Hypothesis: Enhancing lysosomal membrane integrity through neuronal overexpression of LAMP2A prevents cathepsin-mediated eviction of metabolically inefficient neurons, thereby preserving cortical thickness and cognitive function in aged mice without altering basal neuronal activity.

Rationale: It's clear that LMP triggers selective release of cathepsins B, D and L into the cytosol, where they cleave TFAM, raise ROS and initiate apoptosis (see LMP triggers selective cell death)[https://www.tandfonline.com/doi/full/10.1080/15548627.2019.1628538] and Cathepsin D translocation correlates with death)[https://scholars.direct/Articles/neurodegenerative-disorders/jnd-4-017.php?jid=neurodegenerative-disorders]. Cystatin B deficiency lifts inhibition on these proteases, creating a vulnerability checkpoint)[https://pmc.ncbi.nlm.nih.gov/articles/PMC11753708/]. We propose that the decision to evict a neuron hinges on the probability of LMP, which is modulated by LAMP2A levels. Boosting LAMP2A should raise the threshold for LMP, sparing neurons that would otherwise be culled while leaving the proteolytic capacity of lysosomes intact for routine turnover.

Novel mechanistic insight: LAMP2A not only stabilizes the lysosomal membrane but also chaperones cytosolic proteins for lysosomal import via CMA. Increased CMA flux could degrade damaged mitochondrial components before they provoke ROS‑induced LMP, coupling proteostatic and mitochondrial quality control. Thus, LAMP2A overexpression may act at two nodes: (1) physical reinforcement of lysosomal membranes, reducing cathepsin leak; (2) enhanced clearance of ROS‑producing mitochondria, lowering the upstream trigger for LMP.

Testable predictions:

• Neurons with elevated LAMP2A will show reduced cytosolic cathepsin activity after mild oxidative stress compared with controls.
• Cortical sections from aged LAMP2A‑overexpressing mice will retain higher densities of NeuN‑positive cells, especially in layers II/III where metabolic demand is high.
• Behavioral assays (e.g., novel object recognition) will demonstrate improved spatial memory in aged LAMP2A‑overexpressing mice, matching performance of young adults.
• Pharmacological blockade of cathepsins will not provide additional benefit beyond LAMP2A overexpression, indicating that the protective effect acts upstream of protease release.
• If LAMP2A is knocked down in young neurons, lysosomal fragility will increase and cathepsin‑mediated eviction will accelerate, mimicking an aged phenotype.

Experimental approach:

1. Generate AAV9‑syn‑LAMP2A vectors for neuronal delivery; inject into bilateral hippocampus and prefrontal cortex of 18‑month‑old mice. Controls receive AAV9‑syn‑GFP.
2. Two weeks post‑injection, expose cohorts to low‑dose rotenone (complex I inhibitor) to induce proteostatic stress.
3. Measure LMP using galectin‑3 puncta formation; quantify cytosolic cathepsin B activity with a fluorogenic substrate; assess TFAM cleavage by western blot.
4. Stereological counting of NeuN+ cells and layer‑specific density; evaluate lipofuscin accumulation via autofluorescence.
5. Conduct behavioral testing (novel object recognition, Morris water maze) after four weeks.
6. In parallel, treat a subset with cathepsin B/L inhibitor (e.g., CA‑074 Me) to test for additive effects.

Falsifiability: If LAMP2A overexpression fails to reduce LMP, cytosolic cathepsin activity, or neuronal loss relative to controls, the hypothesis is refuted. Conversely, if LAMP2A elevation rescues neurons but does not improve cognition, the link between neuronal eviction and functional decline would be weakened, prompting revision of the model.

This framework transforms the notion of neuronal "eviction" from a passive read‑out of damage into a tunable quality‑control gate that can be bolstered to preserve cortical circuits during aging.