Suppression and Regression of Choroidal Neovascularization by Systemic Administration of an 5 1 Integrin Antagonist

Integrin 5 1 plays an important role in developmental angiogenesis, but its role in various types of pathologic neovascularization has not been completely defined. In this study, we found strong up-regulation of 5 1 in choroidal neovascularization. Implantation of an osmotic pump delivering 1.5 or 10 g/h ( 1.8 or 12 mg/kg/day) of 3-(2-{1-alkyl-5-[(pyridin-2ylamino)-methyl]-pyrrolidin-3-yloxy}-acetylamino)-2-(alkylamino)-propionic acid (JSM6427), a selective 5 1 antagonist, caused significant suppression of choroidal neovascularization; the area of neovascularization was reduced by 33 to 40%. When an osmotic pump delivering 10 g/h of JSM6427 was implanted 7 days after rupture of Bruch’s membrane, there was terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) staining in vascular cells within the neovascularization and significant regression of the neovascularization over the next week. JSM6427 also induced apoptosis of cultured vascular endothelial cells. Fibronectin stimulates phosphorylation of extracellular signal-regulated kinase (ERK) in 5 1expressing cells that is blocked by JSM6427. These data suggest that 5 1 plays a role in the development and maintenance of choroidal neovascularization and provides a target for therapeutic intervention. Vascular endothelial cells participating in angiogenesis get signals from several sources. Surrounding tissue that senses the need for increased blood (oxygen) delivery is an important source of soluble stimulators. In hypoxic tissue, the transcription factor hypoxia inducible factor-1 (HIF-1) is stabilized, resulting in up-regulation of several genes (Manalo et al., 2005). Vascular endothelial growth factor (VEGF) is an important hypoxia-regulated soluble signal produced by surrounding tissue that promotes neovascularization. Things other than hypoxia, such as cytokines, also up-regulate VEGF, and in many instances, this seems to be through HIF-1 or HIF-2 (Bardos et al., 2004; Haddad and Harb, 2005). Although other soluble signals undoubtedly play a role, it is clear that inhibition of VEGF is a useful strategy for treatment of neovascular diseases (Ribatti, 2005). Another source of signals is the extracellular matrix (ECM). The ECM seems capable of providing both stabilizing signals that suppress neovascularization and stimulatory signals (Sottile, 2004). Integrins are cell surface receptors that mediate signaling from the ECM, and signaling from ECM is modulated by alterations in the integrin population on the cell surface. Therefore, it is important to determine in any given situation which integrins mediate stimulation of neovascularization and which seem to suppress it. Integrins v 3 and v 5 are up-regulated on endothelial cells participating in several types of neovascularization, including tumor vessels and retinal neovascularization (Brooks et al., 1994a,b; Luna et al., 1996). Antagonists of v 3 and v 5 suppress tumor neovascularization and retinal neovascularization (Hammes et al., 1996; Luna et al., 1996). However, the same antagonists of v 3 and v 5 that suppress retinal neovascularization, have no identifiable suppressive effect on choroidal neovascularization in mice (P. A. Campochiaro, unpublished data). Gene knockout studies have shown that deletion of v, 3, and/or 5 fails to block developmental angiogenesis and in some cases may enhance anThis work was supported by the Foundation Fighting Blindness, the Macula Vision Foundation, and National Eye Institute grant EY12609. P.A.C. is the George S. and Dolores Dore Eccles Professor of Ophthalmology. Article, publication date, and citation information can be found at http://molpharm.aspetjournals.org. doi:10.1124/mol.105.020941. ABBREVIATIONS: HIF-1, hypoxia inducible factor-1; VEGF, vascular endothelial growth factor; ECM, extracellular matrix; FGF2, fibroblast growth factor 2; PBS, phosphate-buffered saline; TBS, Tris-buffered saline; ERK, extracellular signal regulated kinase; ONL, outer nuclear layer; BSA, bovine serum albumin; PI, propidium iodide; GSA, Griffonia simplicifolia lectin; TUNEL, terminal deoxynucleotidyl transferase dUTP nick-end labeling; DAPI, 4,6-diamidino-2-phenylindole; HRP, horseradish peroxidase; BSA, bovine serum albumin; JSM6427, 3-(2-{1-alkyl-5-[(pyridin-2ylamino)-methyl]-pyrrolidin-3-yloxy}-acetylamino)-2-(alkylamino)-propionic acid. 0026-895X/06/6906-1820–1828$20.00 MOLECULAR PHARMACOLOGY Vol. 69, No. 6 Copyright © 2006 The American Society for Pharmacology and Experimental Therapeutics 20941/3115777 Mol Pharmacol 69:1820–1828, 2006 Printed in U.S.A. 1820 at A PE T Jornals on N ovem er 7, 2017 m oharm .aspeurnals.org D ow nladed from giogenesis (Hynes, 2002). Therefore, it seems that the role of integrins in neovascularization in one situation does not guarantee participation in a different vascular bed and different disease process. Integrin 5 1 is also up-regulated on activated endothelial cells and tumor blood vessels (Collo and Pepper, 1999; Kim et al., 2000; Magnussen et al., 2005; Parsons-Wingerter et al., 2005). Antagonists of 5 1 suppress angiogenesis on chick chorioallantoic membrane induced by FGF2, tumor necrosis factoror interleukin-8 and in murine tumor models (Kim et al., 2000; Stoeltzing et al., 2003; Magnussen et al., 2005). Integrin 5 1 also plays a critical role in developmental angiogenesis (Yang et al., 1993; Hynes, 2002). During vascularization of the central nervous system, angiogenic sprouts express high levels of 5 1, and it is markedly reduced as the vessels mature (Milner and Campbell, 2002). Thus, 5 1 seems to promote angiogenesis in multiple settings, but, as noted above, one cannot assume that it is proangiogenic in all tissues and pathologies. In this study, we have explored the role of 5 1 in choroidal neovascularization, the most common cause of severe vision loss in elderly Americans (Klein et al., 1993). Materials and Methods Mouse Model of Choroidal Neovascularization. Mice were treated in accordance with the Association for Research in Vision and Ophthalmology guidelines for the use of animals in research. Choroidal neovascularization was induced by laser photocoagulation-induced rupture of Bruch’s membrane as described previously (Tobe et al., 1998). In brief, 5to 6-week-old female C57BL/6J mice were anesthetized with ketamine hydrochloride (100 mg/kg body weight), and pupils were dilated with 1% tropicamide. Three burns of 532 nm diode laser photocoagulation (75m spot size, 0.1-s duration, 120 mW) were delivered to each retina with the slit lamp delivery system of an OcuLight GL diode laser (Iridex, Mountain View, CA) using a handheld cover slip as a contact lens to view the retina. Burns were performed in the 9, 12, and 3 o’clock positions of the posterior pole of the retina. Production of a bubble at the time of laser, which indicates rupture of Bruch’s membrane, is an important factor in obtaining choroidal neovascularization; therefore, only burns in which a bubble was produced were included in the study. Immunohistochemistry and Histochemistry. Two weeks after rupture of Bruch’s membrane, mice were euthanized and eyes were removed and fixed for 30 min in 0.1 M phosphate buffer, pH 7.6, containing 4% paraformaldehyde and 5% sucrose. After 30 min, corneas and lenses were removed and then fixation was continued for another hour. After washing overnight with 0.1 M phosphate buffer containing 20% sucrose, the eyecups were frozen in optimum cutting temperature embedding compound (Miles Diagnostics, Elkhart, IN). Ocular frozen sections (10 m) were dried with cold air for 20 min, fixed in freshly prepared 4% paraformaldehyde in 0.05 M phosphate-buffered saline (PBS) at room temperature for 15 min, and rinsed with 0.05 M Tris-buffered saline (TBS) for 10 min. Endogenous peroxidases were inhibited by a 15-min incubation with 0.75% H2O2 in methanol. Sections were washed three times in 0.05 M TBS and nonspecific binding sites were blocked by incubating in 10% normal goat serum in 50 mM TBS for 30 min at room temperature. Sections were incubated with 1:200 rabbit polyclonal antibody directed against 5 integrin subunit (Chemicon, Temecula, CA) in 1% bovine serum albumin (BSA) in TBS at 4°C overnight. For controls, nonimmune IgG (Vector Laboratories, Burlingame, CA) was substituted for primary antibody. After two rinses with TBS, sections were incubated for 30 min at room temperature with secondary antibody, Fig. 1. Immunohistochemical staining for integrin 5 subunit in mice with choroidal neovascularization. Adult C57BL/6J mice had laser-induced rupture of Bruch’s membrane in each eye. Two weeks after laser treatment, mice were euthanized, eyes were removed, and frozen sections were cut through rupture sites. Some sections were stained with Griffonia simplicifolia, which selectively stains vascular cells and allows visualization of choroidal neovascularization (A). Adjacent sections were immunohistochemically stained for integrin 5 subunit (B). Superimposed image from a DAPI-stained section shows the retinal cells in the ONL, inner nuclear layer (INL), and ganglion cell layer (GCL), confirming that the cells expressing 5 integrin are in the subretinal space (C). Retina and choroid remote from Bruch’s membrane rupture sites showed staining of retinal vessels with GSA, but no choroidal neovascularization (D) and no staining for integrin 5 subunit (E and F). Scale bar, 100 m. 5 1 Integrin and Ocular Neovascularization 1821 at A PE T Jornals on N ovem er 7, 2017 m oharm .aspeurnals.org D ow nladed from 1:1000 fluorescein isothiocyanate-conjugated goat anti-rabbit IgG F(ab )2 (Jackson ImmunoResearch Laboratories Inc., West Grove, PA). Sections were counterstained with DAPI (Kirkegaard and Perry Laboratories, Gaithersburg, MD) and mounted with Aquamount (British Drug House, Poole, Dorset, UK). Serial sections were stained with biotinylated Griffonia simplicifolia lectin B4 (GSA; Vector Laboratories) to identify choroidal neovascularization as described previously (Ozaki et al., 1998). In brief, slides were incubated in methanol/H2O2 for 10 min at 4°C, washed with 0.05 M TBS, pH 7.6, and incubated for 30 min in 10% normal porcine serum. Slides were incubated for 2 h at room temperature with biotinylated GSA. After washing, the slides were incubated in streptavidin-phosphatase and developed with HistoMark Red (Kirkegaard and Perry Laboratories) according to the manufacturer’s instructions. Sections were dehydrated and mounted with Cytoseal. Stained sections were examined with a Nikon microscope and captured as digital files with a Nikon digital still camera (DXM1200; Nikon Instruments Inc., New York, NY). Systemic Administration of JSM6427 in Mice with Rupture of Bruch’s Membrane. Two different concentrations of JSM6427, 3 mg/ml in PBS and 20 mg/ml in 100 mM glycine/NaOH, pH 4.0, or the corresponding vehicle alone were loaded into osmotic mini-pumps (model 2002; Alza Corp., Palo Alto, CA) with internal volume of 200 l and mean pumping rate of 0.5 l/h. Pumps were implanted beneath the skin of the back and the following day the mice had laser-induced rupture of Bruch’s membrane at 3 locations in each eye. After 14 days, the mice were perfused with 1 ml of PBS containing 50 mg/ml fluorescein-labeled dextran (2 10 average molecular Fig. 2. Systemic delivery of JSM6427 by osmotic minipump suppresses the development of choroidal neovascularization at Bruch’s membrane rupture sites. A, the osmotic minipumps were approximately 30 mm long. B, the implanted minipumps were visible as humps (arrows) beneath the skin of the back. Mice implanted with pumps containing 3 mg/ml JSM6427 received approximately 1.5 g/h JSM6427 and choroidal flat mounts after perfusion with fluorescein-labeled dextran showed small areas of choroidal neovascularization at rupture sites (C, arrows) compared with areas of choroidal neovascularization seen in mice implanted with pumps containing vehicle (D, arrows). Image analysis confirmed that there was significantly less choroidal neovascularization in mice that received 1.5 g/h JSM6427 compared with those that received vehicle (E). Mice implanted with pumps containing 20 mg/ml JSM6427 received approximately 10 g/h JSM6427 and also seemed to have smaller areas of choroidal neovascularization (F, arrows) than mice that received vehicle (G, arrows). Image analysis showed a statistically significant difference from vehicle (H) in the same range as that seen after infusion of 1.5 g/h of JSM6427. , p 0.0005; †, p 6 10 8 by Mann-Whitney U test. Scale bar, 100 m. 1822 Umeda et al. at A PE T Jornals on N ovem er 7, 2017 m oharm .aspeurnals.org D ow nladed from weight; Sigma-Aldrich, St. Louis, MO) and choroidal flat mounts were prepared as described previously. In brief, eyes were removed and fixed for 1 h in 10% phosphate-buffered formalin. The cornea, lens, and retina were removed, and four radial cuts were made in the eyecup, allowing it to be flat-mounted in aqueous mounting medium. Flat mounts were examined by fluorescence microscopy, and images were digitized using a three-color charge-coupled device video camera and a frame grabber. Image analysis software (Image-Pro Plus; Media Cybernetics, Silver Spring, MD) was used to measure the total area of choroidal neovascularization at each rupture site. To assess the effect of JSM6427 on established choroidal neovascularization, mice had rupture of Bruch’s membrane at three locations in each eye. After 7 days, 9 mice were perfused with fluoresceinlabeled dextran, and the baseline area of neovascularization at each rupture site was measured. The remaining mice had implantation of osmotic minipumps containing 20 mg/ml JSM6427 in 100 mM sodium phosphate buffer, 50 mM NaCl, pH 7.4, or vehicle alone. At 7 days after implantation, some mice were euthanized for terminal terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL). The rest of the mice were perfused with fluorescein-labeled dextran 7 days after implantation, and the area of choroidal neovascularization at rupture sites was measured. Intravitreous Injections of JSM6427. Mice had rupture of Bruch’s membrane at three locations in each eye and were given an intravitreous injection of 1 l of vehicle (PBS or 100 mM phosphate buffer, and 50 mM NaCl, pH 7.4) containing of 3 or 20 g of JSM6427 in one eye and 1 l of vehicle alone in the other eye on days 0 and 7. For an additional control, some mice received no injection. On day 14, the mice were perfused with fluorescein-labeled dextran, and the area of choroidal neovascularization at Bruch’s membrane rupture sites was measured. Identification of Apoptotic Cells in Vivo by TUNEL. Eyes were fixed in 4% paraformaldehyde in 0.1 M phosphate buffer and frozen in OCT. Serial 10m sections were cut through each rupture Fig. 3. Systemic delivery of JSM6427 by osmotic minipump causes regression of choroidal neovascularization. Adult C57BL/6J mice had laser-induced rupture of Bruch’s membrane at three locations in each eye. Seven days after laser treatment, nine mice were perfused with fluorescein-labeled dextran, and the baseline amount of choroidal neovascularization at 7 days (A, arrows) was measured by image analysis. The remainder of the mice had implantation of an osmotic minipump containing 20 mg/ml JSM6427 (B) or vehicle (C), and these mice were perfused with fluorescein-labeled dextran on day 14. In mice that received JSM6427, the area of choroidal neovascularization lesions (B, arrows) appeared smaller than those in mice treated with vehicle (C, arrows) and the baseline amount seen at day 7 (A, arrows). TUNEL (red) of sections also stained with GSA, which stains vascular cells (green), and DAPI, which stains cell nuclei (blue), showed apoptotic cells in the ONL in day 7 baseline eyes (D), a consequence of the laser treatment 7 days before. Sections from day 14 eyes that had been treated with JSM6427 also showed apoptotic cells in the ONL, but in addition there were yellow cells within choroidal neovascularization lesions (E and G, arrows) as a result of colocalization of TUNEL and GSA, indicating apoptosis of cells with the choroidal neovascularization. Sections from eye treated with vehicle showed apoptotic cells in the ONL, but not within the choroidal neovascularization (F). Measurement of the area of choroidal neovascularization by image analysis confirmed that there was a significant reduction in mice treated with JSM6427 compared with the amount seen at baseline or in mice treated with vehicle (H). , p 0.0039 by linear mixed model for comparison with baseline; †, p 0.0283 by linear mixed model for comparison with vehicle. P values were adjusted for multiple comparisons using Dunnett’s method. Scale bars: A–F, 100 m; G, 50 m. 5 1 Integrin and Ocular Neovascularization 1823 at A PE T Jornals on N ovem er 7, 2017 m oharm .aspeurnals.org D ow nladed from site. Sections were fixed with 1% paraformaldehyde for 10 min at room temperature, and TUNEL was done with the ApopTag Red kit (Chemicon International, Temecula, CA) according to the manufacturer’s instructions. The sections were also histochemically stained with GSA as described above. Slides were also stained with DAPI and mounted in aqueous mounting medium. Statistical Analysis. Data were analyzed using a linear mixed model accounting for possible correlations in measurements from the same mice. Dunnett’s adjustment was made for multiple comparisons. Solid Phase Binding Assay. The inhibiting activity and integrin selectivity of the integrin inhibitor was determined in a solid phase binding assay using soluble integrins and coated extracellular matrix protein. Binding of integrins was then detected by specific antibodies in an enzyme-linked immunosorbent assay. Fibronectin and vitronectin were purchased from Sigma (St Louis, MO). The integrin 5 1 extracellular domain Fc-fusion protein was a generous gift from M. Humphries (University of Manchester), and V 3 was purchased from Chemicon (Chemicon Europe, Germany). The integrin antibodies were purchased from Pharmingen, BD Bioscience Europe ( V 3), and Sigma (anti-human-Fc-HRP antibody conjugate and antimouse-HRP conjugate). The detection of HRP was performed using HRP substrate solution 3.3.5.5 -tetramethylethylenediamine (Seramun Diagnostica GmbH, Dolgenbrodt, Germany) and 1 M H2SO4 for stopping the reaction. The developed color was measured at 450 nm. 5 1:Nunc-Immuno maxisorp plates (Nalge Nunc International, Rochester, NY) were coated over night at 4°C with fibronectin (0.25 g/ml) in 15 mM Na2CO3, 35 mM NaHCO3, pH 9.6. All subsequent washing and binding were performed in 25 mM Tris, pH 7.6, 150 mM NaCl, 1 mM MnCl2, and 1 mg/ml BSA. The plates were blocked with 3% BSA in PBS 0.1% Tween 20 for 1 h at room temperature. Soluble integrin 5 1 (0.5 g/ml) and a serial dilution of integrin inhibitor were incubated in the coated wells for 1 h at room temperature. The detection antibody (anti-human-Fc-HRP antibody conjugate) was then applied for 1 h at room temperature, and the binding was visualized as described above. For the V 3 assay, plates were coated with vitronectin (1 g/ml) and blocked as described for 5 1. Soluble V 3 (1 g/ml) was incubated with a serial dilution of integrin inhibitor for 1 h at room temperature. Primary (antiV 3) and secondary antibody (anti-mouse-HRP conjugate) were applied for 1 h at RT, and the binding was visualized as described above. All IC50 measurements were performed at least 30 times. Identification of Apoptotic Cells in Vitro by TUNEL and Fluorescent Activated Cell Sorting. Human umbilical vein endothelial cells were maintained in endothelial cell growth medium (PromoCell, Heidelberg, Germany) and grown to 80% confluence, trypsinized, and preincubated for 20 min in serum-free medium alone or medium containing 100 nM, 500 nM, 1 M, 10 M, 25 M, or 50 M JSM6427 or 10 M camptothecin. Cells (4 10) were plated in six-well plates coated with 10 g/ml fibronectin (Chemicon Europe, Hampshire, UK) and grown for 48 h. Adherent and nonadherent cells were collected, pooled, and washed. Cells were fixed in 1% paraformaldehyde for 30 min on ice, washed, and carefully resuspended in 70% ice-cold ethanol. TUNEL was performed using the APO-Direct kit (BD Pharmingen, San Diego, CA). In brief, cells were incubated with TdT enzyme and fluorescein isothiocyanate-dUTP in reaction buffer for 60 min (positive and negative cells of the kit) or 180 min (probes) at 37°C. Cells were rinsed and incubated with PI-staining solution for 30 min to determine total cell numbers. Fluorescence-activated cell sorting analysis was performed using a FACSCalibur (BD Bioscience, San Diego, CA) and the CellQuest software. Western Blotting. ARPE19 cells were incubated in serum-free Dulbecco’s modified Eagle’s medium/Ham’s F12 (Biochrom AG, Berlin, Germany) culture medium for 16 h. Plates were coated with 10 g/ml fibronectin or 0.002% poly-L-lysine (Sigma-Aldrich, Munich, Germany) and blocked with 1% BSA. Cells were trypsinized and held in suspension for 30 min. Subsequently, 1 10 cells were incubated with compounds for 10 min. Cells were allowed to adhere for 15 min on fibronectin in serum-free medium. Adherent cells and cells from supernatants were lysed in 150 l of lysis buffer (1% Nonidet P-40, 150 mM NaCl, 50 mM Tris-HCl, pH 7.5, protease inhibitor cocktail (Sigma-Aldrich), 25 mM glycerophosphate, 50 mM NaF, 10 mM Na4P2O7, and 1 mM orthovanadate) on ice. Lysates were centrifuged, and equal amounts of protein were separated on a 4 to 15% Tris/HCl gel by SDS-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes by semi-dry Western blotting. Membranes were blocked in 3% BSA. Phosphorylated extracellular signal regulated kinase (ERK) and total ERK were detected with a phospho-specific ERK1/2 (pThr202/pTyr204) antibody (Cell Signaling, Danvers, MA) and a p44/42 mitogen-activated protein kinase kinase antibody (Cell Signaling). Bands were visualized with peroxidase-conjugated anti-rabbit and anti-mouse antibodies (SigmaAldrich) and BM chemiluminescence Western blotting peroxidase substrate (Roche Applied Science, Mannheim, Germany). The luminescence signal was measured and quantified using a Lumi imager (Roche) and corresponding software.