Wnt activation as a potential therapeutic approach to treat partial limbal stem cell deficiency

Abstract

Limbal epithelial stem/progenitor cells (LSCs) are adult stem cells located at the limbus, tightly regulated by their niche involving numerous signaling pathways, such as Wnt. Wnt proteins are secreted morphogens that play critical roles in embryonic development, stem cell proliferation, self-renewal, tissue regeneration, and remodeling in adults. It has been shown that a small molecule Wnt mimic could improve LSCs expansion ex vivo. Damage to the LSCs and/or their niche can lead to limbal stem cell deficiency (LSCD), a condition that can cause corneal blindness and is difficult to treat. This study explored if repopulating residual LSCs in partial LSCD through Wnt activation could be a novel therapeutic approach. To mimic LSCD due to a chemical injury, single cultured LSCs were exposed to various concentrations of sodium hydroxide. A progressive loss of the LSCs phenotype was observed: the percentage of p63^bright cells and cytokeratin (K)14⁺ cells decreased while the percentage of K12⁺ increased. Wnt activation was attained by treating the LSCs with lithium chloride (LiCl) and a small-molecule Wnt mimic, respectively. After 18 h of treatment, LSCs proliferation was increased, and the LSCs phenotype was recovered, while the untreated cells did not proliferate and lost their phenotype. The percentage of p63^bright cells was significantly higher in the Wnt mimic-treated cells compared with untreated cells, while the percentage of K12⁺ cells was significantly lower. These findings suggest that local Wnt activation may rescue LSCs upon alkaline injury.

Figure 1

Slit lamp picture with fluorescein staining and in vivo laser scanning confocal microscopy of three eyes with severe partial LSCD. (A) Slit lamp picture of an eye with severe LSCD from chemical injury. (B) Fluorescein staining and blue cobalt light showing a diffuse vortex staining involving all the cornea corresponding to a partial severe LSCD, stage IIB, score 8 points. (C) IVCM confirming the diagnosis of severe LSCD. The central BCD is decreased but residual corneal and limbal epithelial cells can be seen in the central cornea and limbal quadrants, respectively. (D) Slit lamp picture of an eye with severe partial LSCD from prolonged contact lens wear. (E) Fluorescein staining and blue cobalt light showing a diffuse vortex staining involving all the cornea corresponding to a partial severe LSCD, stage IIB, score 8 points. (F) IVCM confirming the diagnosis of severe LSCD. The central BCD is decreased but residual corneal and limbal epithelial cells can be seen in the central cornea and limbal quadrants, respectively. (G) Slit lamp picture of an eye with severe partial LSCD from chemical injury 36 years prior. (H) Fluorescein staining and blue cobalt light showing a diffuse vortex staining corresponding to a partial severe LSCD, stage III, score 10 points. (I) IVCM confirming the diagnosis of severe LSCD. The central BCD is decreased but residual corneal and limbal epithelial cells can be seen in the central cornea and limbal quadrants, respectively, suggesting that this eye had partial severe LSCD.

Figure 2

In vitro model of limbal stem cell deficiency: p63 immunostaining. Cultured LSCs exposed for one minute to increasing concentrations of sodium hydroxide, progressively lost their p63^bright cells phenotype. Scale bar 20 µm.

Figure 3

In vitro model of limbal stem cell deficiency: K12/K14 immunostainings. Cultured LSCs exposed for one minute to increasing concentrations of sodium hydroxide, progressively lost their K14 + phenotype. The percentage of K12 + positive slightly increased but the difference was not significant. Scale bar 20 µm.

Figure 4

In vitro model of limbal stem cell deficiency: PCK/Vimentin immunostaining. Cultured LSCs exposed for one minute to increasing concentrations of sodium hydroxide, progressively lost their PCK positive phenotype. Scale bar 20 µm.

Figure 5

In vitro model of limbal stem cell deficiency: immunostaining quantification. (A) The decrease in the percentage of p63^bright cells was statistically lower starting at 100 µM NaOH (p < 0.05). (B) The decrease in the percentage of K14 + cells was statistically lower starting at 100 µM NaOH (p < 0.05). The percentage of K12 + positive slightly increased but the difference was not significant. (C) The decrease in the percentage of PCK positive cells was significant starting at 500 µM of NaOH (p < 0.05).

Figure 6

Limbal epithelial cell phenotype after treatment by Wnt activators. (A) Immunostainings of p63^bright cells after treatment by Wnt mimic, LiCl, or SHEM after chemical burn. (B) Treatment with Wnt mimic significantly increased the percentage of p63^bright cells compared with the control and LiCl (p < 0.05). (C) Immunostaining of K12 and K14 cells after treatment by Wnt mimic, LiCl, or SHEM after chemical burn. (D) The percentage of K12 + cells was significantly lower in the Wnt mimic and LiCl group (p < 0.05) (E) The percentage of K14 + cells was comparable in all groups of treatment. (F) Cell proliferation was significantly increased in the Wnt mimic group compared to the control (p < 0.05). Data are expressed as fold change of expression compared with the unburned SHEM control. One way ANOVA with multiple comparisons and Dunnett post-hoc test. Mean ± SEM of 4 individual experiments. Scale bar 20 µm.