Hypothesis

Navitoclax, a Bcl‑2/Bcl‑xL inhibitor, selectively eliminates senescent astrocytes and microglia in the aged brain. By removing these SASP‑secreting cells, navitoclax reduces extracellular matrix metalloproteinase activity and downstream chondroitin sulfate proteoglycan deposition, thereby loosening perineuronal nets (PNNs) that envelop parvalbumin‑positive interneurons. This structural softening restricts less glutamatergic input to PV cells, disinhibits gamma oscillations, and reopens critical‑period‑like plasticity without the off‑target myelin toxicity seen with dasatinib+quercetin.

Mechanistic Rationale

1. Senescent glia up‑regulate Bcl‑2 family proteins to resist apoptosis [3]. Navitoclax antagonizes these proteins, triggering caspase‑dependent clearance specifically in SASP‑rich astrocytes and microglia.
2. SASP from senescent glia includes MMP‑2, MMP‑9 and hyaluronidase‑like enzymes that remodel PNNs toward a hyper‑crosslinked state [2]. Removing the source drops local MMP activity, shifting the balance toward hyaluronan degradation by endogenous hyaluronidases and reducing chondroitin sulfate synthesis.
3. Lower chondroitin sulfate content decreases PNN density around PV interneurons, increasing dendritic spine turnover and rescuing long‑term potentiation (LTP) in layer 2/3 pyramidal cells [1].
4. Enhanced PV‑interneuron activity restores gamma‑band synchrony, which is correlated with improved cognitive flexibility and reversal learning in aged rodents.
5. Because navitoclax is more selective than dasatinib+quercetin, it spares oligodendrocyte precursor cells, avoiding the myelin loss and corpus callosum degeneration reported for broad‑spectrum senolytics [6].

Testable Predictions

• In vivo: Aged mice receiving intermittent navitoclax (e.g., 1 mg/kg intraperitoneally twice weekly for 4 weeks) will show a ≥30 % reduction in senescence‑associated β‑galactosidase‑positive astrocytes and microglia in cortex and hippocampus compared with vehicle.
• PNN quantification: Wisteria floribunda agglutinin staining will reveal a ≥20 % decrease in PNN intensity surrounding PV‑positive cells in navitoclax‑treated mice.
• Plasticity assays: Acute hippocampal slices will exhibit restored LTP magnitude (≥15 % increase over baseline) and increased dendritic spine density on pyramidal neurons.
• Behavioral outcomes: Treated mice will perform better on reversal‑learning tasks (e.g., Barnes maze) and display improved novel object recognition, correlating with PNN remodeling.
• Safety controls: Myelin basic protein immunostaining and corpus callosum thickness will remain unchanged, distinguishing navitoclax from dasatinib+quercetin-induced damage.

Falsifiability

If navitoclax fails to reduce senescent glia numbers, or if PNN density and LTP remain unchanged despite glial clearance, the hypothesis that Bcl‑2‑mediated senolysis restores plasticity via PNN remodeling is falsified. Conversely, observing cognitive benefits without corresponding PNN alterations would suggest alternative mechanisms (e.g., direct neurotrophic effects) and would also challenge the specific mechanistic link proposed.