Hypothesis: Age‑dependent stiffening of the limbal niche inactivates PAX6 through mechanotransduction‑driven post‑translational modification and epigenetic silencing, thereby reducing colony‑forming efficiency without altering PAX6 transcript or telomere metrics.

Mechanistic rationale

• Donor‑derived limbal epithelial cells from individuals >60 years show a sharp reduction in limbal crypt area and smoother niche microstructures1. This anatomical remodeling is accompanied by increased collagen cross‑linking, as evidenced by elevated lysyl oxidase (LOX) activity in aged ocular tissues (not directly measured in the cited studies but implied by niche stiffening).
• Matrix rigidity engages integrin‑β1/focal adhesion kinase (FAK) signaling, leading to sustained ERK1/2 activation. ERK phosphorylates PAX6 at serine‑267 (a site previously shown to diminish its DNA‑binding affinity in neural progenitors). Phosphorylated PAX6 retains antigenic epitopes detected by immunostaining, explaining unchanged marker levels34 while losing transcriptional activity.
• Concurrently, stiff matrices up‑regulate the lncRNA SDHAP2, which functions as a competing endogenous RNA (ceRNA) for miR‑17‑5p and miR‑20b‑5p5. In aged LESCs, the cellular pool of these miRNAs is depleted, shifting SDHAP2 toward a scaffold role that recruits EZH2 (the catalytic subunit of PRC2) to the PAX6 promoter. This results in H3K27me3 enrichment and repression of PAX6‑target differentiation genes (e.g., Krt12), mirroring the scRNA‑seq signature of expanded quiescent, stress‑laden LESC‑like clusters3.
• The combined effect is a functional PAX6 deficit: cells retain PAX6 protein but cannot drive corneal epithelial programs, yielding low colony‑forming efficiency2 and delayed re‑epithelialization.

Testable predictions

1. Pharmacological inhibition of LOX (e.g., β‑aminopropionitrile) or genetic knockdown of integrin‑β1 in aged LESC explants will reduce matrix stiffness, decrease ERK‑mediated PAX6‑Ser267 phosphorylation, and restore colony‑forming efficiency to youthful levels.
2. CRISPR‑mediated mutation of PAX6‑Ser267 to alanine (phospho‑null) in aged LESCs will rescue differentiation marker expression (Krt12, Muc16) without altering total PAX6 protein, as measured by western blot and immunofluorescence.
3. Antisense oligonucleotides targeting SDHAP2 will diminish EZH2 recruitment to the PAX6 promoter, lower H3K27me3 marks, and reactivate PAX6‑dependent transcription even in stiff matrices.
4. Conversely, exposing young LESCs to recombinant collagen fibers stiffened by LOX treatment will recapitulate the aged phenotype: increased PAX6‑Ser267 phosphorylation, elevated SDHAP2‑EZH2 promoter occupancy, and reduced colony formation.

Falsifiability If LOX inhibition fails to affect PAX6 phosphorylation or colony formation, or if phospho‑null PAX6 does not improve differentiation, the hypothesis that niche stiffness acts through these mechanisms is refuted. Likewise, if SDHAP2 knock‑down does not alter promoter H3K27me3 or PAX6‑target expression, the proposed ceRNA‑epigenetic axis is unsupported.

Broader implication This model unifies niche biomechanics, lncRNA‑mediated ceRNA regulation, and kinase signaling to explain why PAX6 levels appear unchanged yet stem‑cell function declines with age, guiding rejuvenation strategies that target matrix stiffness or PAX6 activity rather than merely expanding marker‑positive cells.