Senescent Keratocyte Secretome Drives Age-Related Corneal Stromal Stiffening via Lysyl Oxidase‑Mediated Crosslinking Independent of Advanced Glycation End Products

Pita Silva (PhD)· Mar 24, 2026

Mechanism: Senescent keratocytes secrete SASP factors that activate LOX/LOXL2, leading to excessive collagen crosslinking and corneal stiffening. Readout: Readout: Senolytic clearance or LOX inhibition prevents fibrillar thickening, reduces stiffness, and improves corneal transparency and visual acuity metrics.

Hypothesis: Senescent keratocytes actively promote corneal stromal stiffening by secreting a matricellular secretome that upregulates lysyl oxidase (LOX) and LOXL2 activity, thereby catalyzing non‑enzymatic collagen crosslinking independent of glucose‑derived advanced glycation end products (AGEs). This secretory program precedes and amplifies AGE accumulation, linking keratocyte senescence directly to the biomechanical changes that impair transparency and refraction.

Background

Recent work shows aged keratocytes lose migratory capacity and form weaker collagen networks in vitro[1][2], while stromal fibrils increase in diameter, axial period, and Bragg spacing with age[3]. These structural shifts correlate with heightened stromal stiffness driven by AGE accumulation[4] and are exacerbated by UVA‑induced suppression of keratocyte collagen and proteoglycan synthesis[5]. Open questions include whether senescent keratocytes merely fail to maintain homeostasis or actively provoke aberrant crosslinking.

Mechanistic Insight

Senescent fibroblasts in skin secrete a senescence‑associated secretory phenotype (SASP) rich in TGF‑β1, periostin, and matrix metalloproteinase regulators that stimulate LOX expression[6]. We propose corneal keratocytes undergo a analogous SASP shift, releasing TGF‑β1 and periostin that act in an autocrine/paracrine loop to induce LOX/LOXL2 transcription via Smad3 and AP‑1 pathways. Elevated LOX catalyzes lysine‑derived crosslink formation on collagen fibrils, increasing fibril diameter and axial period without requiring glycation. This mechanism explains why crosslinking advances even in diabetic models where AGE formation is pharmacologically blocked, and why stromal stiffening correlates more tightly with keratocyte senescence markers than with pentosidine levels.

Testable Predictions

In human donor corneas stratified by age, LOX/LOXL2 mRNA and protein levels will correlate positively with p16INK4a+ keratocyte density and negatively with migratory assay outcomes, independent of pentosidine or fluorescence‑based AGE measurements.
Inducible, keratocyte‑specific expression of p16INK4a in young mice will recapitulate age‑like collagen fibrillogenesis (increased diameter and period) within 4 weeks, accompanied by elevated LOX activity and stromal shear‑modulus measurements.
Pharmacological inhibition of LOX (β‑aminopropionitrile) or genetic knockout of LOXL2 in senescent‑keratocyte mice will prevent fibrillar thickening and preserve corneal transparency despite persistent p16INK4a expression.
Senolytic clearance of p16INK4a+ keratocytes will reduce LOX secretion, decrease crosslink density, and improve visual acuity metrics (contrast sensitivity, astigmatism stability) without altering systemic AGE load.

Experimental Approach

Utilize a Krt12‑CreERT2;Rosa26‑LSL‑p16INK4a mouse model to induce senescence selectively in corneal keratocytes. Validate induction via immunofluorescence for p16INK4a and SA‑β‑gal.
Measure LOX/LOXL2 expression by qPCR, Western blot, and zymography; assess crosslink formation via hydroxylysyl pyridinoline (HPLC) and Schiff base fluorescence.
Perform second‑harmonic generation imaging to quantify fibril diameter, axial period, and intermolecular spacing; correlate with atomic force microscopy–derived stromal stiffness.
Apply topical β‑aminopropionitrile or LOXL2‑specific siRNA nanoparticles to test rescue of biomechanical properties.
Evaluate functional outcomes: slit‑lamp photography for opacity, ocular coherence tomography for thickness/profile, and optokinetic tracking for visual acuity.
Parallel human donor studies: laser‑capture microdissection of keratocyte layers from corneas aged <40, 40‑65, >65 years; quantify p16INK4a, LOX, LOXL2, and AGE markers via multiplex immunoassay.

Potential Outcomes and Implications

If LOX inhibition abrogates senescence‑driven fibrillar changes, it would reposition keratocyte senescence from a passive marker to an active driver of stromal biomechanics. This would prioritize senolytics or LOX‑targeted therapies over solely AGE‑focused strategies, aligning corneal anti‑aging approaches with emerging trials in dermal fibrosis. Conversely, a lack of effect would support the hypothesis that AGEs dominate crosslinking, redirecting focus to glycoxidation pathways and keratocyte‑mediated AGE clearance mechanisms.

Key References [1] https://pubmed.ncbi.nlm.nih.gov/11080534/ [2] https://iovs.arvojournals.org/article.aspx?articleid=2161217 [3] https://iovs.arvojournals.org/article.aspx?articleid=2181258 [4] https://www.frontiersin.org/journals/bioengineering-and-biotechnology/articles/10.3389/fbioe.2019.00066/full [5] https://onlinelibrary.wiley.com/doi/10.1111/acel.12324 [6] https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0292791

Background

Mechanistic Insight

Testable Predictions

In human donor corneas stratified by age, LOX/LOXL2 mRNA and protein levels will correlate positively with p16INK4a+ keratocyte density and negatively with migratory assay outcomes, independent of pentosidine or fluorescence‑based AGE measurements.

Inducible, keratocyte‑specific expression of p16INK4a in young mice will recapitulate age‑like collagen fibrillogenesis (increased diameter and period) within 4 weeks, accompanied by elevated LOX activity and stromal shear‑modulus measurements.

Pharmacological inhibition of LOX (β‑aminopropionitrile) or genetic knockout of LOXL2 in senescent‑keratocyte mice will prevent fibrillar thickening and preserve corneal transparency despite persistent p16INK4a expression.

Senolytic clearance of p16INK4a+ keratocytes will reduce LOX secretion, decrease crosslink density, and improve visual acuity metrics (contrast sensitivity, astigmatism stability) without altering systemic AGE load.

Experimental Approach

Utilize a Krt12‑CreERT2;Rosa26‑LSL‑p16INK4a mouse model to induce senescence selectively in corneal keratocytes. Validate induction via immunofluorescence for p16INK4a and SA‑β‑gal.

Measure LOX/LOXL2 expression by qPCR, Western blot, and zymography; assess crosslink formation via hydroxylysyl pyridinoline (HPLC) and Schiff base fluorescence.

Perform second‑harmonic generation imaging to quantify fibril diameter, axial period, and intermolecular spacing; correlate with atomic force microscopy–derived stromal stiffness.

Apply topical β‑aminopropionitrile or LOXL2‑specific siRNA nanoparticles to test rescue of biomechanical properties.

Evaluate functional outcomes: slit‑lamp photography for opacity, ocular coherence tomography for thickness/profile, and optokinetic tracking for visual acuity.

Parallel human donor studies: laser‑capture microdissection of keratocyte layers from corneas aged <40, 40‑65, >65 years; quantify p16INK4a, LOX, LOXL2, and AGE markers via multiplex immunoassay.