Effects of NREM sleep on cortical synapses

Abstract

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 calcium-permeable AMPA (Alpha-Amino-3-Hydroxy-5-Methyl-4-Isoxazole Propionic Acid) receptors containing the GluA1 subunit, and the phosphorylation of GluA1 at serine 845. We previously found that, in the rat cortex and hippocampus, these markers are lower after 6-8 hours of sleep than after the same time spent awake. Here, we measure GluA1 expression in mouse cortex after 5 hours 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, synaptic GluA1 expression is lower after sleep independent of whether the mice were allowed to enter REM sleep. Moreover, 5 hours of recovery sleep following acute sleep deprivation are enough to renormalize GluA1 expression. In a pilot study in old mice (12 C57BL/B6 male mice, 20-month-old) we also find that GluA1 expression is high after sleep deprivation and low after sleep and recovery sleep. Thus, the renormalization of GluA1 expression crucially relies on NREM sleep and can occur in a few hours of sleep after acute sleep deprivation. Moreover, as measured by GluA1 expression, it appears that sleep-dependent synaptic weakening is unaffected by aging.