Hypothesis

Female limbal epithelial stem cells (LESCs) maintain higher PAX6 expression with age because they possess two active copies of X‑linked chromatin regulators (e.g., KDM6A, DDX3X) that escape X‑inactivation and sustain an open chromatin state at the PAX6 locus. Males, with a single copy, experience age‑related decline in these regulators, leading to reduced PAX6, increased stress‑gene activation, and accelerated LESC senescence.

Mechanistic Rationale

• PAX6 transcription is modulated by histone‑demethylase activity; KDM6A (UTX) removes H3K27me3 marks and is known to escape X‑inhibition in humans.[1]
• DDX3X, an X‑linked RNA helicase, promotes transcriptional elongation of stress‑responsive genes and has been shown to interact with PAX6‑containing complexes in neural progenitors.[2]
• In XX cells, biallelic expression of these factors provides redundancy; stochastic silencing of one allele does not diminish total activity. In XY cells, loss‑of‑function mutations or age‑dependent epigenetic silencing of the sole allele produce a haploinsufficient state.
• Reduced PAX6 triggers upregulation of p53‑dependent apoptosis and SASP, diminishing holoclone‑forming capacity (<3% p63‑bright) and compromising long‑term graft survival.[3]

Testable Predictions

1. Expression correlation – In human limbal biopsies from donors aged 20–80, nuclear KDM6A and DDX3X protein levels will positively correlate with PAX6 mRNA and decline faster in males than females (linear mixed‑effects model, sex × age interaction p<0.01).
2. Allelic escape – RNA‑FISH combined with single‑cell RNA‑seq will show ~15‑20% of LESCs in female tissue exhibit biallelic transcription of KDM6A/DDX3X, whereas male cells show only monoallelic signals.
3. Functional rescue – CRISPR‑mediated knock‑down of KDM6A in female‑derived LESC cultures will reduce PAX6 by ~40% and halve holoclone frequency; overexpressing KDM6A in male LESCs will restore PAX6 to female‑like levels and improve colony‑forming efficiency.
4. In vivo aging – Kdm6a heterozygous male mice will exhibit earlier onset of corneal epithelial thinning and reduced label‑retaining LESCs compared with wild‑type males, phenocopying female wild‑type trajectories.

Falsifiability

If sex‑specific differences in PAX6 expression persist despite equalizing KDM6A/DDX3X dosage (e.g., via transgenic overexpression in males), or if X‑linked regulator expression does not correlate with PAX6 across ages, the hypothesis is refuted. Conversely, consistent confirmation across the four predictions would support a dosage‑dependent chromatin mechanism linking the X chromosome to limbal stem cell longevity.

References

[1] https://pmc.ncbi.nlm.nih.gov/articles/PMC9913671/ [2] https://pmc.ncbi.nlm.nih.gov/articles/PMC12859707/ [3] https://pubmed.ncbi.nlm.nih.gov/20573916/