BIBF1120 (Vargatef) Inhibits Preretinal Neovascularization and Enhances Normal Vascularization in a Model of Vasoproliferative Retinopathy

Abstract

Purpose: This study evaluated the effects of BIBF1120, a novel triple angiokinase inhibitor against pathological retinal neovascularization.

Methods: BIBF1120 effect on development of the normal retinal vasculature was evaluated in Sprague-Dawley rat pups. Two models of ischemic oxygen-induced retinopathy (OIR) and the aortic ring assay were used to assess the antiangiogenic effects of BIBF1120. In the vaso-obliteration model (VO), rat pups were exposed to 80% O₂ from postnatal day (P) 5 to P10. In the preretinal neovascularization (NV) model, rat pups were exposed to cycling O₂ (50% and 10%) from P1 to P14, followed by room air until P18. Animals were intravitreally or orally treated with BIBF1120. Retinal vasculature, VO, and NV were evaluated in retinal flat mounts. Retinal expression of VEGF, Delta-like ligand 4 (Dll4), Netrin-1, Ephrin-B2, and EphB4 was analyzed by quantitative PCR and Western blot analysis.

Results: BIBF1120 interfered with normal retinal vascular development and microvessel branching in the aortic assay. However, in VO model BIBF1120 did not accrue VO. On the contrary, in the NV model BIBF1120 accelerated normal retinal vascularization and robustly diminished preretinal neovascularization compared to vehicle (by ∼80%). The expression levels of VEGF negative regulator Dll4 and repulsive cues EphrinB2 and EphB4 mRNA in the retina of vehicle-treated OIR animals were markedly increased compared to normoxia, but were normalized by BIBF1120.

Conclusions: Data reveal efficacy of BIBF1120 on preretinal neovascularization and, of greater interest, on acceleration of normal vascularization, consistent with interference of major repulsive cues expressed in the retina during OIR. Accordingly, BIBF1120 appears to exhibit preferable properties compared to anti-VEGF therapies for the treatment of ischemic retinopathies.

Retinopathy of prematurity (ROP) still remains a serious cause of visual impairment and blindness in premature neonates all over the world.¹ This ocular disease is characterized by an initial phase of retinal vascular constriction, followed by microvascular degeneration that results in ischemia and subsequent predisposition to an abnormal intravitreal neovascularization (NV).^2–4 A large number of proangiogenic factors, including erythropoietin,^5,6 insulin-like growth factor 1,⁷ platelet-derived growth factor (PDGF),⁸ FGF,⁹ and VEGF,^10,11 play a critical role in the development of pathological ocular neovascularization. Until now, VEGF has been recognized as one of the most important factors involved in the pathophysiology of this disease and has become an integral component of treatments for such retinopathies. Anti-VEGF therapy has been approved for the treatment of choroidal neovascularization in AMD^12,13 and for retinal neovascularization in diabetic retinopathy patients.¹⁴ However, to date, much uncertainty exists about its use in ROP¹⁵ because of associated side effects.^16–19 Besides, in some cases VEGF blockade alone cannot completely eliminate pathological angiogenesis, suggesting that other factors could be involved.^20,21 The influence of neuron-derived signaling molecules on endothelial cell function in the retina has been recently highlighted.^22–24 Classic neuronal guidance cues and their receptors, particularly class III semaphorins,^25,26 slits,²⁷ ephrins,²⁸ and netrins,²⁹ could act as repulsive molecules in the hypoxic area of the retina during retinopathy, thus hindering normal vascularization and contributing to the formation of abnormal vascular tufts.²² 

Interestingly, many of the proangiogenic factors involved in development of ocular pathological neovascularization exert their actions through their cell-surface receptors called receptor tyrosine kinase (RTK). Receptor tyrosine kinases are a subclass of cell-surface growth-factor receptors with an intrinsic, ligand-controlled tyrosine-kinase activity.³⁰ Receptor tyrosine kinase activation is associated with several pathways linked to the expression of guidance cues. For instance, FGF receptor (FGFR) signaling induced Netrin-1 expression on lymphatic endothelial cells, contributing to the tumor growth.³¹ A guidance cue, EprhinB2, highly expressed in the neovascular tufts during OIR,³² is controlled by Delta-like ligand 4 (Dll4)³³ and VEGF signaling.^34,35 Conversely, EphrinB2 can control VEGF receptor (VEGFR) and PDGF receptor (PDGFR) internalization and signaling on endothelial and mural cells, respectively.^36,37 Moreover, FGFR activation has been shown to upregulate the expression of slit and Sema3A,³⁸ two neuronal guidance cues strongly implicated in pathological angiogenesis in the retina.^27,34 Therefore, guidance cues mediated by RTKs' activation are associated with regulation of diverse functions in normal cells but also have a crucial role in the development of pathological conditions such as cancer³⁰ and vasoproliferative retinopathies.^8,39 Targeting some of these RTKs would be a logical effective therapeutic approach to limit preretinal neovascularization.^40,41 So far, several multityrosine kinase inhibitors have been validated as effective strategies for the inhibition of new blood vessel formation in ocular pathologies.^42–44 In some cases the combined inhibition of two different tyrosine kinase receptors (such as PDGFR and VEGFR) has exhibited a more effective and potent suppression of aberrant vessel growth than blocking VEGF alone in multiple models of ocular neovascularization.²⁰ 

Recently, a novel, potent triple angiokinase inhibitor named BIBF1120 (Vargatef) was developed. BIBF1120 is a substituted oxindole derivative drug, which competitively binds to the adenosine-5-triphosphate (ATP) binding site of RTKs and inhibits downstream intracellular signaling.⁴⁵ BIBF1120 has been shown to be a potent inhibitor of VEGFR, FGFR, and PDGFR in enzymatic and cellular assays⁴⁵ and has been demonstrated to be effective in several human cancer trials.^46–48 BIBF1120 has displayed potent antiangiogenic effects in vitro by inhibiting endothelial, pericyte, and smooth muscle cell proliferation in culture and antitumor activity in several tumor xenograft models in vivo by reducing tumor microvascular density and number of perivascular cells.^45,49 A recent study showed that BIBF1120 released in the posterior segment of the eye by using a novel nanoparticle drug delivery system sensitive to the UV light inhibited laser-induced choroidal neovascularization in a rodent model.⁵⁰ The aim of the present study is to determine whether the treatment of BIBF1120 can prevent abnormal preretinal neovascularization in a model of ROP in rats, notably oxygen-induced retinopathy (OIR). 

Materials and Methods

Animal Care

All animal experimental procedures were performed with strict adherence to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and approved by the Animal Care Committee of the Hôpital Maisonneuve-Rosemont in accordance with guidelines established by the Canadian Council on Animal Care. 

BIBF1120 (Vargatef) Preparation

The small-molecule inhibitor BIBF1120 was generously provided by Frank Hilberg from Boehringer Ingelheim RCV, Vienna, Austria. To obtain adequate working concentrations, a 0.01 mg/mL stock solution was perfectly diluted in sterile-PBS for intravitreal injections. In the case of oral gavage, BIBF1120 was dissolved in a 3:7 solution of dimethyl sulfoxide (DMSO)/PBS as previously described.⁵¹ 

Oxygen-Induced Retinopathy Model in Rats (50/10 OIR Model)

Within 4 hours after birth, litters of Sprague-Dawley albino rats (Charles River, St. Constant, QC, Canada) were placed with their mothers in an oxygen-regulated environment (OxyCycler A820CV; BioSpherix, Ltd., Redfield, NY, USA) adjusted to alternate between 50% and 10% oxygen every 24 hours for 14 days. At postnatal (P) day 14, pups were transferred to room air (21% O₂) for 4 days.^52,53 Rat pups were divided in three groups and treated as follows: group 1 pups twice received intravitreal injections (1 μL) of BIBF1120 (50 nM, 100 nM, or 200 nM) or vehicle (PBS 1× solution) on P5 and P12. Group 2 pups twice received intravitreal injections (1 μL) of BIBF1120 (50 nM, 100 nM, or 200 nM) or vehicle (PBS 1× solution) on P14 and P16. Group 3 pups twice received BIBF1120 (50 mg/kg)⁴⁵ or corresponding vehicle (3:7 solution of DMSO/PBS) by oral gavage at P14 and P16; two doses of BIBF1120 were chosen based on pilot data with partially effective single dose in line with rapid drug disposition.⁴⁵ In all cases rat pups were anesthetized with isoflurane (2%) and killed by decapitation on P18. Retinas were dissected, and the neovascularization area in retinal flat mounts stained with lectins was analyzed using the SWIFT-NV method.⁵⁴ Control animals were maintained in room air (21% O₂) throughout the 18 days. All other conditions (e.g., light exposure, temperature, feeding, etc.) were similar for both treatment groups. 

Vaso-Obliteration Model

Ischemic retinal vaso-obliteration (VO) was induced in Sprague-Dawley rat pups exposed to hyperoxia (80% O₂) in chambers controlled by a computer-assisted Oxycycler (BioSpherix, Ltd.) from P5 to P10.^24,55 On P5 the pups (n = 6 per experimental condition) were anesthetized and received a single intravitreal injection (1 μL) of BIBF1120 (100 nM or 200 nM) or vehicle (PBS). On P10, the animals were killed and retinas dissected. Vaso-obliteration and peripheral avascular area were evaluated on retinal flat mounts stained with lectins by using the Image J software (http://imagej.nih.gov/ij/; provided in the public domain by the National Institutes of Health, Bethesda, MD, USA). Vaso-obliterated areas were assessed as the retinal area devoid of vasculature over the total retina. 

Developmental Retinal Angiogenesis

Sprague-Dawley albino rat pups (Charles River) were maintained in normoxic conditions (room air 21% O₂) from P1 to P7. Different groups of animals (n = 8 per experimental condition) were treated with a double intravitreal injection (1 μL) of BIBF1120 (50 nM, 100 nM, and 200 nM) or vehicle (PBS) as control on P2 and P4. On P7, the animals were killed, and the eyes were enucleated and retinas processed for immunostaining. Peripheral retinal avascularity was evaluated on retinal flat mounts stained with lectins by using the Image J software. Retinal vessel density was calculated by using the software AngioTool,⁵⁶ which allows a reproducible quantification of vascular networks in retinal flat mounts stained with lectins. 

Retinal Flat Mounts

In all cases the eyes were enucleated and fixed in 4% paraformaldehyde for 1 hour at room temperature and then stored in PBS at 4°C until used. The cornea and lens were removed, and the retina was gently separated from the underlying choroid and sclera under a dissecting microscope. The retinas were incubated overnight at 4°C in 1% Triton X-100, 1 mM CaCl₂/PBS with the tetramethylrhodamine isothiocyanate–conjugated lectin endothelial cell marker Bandeiraea simplicifolia (1:100; Sigma-Aldrich Corp., St. Louis, MO, USA). Retinas were washed in PBS and mounted on microscope slides (Bio Nuclear Diagnostics, Inc., Toronto, ON, Canada) under coverslips with mounting media (Fluoro-Gel; Electron Microscopy Sciences, Hatfield, PA, USA). Retinas were photographed at 10× under an epifluorescence microscope (Zeiss AxioObserver; Carl Zeiss Canada, Toronto, ON, Canada), and the images were merged into a single file using the MosiaX option in the AxioVision 4.6.5 software (Zeiss). 

Intravitreal Injections

Animals were anesthetized with isoflurane (2%), and in the case of pups at P2, P4, and P5, intravitreally injected using a 10-μL Hamilton syringe (Hamilton Co., Reno, NV, USA) attached to a glass capillary of approximately 60 gauge. In the case of rat pups at P12, P14, and P16, the ocular globe was first penetrated within approximately 1 mm posterior to the ora serrata with a 30-gauge needle. The needle was immediately removed and a 10-μL Hamilton syringe attached to a glass capillary of approximately 60 gauge was inserted through the existing hole. Contact with the posterior surface of the lens was avoided by maintaining a steep angle, which delivered the injection bolus near the trunk of the hyaloid artery, just above the posterior pole of the retina. 

Aortic Microvascular Sprouting Assay

The aortic ring explants were performed as previously reported.⁵⁷ Briefly, thoracic aortas were isolated from rat pups, sectioned into 1-mm rings, and placed into growth-factor–reduced Matrigel (BD Biosciences, San Jose, CA, USA) in 24-well plates. Aortic rings were cultured at 37°C for 5 days in supplemented endothelial basal medium (Lonza, Walkersville, MD, USA). At day 3, the culture media was changed and different concentrations of BIBF1120 (50 nM, 100 nM, and 200 nM) were added. Photomicrographs of individual explants were taken at day 5 using an inverted phase-contrast microscopy (AxioObserver; Zeiss), and microvascular sprouting was assessed using Image J. 

Real-Time Quantitative PCR Analysis

Retinas or cell lysates were rapidly isolated and processed using RiboZol RNA extraction reagent (AMRESCO, Solon, OH, USA). Total cellular RNA was isolated by acidic phenol/chloroform extraction followed by treatment with DNase I (Roche Diagnostics, Mannheim, Germany) to remove any contaminating genomic DNA. Approximately 1 μg of total RNA was reverse-transcribed into cDNA using reverse-transcription supermix (iScript; BioRad, Hercules, CA, USA) as described by manufacturer's instructions. The cDNA was analyzed by quantitative real-time PCR using iTaq Universal SYBR Green Supermix (BioRad) with primers targeting for Dll4: (5′-TCAACTTGCTCCAACAGTGG-3′ and 5′-CCAGTGAAGTTAGGGGGACA-3′); Sema3F: (5′-TCGCGCACAGGATTACATCTT-3′ and 5′-ACCGGGAGTTGTACTGATCTG-3′); Sema3A: (5′-GAGTCCCTTATCCACGACCA-3′ and 5′-AATGCTTTCTCCGCTCTGAA-3′); VEGF: (5′-CAATGATGAAGCCCTGGAGT-3′ and 5′-AATGCTTTCTCCGCTCTGAA-3′); Netrin-1: (5′-CCGTGGTGACCAGAGTTTGT-3′ and 5′-ATCACCAGGCTGCTCTTGTC-3′); EphB4: (5′-GACCTGACTTTCGACCCTGG-3′ and 5′-TCTGGGGATAGCCCATGACA-3′); and EphrinB2: (5′-GTGGCCTTATTCGCAGGGAT-3′ and 5′-CCATTGTTGTTGCCACCTCG-3′), were designed using Primer Bank and NCBI Primer Blast software. Quantitative analysis of gene expression was generated by using a sequence detection system (ABI Prism 7500; Applied Biosystems, Foster City, CA, USA) and calculated relative to 18S universal primer pair (Ambion, Austin, TX, USA) expression using the delta cT method. 

Western Blot

Eyes were enucleated and retinas dissected and placed into commercial radioimmunoprecipitation assay buffer (Cell Signaling Technology; Danvers, MA, USA) and homogenized with tissue homogenizer (Precellys 24; Bertin Technologies, Montigny-le-Bretonneux, France). Samples were centrifuged and 50 μg of pooled retinal lysate from two different animals was loaded on an SDS-PAGE gel and subsequently electroblotted onto either polyvinylidene fluoride or nitrocellulose membrane (BioRad). After blocking, the membranes were blotted with 1:200 rabbit antibody to VEGF (sc-152; Santa Cruz Biotechnology, Santa Cruz, CA, USA), 1:1000 goat antibody to Netrin-1 (AF1109; R&D Systems, Minneapolis, MN, USA), 1:400 mouse antibody to Dll4 (AF1389; R&D Systems), 1:1000 rabbit monoclonal antibody to EphrinB2 (EPR10072; Abcam, Toronto, ON, Canada), 1:1000 rabbit antibody to EphB4 (ABC257; Millipore, Temecula, CA, USA), 1:1000 mouse antibody to β-actin (Santa Cruz Biotechnology), 1:1000 rabbit antibody to total p44/42 MAP kinase (Erk 1/2, [4695; Cell Signaling Technology]), phosphorylated p44/42 MAP kinase (Erk 1/2, [4376; Cell Signaling Technology]), 1:1000 rabbit antibody to total AKT (4691; Cell Signaling Technology) or phosphorylated AKT (4060; Cell Signaling Technology). After washing, membranes were incubated with 1:5000 horseradish peroxidase (HRP)–conjugated anti-mouse or 1:2000 HRP anti-goat or anti-rabbit secondary antibodies (Millipore). Membranes were imaged with LAS-3000 imager. 

Stimulation of Retinal Ganglion Cells

Retinal ganglion cells (RGC-5; kindly provided by Neeraj Agarwal University of North Texas Health Science Center, Fort Worth, TX, USA) were cultured in Dulbecco's modified Eagle's medium (Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin at 37°C and 5% CO₂. The RGC-5 cells (1 × 10⁶ cells per well) were seeded in six-well plates and starved 4 hours prior to exposure to treatment with 50 ng/mL of the recombinant human RTK ligand FGFb (100-18B; PreproTech, Rocky Hill, NJ, USA) in presence or absence of 200 nM of BIBF1120 (Boehringer Ingelheim RCV, Vienna, Austria). After 24 hours, the medium was removed and cells were lysed with RiboZol for quantitative PCR (qPCR) analysis. 

Statistical Analysis

Results are expressed as mean ± SEM. Comparisons between groups were made using one-way ANOVA followed by the post hoc Bonferroni's multiple comparison test. Statistical significance was set at P < 0.05. 

RESULTS

Antiangiogenic Effect of BIBF1120 on Aortic Ring Assay

To assess the antiangiogenic efficacy of BIBF1120, we first performed an ex vivo aortic ring assay, a well-established model of angiogenesis.^57,58 Aortic ring explants cultured in a basement membrane matrix containing growth factors began to produce vessel sprouts after 2 days in culture. At day 3, the aortic rings were treated with different concentrations of BIBF1120 (50 nM, 100 nM, and 200 nM). After 48 hours, BIBF1120 significantly inhibited dose-dependently aortic vessel sprouting, respectively, by approximately 52% (P < 0.05, n = 6), 64% (P < 0.01, n = 6), and 83% (P < 0.001, n = 6) (Fig. 1). 

Figure 1

 

BIBF1120 inhibits vascular sprouting in the aortic ring assay. Aortic explants cultured in Matrigel containing growth factors and treated with vehicle (PBS 1×) displayed increased sprouting, whereas explants exposed to 50 nM, 100 nM, and 200 nM of BIBF1120 exhibited a significant diminished sprouting growth compared to vehicle. The histogram depicts the quantification of aortic explant sprouting (*P < 0.05, **P < 0.01, ***P < 0.001, n = 6 per group).

Effect of BIBF1120 on Normal Retinal Vascular Development

Next we determined whether BIBF1120 could inhibit angiogenesis in vivo; for this purpose rat pups at P2 and P4 were intravitreally injected with vehicle or with different concentrations of BIBF1120 (50 nM, 100 nM, and 200 nM) and killed at P7. The development of the vascular superficial plexus in animals treated with vehicle was covered by approximately 86% (n = 6) of the total retinal surface; accordingly, approximately 14% of the retinal area was avascular. Animals treated with 50 nM, 100 nM, and 200 nM of BIBF110 displayed approximately 77%, 68%, and 69%, respectively, in the growth of retinal vascular surface, yielding a significant increase by approximately 23% (P < 0.05, n = 7), 34% (P < 0.001, n = 7), and 31% (P < 0.001, n = 7) in avascular area (Fig. 2). Interestingly, the vascular density in the retinas treated with BIBF1120 was not significantly changed compared to vehicle (Fig. 2). These results reveal that BIBF1120 interferes with retinal angiogenesis in vivo, but does not compromise the established vascular network during development. 

Figure 2

 

Effects of BIBF1120 on normal retinal vascularization. Representative petals from retinal flat mounts stained with lectin at P7 from rats treated with intravitreal injections at P2 and at P4 with vehicle (PBS) or BIBF1120 (50 nM, 100 nM, and 200 nM). The histograms show the quantification of the peripheral avascular area (dotted lines on white) and the vascular density in whole retinas. Values are mean ± SEM, n = 6 per group. *P < 0.05, ***P < 0.001. ns, not significant.

Effects of BIBF1120 on Retinal Vaso-Obliteration

We evaluated the effects of BIBF1120 on retinal VO in rats.²⁴ For this purpose, we exposed rats to hyperoxia from P5 to P10 to cause decay of the central retinal vasculature exacerbated by suppressed levels of growth factors such as VEGF.^59,60 Animals injected intravitreally at P5 with the highest concentrations of BIBF1120 (100 nM and 200 nM) and evaluated at P10 did not exhibit significant changes (∼25%, P > 0.05, n = 4 for 100 nM, and 23%, P > 0.05, n = 3, for 200 nM) in central VO and peripheral avascular area compared with vehicle-treated rats (∼24%, n = 4; Fig. 3). Hence, under hyperoxic conditions BIBF1120 does not aggravate vascular damage and can be used safely prior to the preretinal NV phase of OIR. 

Figure 3

 

Intravitreal administration of BIBF1120 does not affect VO in animals exposed to hyperoxia (80% O₂) from P5 to P10. Representative retinal flat mounts with central VO (blue lines) and peripheral avascular area (dotted lines on green) of rats exposed to 80% oxygen (from P5 until P10) and treated with vehicle (PBS) or BIBF1120 (100 nM and 200 nM) at P5 and killed at P10. The histograms show the quantification of the central vaso-obliterated areas (white bars) and peripheral avascular area (green bars) in whole retinas. Values are mean ± SEM; n = 4 per group. ns, not significant.

BIBF1120 Reduces Pathological Neovascularization and Promotes Retinal Vascularization

BIBF1120 is a tyrosine kinase inhibitor that suppresses proangiogenic intracellular signaling by targeting the proliferative growth factor receptors in smooth muscle cells (FGFR), pericytes (PDGFR), and the vascular endothelium (VEGFR).⁴⁵ We determined whether BIBF1120 inhibits the pathological retinal neovascularization in the 50/10 OIR rat model by administering BIBF1120 at different time points in the pathogenesis of the retinopathy. Pups in group 1 received intravitreal injections of BIBF1120 twice during the first phase of developmental retinopathy (at P5 and at P12) and showed a marked reduction of approximately 74% (P < 0.001, n = 8), 66% (P < 0.001, n = 8), and 76% (P < 0.001, n = 8) in pathological (preretinal) NV after respective treatment with 50 nM, 100 nM, and 200 nM BIBF1120, in comparison to vehicle treatment (Fig. 4). 

Figure 4

 

Intravitreal administration of BIBF1120 (at P5 and at P12) promotes normal vascularization and reduces pathological neovascularization. (A) Representative photomicrographs from isolectin B4–stained retinal flat mounts from P18 rat pups exposed to 50/10 OIR model and treated with vehicle (PBS) or BIBF1120 (50 nM, 100 nM, and 200 nM) at P5 and P12. (Peripheral) avascular and preretinal neovascular (NV) areas are outlined with green dotted lines and solid white lines, respectively. Pups injected intravitreally with BIBF1120 showed significant decrease in NV and avascular areas compared to vehicle injected animals. (B) The histograms show the quantification of retinal NV (white bars) and avascular areas (green bars) in animals treated with vehicle (PBS) or BIBF1120 (50 nM, 100 nM, and 200 nM). Values are mean ± SEM of n = 8 per group; *P < 0.05, **P < 0.01, ***P < 0.001 compared to vehicle.

Animals in group 2 received BIBF1120 during the second (neovascular) phase of OIR (at P14 and P16) and also exhibited a significant reduction in the pathological NV of approximately 88% (P < 0.001, n = 8), 74% (P < 0.001, n = 8), and 86% (P < 0.001, n = 8), for 50 nM, 100 nM, and 200 nM of BIBF1120, respectively, compared to the vehicle treatment (Fig. 5). In both groups 1 and 2, dose-response was not observed, suggesting that maximum efficacy seemed to have been already reached at 50 nM. 

Figure 5

 

Intravitreal administration of BIBF1120 (at P14 and at P16) reduces pathological neovascularization and promotes normal vascularization. (A) Representative photomicrographs from isolectin B4–stained retinal flat mounts from P18 rat pups exposed to 50/10 OIR model and treated with vehicle (PBS) or BIBF1120 (50 nM, 100 nM, and 200 nM) at P14 and P16. (Peripheral) avascular and preretinal NV areas are outlined with green dotted lines and solid white lines, respectively. Pups injected intravitreally with BIBF1120 showed significant decreases in NV and avascular areas compared to vehicle-injected animals. (B) The histograms show the quantification of retinal NV (white bars) and (peripheral) avascular areas (green bars) in the animals treated with vehicle (PBS) or BIBF1120 (50 nM, 100 nM, and 200 nM). (C) Peripheral avascular area (% of total) at P15 and P18 in rats treated or not with BIBF1120 and subjected to OIR. Samples at P18 were compared to injected vehicle-treated animals. Data show that BIBF1120 accelerates normal retinal vascularization. Values are mean ± SEM of n = 8 per group; *P < 0.05, **P < 0.01, ***P < 0.001 compared to vehicle.

Animals in group 3 received BIBF1120 (50 mg/kg) by oral gavage during the second (neovascular) phase of OIR (at P14 and P16) and also exhibited a significant reduction in the pathological NV of approximately 93% (P < 0.0001, n = 6), compared to vehicle treatment (Supplementary Fig. S1A). 

We also quantified the retinal (peripheral) avascular area in the three groups of animals (Figs. 4, 5, and Supplementary Fig. S1A). Interestingly, the retinal peripheral avascular area in the animals intravitreally treated with BIBF1120 was significantly decreased at all the doses used (Figs. 4, 5), while in the orally treated group there was a tendency (albeit not statistically significant) for the peripheral area to diminish in response to BIBF1120 (Supplementary Fig. S1A), suggesting possible need for higher oral doses. For instance, the avascular area in retinas from animals treated with the vehicle in group 1 was approximately 28% (n = 8), while the retinas treated with 50 nM, 100 nM, and 200 nM, was approximately 8% (P < 0.001, n = 8), 12% (P < 0.01, n = 8), and 13% (P < 0.01, n = 8), respectively (Figs. 4A, 4B). For group 2, the retinal avascular area for the vehicle was approximately 36% (n = 8), while for the retinas treated with 50 nM, 100 nM, and 200 nM of BIBF1120 was approximately 12% (P < 0.001, n = 8), 16% (P < 0.001, n = 8), and 8% (P < 0.001, n = 8), respectively (Figs. 5A, 5B). For group 3, the retinal avascular area for the vehicle was approximately 36% (n = 6), while the retinas treated with oral BIBF1120 was approximately 31% (P = 0.1, n = 6; Supplementary Fig. S1A). BIBF1120 was effective (in abolishing pathological neovascularization) and was well tolerated by rat pups, as previously demonstrated.⁴⁵ Interestingly, animals treated by oral gavage with BIBF1120 at P14 and P16 under normoxia did not display compromised superficial and deep vascular networks that are usually completed around P14 in the rat⁶¹ (Supplementary Fig. S1B). Altogether, BIBF1120 accelerates normal peripheral retinal vascularization and diminishes preretinal NV (Fig. 5C). These findings suggest that this enhanced normal vascularization of the retina yielding decreased avascular areas may prevent ischemia-dependent preretinal neovascular tufts after treatment with BIBF1120. 

BIBF1120 Inhibits Phosphorylation of MAPK (Erk1/2) and AKT in OIR Retinas

To demonstrate that BIBF1120 specifically inhibits the proangiogenic RTKs' activity in vivo, whole retinas isolated from animals in normoxia or exposed to the 50/10 OIR model and intravitreally injected with vehicle or 50 nM of BIBF1120 (at P14 and P16) were evaluated with specific antibodies by Western blot at P18. As we expected, MAPK (Erk1/2) and AKT phosphorylation was strongly increased in retinas of animals subjected to OIR (Fig. 6A). Consistent with previously published results,^45,51 the administration of BIBF1120 showed a marked and partial reduction of MAPK (Erk1/2) and AKT phosphorylation levels, respectively (Fig. 6A); immunodensity of total MAPK and AKT was similar among the groups. Hence, BIBF1120 inhibits downstream signals of RTKs involved with pathological angiogenesis in OIR. 

Figure 6

 

BIBF1120 inhibits RTK activity and regulates the expression of angiogenic and angiostatic mediators in the retina during OIR. (A) Western blot analysis on retinal homogenates isolated at P18 from animals exposed to normoxia or to the 50/10 OIR model and intravitreally injected with vehicle (PBS) or 50 nM of BIBF1120 (at P14 and P16). BIBF1120 caused a marked and partial reduction in phosphorylation of MAPK (Erk1/2, 44 kDa, and 42 kDa) and AKT (60 kDa) (n = 3 per group) compared to vehicle-treated animals. Total immunoreactivity to MAPK (Erk1/2) and AKT was comparable in BIBF1120 and vehicle-treated animals. Densitometry quantifications are illustrated in the histograms. Values were normalized to normoxia. ##P < 0.01, ###P < 0.001 versus normoxia or *P < 0.05, **P < 0.01 versus OIR-vehicle. (B) Quantitative real-time PCR analysis of VEGF, Dll4, Netrin-1, EphrinB2, and EphB4 mRNA was performed on whole retinas at P15 from animals exposed to 50/10 OIR model and twice injected intravitreally with vehicle or BIBF1120 (50 nM, 100 nM, or 200 nM) at P5 and at P12; normoxic control (N) values were set at 1. A significant increment in retinal mRNA expression of all factors was observed in retinas from 50/10 OIR animals at P15 compared to the normoxic retinas, while BIBF1120 decreased significantly their mRNA expression. Values are mean ± SEM of n = 4–6 animals per group. The fold changes were normalized to 18S as internal control. Significant differences (P value) in the mRNA levels between vehicle and BIBF1120 treatment are shown in the histograms; *P < 0.05, **P < 0.01, ***P < 0.001 compared to OIR (vehicle). (C) Western blot analysis performed on retinal homogenates at P15 from animals exposed to 50/10 OIR model and injected intravitreally twice with vehicle or BIBF1120 (50 nM) at P5 and P12. An augmentation in protein immunoreactivity for VEGF, Dll4, Netrin-1, EphrinB2, and its receptor EphB4 was seen in retinas from OIR vehicle-treated animals (n = 3 samples) compared to the normoxic retinas (Nor), and BIBF1120 diminished this immunoreactivity. β-actin (42 kDa) was used as internal control. Densitometry quantifications are illustrated in the histograms. Values were normalized to normoxia. #P < 0.05, ##P < 0.01, ###P < 0.001 versus normoxia or *P < 0.05, **P < 0.01, ***P < 0.001 versus OIR-vehicle.

BIBF1120 Interferes With the Expression of Major Angiogenic Factors and Repulsive Cues in the Retina During OIR

The three RTKs known to be inhibited by BIBF1120, VEGFR, PDGFR, and FGFR, are also known to coordinate the expression of numerous angiogenic and repulsive cues implicated in neovascularization.^31,33,62,63 To elucidate the mechanisms of BIBF1120 in modulating pathological angiogenesis, we studied at P15 retinal tissue from animals subjected to OIR and treated or not with different concentrations of BIBF1120 at P5 and P12, the mRNA expression and protein levels of factors that regulate angiogenesis, using real-time qPCR and Western blot analysis. The qPCR analysis showed a marked upregulation in OIR of VEGF (4.0-fold), its negative regulator Dll4 (6.9-fold),⁶⁴ as well as neuronal/vascular guidance cues Netrin-1 (5.5-fold) and repulsive factor EphrinB2 (4.7-fold) and its receptor EphB4 (3.0-fold). BIBF1120 suppressed the expression of all of these factors (Fig. 6B). Similar results were observed by measuring protein expression by Western blot analysis (Fig. 6C). 

Finally, to demonstrate that RTKs' activation can modulate the expression of angiogenic factors and guidance cues, retinal ganglion cells—that abundantly express the RTK FGFR⁶⁵ and are responsible for secreting guidance cues^22,29—were incubated with recombinant human FGFb for 24 hours in presence or absence of BIBF1120. A significant augmentation in the mRNA expression of VEGF (1.3-fold), Netrin-1 (1.6-fold), and Sema3A (1.7-fold), but not Sema3F (1.0-fold), was detected at the end of the incubation period (Supplementary Fig. S2). Cotreatment with BIBF1120 exerted a decrease in expression of VEGF, Netrin-1, and Sema3A, while Sema3F remained unchanged. Taken together, these data suggest that inhibition of the RTK (notably FGFR-coupled, but also VEGFR- and PEDFR-coupled) signaling pathway in the retina using BIBF1120 downregulates the expression of downstream cues known to participate in abnormal retinal vascularization. 

Discussion

Neovascularization is a hallmark feature in vasoproliferative retinopathies and the major cause of irreversible vision loss. Recent efforts in developing new treatment options to counteract aberrant angiogenesis in the eye have been focused on inhibiting the activity of growth factors and its receptors that play a crucial role in the development of pathological ocular neovascularization.^66–68 

Receptor tyrosine kinases are cell-surface growth-factor receptors with an intrinsic ligand-controlled tyrosine-kinase activity³⁰ strongly implicated in the development of ocular neovascular diseases.^8,39 Several studies have suggested that inhibition of RTKs is an effective and potent therapeutic alternative to control ocular angiogenic diseases.^42–44 Recent combinations of inhibitors to ligands of distinct RTKs (specifically VEGFR and PDGFR) appears promising in the clinical setting (Boyer D. IOVS 2013;54:ARVO E-Abstract 2175). In this study, we demonstrate, to our knowledge for the first time, that BIBF1120, a triple angiokinase inhibitor successfully used in several human clinical trials to treat cancer,⁴⁹ abolished pathological retinal neovascularization and enhanced normal vascularization in a model of OIR in rats. 

BIBF1120 is an effective inhibitor of proangiogenic-signaling pathways of VEGFR, PDGFR, and FGFR⁴⁵ expressed in vascular endothelial cells, pericytes, and smooth muscle cells—three cell types contributing to angiogenesis. Its activity has been evaluated in several biochemical assays, showing that BIBF1120 inhibits a narrow range of kinases at pharmacologically relevant concentrations: VEGFR types 1, 2, and 3 (half maximal inhibitory concentration [IC₅₀] 13–34 nmol/L); PDGFR-α and PDGFR-β (IC₅₀ 59–65 nmol/L); and FGFR types 1, 2, and 3 (IC₅₀ 37–108 nmol/L).⁴⁵ Pharmacological concentrations of BIBF1120 used herein were based on these results. BIBF1120 inhibited ex vivo vascular sprouting of aortic rings and normal vascular development in vivo, consistent with effects on tumor blood vessels,⁴⁹ human endothelial cells and skin microvessels,⁴⁵ and corresponding roles of VEGFR,^69,70 FGFR,⁷¹ and PDGFR,⁷² whereas during hyperoxia, BIBF1120 did not alter the vascular area, possibly because VEGF expression is already suppressed by hyperoxia.^59,60 

We used the rat 50/10 OIR model^52,53 to evaluate the antiangiogenic effects of BIBF1120 during pathological conditions because it is a reproducible model and consistently recreates characteristic features akin to severe ROP experienced by human preterm infants.³ The 50/10 OIR model is characterized by a peripheral vascular arrest that occurs during hyperoxia/hypoxia cycling from P0 to P14 (first phase of ROP), followed by a retinal neovessel formation into vitreous seen morphologically from P15 onward and displaying maximal severity at P18 (second phase of ROP). We chose P5 and P12 (group 1) as critical time points to perform our intravitreal injections to inhibit the early actions of VEGFR, FGFR, and PDGFR activation because VEGF and other growth factors are upregulated following the hypoxic phases of oxygen cycling.^73,74 A similar argument applies to choosing P14 and P16 (group 2), since upon return to room air, there is a substantial augmentation of angiogenic factors and their receptors in the retina that initiate an aggressive preretinal neovascular response.^73,74 Accordingly, BIBF1120 inhibited proangiogenic RTK activity in vivo and blocked pathological neovascularization. This last result is consistent with a previous study that showed that BIBF1120 released in the posterior segment of the eye inhibits laser-induced choroidal neovascularization.⁵⁰ Moreover, BIBF1120 diminished the extent of avascular areas, implying that BIBF1120 enhances physiological retinal microvascular repair by promoting intraretinal normal vascularization. Oral administration of BIBF1120 was effective on pathological neovascularization, was well tolerated for the animals, and had a tendency to diminish the avascular area; higher doses may be required to achieve desired local intraocular concentrations. 

A number of studies by us and others report the presence of repulsive cues, notably Sema-3A,^22,24 Slit2,⁷⁵ EphrinB2,⁷⁶ and (bifunctional) Netrin-1,^77,78 in delaying vascularization into ischemic retinal areas. A complex regulation of these guidance factors is governed by VEGFR, PDGFR, and FGFR.^31,33,62,63 For instance, FGFR has been shown to regulate the expression of Netrin-1³¹ (as we demonstrated here on RGC), and along with PDGFR that of VEGF, independent of HIF-1α,⁶² while Dll4 controls expression of EphrinB2/EphB4.³³ Hence, proangiogenic receptors exert counter-regulatory actions to properly sculpt the vascular system; conversely, dysregulation of such counter-regulatory mechanisms preclude appropriate vascular sculpting.^{22,24,75,76,78} BIBF1120 exerted a pronounced suppression of OIR-induced expression of repulsive cue receptor EphB4 and VEGF regulator Dll4, while attenuating the rise in VEGF (and Netrin-1), resulting in enhanced normal retinal vascularization. EphB4 receptor and its surface-bound EphrinB2 ligand are critical regulators of retinal angiogenesis, modulating both endothelial cells and pericyte function.⁷⁹ Accordingly, therapeutic targeting of EphrinB2/EphB4 expression may prove useful for disrupting angiogenesis in vasoproliferative retinopathies²⁸; alternatively, BIBF1120 may interfere with actions of the receptor tyrosine kinase EphB4, but this needs to be separately explored. 

The other factor robustly regulated by BIBF1120 is Dll4, an endothelial-specific ligand that interacts with its cognate receptor Notch1 on the surface of endothelial cells and is strongly implicated in physiological and pathological angiogenesis in the retina.^80,81 In OIR, Dll4 is induced in veins and capillaries adjacent to the zone of VO, where VEGF levels are elevated,⁸² and acts as a negative feedback regulator of VEGF.⁶⁴ Dll4 also regulates the expression of major repulsive factors EphrinB2/EphB4.²¹ Hence, all in all, Dll4-Notch signaling inhibits physiological angiogenesis by reducing the number of tip cells and decreasing angiogenic sprouting in the retina.⁸³ Conversely, Dll4 inhibition has been shown to enhance angiogenic sprouting and regrowth of lost retinal vessels while suppressing ectopic pathological neovascularization in ischemic retinophathy.⁶⁴ Thus, concerted actions of BIBF1120 in inhibiting Dll4 and EphrinB2/EphB4 are consistent with accelerated normal vascularization of the retina. 

In sum, our results reveal that BIBF1120, a triple angiokinase inhibitor that simultaneously acts on three receptor families involved in blood vessel formation,⁴⁵ is an effective and potent inhibitor of the preretinal pathological neovascularization associated with ROP/OIR. BIBF1120 represents a promising alternative to anti-VEGF therapy because, when injected at early stages of ROP, it can accelerate normal vascularization, which is essential to prevent preretinal neovascularization and long-term functional damage.^22,24,84 These properties surpass those of anti-VEGFs and provide a potential alternative to anti-VEGFs through these desirable attributes. Moreover, these properties could be of interest for the treatment of a variety of ischemic retinopathies associated with pathological neovascularization in addition to ROP, notably diabetic retinopathy. Although promising,⁵⁰ future studies with BIBF1120 are needed to evaluate its side effects.^85,86 

Acknowledgments

The authors thank Isabelle Lahaie for technical assistance. This work was supported by Boehringer Ingelheim Pharmaceuticals, Inc. Sylvain Chemtob holds a Canada Research Chair (Vision Science) and the Leopoldine Wolfe Chair in translational research in age-related macular degeneration. 

Disclosure: J.C. Rivera, Boehringer Ingelheim Pharmaceuticals, Inc. (F); B. Noueihed, None; S. Omri, None; J. Barrueco, Boehringer Ingelheim Pharmaceuticals, Inc. (E); F. Hilberg, Boehringer Ingelheim Pharmaceuticals, Inc. (E); S. Chemtob, Boehringer Ingelheim Pharmaceuticals, Inc. (F)