Hypothesis

Aging neurons with low synaptic output and high ATP cost per spike become preferential substrates for lysosomal membrane permeabilization, leading to cathepsin B release that triggers apoptosis. This links energetic inefficiency to selective cathepsin‑mediated culling, reframing LMP as a quality‑control sensor rather than indiscriminate damage.

Mechanistic Basis

Lysosomes in aged neurons accumulate undigested material and show altered lipid composition, making their membranes more prone to stress‑induced permeabilization[4]. TFEB decline reduces lysosomal biogenesis, slowing repair of LMP events[3]. When a neuron fires infrequently, its cytosolic ATP/ADP ratio falls, activating AMPK and inhibiting mTORC1, which further suppresses TFEB transcription. Low mTORC1 activity also diminishes phosphorylation of the lysosomal protein LAMP2, weakening membrane stability. Concurrently, active neurons maintain high Ca2+ buffering and ROS scavenging, preserving lysosomal integrity. In low‑activity cells, the combined effect of impaired repair and heightened membrane fragility allows transient LMP, releasing cathepsin B into the cytosol[1]. Cathepsin B then cleaves pro‑caspase‑3/9 and degrades Bcl‑2, committing the cell to apoptosis[2]. Thus, lysosomal leak becomes a read‑out of metabolic state, coupling energy budget to programmed removal.

Testable Predictions

1. Correlation at single‑cell level: In aged mouse cortex, neurons identified by low c‑Fos or reduced calcium transient amplitude will show higher cytosolic cathepsin B immunoreactivity than highly active neighbors (immunofluorescence + electrophysiology).
2. Metabolic rescue: Chemogenetic activation of low‑firing neurons (hM3Dq) in aged mice should decrease cathepsin B leakage and improve survival, without altering global LMP levels.
3. TFEB dependence: Neuron‑specific TFEB overexpression will blunt the activity‑dependent difference in cathepsin B leakage, making low‑ and high‑activity cells equally resistant.
4. Pharmacological mimic: Acute lysosomal destabilization with low‑dose LLOMe will preferentially kill neurons pre‑selected for low baseline firing in slice cultures, an effect blocked by cathepsin B inhibitor CA‑074Me.

Possible Confounds

Global lysosomal stress could still cause non‑selective death; therefore, experiments must control for overall LMP magnitude (e.g., using galectin‑3 puncta) while assessing cathepsin B distribution. Additionally, activity manipulations may affect lysosomal pH independently, requiring parallel pH‑sensitive probes to isolate the permeabilization effect.

This hypothesis transforms cathepsin‑mediated apoptosis from a stochastic by‑product of lysosomal aging into a regulated mechanism that trims metabolically costly, under‑utilized neurons, offering a concrete framework to test whether cognitive decline reflects loss of inefficient circuits rather than indiscriminate neurodegeneration.