Hypothesis

Astrocytes release extracellular vesicles (EVs) enriched in complement C3 and associated miRNAs that are taken up by neighboring microglia, thereby amplifying microglial CR3‑mediated phagocytosis of synapses. This astrocyte‑to‑microglia EV signaling creates a self‑reinforcing loop that sustains pathological complement activation independent of amyloid plaques, and disrupting EV release blocks the loop and rescues cognitive decline in aged mice.

Rationale

• Complement C3 overexpression in astrocytes impairs insulin signaling and mitochondrial function, leading to memory deficits (C3 overexpression drives memory impairments).
• Global reduction of C3 in adulthood preserves hippocampal synapses and cognition (Global C3 lowering protects against age‑associated hippocampal dysfunction).
• Synaptic elimination requires microglial CR3 binding to iC3b‑tagged synapses (Microglial phagocytosis of synapses via CR3).
• Astrocytes are known to release EVs that carry proteins, lipids, and RNAs influencing microglial phenotype in neuroinflammatory contexts.

We propose that astrocytic stress (e.g., mitochondrial dysfunction) increases EV biogenesis and loads these vesicles with active C3 and miRNAs that suppress microglial complement regulators (e.g., CD55, CD59). Upon microglial uptake, EV‑delivered C3 is secreted or remains surface‑bound, increasing local iC3b deposition on synapses and enhancing CR3‑driven phagocytosis. Simultaneously, EV miRNAs shift microglia toward a pro‑pruning state, establishing a feed‑forward loop that sustains complement‑mediated synapse loss even when amyloid is low.

Novel Mechanistic Insight

The loop links two previously separate observations: astrocyte‑centric metabolic insult (C3 elevation drives memory impairments through dysregulation of astrocytic insulin signaling and mitochondrial dysfunction) and microglial complement phagocytosis (Microglial phagocytosis of synapses occurs via engagement of synapse‑bound iC3b and the microglial CR3 complement receptor). EVs provide the mechanistic conduit that translates astrocytic metabolic failure into amplified microglial synaptic pruning.

Testable Predictions

1. EV C3 is elevated in the hippocampus of aged mice and correlates with synaptic loss.
2. Inhibiting astrocyte EV release (e.g., via conditional knockout of Rab27a) reduces microglial C3 uptake, lowers iC3b tagging on synapses, and rescues memory without affecting developmental pruning.
3. Microglia isolated from aged mice show increased CR3 signaling and phagocytic activity only when co‑cultured with astrocyte EVs from aged, but not young, donors.
4. Blocking EV‑mediated C3 transfer (using anti‑C3 antibodies that cannot cross the plasma membrane or heparan sulfate inhibitors) phenocopies the protective effects of global C3 lowering.

Experimental Design

• Model: Inducible, astrocyte‑specific Rab27a floxed mice (Astro‑Rab27aKO) crossed with aged wild‑type controls; also include a microglial CR3 reporter line (CR3‑GFP).
• Interventions: Tamoxifen induction at 6 months to delete Rab27a in astrocytes; assess at 18 months.
• Readouts:
  • Hippocampal EV concentration and C3 content (Western blot, ELISA).
  • Microglial EV uptake (flow cytometry of GFP‑labelled astrocyte EVs).
  • Synaptic density (synaptophysin, PSD‑95 immunostaining) and dendritic spine counts (Golgi‑Cox).
  • Cognitive performance (Morris water maze, novel object recognition).
  • Developmental pruning control: assess synaptic density in P21 pups to ensure no deficits.
• Pharmacologic arm: Treat aged wild‑type mice with GW4869 (neutral sphingomyelinase inhibitor) to block EV synthesis; compare to Astro‑Rab27aKO.

Potential Outcomes and Interpretation

• If predictions hold: Reduced astrocyte EV release lowers microglial C3 uptake and synaptic tagging, preserving synapses and cognition. This would demonstrate that astrocytes drive microglial complement pruning via EVs, identifying a new therapeutic target that spares developmental microglia‑dependent pruning.
• If EV blockade fails to rescue: Suggests alternative routes (e.g., soluble C3 diffusion or astrocyte‑derived cytokines) dominate, refuting the EV‑mediated loop hypothesis and redirecting focus to other astrocyte‑microglia signaling pathways.

This framework is falsifiable, experimentally tractable, and extends existing complement‑centric models by positioning astrocyte‑derived EVs as the critical conduit linking metabolic dysfunction to maladaptive synaptic pruning in brain aging.