Hypothesis

In the aging brain, neurons that accumulate high burdens of copy number variations (CNVs) or mosaic chromosomal alterations (mCAs) are actively identified and removed by microglia through a phagocytic pathway triggered by extracellular DNA sensing. This eviction is not random damage but an energy‑conserving pruning mechanism that improves cortical efficiency per watt under declining metabolic resources.

Mechanistic Rationale

1. DNA‑damage sensing in neurons – Chromosome instability triggers cGAS‑STING activation, leading to release of cytosolic DNA fragments that can exit the neuron via exosomes or membrane blebs (3).
2. Microglial extracellular DNA detection – Microglia express TLR9 and AIM2 inflammasome components that bind extracellular DNA, prompting a pro‑phagocytic state (1).
3. Metabolic coupling – Low ATP/AMP ratios activate AMPK in microglia, enhancing phagocytic cup formation, while mTORC1 inhibition (mimicking caloric restriction) sustains this state, linking cellular energy status to eviction efficiency.
4. Selective pressure – Neurons with high CNV burden in energetically costly genes (e.g., synaptic vesicle cycling, mitochondrial transporters) produce more extracellular DNA, making them preferential targets (4).
5. Outcome – Removal of these cells reduces overall neuronal CNV load, lowers basal metabolic demand, and preserves network synchrony, which aligns with the observation that centenarians accumulate fewer autosomal mCAs (6).

Testable Predictions

• Prediction 1: In aged mice, pharmacological inhibition of microglial TLR9 or AIM2 will increase neuronal CNV burden (measured by single‑cell CNV sequencing) without affecting overall apoptosis rates.
• Prediction 2: Enhancing microglial AMPK activity (e.g., with AICAR) will accelerate the clearance of CNV‑laden neurons and improve performance on energy‑sensitive cognitive tasks (e.g., delayed alternation).
• Prediction 3: Neurons engineered to overexpress a DNA‑repair factor (e.g., OGG1) will show reduced extracellular DNA release, lower microglial phagocytosis, and retain higher CNV loads, leading to worsened cognitive decline under caloric restriction.
• Falsification: If blocking microglial DNA sensing does not alter neuronal CNV load, microglial phagocytic markers, or cognitive outcomes in aged animals, the hypothesis that microglia mediate active eviction of CNV‑burdened neurons is falsified.

Experimental Approach

1. Use CX3CR1‑CreER;Rosa26‑LSL‑tdTomato mice to label microglia and Camk2a‑Cre;Rosa26‑LSL‑GFP to label excitatory neurons.
2. Apply single‑cell whole‑genome sequencing (scWGS) on sorted neurons from young (3 mo), middle‑aged (12 mo), and old (24 mo) mice to quantify CNV/mCA burden.
3. Treat cohorts with TLR9 antagonist (ODN2088), AIM2 inhibitor, or AMPK activator and repeat scWGS plus behavioral testing.
4. Measure extracellular DNA in CSF via qPCR for Alu repeats and microglial phagocytic activity using pHrodo‑labeled DNA beads.

By linking neuronal genomic quality control to microglial surveillance and metabolic state, this hypothesis reframes age‑related neuronal loss as an adaptive, albeit potentially maladaptive when excessive, pruning process that can be experimentally manipulated.