Corneal and Retinal Effects of Ultraviolet-B Exposure in a Soft Contact Lens Mouse Model

Abstract

Purpose.: To investigate the lipid and DNA oxidative stress as well as corneal and retinal effects after ultraviolet B (UV-B) exposure in mice, with or without silicon hydrogel soft contact lenses (SCL).

Methods.: Twenty-eight C57BL6-strain male mice were divided into four groups: group I, control group with no SCL (SCL [−]) and no UV-B exposure (UV-B [−]); group II, senofilcon A SCL (senofilcon [+]) with UV-B exposure (UV-B [+]); group III, lotrafilcon A SCL (lotrafilcon [+]) with UV-B exposure (UV-B [+]); and group IV, no SCL (SCL [−]), but with UV-B exposure (UV-B [+]). All mice except group I received UV-B exposure for 5 days for a total dose of 2.73 J/cm². All mice underwent tear hexanoyl-lysine (HEL) and tear cytokine ELISA measurements, and fluorescein and rose bengal corneal staining before and after UV-B exposure. Corneal specimens underwent immunohistochemistry staining with CD45, HEL, 4-hydroxynonenal (4-HNE), and 8-hydroxy-2′-deoxyguanosine (8-OHdG) antibodies and evaluation with electron microscopy.

Results.: All mice without SCL but exposed to UV-B developed corneal edema, ulcers, or epithelial damage compared with mice with senofilcon A SCL and exposure to UV-B. Tear HEL and cytokine levels significantly increased in mice without SCL after UV-B exposure. Immunohistochemistry showed a significantly higher number of cells positively stained for CD45, 8-OHdG, HEL, and 4-HNE in the corneas of mice without SCLs compared with those with senofilcon A after UV-B exposure.

Conclusion.: Silicon hydrogel SCL showed corneal and retinal protective effects, owing to UV blocking properties, against oxidative stress-related membrane lipid and cellular DNA damage.

Introduction

It has been well established that UV light exposure can cause many ophthalmic problems depending on the wavelength, duration, and intensity of exposure. While UV-C (100–280 nm) radiation is almost exclusively filtered by the atmosphere, the terrestrial environment is exposed to UV-B (290–320 nm) and UV-A radiation (320–400 nm); and of the wavelengths reaching the earth, 97% are UV-A, whereas only 3% are UV-B radiation. ¹ With the current rates of ozone depletion in the atmosphere, UV radiation exposure will increase in the future and cause a variety of ocular diseases. The mean ocular UV exposure reaching the cornea on a horizontal surface has been estimated to be between 2% and 17% of the total UV reaching the earth. ² Ultraviolet exposure has been reported to have detrimental effects on the cornea, the lens, the retina, and the aqueous humour ^3–6 including spheroidal degeneration, ⁷ UV keratitis, pinguecula, pterygium, climatic droplet keratopathy, lens opacification, vitreous liquefaction, macular edema and degeneration. The human cornea absorbs 60% to 100% of radiation in the UV-B region, in contrast to only 20% to 40% in the UV-A region of the UV spectrum. ⁸ Ultraviolet rays can induce production of reactive oxygen species, ⁹ inactivate corneal antioxidant enzymes, ¹⁰ and induce cell death in cultured corneal cells. ¹¹ Because of these observations, researchers and clinicians have advocated incorporating UV absorbers into sunglasses and contact lenses. Class-II UV-absorbing molecules have been available in contact lenses for many years. Most recently, class-I UV-absorbing silicone hydrogel polymers have been introduced and are claimed to provide a high level of UV protection relative to UV-induced ophthalmic exposure. ^12–17 Soft contact lenses (SCL) cover the cornea and the surrounding perilimbal conjunctiva, which may very well protect the ocular surface, in addition to the retinal structures of the eye that are vulnerable to UV-induced damage. In this study, we investigated the corneal and retinal effects of repeated UV-B exposure, on lipid and DNA oxidative damage, in an SCL mouse model, checking the effects of two types of SCLs with different UV-blocking properties. We also compared the findings in mice without SCLs but exposed to UV-B, as well as control mice without SCLs and not exposed to UV-B radiation. 

Methods

Twenty-eight C57BL6-strain wild-type male mice were divided into four groups (seven mice in each group) according to the type of SCL worn and presence of UV-B exposure as follows: group I, control group with no SCL (SCL [−]) and no UV-B exposure (UV-B [−]); group II, senofilcon A SCL (senofilcon [+], Johnson & Johnson Vision Care, Jacksonville, FL) with UV-B exposure (UV-B [+]); group III, lotrafilcon A SCL (lotrafilcon [+], CIBA Vision, Atlanta, GA) with UV-B exposure (UV-B [+]); and group IV, no SCL (SCL [−]), but with UV-B exposure (UV-B [+]). All mice except group I received UV-B (312 nm) exposure every day for 5 days for a total dose of 2.73 J/cm². All mice were anesthetized during UV-B exposure with intraperitoneal (IP) ketamine (6 mg/mL) and xyladine (4 mg/mL) anesthesia. After injections, the SCLs were trephinated centrally with a 3-mm dermal biopsy punch (Kai Medical, Tokyo, Japan). (See supplemental Figs. 1A–D.) 

Figure 1.

 

Anterior segment and corneal fluorescein staining photographs. Upper left inserts: photographs in mice without SCL and not exposed to UV-B (group I) on days 1 and 5 showing clear corneas without fluorescein staining. Upper right inserts: photographs in mice with senofilcon A SCL and exposed to UV-B (group II) showing clear corneas and lack of epithelial damage as evidenced by absence of fluorescein staining. Lower left inserts: photographs in mice with lotrafilcon A SCL and exposed to UV-B (group III) demonstrate clear corneas on day 1 and development of marked opacity, edema, and ulceration on day 5. Lower right inserts: photographs of mice without SCL and exposed to UV-B (group IV) show development of marked corneal opacity, edema, and ulceration on day 5 after UV-B exposure. (A and B) Changes of fluorescein and rose bengal staining scores with UV-B exposure. Fluorescein and rose bengal mean staining scores were significantly higher in groups III and IV compared with groups I and II after UV-B exposure. Group I: control group with no soft contact lens wear (SCL [−]) and no UV-B exposure (UV-B [−]); group II: senofilcon A SCL (senofilcon [+]) with UV-B exposure (UV-B [+]); group III: lotrafilcon A SCL (lotrafilcon [+]) with UV-B exposure (UV-B [+]); and group IV: no SCL (SCL [−]) with UV-B exposure (UV-B [+]). ANOVA test was performed and error bars represent standard deviations. *P < 0.05 was considered significant. **P < 0.005 was considered very significant. ***P < 0.0001 was considered extremely significant.

After the mice were anesthetized, the trephinated SCLs were held in place with a nontoothed forceps. The eyelids were opened with an additional nontoothed forceps, and the SCLs were placed and centered on the corneas. The centration and fit of the SCLs were then confirmed before the animals were exposed to UV-B (see supplemental Figs. 2A–D). The mice were positioned on a paper towel in the same manner and the same location directly under the UV light source (Nihon Kohden, Tokyo, Japan) (see supplemental Figs. 3A and B). The SCLs were worn only during UV-B exposure, which was 10 min/eye/day. After placing a new lens on the opposite eye, the mice were turned on the opposite side and received identical 10-minute UV-B exposure. 

Figure 2.

 

Tear film cytokine concentration and HEL changes with UV-B exposure. (A–D) Summarized data showing significant elevation of tear IL-10, IL-17, IL-6, and TNF-α concentrations after UV-B exposure in mice without SCLs (group IV). (E) Summarized data showing significant elevation of tear HEL concentrations after UV-B exposure in lotrafilcon A SCL (group III) and mice without SCLs (group IV). Group I: control group with no soft contact lens wear (SCL [−]) and no UV-B exposure (UV-B [−]); group II: senofilcon A SCL (senofilcon [+]) with UV-B exposure (UV-B [+]); group III: lotrafilcon A SCL (lotrafilcon [+]) with UV-B exposure (UV-B [+]); and group IV: no SCL (SCL [−]) with UV-B exposure (UV-B [+]). ANOVA test was performed and error bars represent standard deviations. * P < 0.05 was considered significant.

Figure 3.

 

Corneal H&E and CD-45 antibodies immunohistochemistry staining after UV-B exposure. (A–E) H&E staining in normal corneal architecture in mice (A) without SCL and not exposed to UV-B and (B) in senofilcon A SCL exposed to UV-B, whereas (C) corneal specimens in lotrafilcon A SCL exposed to UV-B and (D) mice without SCL but exposed to UV-B showed corneal thickening, edema, and infiltration with inflammatory cells. In addition, (E) corneal specimens without SCLs but exposed to UV-B also showed corneal ulceration (blue arrows). (F–J) CD-45 immunohistochemistry staining showed variable degrees of infiltration with inflammatory cells (white arrows) in (H) corneal specimens of lotrafilcon A SCLs exposed to UV-B and (I) mice without SCLs but exposed to UV-B. Bar = 100 μm. (K) Summarized data showing CD45 antibody positively stained cell densities were significantly higher in mice without SCLs but exposed to UV-B and lotrafilcon A SCL compared with mice with senofilcon A SCL and control mice. Group I: control group with no soft contact lens wear (SCL [−]) and no UV-B exposure (UV-B [−]); group II: senofilcon A SCL (senofilcon [+]) with UV-B exposure (UV-B [+]); group III: lotrafilcon A SCL (lotrafilcon [+]) with UV-B exposure (UV-B [+]); and group IV: no SCL (SCL [−]) with UV-B exposure (UV-B [+]). ANOVA test was performed and error bars represent standard deviations. * P < 0.05 was considered significant. Bar = 100 μm.

On day 1, before the mice were anesthetized, all underwent tear collection for tear hexanoyl-lysine (HEL) and tear inflammatory cytokine ELISA measurements. After 2 hours, fluorescein and rose bengal corneal-staining assessments were performed with a hand-held slit lamp microscope (Kowa, Tokyo, Japan). The same tests were performed in the same manner and sequence on day 5 after the mice woke up from anesthesia. On day 6, all mice were euthanized. Whole eye globes were retrieved for histopathological analyses. All examinations and procedures were performed in accordance with the Association for Research in Vision and Ophthalmology (ARVO) Statement for the Use of Animals in Ophthalmic and Vision Research. The protocol was prospective, with the histopathological photographs, cellular density, and staining intensity calculations performed in a masked fashion. 

Tear Fluid Collection

Tears for HEL ELISA were initially collected with 2-μL glass capillary tubes. Ten microliters of 0.1 M PBS was then introduced onto the ocular surface with a micropipette and collected with a 10-μL glass capillary tube (Hirschmann Laborgeräte GmbH & Co., Eberstadt, Germany). Tear collection was performed on days 1 and 5, before and at the end of UV-B exposure. Collected tears were also stored at −80°C until tear cytokine concentration assessments. 

Tear ELISA Assessment

A commercially available kit for ELISA (HEL, Japan Institute for the Control of Aging [JaICA], Shizuoka, Japan) was used to determine the tear HEL concentration, as reported previously. ¹⁸ Following the initial collection of tears, ocular surface wash samples were obtained for tear inflammatory cytokine analyses. 

Cytometric Bead Array Assessment of Inflammatory Cytokines in Tears

IL-2, IL-4, IL-6, IL-10, IL-17, INF-γ, and TNF-α levels were measured from the ocular surface wash samples (10 μL) by a mouse inflammation kit (BD Bioscience, Franklin Lakes, NJ) and flow cytometric analysis (FACSCalibur flow cytometer, Becton Dickinson Immunocytometry Systems [BDIS], San Jose, CA). Data were acquired and analyzed using the Becton Dickinson (BD) Cytometric Bead Array (CBA) software (BD Bioscience) as described previously. ¹⁹  

Ocular Surface Epithelial Damage Assessment

Corneal fluorescein staining was evaluated with the hand-held slit lamp 2 minutes after instillation of 2 μL 0.5% sodium fluorescein diluted in 0.9% physiological saline. ²⁰ The mouse cornea was divided into three equal zones (upper, middle, and lower). Each zone had a staining score ranging from 0 to 3 points with the minimum and maximum total staining scores ranging from 0 to 9 points. The presence of scarce staining in 1 zone was scored as 1 point, whereas punctate staining covering the entire zone was scored as 3 points. This was followed by introduction of 2 μL 1% rose bengal solution by micropipette. The rose bengal staining scores ranged from 0 to 9 points. Mouse corneas were photographed before and after staining on days 1 and 5 before and at the end of UV-B exposure experiments. 

Corneal Specimen Collection

Animals were euthanized by IP injection of pentobarbital sodium (64.8 mg) after sedation. Corneal samples were cut into two halves; one-half was fixed in 4% buffered paraformaldehyde for staining, while the other half was stored in 2.5% glutaraldehyde in 0.1 M phosphate for electron microscopy. 

Histopathological Assessment of Corneal Specimens

All corneal specimens were immediately fixed in 4% buffered paraformaldehyde, embedded in paraffin wax, sliced in 4-μm-thick paraffin sections, and processed for hematoxylin and eosin (H&E) staining. ^21,22  

Immunohistochemistry Staining for Oxidative Stress Markers and CD45 Panleukocyte Antigen

Oxidative stress–induced lipid peroxidation was assessed by immunohistochemistry detection of HEL and 4-hydroxynonenal (4-HNE). Oxidative DNA damage was investigated by immunohistochemistry with anti-8-hydroxy-2′-deoxyguanosine (8-OHdG) antibodies. The avidin-biotin-peroxidase complex (ABC) method was used in immunostaining. To block nonspecific background staining, sections were treated with normal horse serum (Vector Laboratories, Burlingame, CA) for 2 hours at 25°C. The tissues were then treated with mouse anti-8-OHdG (10 μg/mL), HEL (10 μg/mL), and anti-4-HNE monoclonal antibodies (25 μg/mL) (JaICA) for 2 hours at room temperature. For the negative controls, the primary antibody was replaced with mouse IgG1 isotype control (MOPC-21, Sigma, St. Louis, MO). The sections were developed in prepared 3,3′-diaminobenzidine (DAB) chromogen solution (Vector Laboratories), lightly counterstained with hematoxylin for 4 minutes at room temperature, washed, dehydrated, and mounted. For CD45 immunohistochemistry staining, we used the purified anti-mouse CD45 antibody solution (10 μg/mL) (BioLegend, San Diego, CA) and the peroxidase system Vectastain ABC kit (rat IgG; Vector Laboratories), treating the tissues for 2 hours at 25°C. For the negative controls, the primary antibody was replaced with rat IgG2B isotype control at the same concentration of the primary antibody (R&D Systems, Minneapolis, MN). 

Inflammatory Cell and Oxidative Stress Marker Staining Intensity Calculations

Five randomly selected nonoverlapping areas in each specimen in 890-μm × 705-μm frames were digitally photographed (Axioplan2imaging, Carl Zeiss, Jena, Germany). The photographer was masked to the mouse group information. Inflammatory cell densities were given as cells per square millimeter (cells/mm²). Using an image processing software (Adobe Photoshop, San Jose, CA) a subset of color that indicated the stained areas (brown color) was selected from the raw pictures and saved as a jpeg image. Another image analysis program (Image J, NIH, Bethesda, MD) was used to measure the intensity of staining for each image, and the area of staining was calculated and expressed in square micrometers (μm²). 

Transmission Electron Microscopic Examination

Corneal specimens were immediately fixed with 2.5% glutaraldehyde in 0.1 M PBS (pH 7.4) for 4 hours at 4°C. The samples were postfixed in 2% osmium tetroxide and embedded in epoxy resin. One-micrometer sections were prepared using an ultratome (LKB, Gaithersburg, MD) with a diamond knife. Sections were collected on 150-mesh grids, stained with uranyl acetate and lead citrate, and photographed using an electron microscope (Model 1200 EXII; JEOL, Tokyo, Japan). 

Statistical Analyses

The ANOVA test (Bonferroni: comparing all pairs of columns) was applied to test the statistical differences between cytokine levels, immunohistochemistry calculations for CD45, 8-OHdG, HEL, and 4-HNE. The differences were considered statistically significant if the P values were less than 0.05. GraphPad Instat 3.0 (GraphPad Software, Inc., San Diego, CA) was used for the analyses. 

Results

Vital Staining Score Changes and Assessment of Corneal Damage with UV-B Exposure

To determine whether a total dose of 2.73J/cm² of UV-B exposure had any adverse effects on the corneas, we carried out slit-lamp examinations under direct light and cobalt blue filter after staining the corneas with fluorescein dye. The mean post–UV-B exposure fluorescein and rose bengal staining scores were significantly higher in groups III (lotrafilcon [+], UV-B [+]) and IV (SCL [−], UV-B [+]) compared with the mean post–UV-B exposure staining scores in groups I (SCL [−], UV-B [−]) and II (senofilcon [+], UV-B [+]). The mean post–UV-B exposure vital staining scores were also significantly higher in group III compared with group II (Fig. 1). Representative anterior segment and corneal fluorescein staining photographs from each group are shown in Figure 1. Eight eyes in group IV and six eyes in group III developed corneal ulcers after UV-B exposure, whereas no ulcers were observed in groups I and II. Our results suggest marked damage to the corneal epithelium after UV-B exposure in group III and group IV. The exposure settings in this study resulted in ocular surface epithelial damage evidenced by significant increases in fluorescein staining scores in group III (3.6-fold) and group IV (4.2-fold). Corneal ulceration was observed in 42% of the eyes in group III and 57.1% of the eyes in group IV. 

Cytometric Bead Array Assessment of Inflammatory Cytokines in Tears

To investigate whether UV-B exposure was associated with changes of inflammatory cytokines in tears, we carried out cytometric bead array, which showed that the mean tear IL-10, IL-17, IL-6, and TNF-α concentrations significantly increased in group IV (2-fold) as shown in Figure 2. These results indicated that UV-B exposure could induce a significant elevation of inflammatory cytokines in tears when SCLs were not worn. 

Tear ELISA Assessment

Since UV-B exposure can stimulate generation of reactive oxygen species and oxidative stress status in ocular tissues and tears, we checked whether cell membrane and tissue associated with early lipid oxidative damage marker HEL increased in the tears of mice exposed to UV-B. The mean tear HEL concentration increased in all groups receiving UV-B exposure. The increase was statistically significant in mice with lotrafilcon A exposed to UV-B (group III) (1.3-fold) and UV-B–exposed mice without SCLs (group IV) (1.5-fold) as shown in Figure 2. The mean post–UV-B exposure tear HEL concentrations were significantly higher in UV-B–exposed mice without SCLs (group IV) compared with the tear HEL concentrations in the other groups (P < 0.01). These results suggested that SCLs with UV-B blocking capabilities might provide protection against elevation of membrane and tissue lipid oxidative stress damage marker, HEL, in tears. 

Corneal Histopathological Alterations with UV-B Exposure

To study the corneal tissue alterations with exposure to UV-B, we performed H&E staining, which showed apparently normal corneal architecture in mice without SCLs and not exposed to UV-B (group I) as well as mice with senofilcon A SCLs with UV-B exposure (group II); whereas corneal specimens in groups III (lotrafilcon A mice exposed to UV-B) and IV (mice without SCLs, but exposed to UV-B) consistently showed corneal thickening, edema, and infiltration with inflammatory cells. In addition, corneal specimens in mice without SCLs but exposed to UV-B (group IV) also showed occasional corneal ulceration as shown in Figure 3. 

CD-45 Immunohistochemistry Staining for Assessment of Corneal Inflammation Status

Since UV-B exposure can induce inflammation in the cornea, we also chose to investigate if UV-B exposure induced inflammation in the cornea and if SCLs could reduce the extent of inflammatory cell infiltrates. CD-45 (leukocyte common antigen) immunohistochemistry staining confirmed infiltration of inflammatory cells (lymphocytes, neutrophils, and monocytes) in the corneal stroma in mice with lotrafilcon A SCLs exposed to UV-B (group III) and mice without SCLs exposed to UV-B (group IV), whereas no inflammatory cells could be confirmed in any of the corneal specimens of the other groups as shown in Figure 3. The mean inflammatory cell counts in corneal specimens of mice with lotrafilcon A SCLs exposed to UV-B (group III) and mice without SCLs exposed to UV-B (group IV) were 61 ± 5 cells/mm² and 184 ± 10 cells/mm², respectively, compared with none in mice without SCLs and not exposed to UV-B (group I), as well as mice with senofilcon A SCLs exposed to UV-B (group II) as shown in Figure 3K. Our observations suggest that UV-B exposure is associated with increased inflammatory infiltrates in mouse corneal tissues but a reduction of inflammatory infiltrates with silicon hydrogel SCL wear. 

HEL, 4-HNE, 8-OHdG Immunohistochemistry Staining for Assessment of Corneal Lipid and DNA Oxidative Damage Status

To test the alterations in lipid oxidative damage status in the corneal tissues associated with UV-B exposure, we carried out immunohistochemistry staining for the early and late lipid oxidative damage markers, HEL and 4-HNE, respectively. Corneal epithelial cells and stromas stained intensely with anti-HEL and anti-4HNE antibodies in mice with lotrafilcon A SCLs exposed to UV-B (group III) and mice without SCLs exposed to UV-B (group IV) compared with mice without SCLs and not exposed to UV-B (group I) as well as mice with senofilcon A SCL exposed to UV-B (group II) as shown in Figures 4A–E and 4G–K, respectively. The mean area of staining for anti-HEL and anti-4HNE antibodies was significantly higher in groups III and IV compared with groups I and II (P < 0.05) (Figs. 4F, 4L). 

Figure 4.

 

Corneal immunohistochemistry staining for lipid and DNA oxidative stress damage (A–E) immunohistochemistry staining with HEL antibodies showed that (C) lotrafilcon A SCL mice exposed to UV-B and (D) mice without SCL but exposed to UV-B had moderate to intense brown staining throughout the entire corneal epithelium and stroma indicating lipid oxidative damage induced by UV-B exposure. (F) Quantitative staining area calculations showed a significantly greater area of staining in groups III and IV compared with groups I and II. (G) Anti–4-HNE immunohistochemistry staining showed inconsiderable brown staining in group I mice. (H) Group II had slight brownish staining in the corneal epithelium and stroma, whereas corneal specimens in groups III (I) and group IV mice (J) had moderate to intense brown staining throughout the entire corneal epithelium and stroma. (K) Shows negative control staining. (L) Quantitative staining area calculations showed a significantly greater area of staining in groups III and IV. The area of staining was also significantly larger in group-III mice compared with group II. Anti–8-OHdG staining (M–P) showed that group-I mice (M) had slight round-oval brownish staining of the cellular nuclei in the corneal epithelium and stroma. (N) Group-II mice had moderate brownish staining of the cellular nuclei in the corneal epithelium and stroma, whereas corneal specimens in group-III (O) and group-IV mice (P) had intense brown nuclear staining throughout the entire corneal epithelium and stroma. (Q) The number of positively stained cells was significantly higher in the group-III mice compared with group-II mice. Group I: control group with no soft contact lens wear (SCL [−]) and no UV-B exposure (UV-B [−]); group II: senofilcon A SCL (senofilcon [+]) with UV-B exposure (UV-B [+]); group III: lotrafilcon A SCL (lotrafilcon [+]) with UV-B exposure (UV-B [+]); and group IV: no SCL (SCL [−]) with UV-B exposure (UV-B [+]). ANOVA test was performed and error bars represent standard deviations. * P < 0.05 was considered significant. Bar = 100 μm.

To further investigate the effect of UV-B exposure and SCL use on the oxidative damage status of the corneas, we performed immunohistochemistry staining for DNA oxidative stress damage marker, 8-OHdG. We observed marked corneal staining with anti-8-OHdG antibodies in mice with lotrafilcon A exposed to UV-B (group III) and more extensive staining consistently in all corneal specimens of mice without SCLs exposed to UV-B as shown in Figures 4M–P. The number of cells positively stained with anti-8-OHdG antibodies was significantly higher in groups III and IV, compared with groups I and II (P < 0.05), respectively, as shown in Figure 4Q. These results suggest that UV-B exposure is associated with increased lipid and DNA oxidative damage in mouse corneal tissues with possible protection from such damage with silicon hydrogel SCL wear. 

4-HNE, 8-OHdG Immunohistochemistry Staining for Assessment of Retinal Lipid and DNA Oxidative Damage Status

Since the effect of SCL wear and UV-B exposure on retinal lipid and DNA oxidative damage status has not been investigated in retinas of mice previously, we made further efforts to investigate such damage through 4-HNE, 8-OHdG immunohistochemistry staining. The retinal specimens showed more intense staining with anti-4HNE antibodies in the inner and outer photo segments, inner and outer plexiform, and ganglion cell layers in specimens of mice with lotrafilcon A exposed to UV-B (group III) and more extensive staining consistently in all retinal specimens of mice without SCLs exposed to UV-B compared with mice without SCLs and not exposed to UV-B (group I) as well as mice with senofilcon A SCLs exposed to UV-B (group II). The mean area of staining was significantly higher in groups III and IV compared with groups I and II as shown in Figure 5. A similar trend was also observed for staining of retinal specimens with anti-8-OHdG antibodies as shown in Figure 6. The number of retinal cells positively stained with anti-8-OHdG antibodies was significantly higher in groups III and IV compared with groups I and II (P < 0.05), as shown in Figure 6. The positively stained cell numbers were also significantly higher in group IV compared with group I (P < 0.001; data not shown). These results suggest that UV-B exposure is associated with increased lipid and DNA oxidative damage in mice retinal tissues as well, with possible protection from such damage with use of silicon hydrogel SCL wear. 

Figure 5.

 

Retinal immunohistochemistry staining with anti–4-HNE antibodies. (A, B) Retinal specimens from groups I and II showed slight brownish staining, respectively. (C, D) Intense brown staining in the inner and outer photo segments, inner and outer plexiform, and ganglion cell layers was observed in group-III (C) and group-IV (D) mice. (E) Insert shows the negative control staining. (F) Quantitative assessment of staining areas showed a significantly greater area of staining in group-III and group-IV mice. The area of staining was also significantly larger in the group-III mice compared with group-II mice. Group I: control group with no soft contact lens wear (SCL [−]) and no UV-B exposure (UV-B [−]); group II: senofilcon A SCL (senofilcon [+]) with UV-B exposure (UV-B [+]); group III: lotrafilcon A SCL (lotrafilcon [+]) with UV-B exposure (UV-B [+]); and group IV: no SCL (SCL [−]) with UV-B exposure (UV-B [+]). ANOVA test was performed and error bars represent standard deviations. * P < 0.05 was considered significant. Bar = 100 μm.

Figure 6.

 

Retinal immunohistochemistry staining with (A) anti–8-OHdG antibody group I and (B) group-II mice had slight round-oval brownish staining of the cellular nuclei in the retinal tissues with 8-OHdG antibodies. (C) Group-III and group-IV mice (D) had intense brown round-oval nuclear staining throughout the entire retina. (E) The number of positively stained cells was significantly higher in group-III compared with group-II mice. Group I: control group with no soft contact lens wear (SCL [−]) and no UV-B exposure (UV-B [−]); group II: senofilcon A SCL (senofilcon [+]) with UV-B exposure (UV-B [+]); group III: lotrafilcon A SCL (lotrafilcon [+]) with UV-B exposure (UV-B [+]); and group IV: no SCL (SCL [−]) with UV-B exposure (UV-B [+]). ANOVA test was performed and error bars represent standard deviations. * P < 0.05 was considered significant. Bar = 100 μm.

Ultrastructural Corneal Alterations with UV-B Exposure

To study the ultrastructural corneal tissue alterations upon exposure to UV-B, we performed transmission electron microscopy (TEM). The corneal epithelium appeared normal in all specimens of mice without SCLs and not exposed to UV-B (group I) as well as mice with senofilcon A exposed to UV-B (group II) (Figs. 7A, 7B). Attenuation of the corneal epithelial basement membrane and disturbance of cytoplasmic organelles were observed in specimens with lotrafilcon A SCLs exposed to UV-B (group III) (Fig. 7C). There was marked damage to the superficial corneal epithelium with irregularity of basal epithelial cells in specimens of mice without SCLs exposed to UV-B (group IV) as shown in Figure 7D. A closer look into the epithelial basement membrane showed regular and normal architecture in specimens of mice without SCLs and not exposed to UV-B (group I) as well as mice with senofilcon A SCLs exposed to UV-B (group II) (Figs. 7E, 7F). Damage to cytoplasmic organelles was more evident under higher magnification in specimens of mice with lotrafilcon A SCLs exposed to UV-B (group III) (Fig. 7G). Corneal specimens in group IV (mice without SCLs exposed to UV-B) showed breaks in epithelial basement membranes with leakage of cytoplasmic organelles and inflammatory cell infiltration near the basement membrane as shown in Figures 7H and 7I. TEM study revealed marked ultrastructural damage to the corneal epithelium, organelles, and basement membrane upon UV-B exposure in mice without SCLs and mice with lotrafilcon A SCL but preservation of the corneal ultrastructural architecture in mice with senofilcon A SCLs exposed to UV-B. 

Figure 7.

 

Transmission electron microscopy observations of mice corneal epithelium and basement membrane Corneal epithelium appeared normal in specimens of mice (A) without SCL and not exposed to UV-B and in (B) senofilcon A SCL exposed to UV-B, whereas (C) corneal specimens in lotrafilcon A SCL exposed to UV-B showed attenuation of the corneal epithelial basement membrane and disturbance of cytoplasmic organelles. (D) Mice without SCL but exposed to UV-B showed marked damage to the superficial corneal epithelium with irregularity of basal epithelial cells. Bar = 10 μm. (E, F) Corneal epithelial basement membrane appeared normal in specimens of mice without SCL and not exposed to UV-B and in senofilcon A SCL exposed to UV-B under high magnification, respectively. (G) Corneal specimens in lotrafilcon A SCL exposed to UV-B showed evident damage to cytoplasmic organelles and vacuolar formations (white arrow). (H) Mice without SCL but exposed to UV-B showed breaks in the epithelial basement membrane with leakage of cytoplasmic organelles into the corneal stroma (white arrow). (I) Insert shows inflammatory cells near the basement membrane in mice exposed to UV-B without SCL. Bar = 2 μm.

Discussion

Ultraviolet exposure in humans has been reported to cause photokeratitis characterized by exfoliation of the corneal epithelium, formation of ulcers, inflammation, edema, and spheroidal degeneration of the cornea. ^7,8,23 Adverse UV effects in the mouse eye include corneal stromal thinning, keratoconus, corneal vascularization, fibrosis, keratitis, and globe rupture, ^24–27 but there have been no reports on lipid and DNA oxidative damage changes in a mouse model with SCL, to our knowledge. We presumed that wearing UV-blocking SCLs might have protective effects against oxidative stress and tissue damage in corneas and retinas of mice exposed to UV-B. We thus performed UV-B exposure experiments on our SCL mouse model comparing the results to mice without SCLs but exposed to UV-B. 

UV-induced corneal lesions in mice vary considerably in severity, depending on UV radiation dose and number of exposures. The daily dose of UV-B used in this study was 2.73 J/cm². Photokeratitis associated with UV-B exposure in humans and experimental animal models has been proposed to be related to increased expression of matrix metalloproteinases (MMP) with corneal stromal degradation and induction of apoptosis. ^24,28  

While MMP expression and cell death have not been investigated in this study, Chandler et al. investigated whether class-I UV light–blocking contact lenses prevented UV-induced pathologic changes in a rabbit model. ²⁸ Exposed corneas in that study (1.667 J/cm² UV-B exposure [0.98 mW/cm²] daily for 5 days) showed a significant increase in MMP-2 and MMP-9, TUNEL-positive cells, and caspase-3 activity in the lotrafilcon A group compared with the senofilcon A group. 

In human corneas, MMP-9 and MMP-2 have been shown to degrade the basement membranes resulting in ulceration. ²⁹ Indeed, after UV radiation, human corneal fibroblasts have been shown to produce MMP- 2. ³⁰ It is likely that UV induction of MMP production may have played a role in initiating a proteolysis in the corneas of UV-exposed mice. ²⁴ Our study showed clear slit-lamp, histopathological and ultrastructural evidence of corneal ulceration, edema, opacification, and basement membrane damage and/or attenuation in mice with lotrafilcon A SCLs exposed to UV-B (group III) and UV-B exposed mice without any SCL (group IV). The corneal architecture appeared to be well preserved in mice with a senofilcon A SCL and exposed to UV-B, which may owe to the higher UV-B blocking properties of the senofilcon A SCL. 

We have a noteworthy observation that mice with lotrafilcon A exposed to UV-B (group III) and UV-B exposed mice without any SCL (group IV) developed marked lipid and DNA oxidative damage in the corneal and retinal tissues. We believe that the UV-B exposure in our experimental settings brought up these findings together with corneal inflammatory changes. Indeed, elevated oxidative stress has been shown to promote inflammation by activating the redox-sensitive transcription factor, nuclear factor-kappa B (NF-κ B), which, in turn, triggers generation of pro-inflammatory cytokines and chemokines, and hence, inflammation. ^31,32 In case of UV exposure–induced skin changes and relevant inflammatory events, one of the earliest detectable response of the skin cells is the activation of multiple cytokine receptors including TNF-α, IL-1 receptors, and epidermal growth factor (EGF) receptors. Activation of cell surface cytokine and growth factor receptors results in recruitment of adaptor proteins and activation of multiple MAP kinase pathways. ^33–35 Mitogen-activated protein (MAP) kinase activation has been reported to result in the induction of transcription factor AP-1, which regulates the expression of many genes involved in the regulation of cellular growth and differentiation. AP-1 is known to tightly regulate the transcription of several MMPs including MMP-9 (gelatinase-b), which degrades collagen fragments generated by collagenases and stromelysin-1, which degrade basement membrane–type collagen. ^36,37  

Of the tear cytokines, IL-6, IL-17, and TNF-α showed a tendency of elevation in mice with senofilcon A and lotrafilcon A SCL exposed to UV-B, with a significant elevation in tears of UV-B–exposed mice without any SCL. The reported roles for TNF-α include induction of inflammation and cell death, and those for IL-6 include induction of inflammation and fibrosis, ^38,39 while IL-17 has been shown to potentiate release of IL-6 and TNF-α and recruitment of inflammatory cells, which may explain the roles for the elevated tear cytokines in our study. ⁴⁰ While the intermediary NF-κ B and MAP kinase pathway signaling has not been investigated in this study, their role in degradation of the corneal tissue through activation of MMPs remains a possibility. It is also possible that inflammatory cells recruited to corneas affected with keratitis might have sustained and amplified MMP responses or exacerbated the oxidative stress changes through release of relevant mediators, causing a vicious cycle. 

Chandler et al. found corneal edema in UV-B–exposed rabbit eyes assigned to either the lotrafilcon A wear or no contact lens wear groups, which they suggested might have been due to lower UV-blocking capacity of the lotrafilcon A SCL. ²⁸ While oxidative stress markers were not investigated in Chandler's study, they looked into the aqueous humor ascorbate levels, an antioxidant substance, which they found to be decreased in eyes allocated to lotrafilcon A wear or UV exposure without any SCL wear. ²⁸ This study did not look into the simultaneous antioxidant status alterations because of the limitations in tear sample quantities, and it remains to be pursued in future studies. 

Of interest, and a first time observation in the literature, was that retinal samples in mice exposed to UV-B but without SCLs or lotrafilcon A SCLs exposed to UV-B appeared to develop comparatively higher lipid and DNA oxidative stress damage. 

The experiments outlined in this study support our initial hypothesis that wearing senofilcon A UV-blocking contact lenses could provide some lipid and DNA oxidative stress protection in the cornea and retinas of the UV-B–exposed mice compared with lotrafilcon A SCLs and non–lens-wearing (exposed) eyes. Further studies providing explanations for mechanisms involved in lipid and DNA oxidative stress damage with UV-B exposure may also result in the development of a new generation of contact lenses that slowly release antioxidants to the tear film and ocular surface in addition to the presence of UV-blocking polymers. 

Supplementary Materials

Supplemental-Fig-new 

Acknowledgments

The authors thank Johnson and Johnson Vision Care Company for their donation to Keio University, which decided to establish an Ocular Surface and Visual Optics Department in the School of Medicine. Neither was Johnson and Johnson Vision Care involved with the scientific content of the research projects run by the department nor were the authors paid employees or received any fees from Johnson and Johnson Vision Care. 

Footnotes

 Presented at the 2010 Tear Film and Ocular Surface Society Meeting, Florence, Italy, September 22–25, 2010.

Footnotes

³  These authors contributed equally to this work and should therefore be regarded as equivalent authors.

Footnotes

 Disclosure: O.M.A. Ibrahim, None; T. Kojima, None; T.H. Wakamatsu, None; M. Dogru, None; Y. Matsumoto, None; Y. Ogawa, None; J. Ogawa, None; K. Negishi, None; J. Shimazaki, None; Y. Sakamoto, None; H. Sasaki, None; K. Tsubota, None