SB-267268, a Nonpeptidic Antagonist of α_vβ₃ and α_vβ₅ Integrins, Reduces Angiogenesis and VEGF Expression in a Mouse Model of Retinopathy of Prematurity

Abstract

purpose. To determine whether SB-267268, a nonpeptidic antagonist of the α_vβ₃ and α_vβ₅ integrins, attenuates angiogenesis in a murine model of retinopathy of prematurity (ROP) and alters the expression of vascular endothelial growth factor (VEGF) and its second receptor (VEGF-R2).

methods. In receptor binding, SB-267268 exhibited nanomolar potency for human, monkey, and murine α_vβ₃ and α_vβ₅. SB-267268 inhibited the attachment of α_vβ₃-transfected HEK293 cells to microtiter plate wells precoated with RGD-containing matrix proteins, and vitronectin-mediated human and rat aortic smooth-muscle–cell migration. At postnatal day (P)12, C57BL/6 mice were exposed to 80% oxygen for 7 days followed by 7 days in room air (angiogenic period). Between P12 and P17, ROP mice were administered sterile saline (vehicle intraperitoneal [i.p.]) or SB-267268 (60 mg/kg bi-daily, i.p.). Shams were exposed to room air from P0 and administered either vehicle or SB-267268 during P12 to 17. In at least 3 randomly chosen paraffin sections from each eye, the number of blood vessel profiles in the inner retina were counted. In situ hybridization for VEGF and VEGFR-2 was performed on at least 8 randomly chosen paraffin sections from each eye.

results. SB-267268 reduced pathologic angiogenesis in ROP mice by approximately 50% and had no effect on developmental retinal angiogenesis in shams. Both VEGF and VEGFR-2 mRNA were upregulated in the inner retina of ROP mice and reduced with SB-267268.

conclusions. Nonpeptidic inhibition of α_vβ₃ and α_vβ₅ integrins is effective in ROP and may be a suitable anti-angiogenic therapy for other ischemic retinal pathologies.

Integrins are a family of multifunctional cell-adhesion molecules composed of noncovalently associated α and β chains. ¹ They are transmembrane receptors that bind extracellular matrix components, including vitronectin, fibronectin, laminin, collagen, fibrinogen, and thrombospondin. Integrins are involved in the regulation of a wide variety of cellular events, including adhesion, migration, invasion, proliferation, and cell survival and apoptosis. ^{2 3} A number of studies have implicated the α_vβ₃ and α_vβ₅ integrins in angiogenesis. α_vβ₃ is located on actively proliferating endothelial cells in human diabetic retinopathy and retinopathy of prematurity (ROP), ⁴ and in human wound granulation tissue and chick chorioallantoic membrane. ⁵ Inhibition of α_v integrins with cyclic peptide antagonists and antibodies reduces angiogenesis in solid tumors and is associated with downregulation of the potent angiogenic and permeability factor, vascular endothelial growth factor (VEGF). These findings have suggested that inhibition of α_v integrins may be an appropriate strategy for organ protection in a variety of angiogenic pathologies. ^{5 6 7}  

Angiogenesis is the hallmark feature of ischemic retinopathies such as ROP ⁸ and proliferative diabetic retinopathy. ⁹ In both diseases, new blood vessels proliferate in the inner retina and penetrate the vitreous cavity. These blood vessels are abnormally formed, leading to sight-threatening hemorrhage and edema. ^{8 9} In these instances, pathologic angiogenesis is linked with upregulation of VEGF and its second receptor, VEGFR-2. ^{10 11} α_vβ₃ and α_vβ₅ integrins may participate in the pathologic angiogenesis that occurs in ROP and proliferative diabetic retinopathy. ⁴ For example, in mouse models of ROP, the administration of cyclic peptide antagonists of α_v integrins and α_vβ₃ attenuates retinal angiogenesis ⁴ and also reduces normal developmental vascularization of the retina in mice. ¹²  

SB-267268 is a small molecule antagonist of both α_vβ₃ and α_vβ₅ integrins whose structure is based on a 2-benzazepine template (Fig. 1) . ¹³ The aim of the present study was to determine whether administration of SB-267268 reduces retinal angiogenesis in mice with ROP and alters retinal VEGF and VEGFR-2 expression. 

Methods

Animals

Pregnant female C57BL/6 mice were obtained from the Animal Resource Centre, Perth, Western Australia. The mothers were randomly divided into 4 experimental groups with 7 to 9 pups per group. These groups consisted of (1) sham treated with vehicle (n = 9), (2) sham treated with the α_vβ₃/α_vβ₅ integrin antagonist SB-267268 (n = 7), (3) ROP treated with vehicle (n = 8), and (4) ROP treated with SB-267268 (n = 9). Shams were mice exposed to room air from birth until P17. The ROP model in mice followed a previously published method. ¹⁴ Seven-day-old pups and their mother were housed in sealed chambers that contained 75 ± 5% O₂ and 2% CO₂, using medical grade O₂ and industrial grade air. Gas levels in the chamber were monitored twice daily by using a gas analyzer (ML 205; AD Instruments, Pty. Ltd, Australia) and a chart recorder (Chart v3.5 program on the MacLab/2E System; AD Instruments, Bella Vista, NSW, Australia). An airflow rate of approximately 2.5 L/min assisted in maintaining adequate levels of metabolically produced CO₂ and drops in O₂ tension. Mice remained in the chamber for 5 days (hyperoxic period, P7 to P12) and were then housed in room air for a further 5 days (hypoxic-induced angiogenic period, P12 to P17). 

SB-267268 was provided by GlaxoSmithKline (CVU CEDD; King of Prussia, PA). SB-267268 was dissolved in sterile water and administered at a dose of 60 mg/kg bi-daily to pups by intraperitoneal (i.p.) injection. Sham pups were administered sterile water (vehicle) bi-daily by i.p. injection. SB-267268 and vehicle were administered at 0830 hours and 0430 hours each day. The injection volume was 100 μL. 

During the experiment, mothers were provided with water and standard mice chow (GR2; Clark-King and Co., Gladesville, Victoria, Australia) ad libitum and were exposed to normal 12-hour light/dark cycles. Pups received nutrition from their mothers. To avoid respiratory distress, each day, mother and pups were removed from the chamber and placed in room air for 2 hours. Experimental procedures were consistent with the guidelines set by the Australian National Health and Medical Research Council Code of Practice for the Care and Use of Animals for Scientific Purposes and adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Receptor Binding Assays

The affinity of SB-267268 for various integrins was determined by using binding assays, as described previously, using purified receptors and a RGD-containing cyclic peptide [³H]-SKF107260 as the ligand. ¹⁵ Briefly, 96-well plates were coated with integrins a_vb₃, a_vb₅, and a₅b₁ purified from human placenta, and a_IIbb₃ purified from human platelets. After blocking of nonspecific sites with Buffer A (50 mM Tris-HCl, 100 mM NaCl, 1 mM CaCl₂, 1 mM MnCl₂, pH 7.4) containing 3% BSA, various concentrations of the compound were added to the wells and followed by the addition of [³H]-SKF107260 (5 nM). After incubation, the wells were aspirated completely, washed twice with Buffer A, and bound [³H]-SKF107260 was solubilized and counted. The binding of SB-267268 to mouse integrins was assessed in an ELISA-type integrin/ligand binding assay by using a commercial platform, MesoScale Discovery (MSD), Gaithersburg, MD. The assay used either native integrin protein or recombinant purified integrin protein. Recombinant protein was derived from a baculovirus expression system and comprised the extracellular domain fused to a fos/jun dimerization cassette and His/FLAG tags. Recombinant fibronectin protein expressed in Escherichia coli as a GST fusion protein (GST-fibronectin) was used as a ligand. In the assay, 5 ng/well integrin protein was bound directly to the electrode (in the case of human α_vβ₃ and α_vβ₆ proteins) or was added to wells of MSD plates coated with 10 ng/well anti-6xhis capture antibody (in the case of 6xhis-tagged mouse α_vβ₃, mouse α_vβ₆, and rat α_vβ₆ proteins). SB-267268, 10 nM biotinylated fibronectin, and 20 nM ruthenylated streptavidin were added, and the plates were incubated for 4 hours. Read buffer was then added, and the plates were read on the MSD instrument. SB-267268 inhibition of integrin/ligand binding was measured as a decrease in the signal in the assay. Data were analyzed with a commercial software (ActivityBase; NovaScreen, Hanover, MD), using a 4-parameter logistic equation for curve fitting. 

Cell Adhesion Assays

The effect of SB-267268 on the adhesion of cells was determined in a cell adhesion assay by using HEK-2893 cells cotransfected with human a_v and b₃. ¹⁶ Briefly, 96-well plates were coated overnight with 0.2 μg/mL of human vitronectin or rat osteopontin. The plates were washed once with PBS and blocked with 3% BSA, followed by the addition of α_vβ₃ expressing HEK293 cells (50,000 cells) in Roswell Park Memorial Institute (RPMI); 20 mM HEPES, pH 7.4, and 0.1% BSA were added to the well. After 1 hour of incubation at 37°C, the cells were fixed by the addition of 25 μL of a 10% formaldehyde solution, pH 7.4, at room temperature for 10 minutes. The plates were washed three times with 0.2 mL of PBS with 0.1% BSA, and the adherent cells were stained with 0.1 mL of 0.5% toluidine blue for 20 minutes at room temperature. Excess stain was removed by extensive washing with deionized water, and toluidine blue incorporated into cells was eluted by the addition of 0.1 mL of 50% ethanol that contained 50 mM HCl and was quantitated by measuring absorbance at 630 nm on a microtiter plate reader (Titertek Multiskan MC, Sterling, VA). 

Cell Migration Assay

Human and rat aortic smooth-muscle–cell (SMC) migration was evaluated as described previously. ¹⁷ Briefly, migration of SMCs was examined in Transwell culture chambers by using a polycarbonate membrane with pores of 8 microns (Costar; Corning, Cambridge, MA). The SMCs were suspended in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 0.2% BSA at a concentration of 2.5 × 10⁶ cells/mL. In the standard assay, 0.2 mL human or rat SMC was placed in the upper compartment of the chamber. The lower compartment contained 0.6 mL DMEM supplemented with 0.2% BSA, vitronectin, and SB-267268 or vehicle. The chambers were incubated at 37°C in an atmosphere of 95% air and 5% CO₂. After incubation, nonmigrated cells on the upper surface of the membrane were scraped gently and washed with PBS. The filters were then fixed with methanol and stained with Giemsa. The number of SMCs per ×150 high-power field that had migrated to the lower surface of the filters was determined microscopically. Four randomly chosen high-powered fields were counted per filter. 

Histopathology

After the 17-day experimental period, the mice were killed by an i.p. injection of 120 mg/kg per body weight pentobarbital sodium (Nembutal; Rhone Merieux, Queensland, Australia). Both eyes from each animal were removed and fixed for 3 hours in 4% paraformaldehyde fixative in 0.1M phosphate buffer. The eyes were processed in graded alcohols before being embedded in paraffin wax. The eyes were then serially sectioned at 3 μm, 90° to the optic nerve and placed on 3 aminopropyl-triethoxysilane (Sigma, St. Louis, MO) coated slides. Approximately 120 sections/eye were collected and incubated overnight at 37°C. 

Three sections from 1 eye from each animal were randomly chosen, deparaffinized, and stained with Mayer’s hematoxylin (5 minutes) and eosin (5 minutes) (Amber Scientific Laboratories, Belmont, Australia), and coverslipped. By using an established technique, ^{11 14 18} blood vessel profiles (BVP) were counted in the inner retina and included vessels adherent to the inner limiting membrane (ILM). The inner retina comprised the ILM, the ganglion cell layer (GCL), the inner plexiform layer (IPL), and the inner nuclear layer (INL). Four nonoverlapping fields per section were evaluated in a masked manner. A BVP was defined as an endothelial cell (stained blue) or a blood vessel with a lumen. Counting was performed on a photomicroscope (Olympus BH-2; Olympus, Tokyo, Japan) at a magnification of ×40, and images were captured on a digital camera connected to an IBM computer (Spot digital camera; SciTECH Pty. Ltd., Preston, Victoria, Australia). Two investigators masked to the experimental groups counted BVPs. 

In Situ Hybridization for VEGF and VEGFR-2

Riboprobes were synthesized from cDNAs encoding mouse VEGF and VEGFR-2 (a gift from Steven Stacker, Ludwig Institute, Parkville, Australia). ¹¹ The cDNAs were cloned into pGEM 4Z (Promega, Madison, WI) and linearized with HindIII to produce antisense probes by using a polymerase (SP6 RNA polymerase; Promega). Three-μm paraffin sections of eye premounted on 1% 3-aminopropyltriethoxysilane coated slides were dewaxed, rehydrated in graded ethanol and milliQ water, equilibrated in P buffer (50 mM Tris-HCL, pH 7.5, 5 nM EDTA), and incubated in 125 μg/mL Pronase E (Bio Scientific, Gynea, NSW, Australia) in P buffer for 10 minutes at 37°C. Sections were then washed in 0.1 M sodium phosphate buffer (pH 7.2), briefly refixed in 4% paraformaldehyde (Crown Scientific Pty. Ltd, Scoresby, Victoria, Australia) for 10 minutes, rinsed in milliQ water, dehydrated in 70% ethanol, and air-dried. Hybridization buffer containing 2 × 10⁴ cpm/ μL riboprobe in 300 mM NaCl, 10 mM Tris-HCL (pH 7.5), 10 mM Na₂HPO₄, 5 mM EDTA (pH 8.0), 1× Denhardt’s solution, 50% formamide, 17 mg/mL yeast RNA, and 10% wt/vol dextran sulfate was heated to 85°C for 5 minutes, and 25 μL of this solution was then added to each section. Hybridization was performed overnight at 60°C in 50% formamide-humidified chambers. Sections hybridized with sense probes for VEGF and VEGFR-2 were used as controls for nonspecific binding. After hybridization, slides were washed in ×2 SSC containing 50% formamide prewarmed to 50°C to remove the coverslips. Sections were then washed in the above-described solution for 1 hour at 55°C, rinsed three more times in RNase buffer (10 mM Tris-HCL, pH 7.5, 1 mM EDTA, pH 8.0, 0.5 M NaCl), and incubated with RNase A (150 μg/mL) for 1 hour at 37°C. Sections were later washed in ×2 SSC for 45 minutes at 55°C, dehydrated in graded ethanol, air-dried, and exposed to autoradiographic film (Kodak X-Omat; Kodak, Rochester, NY) for 5 days. Slides were subsequently dipped in emulsion (Illford LM1; Ilford, Cheshire, UK), stored in a light-free box with desiccant at 4°C for 4 weeks, immersed in developer (Kodak D19; Kodak, Rochester, NY), fixed (Ilford Hypam; Ilford), and stained with hematoxylin and eosin. 

Dark field images were captured by using light microscopy and a digital camera (Fujix HC-2000; Fuji, Tokyo, Japan). The outline of the inner retina (ILM, GCL, IPL, and INL) was defined by interactive tracing. Gene expression was then quantitatively measured to determine the proportion of the area occupied by autoradiographic grains as previously described ¹¹ by using computerized image analysis (AIS; Imaging Research, Ontario, Canada). All sections were hybridized to their respective probes in the same experiment and analyzed in duplicate (n = 8 sections per rat, 6 to 8 rats per group). All analyses were done with the observer masked to the animal study group. 

Statistics

All values are expressed as mean ± SEM. Data were analyzed by ANOVA followed by a Fisher’s post hoc comparison, with a value of P < 0.05 considered statistically significant. A statistics program (Statview for Windows, Version 5.0.1; SAS Institute Inc., Cary, NC) was used to analyze data. 

Results

Receptor Binding, Cell Adhesion, and Cell Migration

In a radioligand displacement assay, SB-267268 potently displaced [³H]-SKF107260 binding to human and monkey α_vβ₃ and human α_vβ₅ with Ki values of 0.9 nM, 0.5 Nm, and 0.7 nM, respectively. However, evaluation of SB-267268 on [³H]-SKF107260 binding to related integrin receptors demonstrated >1000-fold selectivity for the human α_vβ₃ receptor versus the human α_IIbβ₃, α₅β₁, and α_3β1 receptors (see Table 1 ). In further binding assays with native or recombinant nonlabeled integrins and fibronectin as ligand, SB-267268 inhibited the binding of fibronectin to human and mouse α_vβ₃ with IC₅₀ values of 0.68 nM and 0.29 nM, respectively. SB-267268 was much less potent for inhibition of human, mouse, and rat α_vβ₆ integrin (see Table 1 ). Consistent with the binding data, using an in vitro cell adhesion assay format, SB-267268 inhibited the attachment of both α_vβ₃-transfected HEK293 cells to microtiter plate wells precoated with arginine-glycine-aspartic acid (RGD)-containing matrix proteins with IC₅₀ values of 12 nM (see Table 1 ). SB-267268 also inhibited vitronectin-mediated human and rat aortic SMC migration with IC₅₀ values of approximately 12.3 nM and 3.6 nM, respectively. The effect of SB-267268 on α_vβ₃-mediated cell adhesion and migration appeared to be relatively specific, because this compound did not exhibit any appreciable activity when profiled in a screen against a panel of 46 receptors and 9 enzymes (data not shown). 

Body Weight

At P17, there was no significant difference in body weight between the groups of animals; sham+vehicle (7.15 ± 0.11 g), sham+SB-267268 (7.13 ± 0.15 g), ROP+vehicle (7.05 ± 0.08 g), and ROP+SB-267268 (7.10 ± 0.10 g). 

BVPs in the Inner Retina

The results are shown in Figure 2 . In sham mice treated with vehicle, BVPs were observed in the inner retina. SB-267268 treatment to sham mice did not change the appearance of the retina or the density of BVPs in the inner retina. In ROP mice treated with vehicle, the number of BVPs increased in the inner retina, and some vessels penetrated the retinal surface and into the vitreous cavity. In ROP mice, SB-267268 reduced BVPs in the inner retina by 50% compared with ROP+vehicle. 

VEGF and VEGFR-2 Gene Expression

The results are shown in Figures 3 and 4 . Intense VEGF and VEGFR-2 expression was detected in the INL of all retinas from all groups, with less expression observed in the GCL. In sham mice, SB-267268 did not reduce VEGF or VEGFR-2 gene expression in the inner retina. In ROP mice treated with vehicle, both VEGF and VEGFR-2 mRNA increased in the INL and GCL compared with all sham groups. In ROP mice treated with SB-267268, VEGF and VEGFR-2 gene expression in the INL and the GCL was reduced, with VEGFR-2 mRNA levels similar to sham groups. 

Discussion

Anti-angiogenic therapies are currently being evaluated for the treatment of ischemic retinopathies, for example, proliferative diabetic retinopathy. One approach is the development of therapies that prevent the adhesion of vascular endothelial cells to the extracellular matrix. The α_vβ₃ and α_vβ₅ integrins are potential targets, because they are implicated in new blood vessel formation and maturation in a variety of tissues. ^{5 7 12 19} The major results from the present study are that chronic administration of a nonpeptidic combined α_vβ₃/α_vβ₅ antagonist, SB-267268, reduces pathologic but not developmental angiogenesis in neonatal mice. The anti-angiogenic effects of SB-267268 in ROP may involve downregulation of the VEGF/VEGFR-2 system. The effect of SB-267268 in ROP appears to be because of its selective activity on α_vβ₃ and α_vβ₅ integrins, because, in vitro, this compound demonstrated potent and selective binding to α_vβ₃ and α_vβ₅ receptors and inhibition of cell adhesion and migration. SB-267268 exhibits greater than 1000-fold selectivity for other integrins and no activity on a panel of approximately 50 other receptors and enzymes. Overall, these findings have relevance for the development of noninvasive therapies for ischemic retinopathies, for example, proliferative diabetic retinopathy. 

Evidence that α_v integrins are important in retinal angiogenesis is provided by studies in mouse models of ROP, where ligation inhibition of α_v-type integrins with cyclic penta-petid peptide reduces angiogenesis when administrated by either i.p. injection ^{20 21} or topically. ²² Other studies have highlighted the importance of α_vβ₃ integrins in ischemic retinopathies. The α_vβ₃ integrins have been localized to neovascular tissue removed from the retinal surface during vitrectomy from patients with proliferative diabetic retinopathy and in new vessels in mice with ROP. ⁴ Similarly, α_vβ₃ integrin and its major ligand, vitronectin, as well as α_vβ₁ integrins are found in both basal and luminal surfaces of endothelial cells in vascularized tissues from patients with proliferative diabetic retinopathy. ²³ The efficacy of inhibition of α_vβ₃ integrins has been evaluated in animal models of retinal angiogenesis. A cyclic antagonist XJ735 given to mice with ROP reduces retinal angiogenesis when administered by either i.p. or periocular injection. ⁴ In rats with laser-induced choroidal angiogenesis, a cyclic α_vβ₃ antagonist inhibits the progression of the lesion. ²⁴ These findings are consistent with studies in other experimental models in which α_vβ₃ antibodies, for example, LM609, decreased angiogenesis and tumor regression ⁷ and improved arthritic disease. ²⁵ The anti-angiogenic properties of α_vβ₃ is most likely to be caused by the suppression of the activity of p53 and p53-inducibe cell-cyclic inhibitor p21^WAF/CIP1 and increases in the Bcl:Bax ratio, with a consequent anti-apoptotic effect. ²⁶ In more recent years, a humanized monoclonal IgG₁ antibody that binds human integrin α_vβ₃ has been developed. In preclinical studies, a monoclonal antibody (Vitaxin, Gaithersburg, MD) reduces artery size in balloon-injured hypercholesterolemic rabbits, ²⁷ which may be caused by a decrease in transforming growth factor β₁ and an increase in cellular apoptosis. Vitaxin has also been used in a Phase I clinical trial of patients with late stage cancer. Vitaxin was nontoxic and potentially active in patients with progressive tumors with stage IV disease. ²⁸  

Like α_vβ₃ integrins, α_vβ₅ integrins are implicated in angiogenesis. ^{19 29} α_vβ₅ integrins have been identified as the principal α_v integrin associated with endothelial cells in the corneal alkaline burn model of inflammation-mediated angiogenesis. ²⁹ In the retina, α_vβ₅ integrins have been localized to blood vessels and nonvascular areas in proliferative diabetic retinopathy. ¹² To date, few studies have examined the effects of combined α_vβ₃ and α_vβ₅ antagonism on retinal angiogenesis. In rodents, the retinal vasculature develops after birth and is complete by 2 weeks of age. ³⁰ In a study of developmental retinal vasculogenesis in mice, cyclic peptide antagonists to α_vβ₃ and α_vβ₅ administered from P0 to P4, reduced normal retinal vascularization. In the present study, SB-267268 did not alter developmental retinal vasculogenesis in sham animals when administered between P11 to P17. The reasons for the discrepancies between the 2 studies are not clear but may be because of the timing of delivery of the combined α_vβ₃ and α_vβ₅ integrin inhibitors. In rodents, the majority of retinal vasculogenesis occurs within the first week to 10 days of the postnatal period, ³⁰ and, therefore, it is during this period that new blood vessels may be most responsive to combined α_vβ₃ and α_vβ₅ integrin antagonism. Of interest is that both the cyclic peptide antagonist to α_vβ₃ and α_vβ₅ ²⁰ and SB-267268 reduced to a similar extent pathologic angiogenesis in murine ROP. In these studies, the integrin antagonists were administered by either subcutaneous or i.p. injection. ^{20 21} SB-267268 has a high affinity for α_vβ₃ and α_vβ₅ integrins and good pharmokinetics, which include moderate plasma clearance and high oral bioavailability. ¹³ These features suggest that SB-267268 may have more potential as a treatment for ischemic retinopathies than existing integrin antagonists. 

Both VEGF and basic fibroblast growth factor (bFGF) are involved in developmental and pathologic angiogenesis. ^{31 32 33} In mice with ROP, temporal changes in these growth factors occurs after the removal of pups from a hyperoxic environment to room air. ²¹ Retinal VEGF expression increases immediately after the return of mice to room air, whereas bFGF is predominately expressed later when retinal angiogenesis is maximal. ²¹ In earlier studies, 2 growth-factor-dependent angiogenic pathways for α_vβ₃ and α_vβ₅ were described. ¹⁹ α_vβ₃ integrins have been reported to mediate bFGF or TNF-α induced angiogenesis, whereas α_vβ₅ mediated angiogenesis involved VEGF and TGF-α. Subsequent studies have reported that in ROP, administration of inhibitors of α_v-type integrins to mice soon after the return of pups to room air was associated with a 57% reduction in new vessels, whereas a late intervention had no effect. ²¹ These findings were interpreted as indicating that α_v-integrin inhibition is largely VEGF dependent. The results of the present study are consistent with these findings. As previously described in studies by ourselves ^{11 18} and other investigators, ^{34 35} ROP was associated with an increase in VEGF and VEGFR-2 expression in the inner retina. In the present study, SB-267268 reduced VEGF and VEGFR-2 expression in ROP. In support of these findings, there is evidence that platelet-derived growth factor receptor α and VEGFR-2 associate with the extracellular domain of the β₃ integrin subunit. ³⁶ In addition, Soldi et al. ³⁷ have shown that anti-α_vβ₃ antibodies can inhibit VEGF-induced VEGFR-2 phosphorylation and cell migration when endothelial cells are bound to the α_vβ₃ ligand vitronectin. 

In summary, the findings of the present study indicate that SB-267268 is effective in reducing pathologic angiogenesis in a model of ischemic retinopathy and that its beneficial effects most likely involve inhibition of retinal VEGF and VEGFR-2. SB-267268 may be a useful treatment for ischemic retinopathies, for example, proliferative diabetic retinopathy because of its high oral bioavailability ¹³ compared with other integrin antagonists that are administered by either a systemic or intraocular route. 

 Supported by the Juvenile Diabetes Research Foundation.

 Submitted for publication October 6, 2005; revised November 27, 2005; accepted February 13, 2006.

 Disclosure: J.L. Wilkinson-Berka, None; D. Jones, None; G. Taylor, None; K. Jaworski, None; D.J. Kelly, None; S.B. Ludbrook, GlaxoSmithKline (F); R.N. Willette, GlaxoSmithKline (E, F); S. Kumar, GlaxoSmithKline (F); R.E. Gilbert, 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.

 

The chemical structure of the α_vβ₃ and α_vβ₅ integrin antagonist, SB-262728.

Table 1.

 

SB-267268 Binding to Human, Monkey and Mouse Integrins

Receptor Binding Studies	Activity (Ki, nM)*
Human αvβ3 binding	0.9
Monkey αvβ3 binding	0.5
Human αvβ5 binding	0.7
Human αIIbβ3 binding	1000
Human α5β1 binding	2000
Human α3β1 binding	>1000

	Human αvβ3	Mouse αvβ3	Human αvβ6	Mouse αvβ6	Rat αvβ6
Mean IC50 (nM), †	0.68	0.29	41.78	19.76	18.93
SD	0.15	0.10	6.91	2.51	0.99

*  Data represent displacement of [3H]-SKF107260 binding to integrin receptors by SB-267268.

†  Data represent the IC₅₀ of SB-267268 (nM) required for inhibition of fibronectin binding to various integrins.

Figure 2.

 

Three micron paraffin sections of retina from mice with ROP and treated with SB-267268. Quantitation of BVPs in the inner retina. Magnification, ×125. Scale bar, 50 ìm. Sections stained with hematoxylin and eosin. (A) Sham+vehicle. (B) Sham+SB-267268. (C) ROP+vehicle. (D) ROP+SB-267268. Single arrows denote blood vessels in the inner retina. Double arrows denote blood vessels extending from the retina into the vitreous cavity. In ROP+vehicle (B), numerous blood vessels are found in the inner retina and penetrating into the vitreous cavity compared with all other groups. Values are means ± SEM; n = 7 to 9 animals per group. *P < 0.005 compared with sham groups. #P < 0.0001 compared with ROP+SB-267268. †P < 0.0001 compared with sham groups.

Figure 3.

 

Dark field micrographs of retina showing VEGF gene expression in paraffin sections from neonatal mice with ROP and treated with SB-267268. *Retinal pigment epithelium. Magnification, ×125. Scale bar, 50 ìm. (A) Sham+vehicle. (B) Sham+SB-267268. (C) ROP+vehicle. (D) ROP+SB-267268. VEGF mRNA is similar in (A) and (B), with intense expression in the INL and less mRNA in the GCL. VEGF mRNA is increased with ROP (C) and reduced in ROP mice treated with SB-267268 (D). Values are means ± SEM; n = 7 to 9 animals per group. dpm/mm², disintegration per minute of I¹²⁵ per mm² of tissue. *P < 0.005 compared with sham groups. #P < 0.001 compared with ROP+SB-267268. †P < 0.001 compared with untreated ROP.

Figure 4.

 

Dark field micrographs of retina showing VEGFR-2 gene expression in paraffin sections from neonatal mice with ROP and treated with SB-267268. INL, inner nuclear layer; GCL, ganglion cell layer; INL, inner nuclear layer. *Retinal pigment epithelium. Magnification, ×125. Scale bar, 50 ìm. (A) Sham+vehicle. (B) Sham+SB-267268. (C) ROP+vehicle. (D) ROP+SB-267268. VEGFR-2 mRNA is mainly detected in the INL and occurs to a lesser extent in the GCL. In ROP mice, VEGFR-2 mRNA is increased (C) and reduced to sham levels with SB-267268 (D). Values are means ± SEM; n = 7 to 9 animals per group. dpm/mm², disintegration per minute of I¹²⁵ per milimeter² of tissue. *P < 0.0001 compared with all groups.

The authors thank the Juvenile Diabetes Research Foundation International for support and Shing Mei Hwang for help with some experiments.