Hypothesis

Repeated, short‑duration pulses of OSK factors in aged brains do more than reset epigenetic age; they inject stochastic variability into neuronal gene expression that raises cortical entropy and breaks down overly stable predictive maps.

Mechanistic Basis

OSK drives TET1/2‑mediated DNA demethylation at promoters of immediate‑early genes and ion‑channel subunits, increasing transcriptional noise 4. This noise translates into heterogeneous firing patterns and elevated local field potential (LFP) entropy, counteracting the hyper‑synchrony that accompanies over‑consolidation 1. By momentarily destabilizing attractor states, the circuit regains capacity to encode surprise without losing consolidated knowledge.

Experimental Design

1. Animals – 20‑month‑old C57BL/6 mice receive cortical AAV‑OSK (or AAV‑GFP control) with a doxycycline‑inducible system; OSK expressed for 4 h every 72 h for three cycles.
2. Readouts –
   • In vivo LFP entropy (sample entropy) recorded from prefrontal cortex before, during, and after each OSK window.
   • Single‑cell RNA‑seq of excitatory neurons to quantify transcriptional variance (Fano factor) of plasticity‑related genes.
   • Behavioral assessment on a reversal‑learning task (e.g., probabilistic reversal T‑maze) to measure adaptability to surprise.
3. Intervention Controls – Parallel groups receive AAV‑OSK‑ΔTET1/2 (catalytically dead TETs) or a TET inhibitor (Bobcat339) to test dependence on demethylation‑driven noise.

Expected Outcomes & Falsifiability

• If hypothesis is correct: OSK‑treated mice will show a significant rise in LFP entropy during expression windows (p < 0.01, effect size > 0.8), increased transcriptional variance of plasticity genes, and improved reversal‑learning performance compared with controls. Entropy and behavioral gains should peak within 24 h of each pulse and return to baseline before the next dose.
• If hypothesis is false: OSK will either leave LFP entropy unchanged or reduce it (indicating further synchrony), transcriptional variance will not increase, and reversal learning will not improve despite epigenetic age reversal measured by clocks.

The dependence on TET activity provides a clear falsifiable test: blocking TET1/2 should abolish both the entropy boost and the cognitive benefit, even if global DNA methylation is lowered. This links the mechanistic source of noise directly to functional rescue.

By framing rejuvenation as controlled injection of uncertainty rather than mere restoration, the hypothesis predicts that optimal dosing schedules will balance entropy elevation against network stability—a balance trackable via real‑time entropy biomarkers alongside epigenetic clocks.