Hypothesis

In aged human limbus, chronic oxidative stress converts autophagy from a selective cargo transporter into a bulk degradation system that aberrantly targets PAX6 for lysosomal delivery, creating a functional haploinsufficiency without altering total PAX6 levels. This sequestration occurs via a stress‑induced LC3‑interacting region (LIR) on PAX6 that becomes exposed after ROS‑mediated cysteine oxidation, allowing PAX6 to be captured by autophagosomes and degraded, thereby lowering nuclear PAX6 below the threshold needed to sustain corneal epithelial identity.

Mechanistic Basis

1. Stress‑dependent LIR exposure – UV‑A‑generated ROS oxidize specific cysteines in the PAX6 PAI domain, uncovering a cryptic LIR motif (e.g., WxxL) that binds LC3B with high affinity.
2. Shift in autophagy selectivity – Aged limbal niche shows upregulated lysosomal biogenesis (TFEB nuclear) but downregulated selective autophagy receptors (p62/SQSTM1, NBR1) due to chronic mTORC1 inhibition, favoring bulk autophagosome formation.
3. Functional consequence – Cytoplasmic sequestration reduces nuclear PAX6, diminishing transcription of cornea‑specific keratins (K3, K12) and Notch/Wnt targets, while permitting ectopic skin‑like programs (K10, involucrin).

Experimental Plan

Aim 1: Detect ROS‑dependent PAX6‑LC3 interaction in aged limbal tissue

• Obtain limbal explants from donors <30 years (young) and >60 years (aged).
• Treat ex vivo with H2O2 (100 µM, 1 h) or vehicle.
• Perform proximity ligation assay (PLA) for PAX6‑LC3B and quantify puncta per nucleus.
• Use mass spectrometry to identify oxidized cysteines on PAX6 after UV‑A exposure. Prediction: Aged tissue shows basal PLA signal that increases with H2O2; young tissue shows minimal signal unless stressed.

Aim 2: Test whether blocking PAX6 LIR rescues nuclear PAX6 and stemness

• Introduce a cell‑penetrant peptide mimicking the PAX6 LIR (WxxL) to competitively inhibit LC3 binding, or express PAX6‑LIR mutant (W→A) via lentivirus in aged explants.
• Measure nuclear vs cytoplasmic PAX6 by immunofluorescence and subcellular fractionation.
• Assess colony‑forming efficiency (CFE) and expression of corneal keratins (K3/K12) versus epidermal markers (K10). Prediction: LIR blockade restores nuclear PAX6 to young‑like levels and doubles CFE in aged explants without changing total PAX6 protein.

Aim 3: Determine if modulating autophagy flux alters the siege phenotype

• Treat aged explants with rapamycin (mTORC1 inhibitor) to induce bulk autophagy, or with Torin1 plus lysosomal inhibitor chloroquine to block flux.
• Alternatively, overexpress TFEB to increase lysosomal biogenesis while adding the LIR peptide to enforce selectivity.
• Readouts: LC3‑II turnover (Western blot), PAX6‑LC3 PLA, CFE. Prediction: Bulk autophagy induction worsens PAX6 mislocalization and reduces CFE; enhancing lysosomal capacity while preserving selective receptor activity (via LIR peptide) rescues phenotype.

Falsifiability

If aged limbal tissue shows no ROS‑dependent PAX6‑LC3 interaction, or if blocking the LIR fails to restore nuclear PAX6 and CFE despite confirmed peptide delivery, the hypothesis is refuted. Conversely, rescue of nuclear PAX6 and function by LIR blockade would support autophagy as a siege‑mode rationing system that selectively consumes a key transcription factor, offering a reversible target for rejuvenating aged limbal stem cells.