Role of Macrophage Migration Inhibitory Factor in Corneal Neovascularization

Abstract

purpose. To determine the role of macrophage migration inhibitory factor (MIF) in inflammatory corneal neovascularization.

methods. Corneal neovascularization was induced by suturing 10-0 nylon 1 mm away from limbal vessel or limbal scraping after 0.15 M NaOH application in BALB/c mice. MIF expression was evaluated by semiquantitative reverse transcription-polymerase chain reaction (RT-PCR), Western blot analysis, and immunohistochemistry. To investigate the function of MIF in inflammatory corneal neovascularization, the neovascularized area and number of infiltrating F4/80-positive cells (monocytes/macrophages) were compared between wild-type mice and homozygous MIF-deficient mice.

results. MIF mRNA and protein markedly increased in the neovascularized corneas compared with normal corneas by RT-PCR and Western blot analysis, respectively. MIF expression was upregulated immunohistochemically, not only in the corneal epithelium but also in the stromal infiltrating cells of neovascularized corneas. Neovascularized area in corneas of MIF^−/− mice was significantly small compared with that in wild-type mice on day 7 after corneal suture and on day 14 after limbal scrape, and MIF^−/− cornea had ∼30% less neovascularized area than did wild-type cornea in both models. Neovascularized corneas in MIF-deficient mice had significantly fewer monocytes/macrophages than those in wild-type control mice.

conclusions. These findings indicate that MIF, abundantly expressed in neovascularized corneas, has an angiogenic role in inflammatory corneal neovascularization and may be a therapeutic target for suppression of corneal neovascularization.

Corneal neovascularization reflects an important aspect of the pathogenesis of many blinding disorders and a major sight-threatening condition in corneal infection, in chemical injury, and after keratoplasty in which neovascularization adversely affects corneal graft survival. Corneal neovascularization represents a manifestation of inflammatory response. Chemokines, crucial mediators of inflammatory and immune responses, contribute to multiple aspects of corneal neovascularization. ^{1 2} However, the mechanisms of corneal neovascularization have not yet been fully understood in inflammatory situations. 

Macrophage migration inhibitory factor (MIF) is a secretory product of corticotropic and thyrotropic pituitary cells released in response to stress. ³ MIF was first identified as a T-cell derived lymphokine. ⁴ Monocytes and macrophages, originally considered to be the target of MIF action, were found to express MIF in response to various proinflammatory stimuli and to be a significant source of MIF release in vivo. ⁵ MIF exhibits a broad range of immunostimulatory and proinflammatory activities in arthritis, ^{6 7} glomerulonephritis, ⁸ and adult distress respiratory syndrome. ⁹ MIF is also expressed in ocular tissues including cornea. ^{10 11} The expression of MIF in cornea is upregulated in pathologic challenge in animal models such as corneal wound healing and infection, ^{11 12} suggesting that MIF may play important roles in corneal pathology. Recently, the importance of MIF in immune response and inflammation has been confirmed by MIF-deficient mice, which develop a normal phenotype, but are severely deficient in TNF-α response to bacterial endotoxin both in vitro and in vivo. ¹³ Neutralization of MIF by anti-MIF antibody leads to dose-dependent improvement of arthritis in rodent models. ¹⁴ These data demonstrate that MIF is a critical factor in the setting of a pathologic challenge without playing an obvious role in development and normal growth. 

Studies in MIF have also shed some light on the biology of angiogenesis. The blocking MIF activity suppresses vascular formation in several cancer models. ^{14 15 16 17 18} Recently, Amin et al. ¹⁹ clearly demonstrated that MIF has angiogenic effect through MAP kinase and PI3 kinase. Because MIF possesses proinflammatory properties as well as angiogenic ones, we hypothesized that MIF may be involved in inflammatory corneal neovascularization. In this study, we therefore investigated the contribution of MIF to corneal neovascularization. 

Materials and Methods

Animals and Preparation of the Corneal Neovascularization Model

All animal experiments were approved by the University of Tokyo Hospital Animal Care Committee and conformed to the ARVO Statement for the use of Animals in Ophthalmic and Vision Research. The homozygous BALB/c-background MIF-deficient (MIF^−/−) mice were generated as described elsewhere. ²⁰ Wild-type BALB/c mice (Saitama Experimental Animals, Saitama, Japan) were used as controls. All procedures were performed with the animals under general anesthesia by xylazine hydrochloride (5 mg/kg) and ketamine hydrochloride (35 mg/kg). The animals were allowed free access to food and water. A 12-hour day–night cycle was maintained. After general anesthesia, corneal neovascularization was induced by suturing 10-0 nylon 1 mm away from limbal vessel under microscopy. We also investigated the angiogenic responses using another alkali injury of corneal neovascularization models. ²¹ Briefly, after general anesthesia, 2 μL of 0.15 M NaOH was applied to the corneal surface, and then total corneal limbus and epithelium were scraped off with a blade, assisted by a microscope. Erythromycin ophthalmic ointment was instilled immediately after the procedure. 

Semiquantitative RT-PCR

The gene expression levels of MIF on vascularized corneas were investigated by reverse transcription–polymerase chain reaction (RT-PCR). Total RNA was isolated from normal and vascularized corneas (Isogen; Nippon Gene, Tokyo, Japan) and cDNA was produced with reverse transcriptase (SuperScript II; Invitrogen, San Diego, CA). Water was used as a negative control. The PCR conditions were as follows: appropriate cycles of 45 seconds at 94°C, 45 seconds at 55°C, and 45 seconds at 72°C, with an initial 5-minute denaturation step and a final 7-minute elongation step. The PCR primer pair was selected to discriminate between cDNA and genomic DNA by using primers specific for different exons. The primers for MIF (5′ primer, CCA TGC CTA TGT TCA TCG TG; 3′ primer, AGG CCA CAC AGC AGC TTA CT) yielded a 250-bp product and the primers for glyceraldehyde-3-phosphate dehydrogenase (G3PDH) (5′ primer, GGT GAA GGT CGG TGT GAA CGG A; 3′ primer, TGT TAG TGG GGT CTC GCT CCT G) yielded a 223-bp product. Samples were separated in a 2% agarose gel, and the products were visualized with ethidium bromide. An optical scanner was used to determine the densities of the gel bands of the PCR products and to standardize them as to those for G3PDH. The linear amplified curve of the PCR product of each sample was examined at four-cycle intervals. Within the linear range of amplification, four sets of PCR products were prepared under appropriate cycling conditions, and the band densities were compared among the groups. NIH Image (version 1.62, available by ftp at zippy.nimh.nih.gov/ or at http://rsb.info.nih.gov/nih-image; developed by Wayne Rasband, National Institutes of Health, Bethesda, MD) was used for the quantification of band density. 

Western Blot Analysis and Immunohistochemistry

MIF protein expression levels in normal and vascularized corneas were evaluated by Western blot analysis and immunohistochemistry. For Western blot analysis, the eyes were enucleated, and the corneal samples were placed into 150 mL of lysis buffer (20 mM imidazole HCl, 10 mM KCl, 1 mM MgCl₂, 10 mM EGTA, 1% Triton, 10 mM NaF, 1 mM sodium molybdate, and 1 mM EDTA [pH 6.8]) supplemented with a protease inhibitor cocktail (Roche Diagnostics, Indianapolis, IN) and sonicated. The lysate was centrifuged at 14,000 rpm for 15 minutes at 4°C. The samples were boiled for 5 minutes and separated by SDS-polyacrylamide gel electrophoresis under denaturing conditions, and electroblotted to a polyvinylidine difluoride (PVDF) membrane (Bio-Rad, Hercules, CA). The membranes were incubated in blocking buffer, followed by the rabbit anti-MIF polyclonal antibody (Novus Biologicals, Littleton, CO), and then washed and incubated with a horseradish peroxidase-labeled anti-rabbit antibody (GE Healthcare, Piscataway, NJ). The blot was visualized with an ECL Plus kit (GE Healthcare) according to the manufacturer’s instructions. NIH Image was used for the quantification of band density. 

For immunohistochemistry, eyes were enucleated and embedded in OCT compound, snap frozen in liquid nitrogen, and cut into 7-μm-thick sections. The sections were applied with rabbit anti-MIF polyclonal antibody as the primary antibody at 2 μg/mL at room temperature for 1 hour. For the negative control, nonimmunized serum (Vector Laboratories, Burlingame, CA) was used in place of the primary antibody. Immunoreactivity was detected with a kit (Histofine SAB-PO; Nichirei, Tokyo, Japan), according to the manufacturer’s protocol. Briefly, the samples were incubated with biotinylated anti-rabbit goat serum for 15 minutes at room temperature and then rinsed with PBS, after which they were incubated with a streptavidin-biotin-peroxidase complex for 10 minutes at room temperature. The final reaction product was visualized with 3,3′-diaminobenzidine tetrahydrochloride (DAB). Counterstaining was performed with hematoxylin. 

Lectin Angiography and Neovascularization Quantitation

Corneal neovascularization was imaged by lectin angiography. Mice received intravenous BS-1 lectin conjugated with FITC (10 μg/g; Vector Laboratories) and were killed 30 minutes later. The eyes were enucleated and fixed with 1% paraformaldehyde for 15 minutes. After fixation, the corneas were placed on glass slides and studied by fluorescence microscopy (Leica, Deerfield, IL), as described elsewhere. ²² Briefly, NIH Image was used for the image analysis. The neovascularization was quantified by setting a threshold level of fluorescence, above which only vessels were depicted. The neovascularization quantitation was performed in a masked manner. The vascularized area was outlined, with the innermost vessel of the limbal arcade used as the border. 

Monocyte-Macrophage Counts

Eyes were enucleated 2 and 7 days after injury, embedded in OCT compound, snap frozen in liquid nitrogen, and cut into 7-μm-thick sections. After fixation with ice-cold acetone and blocking with normal goat serum, the sections were stained with rat anti-mouse F4/80 (macrophage marker; Biomedicals AG, Rheinstrasse, Switzerland) monoclonal antibody to detect infiltrated macrophage, followed by staining with FITC-conjugated goat anti-rat IgG2b (1:100; Santa Cruz Biotechnology, Santa Cruz, CA). The samples were observed under a fluorescence microscope and counted in a masked fashion. Eight serial sections extending through the wound and limbus were studied per cornea. The number of F4/80-positive cells between the limbus and suture was determined at 50× magnification. 

Statistical Analysis

The Mann-Whitney test was used to compare the band densities on RT-PCR, neovascularized area, and F4/80-positive cell number. P < 0.05 was considered significant. The experiments were repeated twice; representative data are presented. 

Results

MIF Expression Level in Normal and Vascularized Corneas

We investigated MIF expression level in normal (n = 3) and vascularized (n = 3) corneas by semiquantitative RT-PCR and Western blot analysis, respectively. MIF gene expression was detected in the normal mouse corneas (Fig. 1) , but not in negative control samples (data not shown). MIF gene expression levels were not stimulated at 3 and 12 hours after suturing; however, at 24 hours and 3 and 7 days after surgery, the levels were significantly increased compared with those in on normal corneas (P < 0.05, Fig. 1 ). Constitutive MIF protein expression was detected in normal mouse corneas by Western blot analysis, as shown in Figures 2A and 2B . The expression level of MIF protein 7 days after suturing was significantly higher than in normal corneas (P < 0.05, Fig. 2 ). Relatively weak expression of MIF was detected in the corneal epithelium in normal corneas (Fig. 2D) . However, MIF expression was upregulated in the epithelium of vascularized corneas and was also detected in stromal infiltrating cells (Fig. 2E) . 

Corneal Neovascularization in Wild-Type Control and MIF^−/− Mice

The area of corneal neovascularization was compared between the wild-type control and gene-targeting MIF^−/− mice after corneal suture. Figures 3A and 3Bshowed representative neovascularized corneal area in the respective wild-type control and MIF^−/− mice on day 7 after corneal suture. The neovascularized area in the corneas of MIF^−/− mice (n = 8) was significantly smaller than that in wild-type control corneas (n = 9) on day 7, and MIF^−/− corneas had approximately 30% less neovascularized area than did control corneas (P < 0.05; Fig. 3C ). We confirmed those findings in another relevant alkali limbal injury model. As shown in Figures 3D and 3E , MIF^−/− corneas (n = 6) exhibited less corneal neovascularization than did control corneas (n = 7) on day 14 after limbal scraping and had approximately 30% less neovascularized area than did control mice (P < 0.05, Fig. 3F ). 

Macrophage Infiltration in Wild-Type Control and MIF^−/− Mice

Because macrophages are the target for MIF action ^{5 8 9} and are potent inducers of angiogenesis, ^{23 24} we evaluated the number of infiltrating macrophages in the neovascularized areas of the stroma in normal corneas and at days 2 and 7 corneas after suturing. There was no statistically significant difference in the number of F4/80-positive cells between wild-type and MIF^−/− mice on day 2 after suturing (Fig. 4 ; 10.7 ± 7.0 [wild-type] vs. 8.5 ± 3.8 [MIF^−/−], P > 0.05). The number of F4/80-positive cells in MIF^−/− corneas at 7 days after suturing was significantly lower than that in the wild-type control corneas (Fig. 4 ; 25.5 ± 7.0 [wild-type] vs. 10.5 ± 5.2 [MIF^−/−], n = 8 per condition, P < 0.01). 

Discussion

Corneal neovascularization is closely associated with local inflammation and many chemokines and cytokines are involved in this process. ^{1 2 25} In our findings, MIF gene and protein expression was upregulated in vascularized cornea compared with normal cornea. The corneal neovascularization area was significantly reduced in MIF^−/− mice than in wild-type mice. These findings imply that stimulation by suturing or alkali injury increases MIF expression in cornea, and MIF has a critical role in inflammatory corneal neovascularization Moreover, the number of F4/80-positive monocytes-macrophages, not neutrophils or T-lymphocytes (data not shown), decreased in corneas of MIF^−/− mice compared with those in control mice. Because monocyte-macrophage lineage cells participate in the neovascularization process, ^{23 24 26} fewer F4/80-positive cells in the cornea of MIF^−/− mice may at least in part explain the promotion mechanism of neovascularization by MIF in the cornea. 

Involvement of MIF in corneal pathologic conditions has been described elsewhere. ^{11 12} In these reports, MIF expression was markedly upregulated at early time points after pathologic challenge (3–4 hours after challenge), such as corneal injury and Pseudomonas infection. ^{11 12} However, we did not detect the upregulation of MIF until 24 hours after suturing (Fig. 1) , perhaps because the suturing area was very small in our model and it took a longer time to activate corneal epithelial cells and to accumulate MIF-expressing inflammatory cells. Moreover, the difference of invasion of infiltrating cells in different animal models may also explain the discrepancy in the expression pattern of MIF. 

Various potential mechanisms have been proposed for the apparent proinflammatory actions of MIF. MIF has well-characterized macrophage-activating effects, including such responses as cytokine release and augmentation of nitric oxide production. ^{5 9} The pathogenic roles of MIF-induced tissue inflammation have been attributed to monocyte recruitment in a model of glomerulonephritis. ⁸ Gregory et al. ²⁷ showed that MIF directly promotes interactions between leukocytes and endothelial cells in inflamed microcirculation. Monocytes-macrophages can potently induce and enhance angiogenesis through releasing angiogenic factors such as VEGF. ²⁸ Therefore, these studies have suggested that the function of MIF may extend beyond immune responses and that MIF may also play pivotal roles in angiogenesis. ^{29 30 31 32 33 34} Recently, direct mechanisms of proangiogenic activities of MIF have been shown by Amin et al. ¹⁹ They demonstrated that MIF directly promotes migration of endothelial cells and angiogenesis in an ex vivo matrigel assay and in vivo corneal pocket assay. Thus, MIF is recognized as a pleiotropic cytokine possessing angiogenic properties in addition to immune responses. Therefore, it is not surprising that MIF is a potent angiogenic factor in corneal neovascularization. Although previous studies showed MIF upregulation in inflammatory corneal conditions such as wound-healing processes and Pseudomonas infection, ^{11 12} to our knowledge, this is the first report demonstrating that MIF is involved in pathologic inflammatory angiogenesis in vivo. 

The mechanisms underlying the enhanced inflammatory angiogenic response should be complex and mutually related, and a single factor cannot explain the whole angiogenic response, although we showed that MIF is an angiogenic factor in corneal neovascularization. Of note, MIF induces and promotes the expression of angiogenic factors, such as vascular endothelial growth factor (VEGF) and interleukin-8. ^{34 35} VEGF, a potent endothelial growth factor, is also a chemotactic factor for monocyte lineage cells and stimulates the inflammatory responses in corneal neovascularization. ²² These reports and our findings suggest that blockade of MIF expression may directly and indirectly suppress angiogenesis in the cornea. 

In conclusion, our data define the biological significance of MIF in inflammatory corneal neovascularization. Because MIF is a proangiogenic factor in corneal neovascularization, MIF targeting therapy could be a potent strategy for inhibiting inflammatory corneal neovascularization. 

 Submitted for publication June 23, 2006; revised November 23, 2006, and January 9, 2007; accepted June 4, 2007.

 Disclosure: T. Usui, None; S. Yamagami, None; S. Kishimoto, None; Y. Seiich, None; T. Nakayama, None; S. Amano, 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.

 

MIF gene expression level determined by semiquantitative RT-PCR. cDNAs, prepared from untreated, normal corneas and sutured corneas at the indicated time points, were subjected to analysis for MIF gene expression level. Within the linear range of amplification, four sets of PCR products were prepared in 25 cycles, and the band densities were compared. Significantly increased MIF gene expression levels are detected in corneal samples of 3, 12, and 24 hours and 2 and 7 days after operation, compared with those of normal corneas. Representative data are shown. *P < 0.05.

Figure 2.

 

MIF protein by Western blot analysis and immunohistochemical study in normal and neovascularized corneas. (A, B) Proteins prepared from untreated corneas and sutured corneas were subjected to analysis for quantification of MIF. MIF protein was higher in vascularized corneas than in normal corneas. Three sets of results on separate experiments were analyzed with similar findings. Representative data are shown. *P < 0.05. (C–E) Immunohistochemical localization of MIF in normal and vascularized corneas. Corneal sections after treatment with nonimmunized serum (negative control, C) and anti-MIF polyclonal antibody (D, E). In normal cornea (D), relatively weak expression of MIF was detected in corneal epithelium (*). Sections were stained with anti-MIF antibody in the corneal epithelium ( ), showing infiltrating cells in the vascularized corneas (arrows) (E).

Figure 3.

 

Corneal neovascularization in wild-type control and MIF^−/− mice. Angiogenic responses in wild-type and MIF-deficient mice corneas were evaluated. Corneal neovascularization was induced by two types of animal models; suturing and alkali injured. Representative photographs of mouse corneas on day 7 after intrastromal suturing with 10-0 nylon 1 mm away from the limbus (A, B) and on day 14 after alkali injury (C, D). MIF^−/− mice (B, D) developed less neovascularization than control mice (A, C). The surface area (in pixels) of corneal neovascularization is shown 7 days after suturing (E) and 14 days after alkali injury (F). Error bars, mean ± SD. *P < 0.05.

Figure 4.

 

Macrophage infiltration in wild-type control and MIF^−/− mice. Seven days after suturing, the corneas were examined to quantify the number of monocyte lineage cells in the corneal stroma. Immunohistochemical study was performed with antibodies against F4/80, a monocyte-macrophage cell marker. The recruitment of F4/80-positive cells was suppressed in wild-type mice (A) compared with MIF-deficient mice (B). The cells were counted from the limbus to the suture. A fluorescence microscopic study revealed a significantly lower number of F4/80-positive cells in MIF^−/− corneas (•) at 7 days after suturing than in the wild-type control corneas (▪). *P < 0.05.