Hypothesis

Sleep is not merely a clearance window; it is an active triage mechanism in which the glymphatic system selectively tags synaptic complexes for retention or elimination based on recent neuronal activity. We propose that during slow‑wave sleep, astrocytic AQP4‑dependent cerebrospinal fluid influx carries activity‑dependent ubiquitin‑like modifiers (e.g., ISG15) from the interstitial fluid into perivascular spaces, where they label post‑synaptic densities. Synapses bearing high levels of these tags are protected from autophagic removal, whereas low‑tagged synapses are earmarked for lysosomal degradation. This activity‑dependent tagging couples synaptic homeostasis to the restorative functions of sleep, linking glymphatic flux to long‑term memory storage.

Mechanistic Model

• Slow‑wave oscillations drive AQP4 polarization: During SWS, extracellular K+ surges trigger astrocytic depolarization, increasing AQP4‑mediated CSF‑ISF exchange [https://doi.org/10.1101/2024.08.30.610454].
• Activity‑dependent ubiquitin‑like signaling: Neuronal activity during wakefulness promotes calcium‑dependent activation of the TRIM25‑ISG15 pathway, conjugating ISG15 to synaptic proteins (e.g., PSD‑95). ISG15‑modified complexes are more resistant to autophagy [https://pmc.ncbi.nlm.nih.gov/articles/PMC6576168/].
• Glymphatic transport of ISG15: The convective CSF flow carries free ISG15 and ISG15‑conjugated peptides into perivascular lymphatics, where they encounter synaptic debris. High ISG15 occupancy shields synapses from LC3‑mediated phagocytosis; low occupancy permits ubiquitination and lysosomal targeting.
• Feedback to sleep architecture: Accumulation of undegraded ISG15‑conjugated proteins triggers astrocytic release of adenosine, enhancing SWS pressure and thus reinforcing the triage cycle.

Testable Predictions

1. ISG15 levels in cortical interstitial fluid will rise during SWS and correlate with AQP4‑dependent CSF influx measured by intrathecal contrast‑enhanced MRI.
2. Genetic reduction of astrocytic AQP4 will blunt the sleep‑dependent increase in ISG15‑tagged PSD‑95 and lead to preferential loss of high‑activity spines after sleep deprivation.
3. Pharmacological augmentation of ISG15 conjugation (e.g., using a TRIM25 activator) will rescue spine density deficits in AQP4‑knockout mice subjected to chronic sleep fragmentation.
4. Optogenetic enhancement of slow‑wave oscillations will increase perivascular ISG15 flux and improve performance on spatial memory tasks, an effect abolished by ISG15 knock‑down.

Experimental Approach

• Use two‑photon microscopy in Thy1‑YFP mice to track dendritic spine turnover across sleep‑wake cycles while manipulating AQP4 with conditional knockout or pharmacologic agonist (e.g., TGN‑020).
• Collect CSF via cisterna puncture at defined circadian times; quantify free and conjugated ISG15 by immunoblot and mass spectrometry.
• Combine EEG‑EMG recordings to confirm SWS peaks; apply closed‑loop optogenetic drive of thalamic reticular nucleus to boost slow waves.
• Assess memory using Morris water maze; correlate spine retention rates with behavioral performance.
• Statistical analysis: mixed‑effects models testing interaction between genotype/treatment, sleep stage, and spine fate (p<0.05 considered significant).