UVB promotes the initiation of uveitic inflammatory injury in vivo and is attenuated by UV-blocking protection

Animals

All animal procedures were approved by the Animal Care and Use Committee of Chung Shan Medical University and conformed to the guidelines that adhered to the ARVO Statement for use of animals in research. Female CD-1® Mouse, Nomenclature Crl: CD1 (ICR) mice (6–7-weeks-old) were purchased from BioLASCO Experimental Animal Center (Taiwan Co., Ltd, BioLASCO, Taipei, Taiwan). Only mice with normal eyes confirmed by slit lamp microscopy were included in the experiments.

Experimental design

A UVB lamp (peak wavelength at 312 nm: 280–320 nm; power density: 8 mW/cm²; CN-6, Vilber Lourmat, France) was used, and the distance to the mouse cornea was approximately 15 cm (duration: 90 s, once daily for 7 days) [7,10]. Mice were assigned to either a blank control group (n = 12) or one of three UVB treatment groups (Figure 1 A): (1) without protection (UVB group; n = 12), (2) protection with Nelfilcon A contact lens (Nelfilcon A group; n = 12), and (3) protection with Etafilcon A contact lens (Etafilcon A group; n = 12). All contact lenses were used once per eye and discarded daily after UVR. Both eyes of each mouse were administered UVR (Figure 1A). After general anesthesia (Pentobarbital, 50 mg/kg, IP), the mouse eyes were covered individually with the relevant contact lenses and then exposed to UVB. Contact lens characteristics are shown in Figure 1B and the UV transmittance spectra in Figure 1C. The transmittance of UVB (280–320 nm) was approximately 46.8 ± 1.7% and 0.55 ± 0.45% in the Nelfilcon A and Etafilcon A contact lenses, respectively.

Evaluation of the corneal surface and evaluation of uveitis

The corneal surfaces of the left eyes of the anesthetized mice on day 8 were observed macroscopically and imaged. Analysis and scoring system details, including corneal smoothness and corneal lissamine green staining, have been previously described [9,10]. Anterior uveitic inflammatory injury sign and score were evaluated with slit-lamp microscopy on day 8 after UVR. The clinical severity of the uveitic inflammatory injury was graded by two independent observers according to the modification criteria described in Table 1.

Histological analysis and protein analysis

After euthanized by cervical dislocation on day 8, one of the mouse eyes was used for histological analysis and the other for protein analysis. Conventional tissue fixation, paraffin embedding, section preparation, hematoxylin-eosin, and immunostaining were performed as described in our previous studies [9,10]. Leukocyte infiltration was measured parallel to the eye’s vertical meridian. Aqueous humor was collected with anterior chamber puncture with a 30-gauge needle on day 8. The protein concentration was determined using a protein quantification kit. The anterior segments of each eye (including the cornea, iris, and ciliary body) were isolated and homogenized individually in 50 μl lysis buffer (25 mM Tris-HCl, pH 7.5; 100 mM NaCl; and 1% v/v Nonidet P40) for western blotting analysis. The primary antibodies of anti-matrix metalloproteinases MMP9 (1/100 dilution for immunostaining; 1/100 dilution for western blotting; ab38898, Abcam, Cambridge, UK) were used in this study.

Statistical analysis

Data were obtained from individual mouse corneas. For all clinical examination evaluations, data were presented as mean ± standard deviation. Statistical analysis was performed using SPSS (SPSS, Inc., Chicago, IL), and graphs were generated using Microsoft Excel 2010. Data obtained from all groups were analyzed using the Kruskal–Wallis nonparametric ANOVA, and the Mann–Whitney U-test for pairwise comparisons in cases when the Kruskal–Wallis test was statistically significant. Incidence rates were analyzed using the independent-samples t test. Correlation between polymorphonuclear neutrophils (PMNs) counts in the corneal stroma, or the iridocorneal angle, and the transmittance strength of UVB was assessed with Pearson’s correlation test. P values of less than 0.05 and 0.01 were considered statistically significant.

Contact lens UVR protection

None of the control mice showed evidence of abnormalities on the corneal surface (Figure 2A,E). In contrast, the UVB group exhibited serious irregular distortion, shown as irregular white-light ring reflections (Figure 2B) and large dark blue devitalized epithelial staining (Figure 2F). The Nelfilcon A group corneas showed greater distortion (Figure 2C) and staining (Figure 2G) than those in the Etafilcon A group (Figure 2D,H, respectively); this result was statistically significantly different (p<0.01; Figure 2I,K). Contact lens protection reduced the incidence of corneal smoothness (score>2; Figure 2J) and corneal staining (score>2; Figure 2L) after UVR; the effect was statistically significantly greater in the Etafilcon A group compared with the Nelfilcon A group (p<0.01, Figure 2J; p<0.05, Figure 2L, respectively). These findings confirm UVR protective effects of contact lenses on the cornea in vivo.

UVR role in the pathogenesis of uveitic injury in anterior eye segments

Anterior uveitis inflammatory injury can be characterized by hyperemic change in the iris and the limbal vessel, and cellular exudates on the iris anterior surface. These symptoms were severe in the UVB group (Figure 2N), moderate in the Nelfilcon A group (Figure 2O), but absent in the Etafilcon A group (Figure 2P). The uveitis clinical scores were normal in the Etafilcon A group and statistically significantly lower than in the other two UVR groups (p<0.01; Figure 2Q). However, the scoring of the Nelfilcon A group was lower than that of the UVB group (p<0.05; Figure 2Q). After UVR, the aqueous humor protein concentration increased in the UVB group compared with that of the control (Figure 2R). This increase may be due to disruption of the blood–aqueous barrier. However, the reduction in the protein concentration occurred in the Nelfilcon A group after UVR, but this decrease was not statistically significant (p = 0.08; Figure 2R). However, the protein concentration was statistically significantly reduced in the Etafilcon A group (p<0.01; Figure 2R). This decrease confirmed that UVR participates in the pathogenesis of uveitic injury in the anterior segment of the eye.

UVR contribution to the pathogenesis of cellular infiltrates in the iridocorneal angle of eyes

Microscopically, the blank control group showed normal tissue integrity of the cornea (Figure 3A) and the iridocorneal angle (Figure 3E). The UVB group exhibited severe lesions, with a thin corneal epithelium, subepithelial bullae, and friable edematous stroma with infiltrating leukocytes (Figure 3B). Infiltrating leukocytes embedded inside the inferior iridocorneal angle were the predominant feature in the UVB group (Figure 3F). However, the Nelfilcon A group actually showed mild infiltrating leukocytes inside the iridocorneal angle (Figure 3G); in contrast, no evidence of abnormality was found in the Etafilcon A group (Figure 3H).

There were differences in the infiltrating leukocytes in the cornea (Figure 3I) and the iridocorneal angle (Figure 3K) between the study groups. Leukocyte infiltration was statistically significantly lower in the Nelfilcon A group compared with the UVB group (p<0.01); however, the infiltration was higher than in the Etafilcon A group (Figure 3I,K). This result may explained by the decrease in UVR transmittance due to Etafilcon A contact lens protection (Figure 1C). Leukocyte infiltration rates in the cornea (PMNs ≥15 cells/µm² × 10⁵) and the iridocorneal angle (PMNs ≥15 cells/section) are shown in Figure 3J,L; the rate was statistically significantly lower in the Nelfilcon A group compared with the UVB group (p<0.01). The efficacy of the inhibition of leukocyte infiltration was greater in the Etafilcon A group compared with the Nelfilcon A group (p<0.01 Figure 3J; p = 0.067, Figure 3L, respectively). This result suggests a correlation between leukocyte infiltration and UVB radiation intensity, or the efficacy of UV protection.

UVR blocking can retard the initiation and progression of uveitic inflammatory injury

Because reduced UVB transmittance strength improves the inhibitory efficacy of inflammatory injury in vivo, we analyzed the correlation between UVB transmittance strength and inflammatory pathogenesis. There was a statistically significant correlation between the reduction in UVR strength and inhibition of PMN recruitment in the corneal stroma (Pearson’s correlation coefficient; r = 0.851, p<0.05; Figure 3M) and the iridocorneal angle (Pearson’s correlation coefficient; r = 0.754, p<0.05; Figure 3N). This indicates that UVR blocking can retard the initiation and progression of uveitic inflammatory injury.

The Etafilcon A lenses provide better protection of anterior segment of eye from UVB damage compared to the Nelfilcon A lenses

To compare the efficacy of Etafilcon A lenses and Nelfilcon A lenses on UVR protection, an evaluation of both lenses, Etafilcon A (the right eye) and Nelfilcon A (the left eye), contralaterally was performed in the same mice to investigate the relative damage and protection after UVR. None of the Etafilcon A protection eyes (the right eye) showed evidence of abnormalities on the corneal surface (Figure 4A,B) or the iris anterior surface (Figure 4C) and microscopically showed normal tissue integrity of the cornea (Figure 4D) and the iridocorneal angle (Figure 4E). In contrast, the Nelfilcon A protection eyes (the left eye) showed irregular distortion of the cornea surface (Figure 4F), epithelial staining (Figure 4G), hyperemic iris (Figure 4H), and leukocyte influx in the corneal stroma (Figure 4I) and the iridocorneal angle (Figure 4J). Based on the results from the same animal, the evidence indicates that the Etafilcon A lenses provide better protection of the anterior segment of the eye from UVB damage compared with the Nelfilcon A lenses.

Contribution of MMP9 protein produced by infiltrating leukocytes to uveitic inflammatory injury after UVR

MMPs contribute to inflammatory injury. In this study, the MMP9 protein was predominately labeled and localized in infiltrating leukocytes (Figure 4L,N,P,R). Infiltrating leukocytes showed an abundance of MMP9 protein labeling in the iridocorneal angle of the eyes in the UVB group (Figure 4L,P) and the Nelfilcon A protecting eye group (left eye; Figure 4N,R). However, the MMP9 protein can contribute to the degradation of blood barrier structure molecules and leukocyte extravasation from blood vessels. Alternatively, the MMP9 protein may also destroy the blood–aqueous barrier resulting in an over-loading of inflammatory cells and exudate recruitment into the aqueous humor. In contrast, there was no evidence of the MMP9 protein in the control group (Figure 4K,O) and the Etafilcon A protecting eyes group (right eye; Figure 4M,Q). The expression of the MMP9 protein in the anterior segment tissues (including the cornea, iris, and ciliary body) was analyzed and confirmed with western blotting (Figure 4S). The level of the MMP9 protein statistically significantly increased in the corneas of the UVB group compared with those of the control group. In the same animal, the Etafilcon A protecting eye (right eye) showed more inhibition of the level of the MMP9 protein than that of the Nelfilcon A protecting eye (left eye) after UVR (Figure 4S,T).