Background

Recent ATAC‑seq work shows that aged stem cells do not follow a uniform chromatin‑erosion trajectory. In muscle satellite cells, distal enhancers become hyper‑accessible, creating a chronically active state that impairs quiescence{1}. In neural stem/progenitor cells, promoters and enhancers of proliferation genes lose accessibility and H3K4me3/H3K27ac marks, reflecting a depletive mode{2}$. These opposing patterns correlate with heterochromatin loss in satellite cells and MLL complex dysregulation in NSPCs, and RhoA inhibition can reverse erosive opening at retrotransposon‑rich regions in hematopoietic stem cells{3}$. Moreover, transplanting aged HSCs into young niches fails to reset established accessibility states, suggesting that once set, these chromatin configurations are self‑maintaining{5}$.

Hypothesis

We propose that the balance between two antagonistic epigenetic complexes— the nucleosome remodeling and deacetylase (NuRD) complex, which carries HDAC1/2 activity, and the MLL‑containing COMPASS complex, which deposits H3K4me3— determines the direction of age‑related chromatin remodeling at regulatory elements. In tissues where NuRD activity dominates over COMPASS at distal enhancers, deacetylation leads to nucleosome destabilization and increased accessibility (erosive aging). Conversely, where COMPASS activity is suppressed relative to NuRD at promoters, loss of H3K4me3 and reduced nucleosome turnover drives accessibility loss (depletive aging). The tissue‑specific expression or post‑translational modification of complex subunits (e.g., CHD4 for NuRD, WDR5 for MLL) sets this switch, and inflammatory signaling (NF‑κB) biases the balance toward NuRD in mesenchymal lineages while promoting COMPASS inhibition in neurogenic niches.

Predictions

1. Correlative – In aged muscle satellite cells, ChIP‑seq will show enrichment of NuRD subunits (CHD4, MTA2) and reduced H3K27ac at hyper‑accessible enhancers, whereas aged NSPCs will display loss of MLL/WDR5 and H3K4me3 at hypo‑accessible promoters.
2. Causal – Pharmacological inhibition of HDAC1/2 (using a selective HDAC1/2 inhibitor) in aged satellite cells will decrease enhancer accessibility and restore quiescence, while overexpression of WDR5 in aged NSPCs will increase promoter accessibility and rescue proliferative capacity.
3. Reversibility barrier – Forced expression of a NuRD‑resistant HDAC mutant in young HSCs transplanted into aged hosts will lock in an erosive enhancer state that persists after transplantation, recapitulating the failure of niche‑mediated rejuvenation.
4. Specificity – Simultaneous NuRD inhibition and MLL activation in a mixed aged stem‑cell population will preferentially correct erosive loci in satellite cells and depletive loci in NSPCs without affecting the opposite modality, demonstrating context‑dependent outcomes.

Experimental Approach

• Single‑cell multi‑omics (scATAC‑seq + scCUT&Tag for HDAC1, CHD4, WDR5, H3K4me3) on FACS‑purified satellite cells, NSPCs, and HSCs from young (3 mo) and aged (24 mo) mice to map complex occupancy versus accessibility.
• Perturbation – Treat aged mice with isoform‑selective HDAC1/2 inhibitor (e.g., MC1568) or deliver AAV‑WDR5 to the subventricular zone; assess chromatin states, transcriptional noise, and functional readouts (muscle force, neurogenesis, competitive repopulation).
• Rescue test – Transplant HDAC‑mutant HSCs into irradiated young recipients and perform longitudinal scATAC‑seq to test whether the erosive state is maintained independent of niche.
• Readouts – Besides accessibility, measure RNA‑seq for enhancer‑associated eRNAs, heterochromatin marks (H3K9me3), and functional assays (colony‑forming units, EdU incorporation, grip strength).

If the data show that NuRD predominance drives erosive enhancer opening while MLL/COMPASS loss drives depletive promoter closing, and that shifting the balance reverses the respective phenotype, the hypothesis will be supported. Failure to observe these reciprocal changes—or seeing that manipulating one complex alters both accessibility types indiscriminately—would falsify the model and suggest a more global chromatin‑erosion mechanism.