Hypothesis

Age‑associated decline in neuronal NAD+ activates the calcium‑dependent phospholipid scramblase TMEM16F, increasing phosphatidylserine (PS) exposure on synapses. Externalized PS serves as a direct ligand for the complement component C1q, thereby priming synapses for C3 ozonization and microglial phagocytosis independent of classical immune complexes.

Rationale

• Complement‑driven synapse loss escalates with age (1).
• The ligands that recruit C1q to synapses remain unidentified (2).
• PS is a well‑known “eat‑me” signal sensed by phagocytes and can bind C1q directly (studies in apoptotic cells show C1q‑PS interaction).
• NAD+ depletion, prevalent in aged brains, activates calcium signaling and TMEM16F‑mediated PS flipping (3).

Thus, age‑related metabolic stress converts synapses into PS‑rich substrates that hijack the complement cascade.

Testable Predictions

1. Aged mice will show higher synaptic PS exposure (detected by annexin V‑based proximity ligation) than young controls, correlating with hippocampal C3 deposition.
2. Neuronal‑specific overexpression of NAD+‑ biosynthetic enzyme NAMPT will reduce synaptic PS, lower C1q binding, and preserve synapses despite aging.
3. Pharmacological blockade of TMEM16F (e.g., with ANNsynaptic peptide) or extracellular annexin V will decrease C3 fragment deposition on synapses and rescue cognitive performance in aged mice, without altering complement protein levels in CSF.
4. In vitro, isolated synapses from aged neurons will bind recombinant C1q more strongly; this binding will be abolished by pre‑incubation with annexin V or calcium chelation.

Experimental Approach

• Use synaptoneurosome preparations from young (3 mo) and aged (18 mo) WT mice; quantify surface PS via flow cytometry with fluorescent annexin V, and C1q binding via ELISA.
• Generate CamKII‑Cre; NAMPT^fl/fl mice to neuronal‑restrict NAD+ boost; assess PS, C1q deposition, synapse density (Synaptophysin/PSD‑95 immunostaining), and behavior (Morris water maze).
• Treat aged mice intracerebroventricularly with TMEM16F inhibitor (e.g., compound A) or annexin V‑Fc fusion; measure CSF C3 levels, synaptic C3b/iC3b, and cognitive outcomes.
• Perform rescue experiments: add recombinant PS‑blocking peptide to aged synaptoneurosomes to test if C1q binding is PS‑dependent.

Potential Outcomes

• Confirmation would identify PS as the missing synaptic “eat‑me” ligand linking metabolic aging to complement‑mediated pruning.
• Failure to see PS changes would redirect focus to other ligands (e.g., altered glycans, oxidized lipids) and suggest that complement activation proceeds via alternative pathways.

Implications

Targeting the PS‑C1q axis could provide a complementary strategy to C3/C5a inhibition, especially where blood‑brain barrier penetration of large antibodies is limiting; small‑molecule scramblase inhibitors or NAD+ boosters may offer disease‑modifying avenues for age‑related neurodegeneration and progressive MS.