Hypothesis

Neuronal mitochondrial DNA released during aging activates the cytosolic cGAS‑STING pathway, which transcriptionally upregulates complement component C3 and its receptor C3aR within neurons. This intracellular complement (‘complosome’) activation disrupts endo‑lysosomal function, rendering synapses susceptible to C1q tagging and subsequent microglial phagocytosis, thereby linking intrinsic neuronal stress to extracellular complement‑mediated synapse loss.

Mechanistic Rationale

• Aging neurons accumulate damaged mitochondria, releasing mtDNA into the cytosol [1].
• Cytosolic mtDNA engages cGAS, producing STING‑dependent IFN‑stimulated genes that include C3 and C3aR (as demonstrated in immune cells; neuronal expression is plausible) [2].
• Intracellular C3 cleavage generates C3a, which acts in an autocrine manner on neuronal C3aR to alter lysosomal acidification and cathepsin activity, reproducing the complosome‑driven endo‑lysosomal defect described in microglia and astrocytes [4].
• Lysosomal dysfunction reduces degradation of synaptic proteins, exposing phosphatidylserine and decreasing neuronal pentraxin 1 (NP1) levels, making synapses ‘weak’ and preferential targets for C1q tagging [5].
• C1q‑tagged synapses are then engulfed by microglia whose activation is amplified by astrocyte‑microglia C3‑C3aR signaling [3], creating a feed‑forward loop.

Testable Predictions

1. Neuron‑specific deletion of Sting will reduce intracellular C3 levels and complement‑dependent endo‑lysosomal markers in aged mice without altering microglial C3aR expression.
2. Rescue of Sting in neurons will restore synaptic density and improve performance on memory tasks despite sustained microglial complement activation.
3. Conversely, neuron‑specific overexpression of C3 will induce complosome‑mediated synaptic loss and memory deficits even when microglial phagocytosis is pharmacologically blocked (e.g., with Annexin V or microglia‑specific C3aR antagonist).
4. mtDNA release will precede detectable C1q deposition at synapses in longitudinal imaging of aging mice.

Experimental Approach

• Generate Camk2a‑Cre;Sting^fl/fl mice and age‑matched controls; assess intracellular C3 (immunostaining, subcellular fractionation), lysosomal cathepsin activity, synaptic protein levels (synaptophysin, PSD‑95), and behavior (Morris water maze) at 6, 12, 18 months.
• Use AAV‑Camk2a‑C3 overexpression in wild‑type aged mice; treat a subset with microglial inhibitor (PLX5622) or astrocyte‑microglia C3‑C3aR blocker (SB 290157) to isolate intracellular effects.
• Employ in vivo two‑photon imaging of C1q‑GFP reporter mice crossed with mtDNA‑sensor (mt‑Cyto‑GFP) to track temporal order of mtDNA efflux, C1q tagging, and microglial engulfment.
• Measure cytokine profiles to ensure peripheral immune changes are minimal.

Potential Outcomes and Falsifiability

• If neuronal Sting deletion fails to lower intracellular C3 or does not ameliorate synaptic loss, the hypothesis that neuronal cGAS‑STING drives complosome activation is falsified.
• If neuronal C3 overexpression does not cause synaptic deficits when microglia are inhibited, the complosome‑centric mechanism is unsupported.
• If mtDNA efflux does not precede C1q deposition, the proposed upstream trigger is incorrect.

This framework shifts the focus from microglia‑centric complement to a neuron‑intrinsic complosome pathway that couples mitochondrial stress to synaptic vulnerability, offering a precision target upstream of microglial phagocytosis.