Hypothesis

Intermittent fasting activates AMPK, which directly phosphorylates the V‑ATPase subunit ATP6V0D1 to boost lysosomal acidification during sleep. This heightened acid load shifts the autophagic machinery from bulk degradation to a ubiquitin‑dependent selective mode that tags synaptic proteins (e.g., PSD‑95, Synapsin‑1) for removal only when they are marked with K63‑linked ubiquitin chains accumulated during wakefulness. Consequently, sleep becomes an active editing step where the brain decides which synaptic architectures persist based on their ubiquitination state, and fasting sets the proteolytic threshold for this decision.

Mechanistic Basis

• AMPK‑V‑ATPase link: AMPK activation by fasting phosphorylates ATP6V0D1, increasing V‑ATPase assembly and proton pumping into lysosomes (4). Acidic lysosomes enhance cathepsin activity, favoring cleavage of ubiquitin‑linked substrates over long‑lived organelles.
• Ubiquitin code readout: During wakefulness, synaptic activity drives K63‑linked ubiquitination of postsynaptic scaffolds (5). Acidic lysosomes possess ubiquitin‑binding adaptors (e.g., p62/SQSTM1) that exhibit higher affinity for K63 chains at low pH, coupling synaptic tagging to autophagic capture.
• Glymphatic‑astrocyte coupling: Low norepinephrine during slow‑wave sleep expands interstitial space, boosting CSF influx (1, 3). Astrocytic AQP4 channels, regulated by lactate shifts from fasting‑induced ketosis, facilitate perivascular flow that delivers ubiquitin‑tagged synaptic fragments to lysosomal hotspots in perivascular astrocytes.
• Selective outcome: Only synapses bearing sufficient ubiquitin load are engulfed; less‑tagged synapses escape degradation and undergo homeostatic scaling, preserving network integrity while removing maladaptive connections.

Testable Predictions

1. Phospho‑ATP6V0D1 rises in lysosomes during sleep in mice subjected to 16‑h intermittent fasting versus ad libitum controls, measurable by subcellular Western blot or proximity ligation assay (2).
2. Lysosomal pH drops (measured with LysoSensor dyes) specifically during NREM sleep in fasted animals, and this acidification is abolished by AMPK inhibition (Compound C) or V‑ATPase knockout in astrocytes.
3. Ubiquitinated PSD‑95 levels decline during sleep only when fasting‑induced AMPK activity is present; blocking AMPK or lysosome acidification preserves PSD‑95 despite normal glymphatic influx (assessed by CSF‑brain tracer flux).
4. Behavioral correlate: Mice with astrocyte‑specific V‑ATPase mutation show impaired reversal learning after sleep deprivation, reflecting failure to prune outdated synaptic configurations, whereas fasting rescues this deficit only when AMPK is intact.

Falsifiability

If lysosomal acidification does not increase during sleep under fasting conditions, or if blocking AMPK fails to alter the selectivity of synaptic protein clearance (i.e., bulk autophagy markers change but ubiquitin‑dependent synaptic loss remains unchanged), the hypothesis is falsified. Likewise, if glymphatic flow alone accounts for synaptic protein clearance independent of lysosomal pH, the proposed mechanistic link would be unsupported.

Implications

This framework positions fasting not merely as a metabolic prelude to autophagy but as a tuner of lysosomal selectivity that couples extracellular waste flux to intracellular synaptic triage. It suggests that timed feeding regimens could be used to enhance sleep‑dependent cognitive flexibility by promoting the removal of maladaptive synaptic tags while preserving essential circuitry—offering a testable avenue for mitigating neurodegeneration linked to sleep disruption.