Hypothesis

Core idea: Age‑related neuronal loss is not the result of an active, pruning‑like quality‑control system that tags inefficient cells for removal. Instead, progressive lysosomal overload in neurons and microglia impairs autophagic flux, leading to energy failure that disproportionately affects highly active, weakly connected neurons. These cells die passively because they cannot meet their ATP demand, mimicking a selective loss without an instructive elimination signal.

Mechanistic reasoning:

1. Lysosomal burden. Aging neurons accumulate undegraded proteins and lipids in autolysosomes [4]. This storage compromises lysosomal acidification and protease activity, reducing mitophagy and increasing mitochondrial ROS.
2. Energy deficit. Damaged mitochondria lower ATP output. Neurons with high firing rates and sparse synaptic inputs (e.g., layer II cortical pyramidal cells) rely on steady ATP to maintain ion gradients; a shortfall leads to depolarization, calcium influx and activation of calpain‑mediated proteolysis.
3. Secondary disconnection. As ATP fails, dendritic spines retract and synaptic vesicle cycling slows, weakening functional connectivity. Microglia, already occupied with myelin debris [3], cannot efficiently clear these synapses, so the loss appears selective.
4. No complement tag. Unlike developmental C3‑CR3 mediated pruning [2], aged microglia show reduced complement expression and are phagocytically exhausted, so complement blockade will not alter age‑related neuron loss.

Testable predictions:

• Prediction 1: In mouse cortex, lysosomal markers (LAMP1, cathepsin D) will rise with age and correlate spatially with neuronal density loss in supragranular layers, but not with C3 deposition.
• Prediction 2: Genetic activation of TFEB or pharmacological autophagy induction (e.g., spermidine) will restore lysosomal function, increase ATP levels in layer II neurons, and reduce neuronal loss without changing synaptic pruning markers (PSD‑95, synaptophysin) in young adult controls.
• Prediction 3: Antibody‑mediated C3 blockade will prevent developmental synapse elimination in P7 mice but will have no effect on neuronal counts in 18‑month‑old mice.

Falsifiability: If TFEB overexpression fails to improve neuronal survival despite correcting lysosomal markers, or if C3 blockade markedly reduces age‑related neuron loss, the lysosomal‑overload mechanism is insufficient and an active pruning model must be reconsidered.

References (inline): lysosomal accumulation [4]; myelin debris burden on microglia [3]; developmental complement pruning [2]; autophagy‑lipid‑protein association [4]; gene expression reversal [5].