Sleep deprivation leads to an increase in ‘sleep need’ — a homeostatic drive to sleep — but the molecular substrates of sleep need are not known. Wang et al. now show in two different mouse models that increased sleep need is associated with the hyperphosphorylation of a set of 80 proteins in the brain, most of which are synaptic proteins.

Credit: Juniors Bildarchiv GmbH/Alamy Stock Photo

To dissect the effects of increased sleep need from time spent awake, the authors used sleep-deprived mice and ‘sleepy’ mice, which carry a mutation in the gene encoding salt-inducible kinase 3 (SIK3), an AMP-activated protein kinase (AMPK) family member. These Sik3Slp/+ mice spend less time awake than do wild-type mice, reflecting a consistently high sleep need.

The authors analysed proteomic and phosphoproteomic data from whole-brain lysates and used these to make three group comparisons: Sik3Slp/+ mice versus wild-type mice; mice that had been sleep-deprived for 6 hours versus mice that slept for 6 hours; and sleep-deprived mice versus mice that were sleep-deprived for 6 hours and then had 3 hours of recovery sleep. Whereas the levels of most proteins did not seem to be altered with sleep need, there were considerable changes in the levels of different phosphoproteins between the Sik3Slp/+ mice and wild-type mice, and between the sleep-deprived and slept or recovery-slept mice. Changes in the levels of 329 phosphopeptides were common to all three comparisons, suggesting that increased sleep need may be accompanied by changes in phosphorylation in the brain.

Next, the authors investigated sleep-need-associated changes in the phosphorylation state of different proteins. Many more proteins showed increases than decreases in phosphorylation in Sik3Slp/+ mice. Importantly, 80 proteins exhibited increases in phosphorylation state in the sleep-deprived mice and the Sik3Slp/+ mice; the authors term these ‘sleep-need-index phosphoproteins’ (SNIPPs). Strikingly, 69 of these 80 SNIPPs are synaptic proteins, and many SNIPPs showed progressive increases in phosphorylation with time deprived of sleep.

80 proteins exhibited increases in phosphorylation state in the sleep-deprived mice and the Sik3Slp/+ mice

Using immunoprecipitation and mass spectrometry of brain lysates from mice expressing labelled SIK3 variants, the authors found that the SIK3Slp variant interacts with synaptic proteins more than does wild-type SIK3. Moreover, SIK3Slp interacts more than wild-type SIK3 with 28 of the SNIPPs. Thus, SIK3Slp may more readily bind its substrates and thus enhance the phosphorylation of SNIPPs. In support of this, intracerebroventricular injection of Sik3Slp/+ or sleep-deprived mice with a pan-SIK inhibitor reduced SNIPP phosphorylation, and also inhibited slow-wave activity during non-rapid eye movement (NREM) sleep. Thus, lowering SNIPP phosphorylation may decrease sleep need.

Together, these results suggest that SNIPP phosphorylation may represent a molecular substrate or promoter of sleep need.