Hypothesis

Restoring precise nucleosome positioning and histone modification balance at enhancer regions of master regulators (e.g., Ppara, Esrra, Ppargc1b) reduces epigenetic noise, thereby rescuing TF‑target information flux in aged skeletal muscle without needing TF overexpression.

Rationale

• Age‑related broadening of H3K4me3 domains and loss of H3K27ac at enhancers increases stochastic chromatin variability [2] .
• This epigenetic noise propagates hierarchically through GRNs, amplifying transcriptional variability [3] .
• NAD+‑dependent SIRT1/6 decline elevates acetylation, creating a feedback loop that worsens noise [4] .
• Preliminary data show that exogenous expression of metabolic TFs recovers ~10% of lost mutual information [1], suggesting that upstream chromatin states limit TF efficacy.

We propose that fixing the chromatin landscape at master regulator enhancers reduces noise at its source, allowing endogenous TFs to transmit information more faithfully.

Predictions

1. In aged mouse muscle, targeted CRISPR‑dCas9‑KRAB (to compact nucleosomes) or dCas9‑p300 (to deposit H3K27ac) at Ppara/Esrra/Ppargc1b enhancers will decrease cell‑to‑cell variance in H3K4me3 width and H3K27ac signal (measured by scCUT&Tag).
2. This reduction in epigenetic noise will increase the mutual information between TF binding and target gene expression by at least 15% compared with aged controls, surpassing the ~10% rescue seen with TF overexpression.
3. Single‑cell RNA‑seq will show a shift of myonuclear subpopulations toward a younger transcriptional profile, specifically reduced interferon/immune response and endogenous retrovirus activation.
4. The rescue will be absent if SIRT1 activity is pharmacologically inhibited, confirming the NAD+‑SIRT1‑dependent mechanism.

Experimental Design

• Animals: 24‑month‑old C57BL/6 mice; young (3‑month) controls.
• Interventions: AAV‑delivered dCas9‑KRAB or dCas9‑p300 guided to enhancer regions identified by bulk H3K27ac loss; control AAV‑dCas9 (no effector).
• Readouts (4 weeks post‑treatment):
  • scATAC‑seq and scCUT&Tag for H3K4me3/H3K27ac to quantify noise (variance across cells).
  • scRNA‑seq to compute TF‑target mutual information using the information‑flux framework [1].
  • Flow cytometry for interferon‑stimulated genes and ERV RNA (FISH).
• Validation: Treat a subset with EX-527 (SIRT1 inhibitor) to test dependence on SIRT1.

Potential Outcomes & Falsifiability

• If enhancer stabilization reduces chromatin variance and boosts TF‑target mutual information ≥15% and SIRT1 inhibition abolishes the effect, the hypothesis is supported.
• If noise remains unchanged or mutual information does not improve despite successful epigenetic editing, the hypothesis is falsified, indicating that epigenetic noise at master regulator enhancers is not the primary driver of information loss.
• If TF overexpression still outperforms epigenetic editing, it would suggest that TF concentration, not chromatin noise, is limiting.

This experiment directly tests whether fixing epigenetic noise at the source can rewire aged GRNs, addressing the open question of whether chromatin accessibility rescue alone suffices without TF overexpression.