The reason we need sleep seems clear: without sleep, we become tired, and irritable and our brain functions less well. Conversely, after a good night's sleep, the brain and body feel refreshed. But why does the brain need to disconnect from the environment for hours every day? What is restored by sleep has proven difficult to explain.
Sleep efficiency in older adults. Sleep deprivation increases amyloid-β (Aβ) concentrations in the interstitial fluid of experimental animal models and in cerebrospinal fluid in humans, while increased sleep decreases Aβ. Sleep abnormalities may therefore represent a risk factor for neurodegeneration.
It has recently been shown that sleep is likely a time for clearing waste in the brain or repairing damaged cells. Circadian sleep-wake rhythm disorders are strong predictors of institutionalization.
During a period of wakefulness, coping with the environment requires increasing the number and strength of connections at the synapses between neurons in the brain. This increased activity increases cellular requirements for energy and materials, leading to cellular stress, a major factor in neurodegenerative diseases, and forcing changes in supporting cells such as glial cells, while hindering learning. During sleep, this synaptic activity is decreased which helps restore cellular health and increase plasticity through negative selection of synapses. This may also explain the benefits of sleep for memory acquisition, consolidation, and integration.
In other words, for the theory of synaptic homeostasis, sleep is “the price we pay for our learning and memory abilities”. Increased synaptic activity reduces the selectivity of neuronal responses and limits the ability to learn. By renormalizing synaptic activity, sleep reduces the plasticity burden of neurons and other cells while restoring neuronal selectivity and the ability to learn, and the consolidation and integration of memories.
A new study has just confirmed this, at least in translucent zebrafish larvae. However, these results obtained with zebrafish larvae should only be extrapolated with great caution to humans, but it is nevertheless an interesting discovery for fundamental neuroscience. The study authors used in vivo synaptic labeling tools in larval zebrafish to image the same neurons and their synapses repeatedly over long periods, allowing them to map the synapse changes of a single neuron in states of sleep and wakefulness. In effect, this meant genetically modifying these neurons to allow fluorescence upon firing.
By tracking the synapses of single tectal neurons across sleep-wake states and circadian time, scientists resolve several outstanding questions about the magnitude, universality, and mechanisms of sleep-related plasticity.
They show that synaptic dynamics are present in many cells on average, but when examined neuron by neuron, more diverse patterns of synaptic changes are revealed. These observations may explain some discrepancies between previous studies on the synaptic homeostasis hypothesis, as such single-neuron synaptic dynamics were not captured by one-time snapshots of synapse number or function at the population level.
The authors also found that sleep-related synapse loss depends on molecular signals related to elevated sleep pressure and, notably, also reflects slow-wave activity by occurring primarily at the beginning of the sleep period. This finding raises the question of whether sleep periods associated with low sleep pressure, such as in the second half of the night, play an additional role in non-synaptic remodeling.
Despite all this work was carried out on zebrafish larvae equipped with genetically modified neurons. Thus, any extrapolation to mammals, without even thinking about humans, is highly speculative. Still, it's another clue that sleep is an integral part of good neuronal health, probably along with vascular health and physical activity.