Anti-Angiogenic Effect of KR-31831 on Corneal and Choroidal Neovascularization in Rat Models

Abstract

Purpose.: We attempt to determine the effect and mechanism of KR-31831 in rat models of corneal neovascularization and choroidal neovascularization (CNV).

Methods.: Corneal neovascularization was induced by silver nitrate cauterization. Balanced salt solution (for control), KR-31831 (0.1 mg/mL), and bevacizumab (10 mg/mL) were applied topically with or without subsequent subconjunctival injection (10 μL). The degree of corneal neovascularization was compared among treatments. The effects of intravitreal (0.1 and 0.3 mg/mL) and intraperitoneal (25 mg/kg) of KR-31831, and intravitreal injection of bevacizumab (2.5 mg/mL) were compared in a laser-induced CNV model. FITC-dextran was used to observe the choroid vessels and to evaluate vessel leakage by fluorescence intensity.

Results.: In the silver nitrate cauterized rat, topical KR-31831 (P = 0.008) or bevacizumab (P = 0.008) reduced effectively the area of corneal neovascularization compared to control on day 14. This was reduced further by additional subconjunctival injection of KR-31831 (P = 0.024) and bevacizumab (P = 0.016). After KR-31831 application, vascular endothelial growth factor (VEGF) receptor 2 and matrix metalloproteinase (MMP)-2 expression was decreased in the cornea. In the CNV model, intravitreal (0.3 mg/mL) and intraperitoneal KR-31831 inhibited significantly the CNV area (P = 0.008 and P = 0.008, respectively) and fluorescence leakage (P = 0.008 and P = 0.032, respectively). This effect was more significant compared to intravitreal bevacizumab in terms of the CNV area (P = 0.032 and P = 0.008, respectively) and fluorescence leakage (P = 0.016 and P = 0.008, respectively).

Conclusions.: The anti-angiogenic effect of KR-31831 was comparable in the cornea and more effective in the choroid compared to that of bevacizumab, and it may exert its effect by VEGF signaling and MMP-2.

Introduction

Angiogenesis involves enzymatic degradation and remodeling of the extracellular matrix, and the migration and proliferation of capillary endothelial cells. ¹ These processes are regulated tightly by the balance among many stimulatory and inhibitory factors. ² When this balance is disrupted, stimulation by angiogenic factors, such as vascular endothelial growth factor (VEGF) and fibroblast growth factor (FGF), triggers the proliferation and migration of vascular endothelial cells into the surrounding tissues. The newly formed dysfunctional blood vessels are leaky, fragile, and prone to rupture and hemorrhage, a condition that is associated with fibrous proliferation. 

Pathologic angiogenesis is found in various diseases of the eye, including corneal neovascularization, diabetic retinopathy, retinopathy of prematurity, neovascular glaucoma, and choroidal neovascularization (CNV) in age-related macular degeneration. VEGF has a key role in the angiogenic process of neovascularization in the eye, ^3–6 and its secretion and receptor expression are increased in neovascularization of the cornea and retina. ^7–9  

KR-31831, a 4-(N-imidazol-2-yl-methyl)aminobenzopyran analogue, was designed originally as an ATP-sensitive potassium (K_ATP)-channel opener, and is used in the treatment of ischemic diseases, such as myocardial infarction and stroke. ^6,10 It has an antioxidant effect and anti-apoptotic activity against ischemic brain injury in rats in vitro. ^8,10 These results suggest that K_ATP channels in the mitochondrial or plasma membrane confer protection against ischemia. ^2,4 In several studies, KR-31831 was shown to exert anti-angiogenic activity, as measured by the inhibition of human umbilical vein endothelial cells. ^11–13 It also inhibited VEGF-activated cell proliferation, migration, and invasion. ¹⁴ Potassium channel activators increase potassium conductance, leading to hyperpolarization of the plasma membrane, which consequently inhibits the activation of voltage-sensitive Ca²⁺ channels. ¹⁵ The inhibition of Ca²⁺ channels mediates signal transduction, and inhibits the proliferation and invasion of endothelial cells. ^16,17 Also, KR-31831 is thought to exert its anti-angiogenic effect by downregulation of VEGF receptor (VEGFR)-2 and fetal liver kinase (Flk)-1, and reduction of matrix metalloproteinase-2 (MMP)-2.^12,18  

We investigated the effects of KR-31831 on angiogenesis in rat models of corneal neovascularization and CNV delivered at different concentrations, and by different methods. Changes in the expression levels of Flk-1 and MMP-2 were observed to determine the anti-angiogenic mechanism by KR-31831. The results were compared to those using the common anti-VEGF agent, bevacizumab (Avastin; Roche, Basel, Switzerland). 

Materials and Methods

Materials

KR-31831, (2R,3R,4S)-6-amino-4-[N-(4-chloropheyl)-N-(1H-imidazol-2ylmethyl)amino]-3-hydroxyl-2-methyl-2-dimethoxymethyl-3,4-dihydro-2H-1-benzopyran, was synthesized by the Korea Research Institute of Chemical Technology (Daejeon, Korea). Bevacizumab (Avastin; Roche) was used as a positive control for comparing anti-angiogenic effects. 

Animals

Male Long-Evans rats, 8 weeks old and weighing 250–300 g each, were used in this study. The test and control groups each comprised five animals for each experimental procedure, for a total of 105 animals used. The animals were treated according to the regulations of the Ethics Committee of the Catholic University of Korea, Seoul, and the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals (NIH Publication no. 80-23, revised 1996). All animals were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. All efforts were made to minimize suffering and the number of animals used for this study.  

In Vivo Rat Model of Corneal Neovascularization Induced with Silver Nitrate Cauterization

To establish the corneal neovascularization model, general anesthesia was induced in rats by an intramuscular injection of ketamine (40 mg/kg) and xylazine (8 mg/kg), supplemented with topical anesthesia (0.4% proparacaine hydrochloride). Under a surgical microscope, a stick (end diameter 2.5 mm) coated with silver nitrate (75%) and potassium nitrate (25%, Graham Field Health Products, Atlanta, GA), was applied to the central cornea of the right eye of each rat for 10 seconds to create an acute chemical burn over an area of approximately 2 × 2 mm. The degree of chemical burn was not scored, but the area of the burn was measured with calipers. Eyes with burn areas larger or smaller that 2 × 2 mm were discarded. The eyes were rinsed immediately with 5 mL of saline solution. 

First, to examine the dose-dependent response, KR-31831 at concentrations of 0.01, 0.05, 0.1, and 0.3 mg/mL was applied topically to the ocular surface. The lowest dose showing effective inhibition of neovascularization was 0.1 mg/mL for KR-31831. This dose was used in subsequent studies (Fig. 1). For bevacizumab, previous studies demonstrated that topical application of 10 mg/mL bevacizumab showed the most effective inhibition of corneal neovascularization, and this dose was used in subsequent studies. ^19–22 To compare the effects of KR-31831 and bevacizumab, the rats were randomized into two groups: topically-treated only (Group T), and topically- and subconjunctivally-treated (Group TS). The control group was treated with balanced salt solution (BSS), and subdivided into topically-treated only (Group T-1) or topically- and subconjunctivally-treated (Group TS-1) groups. The bevacizumab group was treated at a concentration of 0.1 mg/mL, and also subdivided into topically-treated only (Group T-2), or topically- and subconjunctivally-treated (Group TS-2) groups. The KR-31831 group was treated at a concentration of 10 mg/mL, and likewise subdivided into topically-treated only (Group T-3), or topically- and subconjunctivally-treated (Group TS-3) groups. The subconjunctival injection of each drug was done with 10 μL at 1.0 mm from the superior limbus. 

Figure 1. 

 

Corneal neovascularization in the rat silver nitrate cauterization model. In the control group, significant corneal neovascularization developed around the entire cornea. Topical application of KR-31831 produced dose-dependent inhibition of new vessels, and 0.1 mg/mL was the lowest dose showing the most effective inhibition of the area of corneal neovascularization and FLk-1 expression (n = 5 in each group).

Evaluation of Corneal Neovascularization

Corneal neovascularization was observed at 7 and 14 days after silver nitrate cauterization, and digital photographs were taken using a digital camera (FinePix S602 Zoom, Minatoku, Tokyo, Japan) at the same focus setting (×25). Each photograph was analyzed at the same magnification using ImageJ 1.40g image analysis software (supplied by Wayne Rasband, Research Services Branch, National Institute of Mental Health, Bethesda, MD), and the area of neovascularization was measured in pixels. The photographs were numbered randomly, and two investigators were double-blinded for the data analysis to minimize observer bias. 

Western Blot Analysis of Cornea Proteins

At 14 days after silver nitrate cauterization, rats were euthanized with an intramuscular injection of ketamine (40 mg/kg) and xylazine (8 mg/kg), and the corneas from the treated eyes were harvested. To quantify VEGF and Flk-1 protein expression, neovascularized areas of equal size (3.0 × 3.0 mm) were dissected from each cornea, frozen at −70°C, and then homogenized in 200 mL of ice-cold buffer solution containing 1 mM EDTA, 10 mM Tris, pH 7.6, 0.1 M NaCl, 1 mg/mL aprotinin, and 100 mg/mL phenylmethylsulfonyl fluoride. The samples were clarified by centrifugation for 5 minutes at 13,000 revolutions per minute, and the supernatants were collected. The protein concentration was measured with Bio-Rad Bradford total protein assay reagent (Bio-Rad, Hercules, CA) and adjusted to 1.0 mg of protein in 200 mL of buffer. Then, the protein was rinsed with cold PBS and lysed with lysis buffer (20 mM Tris, pH 7.4, 150 mM NaCl₂, 1 mM EGTA, 1 mM EDTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM orthovanadate, 1 mM phenylmethylsulfonyl fluoride, 1 μg/mL leupeptin, and 10 μg/mL aprotinin). Identical amounts of the protein lysates were resolved by 4–12% SDS-PAGE and electroblotted onto nitrocellulose membranes (Invitrogen, San Diego, CA). The membranes were blocked with 5% skim milk in PBS and then incubated overnight with anti-VEGF antibody (1:5000). For MMP-2 evaluation, anti-mmp2 (Santa Cruz, Santa Cruz, CA; 1:500) , and anti-α-tubulin (Santa Cruz; 1:500) were used as primary. The following secondary antibodies and dilutions were used to detect primary antibodies: anti-goat (Santa Cruz; 1:2000), anti-rabbit (Santa Cruz; 1:1000), and anti-mouse (Santa Cruz; 1:1000). Bound antibodies were visualized using enhanced chemiluminescence solution (ECL reagent; Santa Cruz, CA).  

In Vivo Rat Model of Choroidal Neovascularization after Laser Injury

For the choroidal neovascularization model, general anesthesia was induced in the rats with an intramuscular injection of ketamine (40 mg/kg) and xylazine (8 mg/kg), supplemented with topical anesthesia (0.4% procaine hydrochloride). A warming pad maintained the body temperature of the rat during the procedure. The pupils were dilated with topical 1% tropicamide (Santen, Osaka, Japan), and four burns were made in each retina using 647.1-nm krypton red laser photocoagulation (260 mW; spot size, 0.1 mm; duration, 0.05 seconds) via a slit-lamp delivery system (Coherent Novus2000 laser, Oberkochen, Germany ) with a 5.4-mm hand-held contact fundus lens (Ocular Instruments, Bellevue, WA). The burns were grouped at the 3 o'clock position relative to the optic disc. Production of a bubble, indicating rupture of Bruch's membrane, without bleeding at the time of laser-hitting, was considered a valid burn. Lesions in which the four laser spots became confluent were excluded from subsequent study. At the third day after laser injury, a 10-μL volume of drug was injected into the vitreous cavity with a 30-gauge needle. Both eyes of each animal received the same agent to avoid any potential crossover effect. A careful examination using a slit-lamp was performed after injection, to confirm that no animal had a traumatic cataract or vitreous hemorrhage. Levofloxacin ophthalmic solution (Oculevo; Samil, Seoul, Korea) was applied to the ocular surface before and immediately after injection. The control group had intravitreal injection of 10 μL of BSS (Group IV-1). Bevacizumab was injected intravitreally with a concentration of 2.5 mg/mL (Group IV-2). This concentration of bevacizumab was selected based on previous dose-dependent studies that showed 2.5 mg/mL bevacizumab resulted in the most effective inhibition of CNV. ^23,24 KR-31831 was injected intravitreally with a concentration of 0.1 mg/mL (Group IV-3) or 0.3 mg/mL (Group IV-4). KR-31831 was injected intraperitoneally at 25 mg/kg for 1 week from the third day after laser injury (Group IP).  

FITC-Dextran Angiography

A solution of FITC-dextran (2 × 10⁶ molecular weight; Sigma, St. Louis, MO) was prepared at a concentration of 10 mg/mL in PBS. Rats were anesthetized, 0.5 mL of FITC-dextran solution was injected into the tail vein of each rat, and a cover glass was placed on the cornea as a contact lens. Angiography was performed on day 14 after laser burn, and digital images at a magnification of 40× were captured using a charge-coupled device (CCD) camera (DC500; Leica, Switzerland) attached to a fluorescence stereomicroscope (MZ-III; Leica, Wetzlar, Germany).  

Digital Image Processing

The digital images obtained with the CCD camera were imported into public domain (free) image analysis software, Image J program (available from http//rsb.info.nih.gov/ij/). The hyperfluorescent area (pixels) on FITC-dextran angiography was evaluated at three laser burn sites. Bridging neovascularization was included by measuring the total area of the bridging CNV membrane and then dividing by the number of underlying laser lesions from which the network originated. The intensity of the fluorescence leakage was measured at the major vessels at each laser burn site. The CNV parameters of each sample were measured three times in a masked fashion, and the average values were analyzed. 

Microscopic Evaluation

After 1 week, 25 experimental rats (five subgroups of five rats each) that had been treated with KR-31831 or bevacizumab using different routes of administration were anesthetized by an intramuscular injection of ketamine (40 mg/kg) and xylazine (8 mg/kg), and their eyes were enucleated and harvested. The eyes of two rats in each subgroup were analyzed histologically, and the eyes of three rats were evaluated by transmission electron microscopy. 

Statistical Analysis

Statistical analyses were performed using the Kruskal-Wallis test for comparison among several groups and the Mann-Whitney U test for comparison between two groups. Comparison between days 7 and 14 in each group was performed with the Wilcoxon signed-rank test. SPSS 11.0 (SPSS, Chicago, IL) was used for all analyses, with P < 0.05 taken to indicate significance. 

Results

In Vivo Rat Model of Corneal Neovascularization after Silver Nitrate Cauterization

Topical application of bevacizumab or KR-31831 in silver nitrate cauterized rats reduced corneal neovascularization, which was reduced further by subsequent subconjunctival administration of each drug (Fig. 2). The mean area of corneal neovascularization was 68.0 ± 4.5 on day 7 and 139.0 ± 3.5 on day 14 in Group T-1 (control). This was decreased significantly with topical bevacizumab in Group T-2 to 37.2 ± 5.3 on day 7 (P = 0.008) and 115.8 ± 3.8 on day 14 (P = 0.008). With topical application of KR-31831, the mean area of corneal neovascularization was decreased significantly to 32.8 ± 4.4 on day 7 (P = 0.008) and 113.4 ± 5.9 on day 14 (P = 0.008) in Group T-3. Comparison between Groups T-2 and T-3 did not show significant difference for the inhibition of corneal neovascularization on day 7 (P = 0.222) and day 14 (P = 0.332). With topical and subconjunctival treatment, the mean area of corneal neovascularization was 68.6 ± 4.2 on day 7 and 139.6 ± 4.0 on day 14 in Group TS-1 (control). With bevacizumab treatment in Group TS-2, this was decreased significantly to 34.4 ± 3.8 on day 7 (P = 0.008) and 104.2 ± 5.5 on day 14 (P = 0.008). With KR-31831 in Group TS-3, the area of corneal neovascularization also was decreased significantly to 31.2 ± 4.2 on day 7 (P = 0.008) and 101.2 ± 6.5 on day 14 (P = 0.008). Comparison between Groups TS-2 and TS-3 showed that the effect was similar on day 7 (P = 0.222) and day 14 (P = 0.421). 

Figure 2. 

 

Topical and subconjunctival application of KR-31831 compared to bevacizumab (Avastin) in the rat silver nitrate corneal neovascularization model on days 7 and 14 (n = 5 in each group). Topical application of KR-31831 (Group T-2) and bevacizumab (Group T-3) suppressed new vessel formation compared to control (Group T-1). With additional subconjunctival application, KR-31831 (Group TS-2) and bevacizumab (Group TS-3) further suppressed neovascularization of the cornea compared to control (Group TS-1). *P < 0.05, Mann-Whitney U test.

Comparison between topical treatment only and combined treatment by topical and subconjunctival application of the drug showed that there was no significant difference on day 7 with bevacizumab (Group T-2 versus Group TS-2, P = 0.421) and KR-31831 (Group T-3 versus Group TS-3, P = 0.421). However, combined treatment showed stronger inhibition of the area of corneal neovascularization on day 14 with bevacizumab (Group T-2 versus Group TS-2; P = 0.016) and KR-31831 (Group T-3 versus Group TS-3, P = 0.024). 

Based on Western blot analysis, topical bevacizumab and KR-31831 significantly decreased Flk-1 protein expression and MMP-2 expression compared to control (Fig. 3). 

Figure 3. 

 

With topical application of KR-31831 and bevacizumab, the expression of Flk-1 and MMP-2 was inhibited by Western blot analysis of the cornea. Ava, bevacizumab group (10 mg/mL); KR, KR-31831 group (0.1 mg/mL); n = 5 in each group. *P < 0.05, Mann-Whitney U test.

In Vivo Rat Model of Choroidal Neovascularization with Laser Injury

To induce CNV, four burns were made in the retina using krypton red laser photocoagulation (Fig. 4A). Histologically, the laser injury caused a break in the retinal pigment epithelial layer with retinal degeneration around the burn (Fig. 4C), compared to the normal parts of the retina in the same rat (Fig. 4B). One week after the laser injury, new vessel formation in the retina was detected on histologic sections and by transmission electron microscopy (Fig. 4D). 

Figure 4. 

 

Choroidal neovascularization induced by laser injury in a rat. (A) The laser-burned rat retina. (B, C) Compared with the control retina (B), the laser-injured retina showed disrupted retinal pigment epithelium and accompanying retinal degeneration (C). (D) One week after the laser injury, new vessel formation was seen in the retina. Electron microscopy showed red blood cells in the new vessels.

The area of new vessels was measured 14 days after drug administration (Fig. 5A). CNV area was 4210.0 ± 233.5 in Group IV-1 (control). This was reduced significantly to 2061.6 ± 145.3 (P = 0.008) after intravitreal injection of bevacizumab in Group IV-2. After intravitreal injection of 0.1 mg/mL of KR-31831, the CNV area was reduced significantly to 2207.0 ± 125.7 (P = 0.008) in Group IV-3. With intravitreal injection of 0.3 mg/mL of KR-31831, the CNV area was reduced significantly to 1737.8 ± 164.2 (P = 0.008) in Group IV-3. With intraperitoneal injection of KR-31831 in Group IP, CNV area also was decreased significantly compared to Group IV-1 (1665.8 ± 210.0, P = 0.008). Comparison between Groups IV-2 and IV-3 showed that bevacizumab injection and 0.1 mg/mL of KR-31831 had similar effects in reducing the CNV area (P = 0.151). However, comparison between Groups IV-2 and IV-4 showed that 0.3 mg/mL of KR-31831 had better effects in reducing the CNV area than intravitreal bevacizumab (P = 0.032). Comparison between Groups IV-2 and IP demonstrated that intraperitoneal injection of high dose KR-31831 showed better effects in inhibition of the CNV area than bevacizumab (P = 0.008). Comparison between Groups IV-4 and IP showed that intraperitoneal injection of high dose of KR-31831 had no additive effect in reducing the CNV area when compared to 0.3 mg/mL of KR-31831 (P = 0.548). 

Figure 5. 

 

After inducing CNV in the rat retina, FITC-dextran angiography was used to observe the CNV area and fluorescein leakage from the CNV (40× magnification; n = 5 in each group). With administration of 2.5 mg/mL intravitreal bevacizumab (Group IV-2) or 0.1 mg/mL intravitreal KR-31831 (Group IV-3), 0.3 mg/mL intravitreal KR-31831 (Group IV-4), and 25 mg/kg intraperitoneal KR-31831 (Group IP), the CNV area (A) and fluorescein intensity (B) were decreased significantly. Both 0.3 mg/mL intravitreal (Group IV-4) and intraperitoneal (Group IP) injections of KR-31831 showed more effective inhibition of CNV area and fluorescein intensity compared to bevacizumab (Group IV-2). *P < 0.05, Mann-Whitney U test.

The fluorescence intensity of injected FITC-dextran, which reflects leakage from new vessels, was compared among the groups and showed similar results with the CNV area (Fig. 5B). After CNV induction, fluorescence intensity was increased to 68.6 ± 2.3 in Group IV-1 (control). In Groups IV-2, IV-3, IV-4, and IP the fluorescence intensity was decreased significantly to 45.0 ± 3.5, 50.0 ± 1.6, 38.8 ± 4.2, and 40.0 ± 1.6, respectively, compared to Group IV-1 (P = 0.008, P = 0.008, P = 0.008, and P = 0.032, respectively). Comparison between Groups IV-2 and IV-3 showed that bevacizumab injection and 0.1 mg/mL of KR-31831 had similar effects in reducing the fluorescence intensity (P = 0.056). However, comparison between Groups IV-2 and IV-4 showed that 0.3 mg/mL of KR-31831 had better effects in reducing the fluorescence intensity than bevacizumab (P = 0.016). Comparison between Groups IV-2 and IP demonstrated that intraperitoneal injection of high dose KR-31831 showed better effects in inhibition of the CNV area than bevacizumab (P = 0.008). 

Discussion

Anti-angiogenic drugs currently under development can be categorized based on their mechanisms of action, which include blocking of extracellular matrix degradation, direct inhibition of endothelial cell proliferation or migration, inhibition of endothelial cell-specific integrin/survival signaling, blocking of pro-angiogenic gene promoters, and unknown mechanism of action. The drugs showing the most promising results are those that target VEGF signaling. ^25,26 The angiogenic activity of VEGF is thought to be mediated by two high-affinity receptor tyrosine kinases, VEGFR-1 (fms-like tyrosine kinase, Flt-1) and Flk-1. Flk-1 is reported to dominate the angiogenic activity of VEGF, ²⁷ and thus many therapeutic approaches have targeted the VEGF pathway. 

In our study, KR-31831 significantly inhibited neovascularization of the cornea and retina, and decreased the expression of Flk-1 and MMP-2 protein in the cornea. The anti-angiogenic activity of KR-31831 has been shown to involve VEGF signaling. ^12,13 Physiologically, upon binding of VEGF, Flk-1 undergoes dimerization and is activated by ligand-dependent tyrosine phosphorylation, resulting in mitogenic, chemotactic, and prosurvival signals. ^28,29 In endothelial cells, VEGF induces the phosphorylation of several proteins and activates the Raf-Mek-Erk pathway, which promotes endothelial cell growth. KR-31831 has been reported to downregulate VEGF-induced phosphorylation of Erk1/2 and, thereby, to inhibit endothelial cell growth. ¹³ In addition, KR-31831 downregulates the transcription and translation of Flk-1. In another study, KR-31831 reduced the production and activation of pro- and active-MMP-2, ¹² which may result in decreased extracellular matrix degradation. KR-31831 acts via these anti-angiogenic mechanisms mediated by Flk-1 and MMP-2 to inhibit broadly neovascularization. In our study, KR-31831 showed anti-angiogenic effects in the cornea and retina that were comparable to or more effective than those of bevacizumab, which is used widely to treat neovascular diseases of the eye. 

The healthy cornea is avascular and alymphatic, owing to a balance between pro-angiogenic and anti-angiogenic regulators. Several anti-angiogenic factors in the cornea have been studied, including angiostatin, restin, endostatin, and soluble VEGFR-1, which acts as an endogenous VEGF trap. Pro-angiogenic factors include VEGF, bFGF, and MMP-2. ¹⁸ Corneal neovascularization may be caused by immunologic, degenerative, or infectious processes, or chemical injury. Recently, several reports have shown that subconjunctival bevacizumab inhibits corneal neovascularization in animal models, and that topically applied bevacizumab limits corneal neovascularization after chemical burns. ^30,31 In our study, topical application of bevacizumab or KR-31831 effectively inhibited neovascularization in silver nitrate cauterized rat corneas. The combination of topical administration and subconjunctival injection of bevacizumab or KR-31831 further decreased corneal neovascularization. KR-31831 exhibited dose-dependent inhibition of corneal neovascularization. The degree of the inhibition of corneal neovascularization was not different between KR-31831 and bevacizumab, showing that KR-31831 was as effective as bevacizumab. Our results suggest that the effects of KR-31831 are mediated by VEGF/Flk-1 and MMP-2; however, previous results have suggested that KR-31831 may act via a different mechanism. 

Recently, drugs targeting VEGF have been made available for treatment of CNV, and are used widely to treat various neovascular diseases of the retina. They include an anti-VEGF aptamer (pegaptanib sodium), a humanized recombinant anti-VEGF monoclonal antibody (bevacizumab; Avastin), and a recombinant anti-VEGF antibody fragment (ranibizumab; Lucentis). ^32–34 While findings suggest strongly that VEGF acts as a major stimulator of CNV, non-VEGF pathways and other growth factors that signal via its receptors also may be involved in neovascularization. Some studies have suggested that the inhibition of VEGF signaling alone is sufficient to decrease CNV, whereas others have demonstrated even more potent suppression of angiogenesis when drugs targeting multiple pathways are used. In this regard, a pharmacologic strategy to inhibit multiple angiogenic pathways may be a more desirable therapeutic approach. ³⁵ KR-31831 is a promising agent that may inhibit CNV via multiple angiogenic pathways. Other than VEGF/Flk-1 signaling, reduced MMP-2 may result in decreased extracellular matrix degradation and exert additional anti-angiogenic mechanism. In addition, current anti-VEGF therapies, although efficacious, require prolonged treatment regimens with frequent intravitreal injections, and consequently carry some risks. In our study, high dose of KR-31831 administered by intraperitoneal injection, which is a convenient and less invasive delivery route, showed better results for inhibiting CNV when compared to intravitreal bevacizumab. Compared with intravitreal bevacizumab injection, intraperitoneal KR-31831 injection significantly reduced the area of CNV and vessel leakage. Also, our results showed that the efficacy of KR-31831 in suppressing CNV was better than that of bevacizumab, with intravitreal (0.3 mg/mL) and high dose intraperitoneal administration. Further studies are required to compare the safety of repeated KR-31831 and bevacizumab injections to confirm this advantage of KR-31831. 

In conclusion, our data suggest that KR-31831 blocks VEGF signaling and MMP-2 mediated pathways in newly developing vessels, and significantly reduces neovascularization of the cornea and retina. The inhibitory effect of KR-31831 may be of great value in the treatment of corneal neovascularization. Intravitreal or intraperitoneal injection of KR-31831 also was effective in suppressing CNV. KR-31831 may have therapeutic potential comparable to that of bevacizumab in angiogenic disorders of the eye. 

Footnotes

 Supported by Grant CBM34-B3003-01-00-00 (CKJ) from the Center for Biological Modulators of the 21C Frontier R&D Program, Ministry of Education, Science and Technology, Korea. Supported in part by the Martin and Toni Sosnoff Foundation.

Footnotes

 Disclosure: I.-T. Kim, None; H.-Y.L. Park, None; J.-S. Choi, None; C.-K. Joo, None