CSF diffusion is much increased during wake time
The rate of flow of CSF diffusion during wake is 5% that of in sleep
(Xie et al, 2015)
Q: How does sleep contribute to the clearance of toxic proteins in the brain?
Counter-argument
Sleep enhances the function of the glymphatic system, which clears waste from the brain.
During sleep:
Interstitial space increases by >60%.
Cerebrospinal fluid (CSF) flow improves, leading to more effective clearance of metabolites, including β-amyloid.
β-amyloid is cleared twice as fast during sleep compared to wakefulness.
In Alzheimer’s disease, sleep disruption may impair glymphatic clearance, contributing to protein accumulation and neurodegeneration.
HOWEVER
- A new paper published recently (Miao et al, 2024) found that brain clearance was reduced during sleep and anestesia
:
Tracer was injected directly into the brain parenchyma (not CSF), to precisely track clearance.
Findings showed that clearance of metabolites was faster during wakefulness compared to sleep or anaesthesia.
Under anesthesia (mimicking NREM sleep), clearance was slower (shown using DEX, KET-XYL, PENTO).
During natural sleep, clearance was also reduced compared to wake (panel g).
Challenges prior assumptions that sleep always enhances glymphatic clearance; state-dependent mechanisms may differ by context or clearance pathway.
how do you measure sleep?
EEG
During rem sleep. EEG trace is similar to awake stage. It is paradoxical because it sugests the person is awake. To measure REM sleep we must measure the EMG which measures signals from skeletal muscles. During REM sleep there is muscle atonia so there is very little signal from skeletal muscles.
NREM sleep:
N1: transient, half alseep
N2: K-complexes, spindles. Entering into deeper sleep
N3: Slow wave sleep, high amplitude low frequency oscillations.
How is sleep measured, and what defines different sleep stages?
Sleep is assessed using:
EEG (electroencephalogram): measures brain waves.
EMG (electromyogram): measures muscle tone.
EOG (electrooculogram): tracks eye movements (especially for REM).
Sleep stages:
NREM 1: transition to sleep; theta waves.
NREM 2: light sleep; sleep spindles and K-complexes.
NREM 3 (SWS): deep sleep; slow delta waves (<4 Hz).
REM: resembles wake EEG; associated with vivid dreams and PGO waves (ponto-geniculo-occipital).
whats the difference between human and rodent sleep stage cycles?
cycles are much shorter
how does sleep stages change throguhout the night
We experience more NREM sleep in the first half of the night, and more REM sleep in the second half of the night.
how does sleep impact memory?
sleep is beneficial for all types of memory
consolidating existing memorys and preparing the brain to learn the following day
how does sleep impact declerative memory
lehal et al 2007
positive correlation between words recalleed and sleep
even ultra short naps improved performance
what is the dual process hyptohesis?
plihal and born 1999
NREM facillllitates hippocampus dependent declerative memory consolidation
REM sleep facilitates non-declarative hippocampus independent memory
describe study plihal and born
what is the cost of staying awake
learning comes at a cost
sleep may function to restore the learning potential of the brain
what is the impact of sleep deprivation on ltp and memory?
read abel et al 2013
cAMP-PKA-CREb signalling required for long lasting LTP
reduced levels of CREB-containing genes transcribed during sleep deprivation.
cAMP-degrading phosphodiesterase (pdef_ upregulated by Sleep deprivation
Blocking PDE4 eliminates effect of Sleep deprivation on LTP and restores learning and memory
applied fear conditioning paradigm to rodents that had sleep deprivation
what happens to cortical neurons during sleep deprivation
read vyazovisky et al 2011
when animals are sleep deprived individual neurons in the cortex go into slow oscillatory sleep like states
what is the synaptic homeostasis hyptohesis
tononi and cirelli 2014
sleeps main function is the restoration of learning potential via synaptic homeostasis
when we are awake we make ‘noise synapses’ and when we are asleep we destrengthen these synapses to improve signal to noise ratio
waking = synaptic potentiation. Sleep = synaptic downscaling (de vivo et al, 2017)
what happens to synaptic proteins during sleep
evidence from flies and rodents suggest that synaptic proteins required for memory consolidated are downregulated during sleep - Gilestro et al., Science 2009, vyazovskiy et al, 2008
mechanism for synaptic plasticity during sleep according to SHY - tononi and cirelli 2018
Net (global) synaptic downscaling during NREM
Achieved by removal of AMPAR from spines
Sleep-dependent renormalization seems to spare neurons and/or synapses that are most active during sleep
UP states in slow oscillation of NREM can result in weakening or preservation of spines by a STDP mechanism.
Presence of co-active postsynaptic neurons crucial for preservation (and strengthening?).
Candidate molecules for positive tagging (preservation of spines): GSK3b phosphorylation (inactivates LTD)
Arc and Homer1a may localize in weaker spines and tag those for downscaling during sleep.
What is the glymphatic system, and how does it relate to sleep?
The glymphatic system is a brain-wide waste clearance pathway using glial channels.
During sleep, especially NREM, interstitial space increases by ~60%, enhancing CSF influx and clearance of solutes like β-amyloid (Xie et al., 2013).
Clearance may help prevent neurodegenerative diseases like Alzheimer’s.
But recent work (Miao et al., 2024) suggests state-dependent differences, where clearance can be faster during wake for some solutes—suggesting complexity in clearance mechanisms.
Evidence that sleep improves memory:
- Hippocampal Replay during Sleep
Wilson & McNaughton (1994) recorded from place cells in the hippocampus of rats.
After navigating a maze, those same neurons reactivated in the same sequence during NREM sleep.
This “replay” supports the idea that sleep consolidates spatial memory.
- Human EEG Studies – Spindle and Slow-Wave Activity
Gais et al. (2002): After word-pair learning, sleep spindles (12–15 Hz) during NREM were increased.
The more spindle activity, the better memory retention.
Marshall et al. (2006): Used transcranial stimulation to enhance slow oscillations during sleep → improved word recall.
This shows causality, not just correlation.
- Sleep Deprivation Impairs Memory Consolidation
Yoo et al. (2007): Sleep-deprived participants showed reduced hippocampal activity and impaired memory for learned word lists.
Sleep after learning restored hippocampal-cortical connectivity, aiding consolidation.
- Reactivation Enhances Learning (Targeted Memory Reactivation)
Rasch et al. (2007): Played odour cues during SWS that were associated with prior learning → improved recall.
Suggests that cue reactivation during sleep enhances consolidation.
Flashcard 5: Memory Types and Sleep Stages (Expanded)
NREM (especially slow-wave sleep) consolidates declarative memory (facts and experiences).
REM sleep supports procedural memory (skills, habits) and emotional memory.
The sequential hypothesis (Giuditta, 1995) suggests optimal consolidation when SWS is followed by REM, facilitating both memory types.
How does sleep deprivation affect molecular mechanisms of memory?
Sleep deprivation reduces LTP via disrupted cAMP-PKA-CREB signalling.
Increases PDE4, which breaks down cAMP, impairing plasticity.
Abel et al. (2013): Inhibiting PDE4 (using rolipram) restored memory and LTP in sleep-deprived mice.
Highlights a molecular link between sleep and long-term memory formation.
Q: What is ‘local sleep’ and how does it affect behaviour during wakefulness?
Local sleep refers to the phenomenon where small groups of neurons enter sleep-like states while the animal remains behaviourally awake.
Vyazovskiy et al. (2011): Rats showed reduced performance in a sugar pellet-reaching task when some cortical neurons were offline.
Suggests that cognitive lapses during sleep deprivation may stem from patchy, local neuronal rest.
What is the Synaptic Homeostasis Hypothesis (SHY) and what evidence supports it?
SHY proposes that wakefulness leads to global synaptic potentiation, which increases energy consumption and noise.
Sleep, particularly slow-wave activity, promotes synaptic downscaling:
Reduces synaptic strength without erasing key information.
Restores learning capacity for the next day.
EVIDENCE:
- Molecular and Structural Changes
Vyazovskiy et al. (2008, Science):
Measured miniature EPSCs (mEPSCs) in the cortex after sleep and wake.
Found that mEPSC amplitudes and frequencies were higher after wake and lower after sleep, indicating reduced synaptic strength.
Diering et al. (2017, Science):
In mice, sleep reduced the levels of AMPA receptor subunits (GluA1) at the synapse.
Also observed increased Homer1a, a protein that helps remove AMPARs from silent synapses.
Suggested that sleep triggers a molecular program for synaptic weakening.
- Structural Evidence from Electron Microscopy
de Vivo et al. (2017, Nature Communications):
Used 3D electron microscopy to measure synapse size in mouse cortex after wake and sleep.
Found a ~18% reduction in synaptic spine volume after sleep compared to wake.
This structural shrinkage supports the idea of synaptic downscaling during sleep.
- Functional Evidence from Behaviour and Learning Capacity
Tononi & Cirelli (2014): Review of SHY argues that:
Waking leads to synaptic saturation — more energy use, less signal-to-noise.
Sleep downscaling improves signal-to-noise ratio, allowing new learning the next day.
Rats and humans learn better after sleep, consistent with a reset in synaptic readiness.
- EEG Evidence – Slow Wave Activity (SWA)
SWA (~0.5–4 Hz) is a hallmark of deep NREM sleep.
Tononi and colleagues suggest that higher SWA reflects a greater need for synaptic downscaling, and SWA decreases across the night as downscaling proceeds.
SWA increases after intense learning tasks, suggesting it scales with prior synaptic load.
✅ Summary of SHY Evidence
Type of Evidence Finding
Electrophysiology ↓ mEPSC amplitude after sleep (Vyazovskiy et al.)
Molecular biology ↓ AMPARs and ↑ Homer1a during sleep (Diering et al.)
Anatomy (EM) ↓ Synaptic spine size after sleep (de Vivo et al.)
EEG SWA tracks homeostatic sleep pressure & learning load
Behaviour Learning improves after sleep, impaired after deprivation
Is synaptic downscaling during sleep uniform across all synapses?
No. Evidence suggests selective downscaling:
Weaker or inactive synapses are more likely to be weakened.
Active synapses during NREM UP states may be preserved via spike-timing dependent plasticity (STDP).
Molecular signals like:
Arc: tags weak synapses for removal.
Homer1a: disrupts scaffold proteins at silent synapses.
GSK3β: its phosphorylation state modulates AMPAR retention.
This selectivity supports refined memory retention, not global erasure.
