Effects of 1α,25-Dihydroxyvitamin D₃ on Langerhans Cell Migration and Corneal Neovascularization in Mice

Abstract

purpose. To examine the effects of 1α,25-dihydroxyvitamin D₃ (1α,25[OH]₂D₃), a hormone that has immunosuppressive properties, on Langerhans cell (LC) migration and corneal neovascularization in mouse corneas.

methods. Two 10-0 nylon interrupted sutures were placed in the center of 50 BALB/c mouse corneas to induce LC migration and corneal neovascularization. The mice were then randomly assigned to one of five groups. Three groups (n = 11, n = 11, n = 6) received topical 1α,25(OH)₂D₃ (at concentrations of 10⁻⁷ M, 10⁻⁸ M, 10⁻⁹ M), one group (n = 11) received vehicle only, and one group (n = 11) received no eye drops. Instillation (three times a day) began on the first day after suturing. Corneal neovascularization was assessed by slit lamp microscopy and scored according to the length of newly formed corneal vessels. Fourteen days after suturing, the number of LCs that had migrated into the central corneal epithelium was counted by an immunofluorescence assay using an anti-Ia antibody.

results. The number of LCs in the central cornea was 21.9 ± 2.8 cells/mm² in the nontreated group and 17.8 ± 3.9 cells/mm² in the vehicle-only group. Significantly fewer LCs were detected in all groups that had received 1α,25(OH)₂D₃ compared with the vehicle only and nontreated groups (10⁻⁷ M: 7.4 ± 1.2 cells/mm², 10⁻⁸ M: 7.2 ± 2.0 cells/mm², 10⁻⁹ M: 6.2 ± 0.7 cells/mm²). Moderate inhibition of corneal vascularization was observed in the 10⁻⁷ M 1α,25(OH)₂D₃ group, but not the other groups.

conclusions. Topical administration of 1α,25(OH)₂D₃ can be effective in suppressing ocular surface inflammation by inhibiting LC migration into mouse corneas.

The normal cornea is a well known immunologically privileged tissue that has no blood or lymphatic vessels or antigen-presenting cells such as Langerhans cells (LCs) and macrophages. Despite this immune privilege, a wide range of factors (e.g., bacterial infection, viral infection, and trauma) can cause severe inflammation accompanied by corneal neovascularization and cellular infiltration. Often, this results in significant visual loss. Conventional therapy for corneal inflammation relies mainly on anti-inflammatory corticosteroids such as hydrocortisone, dexamethasone, and prednisolone. However, the use of these may contribute to cataract formation, microbial keratitis, and/or glaucoma. In view of this, new anti-inflammatory drugs to suppress corneal inflammation are desired. 

1α,25-dihydroxyvitamin D₃ (1α,25[OH]₂D₃), the active form of vitamin D₃, is a hormone involved in the regulation of calcium homeostasis. It exerts its effect by binding to an intracellular receptor that belongs to the family of nuclear receptors for steroids, thyroid hormones, and retinoic acid. ¹ The receptors for 1α,25(OH)₂D₃ have been found in almost all tissues and cells. Besides the regulation of calcium homeostasis, a wider biologic role for 1α,25(OH)₂D₃ has been reported that includes immune regulation and an influence on cell proliferation, differentiation, and gene transcription. ¹ Moreover, antiangiogenic effects of 1α,25(OH)₂D₃ have been reported in a transgenic murine retinoblastoma model, ² and it has been shown that 1α,25(OH)₂D₃ has a direct effect on LCs in the skin, because it suppresses their antigen-presenting ability in vitro. ³  

Because neovascularization and LC migration in the central cornea are often associated with severe corneal inflammation, we set out to examine whether topical administration of 1α,25(OH)₂D₃ inhibits LC migration and corneal neovascularization during corneal inflammation in mice. 

Materials and Methods

Animals

Six-week-old BALB/c mice (n = 50) were obtained from Shimizu Laboratory Supplies (Kyoto, Japan). All experimental procedures conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Induction of Corneal Inflammation

Mice were anesthetized and placed under the dissecting microscope where two interrupted sutures (10-0 nylon; Alcon, Tokyo, Japan) were placed in the centers of the right corneas. As described previously, ⁴ these sutures cause corneal inflammation, corneal neovascularization, and LC migration as early as 5 days after placement, and after 2 weeks, newly formed vessels occupy almost all quadrants of the cornea. 

Topical Administration of 1α,25(OH)₂D₃

1α,25(OH)₂D₃ (Chugai, Tokyo, Japan) was suspended in absolute ethanol at 10⁻⁴ M and stored at −20°C in a shielded tube. Various amounts of phosphate-buffered saline (Nissui Pharmaceutical, Tokyo, Japan) were added so that the final concentration of 1α,25(OH)₂D₃ was brought to 10⁻⁷ M, 10⁻⁸ M, and 10⁻⁹ M in a final ethanol concentration of 0.1%. One drop (40 μl) of each preparation was applied to a sutured mouse eye three times a day from the day after suturing until the end of the experiment. Control mice received vehicle only or no treatment. All experiments were performed in a double-masked fashion. 

Langerhans Cell Infiltration

LCs in sutured BALB/c corneas were examined by immunohistochemistry using immunofluorescence secondary to an I-A^d–specific antibody, as described previously. ⁵ The assay was performed on wholemounts of pure corneal epithelium, separated from the underlying stroma after incubation in 20 mM EDTA. 

Corneal Neovascularization

Corneal neovascularization was assessed by slit lamp microscopy every other day after the placement of sutures. To give an overall assessment of the extent of the neovascularization, it was scored 0 to 3, according to the length of newly formed corneal vessels as follows: 0, no vessels; 1, vessels only in the peripheral cornea (within 1 mm of the limbus); 2, vessels extending over 1 mm from the limbus; 3, vessels reaching the center of the cornea. In each eye, corneal neovascularization was scored independently in each quadrant of a cornea, and the sum of these four scores was recorded. 

Statistical Analyses

The numbers of LCs in the central cornea were evaluated with a two-tailed Student’s t-test. Corneal neovascularization scores were evaluated by the Mann–Whitney test. P < 0.05 was deemed significant. 

Results

LC Infiltration after 1α,25(OH)₂D₃ Administration

It has been reported that sutures placed in the central cornea induce inward Ia⁺ LC migration within 1 week of suturing and that the maximum number of LCs are detected in the cornea 2 weeks later. ⁴ To best examine the effect of 1α,25(OH)₂D₃ on LC infiltration, mice were killed 2 weeks after suturing, and their eyes enucleated to count LCs in the central cornea using a fluorescein isothiocyanate (FITC)–labeled anti I-A^d antibody. Whereas a large number of Ia⁺ cells were detected in the central corneas of untreated mice (21.9 ± 2.8 cells/mm²; Fig. 1A ) and mice treated with vehicle only (17.8 ± 3.9 cells/mm²; Fig. 1B ), significantly fewer Ia⁺ cells were detected in the central corneas of mice treated with 1α,25(OH)₂D₃ at all concentrations used (10⁻⁷ M: 7.4 ± 1.2 cells/mm²; 10⁻⁸ M: 7.2 ± 2.0 cells/mm²; 10⁻⁹ M: 6.2 ± 0.7 cells/mm²; Fig. 1C ). The Ia⁺ cells that had migrated into the central cornea contained fewer dendrites. Therefore, the Ia⁺ cells in Figure 1C were probably LCs; however, we cannot exclude the possibility that they may have been macrophages or non-LC cells because of their immature morphology. The mean densities of Ia⁺ cells in the central corneas of each group of animals is shown in Figure 2 . The values clearly indicate that the topical administration of 1α,25(OH)₂D₃ inhibits LC migration during corneal inflammation caused by the placement sutures in the central mouse cornea. 

Corneal Neovascularization after 1α,25(OH)₂D₃ Administration

Sutures placed in the central corneas of mice not receiving 1α,25(OH)₂D₃ induced corneal neovascularization as early as 2 days after their placement. Vessels grew into the central cornea so that after the sutures had been in place for 2 weeks all untreated mice had neovascularization scores of 8 or higher (9.17 ± 0.48, mean ± SD). To examine whether 1α,25(OH)₂D₃ has any antiangiogenic effects on corneal neovascularization, we plotted the time course of the mean neovascularization scores for each group of 1α,25(OH)₂D₃-treated mice (Fig. 3) . This disclosed no inhibitory effects in mice treated with 10⁻⁸ M and 10⁻⁹ M 1α,25(OH)₂D₃. However, 1 week after suturing, topical administration of 10⁻⁷ M 1α,25(OH)₂D₃ was found to have slightly inhibited the development of corneal neovascularization. This trend continued, and 2 weeks after suturing, only 4 of 11 mice treated with 10⁻⁷ M 1α,25(OH)₂D₃ had neovascularization scores of 8 or higher. In contrast, 9 of 11 mice treated with vehicle only had scores of 8 or higher, as did all 11 untreated mice. These results indicate that, if the concentration is high enough, topical administration of 1α,25(OH)₂D₃ has a mild antiangiogenic effect on corneal neovascularization induced by sutures in mouse eyes. Our present data are limited to a 2-week observation period. We have next to examine animals treated with 1α,25(OH)₂D₃ for longer than 2 weeks to examine the longevity of this effect. 

Discussion

In this series of experiments, we show that the topical administration of 1α,25(OH)₂D₃ inhibits the LC migration and corneal neovascularization that are ordinarily seen when sutures are placed in the center of a mouse cornea. The inhibitory effect was greater on LC migration than on neovascularization because the lowest concentration of 1α,25(OH)₂D₃ (i.e., 10⁻⁹ M) was enough to inhibit LC migration, whereas only the highest concentration (i.e., 10⁻⁷ M) effectively suppressed corneal neovascularization. The longevity of this effect is unknown at present, because we have yet to examine animals treated with 1α,25(OH)₂D₃ for longer than 2 weeks. 

The effects of 1α,25(OH)₂D₃ on LCs have been demonstrated recently by several investigators. For example, Dam et al. ⁶ showed that topical administration of 1α,25(OH)₂D₃ suppressed the number and antigen-presenting function of LCs in human skin both in vitro and in vivo. In their experiments, the application of calcipotriol cream (an analog of 1α,25[OH]₂D₃) to normal human skin for 4 days resulted in a decrease in the number of CD1a-positive cells with a dendritic morphology, and also in the number of dendrites per cell. Our experiments also disclosed a decrease in the number of LCs in the cornea after topical application of 1α,25(OH)₂D₃ and, similar to the work of Dam et al., ⁶ document changes in cell morphology suggesting that 1α,25(OH)₂D₃ may have similar effects on LCs both in skin and on the ocular surface. 

In another investigation of LCs in skin, Bagot et al. ³ showed that 1α,25(OH)₂D₃ decreased the ability of LCs to stimulate allogeneic lymphocytes in vitro. In their experiments 1α,25(OH)₂D₃ seemed to act directly on LCs because freshly isolated LCs were used. Also, 1α,25(OH)₂D₃ was found to inhibit the ability of LC-depleted keratinocytes to stimulate allogeneic lymphocytes, suggesting that, in vivo, the suppression of the immune response in skin may result from both a direct effect on LCs and indirect effects (for example, by modulating the production of cytokines by keratinocytes). In our experiments we observed the in vivo inhibitory effect of 1α,25(OH)₂D₃ on LC migration into the central cornea during inflammation. As in skin, this can conceivably be explained by both direct and indirect effects of 1α,25(OH)₂D₃ on LCs. For example, 1α,25(OH)₂D₃ may immobilize LCs directly through their receptors. It may, however, act on corneal epithelial cells and inhibit the production of cytokines such as interleukin (IL)-1, granulocyte-macrophage colony-stimulating factor (GM-CSF), and tumor necrosis factor (TNF)-α known to induce LC migration. Even though we observed fewer LCs in the central cornea after administration of 1α,25(OH)₂D₃, it is not clear whether 1α,25(OH)₂D₃ inhibits LC migration into the cornea or suppresses the major histocompatibility complex (MHC) class II expression of LCs. However, as we could not find any differences between the number of LCs in conjunctiva treated with 1α,25(OH)₂D₃ and conjunctiva treated with vehicle only (data not shown), and because 1α,25(OH)₂D₃ does not affect MHC class II expression of LCs in human skin in vitro, ⁶ it is likely that 1α,25(OH)₂D₃ inhibits LC migration into the central cornea from conjunctiva. 

Many investigators have reported the antiangiogenic effect of the systemic administration of 1α,25(OH)₂D₃ in mice. ^{2 7 8} Most of these experiments have shown 30% to 60% suppression of vessel formation in mice treated with 1α,25(OH)₂D₃ compared with control mice. Our results showed a similar suppressive effect of 1α,25(OH)₂D₃ on corneal neovascularization, except that it was noted only at the highest concentration chosen (i.e., 10⁻⁷ M), not at lower concentrations (i.e., 10⁻⁸ M, and 10⁻⁹ M). The reasons that lower concentrations of 1α,25(OH)₂D₃ inhibit LC migration but not corneal neovascularization in our experiments are not readily evident. However, it has been reported that some cytokines induce LC migration (IL-1, GM-CSF, and TNF- α, for example), whereas others induce corneal neovascularization (vascular endothelial growth factor [VEGF], basic-fibroblast growth factor [b-FGF], and IL-8, for example). ^{9 10} Our preliminary experiments using human corneal epithelial cells show that 1α,25(OH)₂D₃ inhibits IL-1α and IL-1β production more than IL-8 production. ¹¹ Therefore, it is possible that 1α,25(OH)₂D₃ acts by regulating cytokine or growth factor production by corneal epithelial cells in vivo and that it inhibits cytokines that induce LC migration more than it inhibits those that induce corneal neovascularization. 

We have demonstrated that the topical administration of 1α,25(OH)₂D₃ inhibits ocular surface inflammation in mice examined 2 weeks after the placement of sutures in their central corneas. In skin, 1α,25(OH)₂D₃ is used clinically to treat psoriasis and certain other diseases. We are currently examining the toxicity of 1α,25(OH)₂D₃ on corneal epithelial cells to assess its suitability for use as an agent to inhibit corneal neovascularization. 

 Supported in part by Research Grants 10470365 and 10770946 from the Japanese Ministry of Education, Culture and Science; a Health Sciences Research Grant from the Japanese Ministry of Health and Welfare; and a research fund grant from Kyoto Foundation for the Promotion of Medical Science.

 Submitted for publication February 9, 1999; revised May 18 and July 22, 1999; accepted August 2, 1999.

 Commercial relationships policy: N.

 

Figure 1.

 

Immunohistochemical staining of Ia⁺ Langerhans cells in central mouse corneas near where sutures had been placed. Numerous Ia⁺ dendritic cells are seen in the corneal epithelium of untreated mice (A) and mice treated with vehicle only (B). Only a few Ia⁺ dendritic cells with less dendrites are found in the corneal epithelium of mice treated with 1α,25(OH)₂D₃ at a concentration of 10⁻⁹ M (C). Bar, 50 μm.

Figure 2.

 

Density of Ia⁺ Langerhans cells in the central cornea 14 days after suturing. Wholemounts of corneal epithelium from mice after topical administration of vehicle or 1α,25(OH)₂D₃ were stained with an anti I-A^d antibody. I-A^d–positive cells in the central cornea were counted in each group, and their means ± SEM are presented. ∗Mean values significantly lower than in vehicle-treated controls, P < 0.005 (Student’s t-test).

Figure 3.

 

Time course of the development of corneal neovascularization after suturing of mouse corneas. In each group (n = 6 or 11), corneal neovascularization was scored independently in each mouse. The mean scores at each time point are presented.

The authors thank Andrew J. Quantock for helpful comments.