Hypothesis

Glutamine availability controls the acetylation state of the ARID domain in KDM6A/B demethylases, which in turn sets their chromatin residence time and demethylase activity at bivalent promoters. In cancer, high glutamine uptake fuels α‑ketoglutarate production and maintains high acetyl‑CoA levels, promoting p300/CBP‑mediated acetylation of specific lysine residues on the ARID domain. This acetylation increases ARID affinity for nucleosomes bearing H3K27me3, stabilizing KDM6 chromatin binding and allowing robust demethylation even when competing metabolites fluctuate. Consequently, bivalent loci lose H3K27me3 while retaining H3K4me3, sustaining an open, transcriptionally poised state that supports cancer stem cell expansion and therapy resistance.

In aging tissues, declining SLC1A5‑mediated glutamine import lowers intracellular α‑ketoglutarate and acetyl‑CoA, reducing ARID acetylation. Deacetylated ARID exhibits weaker nucleosome interaction, shortening KDM6 residence time and permitting EZH2‑driven H3K27me3 to accumulate. The resulting shift toward repressive bivalent resolution silences developmental genes and contributes to loss of cellular plasticity.

Testable Predictions

1. ARID acetylation status correlates with glutamine levels – Quantify acetylation of KDM6A ARID lysines (e.g., Kxxx) by immunoblot with acetyl‑lysine antibodies in cultured cells treated with glutamine deprivation or supplementation. Expect decreased acetylation under low glutamine and rescue by adding cell‑permeable α‑ketoglutarate or acetyl‑CoA precursors.

2. Manipulating ARID acetylation alters bivalent chromatin – Use CRISPR‑knockin to replace target ARID lysines with arginine (non‑acetylatable) or glutamine (acetyl‑mimic) in mouse models. Predict that arginine mutants will show increased H3K27me3 at bivalent promoters in cardiac tissue (mimicking aging) and reduced tumorigenesis in xenograft models, whereas glutamine mimetic mutants will retain low H3K27me3 despite glutamine restriction.

3. Pharmacological modulation of acetylation impacts disease phenotypes – Treat aged mice with SIRT1 inhibitors (to increase acetylation) or cancer cells with p300/CBP inhibitors (to decrease acetylation) and assess:

   • Cardiac aging: H3K27me3 levels at bivalent domains, fibrosis, and functional echocardiography.
   • Cancer: tumor growth, cancer stem cell frequency, and sensitivity to chemotherapeutic agents. Expect that boosting ARID acetylation in aged hearts reduces H3K27me3 accumulation, while lowering it in tumors increases repression of oncogenic bivalent loci and attenuates stem‑like phenotypes.

4. Chromatin residence time measured by FRAP or live‑cell imaging – Compare fluorescence recovery of GFP‑tagged wild‑type, acetylation‑dead, and acetylation‑mimic KDM6A in nuclei under varying glutamine concentrations. Anticipate faster recovery (shorter residence) for acetylation‑dead mutants, especially when glutamine is low.

Falsifiability

If ARID acetylation does not change with glutamine manipulation, or if altering ARID lysine acetylation fails to shift H3K27me3/H3K4me3 balance at bivalent promoters, the hypothesis would be refuted. Likewise, if SIRT1 or p300/CBP inhibition produces opposite effects to those predicted in aging versus cancer contexts, the mechanistic link between glutamine‑driven acetylation and opposing bivalent resolution would be untenable.

Relevance to Open Questions

This hypothesis directly addresses the lack of comparative KDM6 cancer biology data by proposing a metabolic‑post‑translational switch that explains why the same bivalent loci resolve toward activation in cancer but repression in aging. It also introduces the ARID domain as a chromatin‑anchoring regulator whose activity is tuned by glutamine‑dependent acetylation, a layer not yet examined in existing PHD/ARID domain studies.