Hypothesis

Aged, senescent microglia secrete extracellular vesicles enriched in miR-34a that are taken up by neurons in the medial prefrontal cortex (mPFC) and basolateral amygdala (BLA). This miR‑34a directly inhibits key autophagy genes (e.g., Atg5, Becn1), impairing neuronal clearance of damaged mitochondria and protein aggregates. The resulting cellular stress amplifies local inflammaging, fuels further microglial senescence, and promotes excessive phagocytic pruning of excitatory synapses, thereby locking fear extinction circuits into a hyperreactive state.

Mechanistic Rationale

• Senescent microglia exhibit a SASP that includes not only cytokines but also bioactive RNA cargo in exosomes 2.
• miR‑34a is upregulated in senescent glial cells and known to target autophagy regulators in neurons 1.
• Defective autophagy in mPFC and BLA neurons leads to mitochondrial ROS accumulation, which activates NF‑κB in neighboring microglia, creating a feed‑forward loop of senescence and inflammation 3.
• Excessive microglial phagocytosis of synapses, driven by TREM2‑dependent pathways, underlies extinction deficits after fear recall 45.
• Senolytics reduce microglial burden but do not block exosome release; inhibiting exosome biogenesis should break the loop upstream of synaptic pruning 6.

Testable Predictions

1. Aged mice will show elevated miR‑34a levels in exosomes isolated from cerebrospinal fluid and in mPFC/BLA neuronal lysates compared with young controls.
2. Neuronal overexpression of miR‑34a will recapitulate autophagy deficits, increased mitochondrial ROS, and impaired fear extinction without altering baseline fear memory.
3. Inhibition of neutral sphingomyelinase 2 (nSMase2) with GW4869 to block exosome release will normalize neuronal miR‑34a, restore autophagy flux, reduce synaptic pruning, and rescue extinction learning in aged mice.
4. Combined senolytic (ABT‑263) + exosome blockade will produce synergistic improvement in extinction performance beyond either treatment alone.

Experimental Approach

• Animal cohorts: Young (3 mo) and aged (20 mo) C57BL/6 mice; subgroups receive vehicle, ABT‑263 (senolytic), GW4869 (exosome inhibitor), or both.
• Fear conditioning & extinction: Standard tone‑footshock pairing; extinction sessions 24 h later; measure freezing and spontaneous recovery.
• Exosome isolation: Ultracentrifugation of CSF; quantify miR‑34a by qPCR.
• Neuronal assays: Laser‑capture microdissection of mPFC/BLA; assess LC3‑II/I ratio, p62, mitochondrial ROS (MitoSOX), and synaptic marker (PSD‑95, synaptophysin) via immunoblot and immunohistochemistry.
• Microglial profiling: Flow cytometry for senescent markers (p16^INK4a^, SA‑β‑gal) and TREM2; ELISA for IL‑1β, TNF‑α.
• Statistical plan: Two‑way ANOVA (age × treatment) with post‑hoc Tukey; n ≥ 10 per group to detect 20 % effect size with α = 0.05, power = 0.8.

Falsifiability

If aged mice do not exhibit increased CSF‑derived exosomal miR‑34a, or if neuronal miR‑34a manipulation fails to alter autophagy or extinction, the core mechanistic link is refuted. Likewise, if exosome blockade does not reduce synaptic pruning or improve extinction despite confirmed miR‑34a decrease, the hypothesis that microglial exosomes mediate the senescence‑to‑circuit pathology is invalidated. Conversely, confirmation of the predicted molecular and behavioral changes would support the model and suggest a combinatorial therapeutic strategy targeting both senescent cell clearance and exosomal signaling.