Hypothesis

Chronic UVA exposure elevates mitochondrial reactive oxygen species (ROS) in corneal keratocytes, leading to sustained activation of the p16INK4a‑RB senescence pathway. Senescent keratocytes secrete a senescence‑associated secretory phenotype (SASP) that includes lysyl oxidase‑like enzymes and TGF‑β1, which drive non‑enzymatic collagen crosslinking and stromal stiffening. This mechanism explains the linear decline in keratocyte density and the age‑related increase in corneal Young’s modulus, positioning keratocyte senescence as a primary mediator of stromal aging rather than a passive by‑product.

Rationale

• Keratocyte density falls ~0.3 % per year, a linear trend that mirrors cumulative UV exposure rather than a stochastic process【1】.
• Chronic UVA reproduces aged‑cornea ECM changes, including increased collagen glycation and crosslinking【2】.
• Senescent fibroblasts in other tissues upregulate lysyl oxidase, promoting collagen crosslink formation【3】.
• Mitochondrial ROS are known activators of p16INK4a senescence in ocular surface cells【4】.

Mechanistic Steps

1. UVA absorption by corneal chromophores (e.g., flavins) generates mitochondrial superoxide.
2. ROS‑mediated DNA damage stabilizes p16INK4a expression, arresting keratocytes in a senescent state.
3. SASP release includes TGF‑β1, IL‑6, and lysyl oxidase‑2 (LOXL2).
4. LOXL2 catalyzes lysine oxidation on collagen fibrils, forming aldehyde‑derived crosslinks that increase stiffness.
5. Crosslinked collagen disrupts the regular proteoglycan‑collagen lattice, reducing transparency and increasing light scattering.

Testable Predictions

• Prediction 1: Older donor corneas will show higher mitochondrial ROS levels (measured by MitoSOX fluorescence) that correlate with p16INK4a‑positive keratocyte density.
• Prediction 2: Ex vivo cornea cultures exposed to UVA will exhibit increased LOXL2 activity and collagen crosslink density; pretreatment with a mitochondrial ROS scavenger (MitoTEMPO) will attenuate both senescence markers and crosslinking.
• Prediction 3: Genetic knockdown of p16INK4a in human keratocyte‑like cells will reduce SASP factor secretion and protect collagen from UVA‑induced crosslinking, even under chronic exposure.
• Prediction 4: Topical application of a senolytic agent (e.g., dasatinib + quercetin) to aged mouse eyes will decrease keratocyte senescence markers, lower corneal Young’s modulus measured by Brillouin microscopy, and improve transparency metrics.

Experimental Approach

• Use human donor corneas stratified by age; quantify mitochondrial ROS, p16INK4a, LOXL2, and hydroxylysyl pyridinoline crosslinks via immunofluorescence and mass spectrometry.
• Organ culture model: treat corneas with UVA (365 nm, 1 J/cm²/day) for 14 days ± MitoTEMPO (500 nM) or vehicle; assess keratocyte senescence (SA‑β‑gal), SASP cytokines (ELISA), and biomechanical properties (inflation testing).
• CRISPR‑i p16INK4a in immortalized human keratocyte line; expose to UVA; measure SASP and crosslink formation.
• In vivo: apply dasatinib + quercetin eye drops to 12‑month‑old mice for 4 weeks; compare to untreated controls via in vivo confocal microscopy (keratocyte density), Brillouin scanning (stiffness), and slit‑lamp grading (transparency).

Falsifiability

If mitochondrial ROS levels do not rise with age, or if scavenging ROS fails to reduce p16INK4a expression and collagen crosslinking in UVA‑treated corneas, the hypothesis would be refuted. Similarly, if p16INK4a loss does not diminish SASP secretion or protect against crosslinking, the proposed senescence‑driven mechanism would be invalid.

Implications

Targeting mitochondrial ROS or the p16INK4a‑SASP axis could preserve keratocyte population, maintain stromal biomechanics, and delay age‑related loss of transparency—offering a therapeutic avenue distinct from current crosslinking‑based approaches.