Age‑dependent stromal stiffening drives PAX6 cytoplasmic sequestration and YAP‑mediated loss of limbal stem cell plasticity via macrophage‑derived TNF‑α signaling

Vera Akintola (DPhil)· Mar 26, 2026

Mechanism: Age-related limbal stromal stiffening triggers Src-mediated PAX6 phosphorylation and cytoplasmic retention, leading to YAP-TEAD activation and a macrophage-amplified TNF-α/NF-κB feedback loop that further sequesters PAX6. Readout: Readout: This cycle diminishes LSC plasticity and de-differentiation capacity, which can be rescued by pharmacological inhibition of Src, YAP, or NF-κB, or by stromal softening.

Hypothesis

Background

Aging compromises limbal stem cell (LSC) function through UV‑induced DNA damage, altered Wnt/BMP signaling, and disrupted PAX6 regulation 1. PAX6 is a nuclear master transcription factor that maintains corneal epithelial identity and suppresses epidermal fate 2 3. In Pax6 heterozygous mice, reduced PAX6 expands quiescent LESC‑like clusters, increases apoptosis (Lcn2) and impairs differentiation (↓Krt12, ↑Krt4) 4. Corneal macrophages secrete TNF‑α, IL‑6, TGF‑β and IL‑10 that modulate LSC behavior in a context‑dependent manner 5. Differentiated corneal epithelial cells can de‑differentiate into functional LSCs via Hippo/YAP after injury 6, but whether this plasticity declines with age is unknown.

Core Hypothesis

We hypothesize that age‑related increase in limbal stromal stiffness triggers mechanotransductive signaling that sequesters PAX6 from the nucleus to the cytoplasm, thereby diminishing its transcriptional repression of epidermal programs and permitting sustained YAP activation. Cytoplasmic PAX6 loses its ability to inhibit YAP‑TEAD transcriptional activity, leading to a feedback loop where YAP drives expression of pro‑fibrotic and inflammatory cytokines (e.g., TNF‑α) from resident macrophages, which in turn activate NF‑κB pathways that further phosphorylate PAX6, promoting its nuclear export. This creates a self‑reinforcing cycle that locks aged LSCs into a senescent, epigenetically restricted state, impairing their capacity to undergo YAP‑mediated de‑differentiation after injury.

Mechanistic Model

Stiffness sensing – Elevated collagen cross‑linking in the aged limbal stroma increases tensile stress transmitted to basal epithelial cells via integrin‑FAK signaling.
PAX6 cytoplasmic shift – Mechanical stress activates Src family kinases that phosphorylate PAX6 on serine residues, creating a binding site for 14‑3‑3 proteins that retain PAX6 in the cytoplasm 2.
YAP disinhibition – Cytoplasmic PAX6 can no longer sequester YAP in the cytoplasm or impede its nuclear translocation, allowing YAP‑TEAD to drive transcription of CTGF, CYR61 and pro‑inflammatory mediators.
Macrophage amplification – YAP‑dependent epithelial secretion of CCL2 and IL‑1α recruits and activates limbal macrophages, which release TNF‑α. TNF‑α triggers IKK‑mediated NF‑κB signaling that phosphorylates PAX6, enhancing its cytoplasmic retention and reducing its DNA‑binding affinity.
Loss of plasticity – The combined PAX6 loss of nuclear function and sustained YAP activity locks epithelial cells in a transitional state that resists the Hippo/YAP‑driven de‑differentiation program required for LSC regeneration after ablation.

Predictions and Experimental Design

Prediction 1

Human limbal biopsies from donors >60 yr will show a significantly higher cytoplasmic:nuclear PAX6 ratio compared with donors <30 yr, correlating with increased stromal Young’s modulus measured by atomic force microscopy (AFM).

Prediction 2

Ex‑vivo aged limbal explants subjected to controlled stiffening (e.g., lysophosphatidylcholine‑induced cross‑linking) will exhibit reduced PAX6 nuclear localization and diminished colony‑forming efficiency (CFE) after epithelial injury, which can be rescued by pharmacological inhibition of Src (PP2) or YAP‑TEAD (verteporfin).

Prediction 3

Conditioned medium from TNF‑α‑treated macrophages will decrease PAX6 nuclear import in young LSCs, an effect blocked by NF‑κB inhibitor (BAY 11‑7082) or PAX6‑S2A phospho‑mutant rescue.

Experimental Approach

Collect human limbal rims from eye banks stratified by age (n = 10 per group). Perform immunofluorescence for PAX6, YAP, and collagen; quantify nuclear/cytoplasmic intensity ratios.
Measure stromal stiffness via AFM indentation mapping.
Establish primary LSC cultures; treat with stiffness‑modulating agents (LOX inhibitor β‑aminopropionitrile to soften, LOX to stiffen). Assess PAX6 localization, YAP target gene expression (qPCR), and CFE after 5 Gy UV‑C or mechanical scrape.
Co‑culture LSCs with autologous limbal macrophages; manipulate TNF‑α/NF‑κB pathways and evaluate PAX6 shuttling via live‑cell imaging of PAX6‑GFP.
Use CRISPR‑edited PAX6 mutants (S2A phospho‑deficient, S2D phospho‑mimetic) to test causality.

Potential Outcomes and Interpretation

If aged tissue demonstrates elevated stromal stiffness, cytoplasmic PAX6, and impaired YAP‑mediated plasticity that is rescued by softening the matrix or blocking Src/YAP/NF‑κB, the hypothesis will be supported. Conversely, if PAX6 remains nuclear despite stiffness, or if macrophage‑derived TNF‑α fails to alter PAX6 localization, the model would be refuted, prompting investigation of alternative age‑dependent regulators (e.g., epigenetic silencing, microRNA‑mediated decay).

References

[1] https://pmc.ncbi.nlm.nih.gov/articles/PMC12785831/ [2] https://pmc.ncbi.nlm.nih.gov/articles/PMC9913671/ [3] https://pmc.ncbi.nlm.nih.gov/articles/PMC4536450/ [4] https://pmc.ncbi.nlm.nih.gov/articles/PMC12859707/ [5] https://www.tandfonline.com/doi/full/10.1080/07853890.2026.2616082 [6] https://doi.org/10.1101/2024.06.11.596348

Background

Core Hypothesis

Mechanistic Model

Stiffness sensing – Elevated collagen cross‑linking in the aged limbal stroma increases tensile stress transmitted to basal epithelial cells via integrin‑FAK signaling.

PAX6 cytoplasmic shift – Mechanical stress activates Src family kinases that phosphorylate PAX6 on serine residues, creating a binding site for 14‑3‑3 proteins that retain PAX6 in the cytoplasm 2.

YAP disinhibition – Cytoplasmic PAX6 can no longer sequester YAP in the cytoplasm or impede its nuclear translocation, allowing YAP‑TEAD to drive transcription of CTGF, CYR61 and pro‑inflammatory mediators.

Macrophage amplification – YAP‑dependent epithelial secretion of CCL2 and IL‑1α recruits and activates limbal macrophages, which release TNF‑α. TNF‑α triggers IKK‑mediated NF‑κB signaling that phosphorylates PAX6, enhancing its cytoplasmic retention and reducing its DNA‑binding affinity.

Loss of plasticity – The combined PAX6 loss of nuclear function and sustained YAP activity locks epithelial cells in a transitional state that resists the Hippo/YAP‑driven de‑differentiation program required for LSC regeneration after ablation.

Predictions and Experimental Design

Prediction 1

Prediction 2

Prediction 3

Conditioned medium from TNF‑α‑treated macrophages will decrease PAX6 nuclear import in young LSCs, an effect blocked by NF‑κB inhibitor (BAY 11‑7082) or PAX6‑S2A phospho‑mutant rescue.

Experimental Approach

Collect human limbal rims from eye banks stratified by age (n = 10 per group). Perform immunofluorescence for PAX6, YAP, and collagen; quantify nuclear/cytoplasmic intensity ratios.

Measure stromal stiffness via AFM indentation mapping.

Establish primary LSC cultures; treat with stiffness‑modulating agents (LOX inhibitor β‑aminopropionitrile to soften, LOX to stiffen). Assess PAX6 localization, YAP target gene expression (qPCR), and CFE after 5 Gy UV‑C or mechanical scrape.

Co‑culture LSCs with autologous limbal macrophages; manipulate TNF‑α/NF‑κB pathways and evaluate PAX6 shuttling via live‑cell imaging of PAX6‑GFP.

Use CRISPR‑edited PAX6 mutants (S2A phospho‑deficient, S2D phospho‑mimetic) to test causality.

Potential Outcomes and Interpretation