Protective Role for CD1d-Reactive Invariant Natural Killer T Cells in Cauterization-Induced Corneal Inflammation

Abstract

purpose. Corneal inflammation can be induced by various stimuli, such as chemical burns, trauma, and acute bacterial infection, and directly impairs visual acuity. Natural killer T (NKT) cells belong to a specialized population of leukocytes that coexpress the T-cell receptor and NK markers. This study examined the role of CD1d-reactive invariant NKT cells in cauterization-induced acute corneal inflammation.

methods. The corneas of CD1d-knockout (KO) mice and Jα18-KO mice (both of which are NKT cell deficient) and control mice were cauterized with silver nitrate. Corneal edema and opacity were examined, and the phenotypes of the corneal-infiltrating cells were analyzed histologically at 24 hours and by flow cytometry at 96 hours. Reverse transcription–polymerase chain reaction (RT-PCR) was used to determine the expression of vascular endothelial growth factor (VEGF), interferon (IFN)γ, and tumor necrosis factor (TNF)α in the cauterized corneas.

results. The CD1d-KO and Jα18-KO mice had significantly greater levels of corneal edema and opacity than did the control mice. Although the number of infiltrating cells was not significantly different at 96 hours, both groups of NKT cell-deficient mice demonstrated increased early neutrophil accumulation at 24 hours and early expression of VEGF, IFNγ, and TNFα. There was no difference in the level of VEGF-induced corneal neovascularization.

conclusions. NKT cells appear to regulate the early accumulation of neutrophils, protect the cornea from excessive inflammation, and maintain corneal clarity. However, in this study, they did not affect the corneal revascularization process induced by VEGF.

The eye is a specialized organ that is capable of limiting excessive inflammation to avoid loss of vision. This is known as immune privilege of the eye and is attributed to a variety of local factors, including the lack of lymphatic drainage, ¹ Fas-ligand expression, ² and multiple immunosuppressive factors in the aqueous humor. ^{3 4} Because the cornea is located at the front of the eye and is exposed to many dangerous stimuli, excessive reactive inflammation might directly induce irreversible opacity. ^{5 6} It is, therefore, important to elucidate the cellular and molecular mechanisms responsible for controlling inflammation to maintain the clarity of the cornea. 

In general, nonspecific corneal inflammation is initiated by innate immunity. ⁷ The corneal expression of proinflammatory cytokines and chemokines in response to injury or infection leads to the recruitment of cells of the innate immune system. We previously demonstrated the critical role of neutrophils in experimental corneal inflammation. ⁸  

Natural killer T (NKT) cells belong to a specialized population of leukocytes that coexpress the T-cell receptor (TCR) αβ chain and NK markers, ^{9 10} and they are classified as mediators of the innate immune response. Approximately 85% of the mouse NKT cell population expresses a restricted TCR repertoire consisting of an invariant TCRα chain (Vα14 Jα18). ^{11 12} Similarly, human NKT cells express the invariant Vα24 JαQ TCR chain. ¹³ NKT cells are restricted by major histocompatibility complex (MHC) class I-like CD1d molecules, ^{14 15} and, because the CD1d molecule is also required for the development of NKT cells, CD1d-knockout (KO) mice selectively lack NKT cells. ^{16 17 18} The role of NKT cells in disease models has been reported in terms of tumor rejection, ¹⁹ abortion, ²⁰ and infection. ^{21 22} Several reports also imply a role for NKT cells in preventing certain autoimmune diseases ^{23 24} and inducing transplantation tolerance. ^{25 26}  

In addition to local mechanisms, ocular immune privilege is associated with the development of antigen-specific systemic immunologic tolerance. The mechanisms of ocular-type immunologic tolerance have been well studied in an experimental system, known as the anterior chamber-associated immune deviation (ACAID) animal model. ¹ ACAID is a critically important dimension of the ability of the immune system to protect the eye from various types of inflammations, including postoperative injury. For example, a host bearing a long-term clear corneal allograft displays an antigen-specific downregulation of the delayed-type hypersensitivity response to donor alloantigens that is reminiscent of ACAID. ²⁷ Sonoda et al. ^{28 29} demonstrated that CD1d-reactive NKT cells are critical for ACAID and transplantation tolerance after corneal allograft. Indeed, NKT cell-deficient mice failed to induce donor-specific tolerance and were unable to accept corneal grafts in the long term. ²⁸  

In the present study, we developed a cauterization-induced corneal inflammation model and compared the extent of local inflammation between control mice and NKT cell-deficient mice to elucidate the role of NKT cells in nonspecific corneal inflammation. 

Methods

Mice

Female mice (8–10 weeks old) were used in all experiments. C57BL/6 (B6) mice were obtained from SLC Japan (Shizuoka, Japan). CD1d-KO mice were generated in the Transgenic Facility of Harvard Medical School (Boston, MA) and were backcrossed to B6 mice for six generations. ²⁹ Jα18-KO mice (NKT-KO mice) were generated at Chiba University (Chiba, Japan) and were backcrossed eight times to B6 mice. ¹⁹ All mice were maintained on food and water ad libitum until they reached the desired weight (20–24 g). All animals were treated humanely and were housed in specific pathogen-free conditions at Kyushu University (Japan). Treatment of the animals conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Induction and Evaluation of Corneal Inflammation

Corneal inflammation was induced as described previously. ³⁰ The center of the right cornea of each mouse was cauterized with 1-mm diameter plastic sticks containing 0.4 g/mL silver nitrate. After 96 hours, corneal edema and opacity were evaluated by stereoscopic microscopy with a slit lamp. Cauterization and evaluation were performed in a masked fashion. Lesions were scored for corneal edema and opacity on scales of increasing severity from 0 to +4. For edema, the scores corresponded to the following descriptions: 0, completely normal cornea; +1, no evidence of ongoing edema, corneal thickness less than normal; +2, mild edema limited to the cauterized area, corneal thickness increased but less than two times that of the normal cornea; +3, severe edema limited to the cauterized area, corneal thickness more than two times that of the normal cornea; +4, extensive edema extended to the whole cornea, corneal perforation in some cases. For opacity, the scores corresponded to the following descriptions: 0, completely clear cornea; +1, slight opacity; +2, mild opacity, iris and lens visible; +3, severe opacity limited to the cauterized area, iris and lens invisible; +4, extensive opacity extended to the whole cornea. 

Corneal neovascularization was evaluated according to length and extension and occurred in particular areas of the limbal vascular plexus where neovascularized contiguous limbal circumferential zones formed. Neovascularization often occurred in the total limbal vascular plexus in severely damaged corneas. Maximum vessel length was measured from the limbs toward the cauterized area within the total neovascular zones using a linear reticule through a slit lamp. To assess vessel extension, the central angle of each contiguous circumferential zone of neovascularization was measured with a 360° reticule. All results are expressed per cornea. 

Reverse Transcription–Polymerase Chain Reaction

The right eyes were removed from mice (n = 3, pooled samples of three eyes) at 24, 48, and 96 hours after cauterization under deep anesthesia. Whole corneas, including neovascular invasion, were isolated and dissected. Total retinal messenger RNA (mRNA) was extracted (Trizol; Life Technologies, Grand Island, NY) and reverse transcribed (RT; Gene Amp PCR System 9600; Perkin-Elmer, Norwalk, CT). First-strand complementary DNA (cDNA) was synthesized using AMV reverse transcriptase (Boehringer Mannheim, Indianapolis, IN) according to the manufacturer’s guidelines. Incubations were carried out for 10 minutes at 25°C and 60 minutes at 42°C, and reverse transcriptase (Boehringer Mannheim) was denatured at 99°C for 5 minutes before PCR amplification. PCR of the cDNAs was carried out in 20-μL volumes containing 10 pmol primer pair and 2 μL thermal cycling (LightCycler; Roche, Indianapolis, IN). In total, 30 amplification cycles were performed, and the PCR products were separated on 2% agarose gels. Band intensities were measured with an image sensor (Densitograph; Atto, Tokyo, Japan) with a computer-controlled display. Primers used in these experiments were as follows: β-actin, sense 5′-GTG GGC CGC TCT AGG CAC CAA-3′, antisense 5′-CTC TTT GAT GTC ACG CAC GAT TTC-3′ (product size, 539); VEGF, sense 5′-TTA CTG CTG TAC CTC CAC C-3′, antisense 5′-ACA GGA CGG CTT GAA GAT G-3′ (product size, 189 base pairs [bp]); tumor necrosis factor (TNF)α, sense 5′-GGC AGG TCT ACT TTG GAG TCA TTG-3′, antisense 5′-ACA TTC GAG GCT CCA GTG AAT TCG G-3′ (product size, 309 bp); interferon (IFN)γ, sense 5′-AGCGGCTGACTGAACTCAGATTGTAG-3′, antisense 5′-GTCACAGTTTTCAGCTGTATAGGG-3′ (product size, 213 bp); and Vα14, sense 5′-GTTGTCCGTCAGGGAGAGAA-3′, antisense 5′-TCCCTAAGGCTGAACCTCTATC-3′ (product size, 268 bp). 

mRNA Quantification by Reverse Transcription–Polymerase Chain Reaction

Total RNA was extracted from whole eyes 12 or 24 hours after photocoagulation (Trizol; Life Technologies). An aliquot (approximately 1 μg) of the total RNA was reverse-transcribed using a first-strand cDNA synthesis kit for RT-PCR (Roche Diagnostics) according to the manufacturer’s instructions. The reverse-transcribed cDNAs were then subjected to real-time RT-PCR (SYBR Premix Ex Taq; TaKaRa Bio Inc., Otsu, Japan) and thermal cycling (LightCycler; Roche). Reaction conditions were as follows: denaturing at 95°C for 10 seconds followed by 40 cycles of denaturing at 95°C for 5 seconds and annealing and extending at 60°C for 20 seconds. The level of mRNA expression for IL-18 was estimated from the fluorescence intensity relative to β-actin. Primers used were 5′-CTAAGCACAGCACGCTGCACA-3′ and 5′-AGGTATGACAATCAGCTGAGTCCC-3′ for Vα14, 5′-CCCTTTTGGGCCTATGCAG-3′ and 5′-ATGCCTCGAAGAGTTTTGCAC-3′ for CXCR6, and 5′-GATGACCCAGATCATGTTTGA-3′ and 5′-GGAGAGCATAGCCCTCGTAG-3′ for β-actin. Each experiment was repeated at least twice, and representative data are shown. 

Isolation of Corneal-Infiltrating Cells

Inflammatory cells were isolated from the corneas, as described for hepatic lymphocytes, with some modifications. ³⁰ Three corneas were pooled to obtain enough viable cells for flow cytometry. Corneas were teased away with scissors and shaken at 37°C for 40 minutes with 0.5 mg/mL collagenase (Collagenase Type D; Boehringer Mannheim, Germany). Supernatants were collected, and viable cells were counted using trypan blue dye exclusion. Eighteen corneas were examined as six pooled samples of three individual eyes. 

Antibodies and Reagents

The following reagents were used for flow cytometry: phycoerythrin (PE)-conjugated anti-CD11b monoclonal antibody (mAb; clone M1/70.15) and fluorescein isothiocyanate (FITC)-conjugated anti-F4/80 mAb (clone A3–1) from Caltag Laboratory. Inc. (San Francisco, CA); FITC-conjugated anti–Gr-1 mAb (clone RB6–8C5), Cy-Chrome 5-conjugated anti-TCRβmAb (H57–597), PE-conjugated CD4 (RM4–5) and propidium iodide staining solution from BD Bioscience (San Jose, CA); and FITC-conjugated anti-mouse 7/4 mAb from Cedarlane Laboratory. (Ontario, Canada). Antibodies used for labeling sections were anti-F4/80 mAb (clone A3–1; Serotec, Oxford, UK) and Cy5-conjugated goat anti–rat immunoglobulin G (IgG) antibody (Zymed Laboratory, San Francisco, CA). 

Flow Cytometry

Corneal-infiltrating cells were adjusted to the required concentrations for two-color labeling. Macrophages within the intraocular infiltrate were identified by double labeling with PE-anti–CD11b mAb and FITC-conjugated anti–F4/80 mAb. Cells stained with PE-anti–CD11b mAb and FITC-anti–7/4 mAb were identified as neutrophils. Analysis was performed on a flow cytometer (EPICS XL; Beckman Coulter, Mannheim, Germany) with data analysis software (FlowJo; Tree Star, San Carlos, CA). Propidium iodide staining was used to discriminate dead cells. The numbers of macrophages and neutrophils infiltrating the cornea were calculated from the percentage of each population measured by flow cytometry, and the total number of viable cells was counted by trypan blue dye exclusion. The gate within which the F4/80+ and 7/4+ cells were counted was set using an isotype-matched control antibody (FITC-conjugated rat IgG2b; Caltag Laboratory, Inc., San Francisco, CA). CD11b-F4/80− (double negative) cells, possibly containing keratocytes, ³¹ constantly made up less than 2% to 3% of the total isolated viable cells. 

Corneal Micropocket Assay

The corneal micropocket assay and the quantification of corneal neovascularization were performed as described previously. ^{32 33} Briefly, 0.3 μL poly(hydroxyethyl) methacrylate pellets (Hydron; Interferon Sciences, New Brunswick, NJ) containing 30 ng human or murine IL-1β or 200 ng murine VEGF (493-MV; R&D Systems, Minneapolis, MN) were prepared and implanted into the corneas of C57BL/6 mice, CD1d-KO mice, or Jα18-KO mice. After implantation, ofloxacin eyedrops (Santen Pharmaceuticals, Osaka, Japan) were applied to each eye to prevent infection. Six days after implantation, digital images of the corneal vessels were obtained and recorded (Viewfinder 3.0; Pixera, Los Gatos, CA), with standardized illumination and contrast, and were saved to disks. Quantitative analysis of neovascularization was performed using digital imaging software (Image, version 4.0.2; Scion Corp., Frederick, MD). 

Histologic Examination

Freshly enucleated eyes were fixed in 10% paraformaldehyde and embedded in paraffin. Sections (4 μm) were then prepared and were stained with hematoxylin and eosin solution. The numbers of neutrophils and macrophages were counted under the microscope in three independent visual fields (magnification, ×200) in the cornea and in the anterior chamber, and the average was calculated. 

Statistical Analysis

Significant differences in the grade of corneal inflammation were assessed using the Student’s t-test. P ≤ 0.05 was considered significant. 

Results

NKT-Deficient Mice Showed Augmentation of Corneal Edema and Opacity

Initially, we examined corneal inflammation in two independently derived NKT cell-deficient mice (CD1d KO and Jα18 KO). At 24 hours after cauterization, opacity scores had increased in both types of NKT-deficient mice, though the edema scores were not significantly different. At 96 hours after cauterization, both types of NKT-deficient mice showed significant increases in opacity and edema compared with control B6 mice (Figs. 1A 1B) . No differences were observed in corneal edema or opacity between CD1d-KO mice and Jα18-KO mice during the entire experiment. 

NKT-Deficient Mice Enhanced the Early Production of VEGF, IFNγ, and TNFα

To understand the functional changes that occurred in the cornea after cauterization in NKT cell-deficient mice, we examined VEGF, IFNγ, and TNFα expression at multiple time points (24, 48, and 96 hours) using RT-PCR. VEGF expression peaked 24 hours after cauterization and regressed by 48 hours in NKT cell-deficient mice (Fig. 2) . By contrast, VEGF expression peaked at 48 hours in control mice. Both types of NKT cell-deficient mice expressed higher levels of IFNγ and TNFα at 24 and 48 hours compared with control mice. 

Early Infiltration of Neutrophils into the Cornea Was Observed Histologically

We next examined the phenotype of corneal infiltrating cells during the early phase. We have previously demonstrated a role for neutrophils as the final effector cells mediating cauterization-induced corneal inflammation. ⁸ In addition, F4/80+ macrophages may contribute to the activation of accumulating neutrophils during inflammation. ³⁰ We thus examined the numbers of corneal neutrophils and macrophages. The severity of corneal inflammation was insufficient to achieve this by flow cytometry, and it was difficult to obtain enough corneal-infiltrating cells for accurate analysis. Therefore, we evaluated early changes of the cornea by histologic analysis. 

As shown in Figure 3 , marked infiltration of viable neutrophils, with clover-like nuclei (Fig. 3B) , was observed in the corneas and anterior chambers of both types of NKT cell-deficient mice but not in control mice. The thickness of the cauterized area was prominently increased in Jα18-KO mice (Fig. 3B) . 

Recruited Neutrophils and Macrophages Did Not Decrease in Number in NKT-Deficient Mice during the Later Phase

Next we examined the number of corneal-recruiting neutrophils and macrophages by flow cytometry during the effector phase (96 hours). Figure 4A(left panel) demonstrates the typical flow data of macrophages and neutrophils in B6 mice, which can clearly be discriminated by F4/80 and 7/4 antibody staining. Although the corneal edema and opacity scores were significantly increased at 96 hours, the numbers of ocular infiltrating neutrophils and macrophages did not increase (Fig. 4A , right panel). Despite the differences in the numbers of neutrophils in the early phase, the absence of NKT cells did not appear to affect the number of neutrophils or macrophages at 96 hours. As shown in Figure 4B , histologic examination revealed that B6 mice and Jα18KO mice (same as CD1d KO mice; data not shown) had equivalent levels of damage to the anterior segment of the eye. 

In addition to numbers of neutrophils and macrophages, we also determined the ratio of T lymphocytes in the corneal infiltrating cell at 96 hours. As shown in Figure 4C , the ratio of TCR⁺ T lymphocytes was approximately 1% of total infiltrating cells. We thus speculated that conventional T cells play a minimum role in this model. 

NKT Cell-Dependent Suppression Was Not Caused by VEGF-Mediated Vascularization

We compared the lengths and areas of corneal neovascularization in control mice and NKT-deficient mice and found no significant differences at 24 or 96 hours after cauterization (Fig. 5) . However, the cauterized corneas of the NKT cell-deficient mice expressed high levels of VEGF at 24 hours (Fig. 2) . To determine whether the increased inflammation at 96 hours was initiated by a VEGF-associated vascularization process, we compared the angiogenic ability in response to VEGF between both types of NKT cell-deficient mice and control mice. As shown in Figure 6 , there was no significant difference in the ability to form new vessels, suggesting that NKT cells can regulate corneal inflammation (edema and opacity) but not VEGF-mediated corneal angiogenesis. 

NKT Cells Accumulated to the Cornea in the Early Phase

To investigate the mechanism of NKT-dependent early neutrophil accumulation, we used RT-PCR to determine the location of NKT cells. We hypothesized that NKT cells would localize to the cornea during the early phase, where they would produce specific chemokines to recruit neutrophils. However, in contrast to our hypothesis, we could not detect Vα14 gene expression in the inflamed corneas until 96 hours after cauterization (Fig. 7A) . 

As shown in Figure 4C , the number of corneal infiltrating TCR⁺ cells was extremely low, so we knew there was a possibility we could not detect changes by conventional RT-PCR. In addition, it has reported that CXCR6 is another marker for NKT cells (the ligand is CXCL16). ³⁴ We thus attempted to detect the difference in the expression of the Vα14 and CXCR6 genes in the cauterized cornea in the early phase by quantitative real-time PCR. As shown in Figure 7B , we observed a significant increase in the expression of the Vα14 and CXCR6 genes in the cauterized cornea (24 hours). Compared with the positive control (spleen), the expression levels of Vα14 and CXCR6 were extremely low, and the difference could not be detected by conventional RT-PCR. However, we detected significant differences with quantitative real-time PCR. 

Discussion

NKT cells promote systemic tolerance associated with the eye and ACAID ²⁹ and are required for the generation of allospecific regulatory T cells after orthotropic corneal transplantation. It is important to distinguish the nonspecific acute corneal inflammation that occurs after injuries from corneal graft rejection. The former is mediated by macrophages and neutrophils, ^{7 8} whereas the latter are mediated by allospecific T cells. ²⁷ We therefore focused on the role of NKT cells in acute corneal inflammation after cauterization in the present study. 

Our data indicate that CD1d-restricted NKT cells play an important role in the formation of acute nonspecific corneal inflammation induced by exogenous, dangerous stimuli, such as cauterization. Independent NKT cell-deficient (CD1d-KO and Jα18-KO) mice demonstrated accelerated corneal inflammation and enhanced early (24-hour) corneal accumulation of neutrophils. Both strains of mice also upregulated VEGF, TNFα, and IFNγ in the cornea 24 hours after cauterization. Thus, we concluded that NKT cells regulate the early accumulation of neutrophils and protect the cornea from excessive inflammation. 

The increase in the number of neutrophils was temporary and was observed at 24 hours but not at 96 hours after cauterization. We postulate that early-accumulated neutrophils must be important for initiating functional changes and for activating infiltrating macrophages and corneal resident cells (such as keratocytes) critical for the further development of corneal inflammation. ^{31 35} This might explain why both types of NKT cell-deficient mice showed severe corneal edema and opacity 96 hours later, even though the number of neutrophils and macrophages was not significantly reduced at this time. 

Our data contrast with the finding that NKT cells promote neutrophil accumulation in a lipopolysaccharide-induced hepatitis model ³⁶ and in corneal Pseudomonas aeruginosa infection. ³⁷ These studies also showed that NKT cell-deficient mice had reduced areas of hepatic injury and corneal IFNγ production. The reason for this discrepancy in our own results is unclear. However, NKT cells have the multipotential to regulate the immune response, and the type of reaction in vivo varies according to the systems or organs. Indeed, in vivo treatment of αGalCer, which is an NKT cell ligand, induced the production of either T-helper 1 (Th1) or Th2 cytokines, depending on the dose and timing of the inoculations. ³⁸ The environmental differences between the eye and the liver may be reflected by the type of NKT cell activation. 

Although corneal angiogenesis is an important step in the development of corneal edema and opacity, the two processes could be separate. Our study focused on corneal edema and opacity, which are directly associated with loss of visual acuity in patients, rather than on corneal angiogenesis. Even though both types of NKT cell-deficient mice expressed higher levels of VEGF in the cornea, early VEGF expression did not augment corneal neovascularization in these animals (Fig. 5) , which we confirmed with the corneal micropocket assay embedded with exogenous VEGF (Fig. 6) . The difference in corneal inflammation between wild-type mice and NKT cell-deficient mice was not dependent on the growth of new corneal vessels induced by VEGF. However, it is known that VEGF is not merely an angiogenic factor but that it is also a proinflammatory cytokine. ³⁹ Early VEGF can, therefore, still affect the subsequent inflammatory process in the cauterized cornea. 

Although the number of recruited NKT cells must have been very small (Fig. 4B) , we could detect significant differences in the expression of the Vα14 and CXCR6 genes with the use of quantitative real-time PCR. NKT cells can be effective, even in low numbers, in initiating immune reactions. ^{9 10} NKT cells might regulate neutrophil function in their natural state or development level. Indeed, Hwang et al. ⁴⁰ previously demonstrated the hyperactivity of neutrophils in NKT-deficient mice, which could be the reason for early neutrophil accumulation after corneal cauterization. 

Our data indicate a protective role for CD1d-restricted invariant NKT cells against nonspecific corneal inflammation. In vivo stimulation of NKT cells by their specific ligands may have potential for use in therapeutic intervention. Further study is required to elucidate the role of NKT cells in various corneal disorders associated with inflammation. 

 Supported in part by grants from the Ministry of Education, Science, Sports and Culture, Japan (C2 No.16571757: K-HS).

 Submitted for publication March 8, 2007; revised August 18, 2007; accepted November 20, 2007.

 Disclosure: T. Oshima, None; K.-H. Sonoda, None; S. Nakao, None; K. Hijioka, None; M. Taniguchi, None; T. Ishibashi, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Corneal inflammation in NKT-deficient mice. Corneas (n = 10) were cauterized with silver nitrate in control (C57BL/6J) and NKT cell-deficient (CD1d KO and Jα18 KO) mice. Corneal edema and opacity were examined at 24 hours and 96 hours after cauterization (A) and were scored according to severity (B). *P < 0.05, **P < 0.01, compared with control mice. The experiment was performed three times with similar results.

Figure 2.

 

VEGF, TNFα, and IFNγ expression after cauterization. The right eyes (n = 3) were removed at 24, 48, and 96 hours after cauterization from control (B6) and NKT cell-deficient (CD1 KO and Jα18 KO) mice, and total corneal mRNA was extracted for the examination of VEGF, TNFα, and IFNγ expression. Three corneas were pooled to obtain sufficient mRNA. The experiment was performed three times with similar results.

Figure 3.

 

Corneal inflammation in NKT-deficient mice. Right eyes (n = 5) were removed at 24 hours after cauterization. Sections (4 μm) were stained with hematoxylin and eosin, numbers of neutrophils and macrophages were counted in three independent visual fields (magnification, ×200) in the cornea and anterior chamber, and the averages were calculated. (A) Early infiltration of neutrophils (solid black bars) and macrophages (hatched bars) in control (B6; n = 10), Jα18-KO (n = 10), and CD1d-KO (n = 10) mice. The experiment was performed twice with similar results. (B) Histologic sections of cauterized corneas from control (B6) and Jα18-KO mice, with an enlarged area of the Jα18-KO sample shown on the right.

Figure 4.

 

Corneal-infiltrating neutrophils and macrophages after cauterization. Inflammatory cells were isolated from corneas 96 hours after cauterization. Three corneas were pooled to obtain enough viable cells for flow cytometry. Eighteen corneas were examined as six pooled samples of three individual eyes. (A, left) Typical discriminative plots of corneal-infiltrating neutrophils (7/4) and macrophages (F4/80) by flow cytometry in control C57BL/6 mice. Right: analysis of corneal cell infiltrates in control (C57BL/6J) and NKT cell-deficient (CD1 KO and Jα18 KO) mice. The experiment was performed three times with similar results. No significant differences were detected. (B) Histologic sections of cauterized corneas from control (B6) and Jα18-KO (96 hours) mice. (C) Analysis of TCR⁺ cells in cornea (96 hours).

Figure 5.

 

Evaluation of new corneal vessels after cauterization. Cauterized corneas (n = 5) from control (C57BL/6J) and NKT cell-deficient (CD1 KO and Jα18 KO) mice were evaluated for corneal neovascularization, by length and extension, 24 and 96 hours after cauterization. The experiment was performed twice with similar results. No significant differences were detected.

Figure 6.

 

Comparison of new vessel-forming ability against exogenous VEGF expression. (A) Poly(hydroxyethyl) methacrylate pellet-embedded corneas of control (C57BL/6J) and NKT cell-deficient (CD1 KO and Jα18 KO) mice after 96 hours on corneal micropocket assay. (B) Evaluation of corneal neovascularization induced by exogenous VEGF showed no significant difference between control (C57BL/6J) and NKT cell-deficient (CD1 KO and Jα18 KO) mice. The experiment was performed twice with similar results.

Figure 7.

 

Vα14 gene expression in cauterized corneas. (A) The right eyes (n = 3) were removed at 24, 48, and 96 hours after cauterization from control (B6) and NKT cell-deficient (CD1 KO and Jα18 KO) mice, and total corneal mRNA was extracted for the examination of Vα14 mRNA. Three corneas were pooled to obtain sufficient mRNA. The experiment was performed twice with similar results. (B) Nontreated (n = 3) and cauterized (n = 3) corneas at 24 hours were removed, and levels of V α 14 and CXCR6 mRNA were examined by quantitative real-time PCR. The experiment was performed twice with similar results.

The authors thank Mari Imamura and Michiyo Takahara for their excellent technical support.