Hypothesis

Senolytics restore synaptic density in aged brains not only by clearing senescent microglia but by reprogramming the metabolic state of residual microglia to a neuroprotective phenotype that suppresses complement-mediated pruning. We propose that senescent microglia exhibit a shift toward aerobic glycolysis and heightened mTORC1 activity, which drives upregulation of the complement component C1q and downstream opsonization of synapses [3][4]. Senolytic treatment removes these cells and, through transient exposure to SASP‑depleted milieu, activates AMPK signaling in neighboring microglia, promoting oxidative phosphorylation and inhibiting mTORC1. This metabolic switch reduces C1q transcription via FOXO‑dependent repression and increases surface CD47, rebalancing the "eat me/don’t eat me" signaling axis [1]. Consequently, complement tagging returns to developmental levels, preserving synapses that are weakly connected but still functionally relevant, thereby improving cognitive performance without causing global hyperconnectivity.

Testable Predictions

1. In aged mice, senolytic (D+Q) treatment will decrease the proportion of microglia expressing glycolytic markers (HK2, LDHA) and increase oxidative phosphorylation markers (PGC-1α, COXIV) within 7 days, preceding changes in synaptic density [5][6].
2. Pharmacological inhibition of AMPK (Compound C) administered together with senolytics will block the metabolic shift, sustain high C1q expression, and prevent synaptic rescue, despite equivalent senescent cell clearance.
3. Conversely, direct activation of AMPK in microglia (using a microglia‑targeted agonist) in the absence of senolytics will mimic the senolytic effect: reduced C1q, increased CD47, and rescued hippocampal‑dependent memory.
4. Metabolomic profiling of isolated microglia will reveal a rise in NAD+/NADH ratio and a fall in succinate levels correlating with synaptic preservation.

Experimental Approach

• Use 24‑month‑old C57BL/6 mice; treat groups with vehicle, D+Q, D+Q+Compound C, and microglia‑specific AMPK activator (via CX3CR1‑Cre‑driven construct).
• Quantify senescent microglia via p16^Ink4a^ immunostaining to confirm clearance [3][4].
• Measure microglial metabolic state (Seahorse OCR/ECAR, HK2, PGC-1α) and complement components (C1q, C3, CR3) by flow cytometry and qPCR [2].
• Assess synaptic density in hippocampus (synaptophysin/PSD‑95 immunofluorescence) and cortical areas.
• Evaluate cognition with Morris water maze and novel object recognition.
• Perform metabolomics on FACS‑sorted microglia.

Falsifiability

If senolytics improve synaptic density and cognition without altering microglial metabolism, or if metabolic manipulation fails to affect complement signaling, the hypothesis is refuted. Conversely, demonstration that metabolic reprogramming alone is sufficient to normalize pruning would strongly support the proposed mechanism.