Hypothesis

Aging neurons with declining oxidative phosphorylation release metabolic cues that remodel the local extracellular matrix, reducing its capacity to sequester complement proteins and thereby marking synapses for microglial phagocytosis.

Mechanistic Model

1. Metabolic shift – In aged neurons, mitochondrial efficiency falls, increasing the cytosolic lactate‑to‑pyruvate ratio and decreasing NAD⁺/NADH. This metabolic state drives upregulation of glycolytic enzymes and secretion of lactate into the synaptic cleft.
2. Matrix modification – Lactate reacts with extracellular matrix proteoglycans (e.g., brevican, aggrecan) via non‑enzymatic glycation, altering their glycosaminoglycan chains. Modified proteoglycans bind C1q with lower affinity, diminishing their natural inhibitory effect on complement activation.
3. Complement exposure – With less C1q sequestered, free C1q accumulates at synapses, triggering the classical cascade and C3 deposition. Microglia recognize C3‑opsonized synapses via CR3 and engulf them, a process that parallels developmental pruning but is now gated by neuronal metabolism.
4. Feedback loop – Synaptic loss reduces neuronal activity, further lowering ATP production and reinforcing the metabolic shift, creating a self‑amplifying cycle of elimination.

This model extends the seed idea by specifying a metabolic‑matrix‑complement axis that converts energetic inefficiency into an “eat‑me” signal, rather than invoking nonspecific damage.

Testable Predictions

• Prediction 1: Neuronal overexpression of lactate dehydrogenase (LDH) to increase lactate export will elevate C1q deposition on synapses in young mice, mimicking the aging phenotype.
• Prediction 2: Genetic or pharmacological boost of neuronal mitochondrial function (e.g., PGC‑1α activation) will reduce lactate release, preserve proteoglycan integrity, lower synaptic C3 tagging, and rescue age‑related memory deficits.
• Prediction 3: Enzymatic removal of lactate‑induced glycation products from hippocampal ECM (using deglycase enzymes) will restore C1q binding to proteoglycans and decrease microglial phagocytosis without altering overall complement levels.
• Prediction 4: In vivo two‑photon imaging will show a temporal sequence: rising extracellular lactate → decreased brevican‑C1q colocalization → increased C3‑tagged synapses → microglial engulfment.

Experimental Approach

Use aged (18‑month) wild‑type mice and complementary interventions:

1. Viral delivery of AAV‑LDH‑A or AAV‑PGC‑1α to hippocampus.
2. Intracerebral injection of recombinant brevican pre‑treated with lactate or glycation inhibitors.
3. Measure lactate (microdialysis), proteoglycan modification (lectin blotting), C1q/C3 immunofluorescence (confocal), microglial phagocytosis (Iba1+/synaptophysin+ puncta), and behavior (Morris water maze).

Include appropriate controls (scrambled virus, saline injection) and complement‑deficient (C3⁻/⁻) mice to confirm dependence on the cascade.

Potential Confounds and Mitigations

• Off‑target metabolic effects: Monitor global ATP levels and ensure manipulations do not cause widespread neurodegeneration; restrict analysis to synapse‑rich layers.
• Inflammatory confound: Assess cytokine profiles to distinguish complement‑specific phagocytosis from general microglial activation.
• Age‑related BBB permeability: Use intracerebral delivery to avoid peripheral lactate contributions.

If the predictions hold, the data would support a mechanism where neuronal metabolic state directly gates complement‑mediated synaptic elimination, reframing age‑related synapse loss as an adaptive, albeit maladaptive, response to energetic decline rather than indiscriminate damage.