Background

During slow‑wave sleep the glymphatic system expands extracellular space and drives convective CSF‑ISF exchange, clearing soluble amyloid‑β, tau and α‑synuclein12. Recent work shows this clearance is not a blunt drain but an active triage that decides which molecular substrates are retained for reuse and which are expelled3. However, the cellular logic that tags synapses or protein complexes for preservation versus removal remains unknown.

Hypothesis

We propose that glymphatic triage is coupled to astrocytic lactate signaling: neurons that release lactate during prior waking activity activate astrocytic monocarboxylate transporter 1 (MCT1) and trigger a rise in intracellular cAMP, which phosphorylates aquaporin‑4 (AQP4) at serine‑111 and biases perivascular CSF influx toward those perivascular domains. Consequently, the extracellular space surrounding lactate‑active synapses experiences reduced convective flow, shielding resident proteins from clearance. In contrast, synapses with low lactate efflux leave AQP4 in a dephosphorylated state, permitting high‑flow glymphatic washing and removal of associated protein aggregates. Chronic sleep restriction blunts the slow‑wave‑dependent norepinephrine drop, dampening the lactate‑AQP4 phosphorylation cascade and causing a global reduction in selective protection; the result is a non‑specific increase in clearance that removes both toxic aggregates and essential synaptic proteins, or alternatively a failure to clear due to AQP4 mislocalization, depending on the region’s metabolic baseline.

Testable Predictions

1. In vivo two‑photon imaging of CSF tracer flux in mice will show inverse correlation between local lactate levels (measured with lactate‑specific biosensors) and glymphatic influx rate during NREM sleep, specifically in peri‑synaptic extracellular space.
2. Pharmacological blockade of astrocytic MCT1 (with AR‑C155858) will abolish the lactate‑dependent reduction in glymphatic flow, leading to uniform clearance across high‑ and low‑activity cortical columns.
3. Phospho‑specific antibodies against AQP4‑S111 will reveal enriched phosphorylation in perivascular astrocytic endfeet surrounding lactate‑active synapses after a waking period, and this enrichment will be lost after sleep deprivation.
4. Mice expressing a non‑phosphorylatable AQP4 (S111A) will lose the sleep‑dependent protection of lactate‑tagged synapses, exhibiting accelerated loss of dendritic spine density and heightened susceptibility to Aβ oligomer induced toxicity despite normal total glymphatic volume.
5. Human PET‑MRI studies using lactate‑hyperpolarized 13C‑pyruvate and intrathecal CSF tracer will demonstrate that individuals with high regional lactate efflux during wakefulness have slower CSF clearance rates in the same regions during subsequent sleep, and that this relationship predicts slower accumulation of tau PET signal over 2 years.

Experimental Approach

• Use lactate‑binding FRET sensors (Laconic) expressed in excitatory neurons of layer 2/3 cortex, combined with caged‑CSF fluorescent tracer and two‑photon microscopy to quantify real‑time glymphatic influx during natural sleep cycles.
• Apply cell‑type specific chemogenetic activation (hM3Dq) of astrocytes to manipulate MCT1 activity while measuring AQP4‑S111 phosphorylation via immunofluorescence and proximity ligation assay.
• Generate AAV‑mediated knock‑in of AQP4‑S111A in astrocytes and assess synaptic integrity (synaptophysin, PSD‑95) and neurodegeneration after chronic sleep fragmentation.
• Translate to humans with simultaneous hyperpolarized 13C‑pyruvate PET and intrathecal Gadolinium‑based CSF tracer MRI during a night of polysomnography.

Potential Confounds and Controls

• Ensure that changes in arterial CO₂ or blood pressure do not drive CSF flow variations; monitor physiological parameters and include acetazolamide controls.
• Verify that lactate sensor signals reflect extracellular lactate and not intracellular metabolism by co‑expressing a cytosolic lactate dehydrogenase mutant.
• Account for regional differences in vascular density by normalizing tracer influx to local cerebral blood flow measured via laser Doppler.

If validated, this hypothesis reframes the glymphatic system as a metabolism‑dependent selective filter that couples neuronal activity history to synaptic survival, providing a mechanistic bridge between sleep‑dependent metabolic homeostasis and the selective vulnerability of neural networks in neurodegeneration.