Antiangiogenic effect of dasatinib in murine models of oxygen-induced retinopathy and laser-induced choroidal neovascularization

Cell culture

Human retinal microvascular endothelial cells (HRMECs), endothelial growth medium-2 (EGM-2), and endothelial basal medium (EBM) were purchased from Lonza (Walkersville, MD). The HRMECs were plated in gelatin-coated plates and cultured in EGM-2 at 37 °C in a humidified atmosphere of 5% CO₂ in air. The HRMECs used for the experiments were between passages 3 and 6.

In vitro angiogenesis assays

The tube formation assay was performed by seeding cells (3 × 10⁴/well) in 24-well plates coated with Matrigel (BD Bioscience, Bedford, MA) and cultured in EBM containing 1% fetal bovine serum (FBS, Lonza) in the absence or presence of 100 ng/ml recombinant human VEGF (rhVEGF; R&D Systems, Minneapolis, MN) and 25 nM dasatinib (Selleckchem, Houston, TX). After a 6-h incubation period, tube lengths were measured in four random microscope fields.

For the scratch wounding migration assay, confluent cell monolayers grown in gelatin-coated four-well plates were scratched using a micropipette tip. The plates were washed with PBS (1X; 137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 1.5 mM KH₂PO₄; pH 7.4) to remove dislodged cells and media, and incubated with EBM containing 1% FBS in the absence or presence of rhVEGF (100 ng/ml) and dasatinib (25 nM). After a 24-h incubation period, the plates were observed using optical microscopy, and cell migration was assessed by measuring the area covered by cells that had migrated from the wound edges.

For the cell proliferation assay, cells (1 × 10⁴/well) were seeded in gelatin-coated 96-well plates and incubated with EBM containing 1% FBS in the absence or presence of rhVEGF (100 ng/ml) and dasatinib (25 nM) for 3 days. Cell proliferation was assessed using the Cell Counting Kit-8 (CCK-8; Dojindo Molecular Technology, Rockville, MD) assay in accordance with the manufacturer’s protocol. To determine the number of viable cells, the optical density of each well was read at a wavelength of 450 nm using a microplate reader (BioTek Instruments, Seoul, Korea).

Animals

All mice were cared for in accordance with the Association for Research in Vision and Ophthalmology statement for the use of animals in ophthalmic and vision research, and the protocols were approved by the Institutional Animal Care and Use Committee. Pregnant C57BL/6J mice and 9- to 10-week-old male C57BL/6J mice (Orient, Seoul, Korea) were housed at room temperature with a 12 h:12 h light-dark cycle. The mice were anesthetized with an intraperitoneal injection of ketamine (79.5 mg/kg) and xylazine (9.1 mg/kg), and the pupils of the anesthetized mice were dilated with topical 1% tropicamide (Santen, Osaka, Japan).

Mouse model of OIR

OIR was induced in mice using the protocol reported by Smith et al. [10]. Briefly, newborn mice were reared in a hyperoxic (75% oxygen) chamber connected to an oxygen controller (ProOx P110; BioSpherix, Parish, NY) from postnatal day (P)7 to P12 (Phase I: vaso-obliteration). On P12, the mice were returned to normoxia and received a single intravitreal injection of dasatinib (1 µg in 1 µl dimethyl sulfoxide [DMSO]) or 1 µl DMSO (contralateral control). After exposure to normoxic air for 5 days (P12-P17), the mice were euthanized using carbon dioxide inhalation, and both eyes were harvested for the analysis of retinal neovascularization (Phase II; neovascularization). In Phase I, hyperoxia inhibits central vessel growth in the retina and causes significant vessel loss; in Phase II, hypoxia upregulates the expression of angiogenic factors such as VEGF, thus triggering pathological neovascularization, which reaches a maximum at P17.

Quantification of retinal neovascularization

The eyes were enucleated and fixed overnight in 4% paraformaldehyde. The retinas were isolated, cut four times from the edge to the center, soaked in blocking solution (5% BSA [BSA], 5% normal donkey serum, and 0.5% Triton X-100 in PBS), and then incubated overnight at 4 °C with Alexa Fluor^® 594-conjugated Griffonia simplicifolia isolectin B4 (1:100 dilution; Invitrogen, Carlsbad, CA) in PBS containing 1 mM CaCl₂. Retinas were rinsed three times with PBS, flatmounted on microscope slides, and embedded in fluorescent mounting medium (Dako, Carpinteria, CA). Images of the whole-mounted retina were captured using a fluorescence microscope (Olympus, Tokyo, Japan); the exposure and gain were kept constant for all samples. In each image, the number of pixels in the preretinal neovascular tufts was determined using ImageJ software (National Institutes of Health, Bethesda, MD) and was expressed as a percentage of the number of pixels in the entire retinal area.

Mouse model of laser-induced CNV

CNV was induced by laser photocoagulation of Bruch’s membrane with a Micron IV image-guided laser system (532 nm; IRIS Medical, Mountain View, CA). Briefly, 9- to 10-week-old male C57BL/6J mice were anesthetized with an intraperitoneal injection of 79.5 mg/kg ketamine and 9.1 mg/kg xylazine, and the pupils were dilated with topical 1% tropicamide (Santen). Laser photocoagulation (spot size, 75 µm; intensity, 130 mW; duration, 0.1 s) was performed with four laser burns in the 3, 6, 9, and 12 o’clock positions of the posterior pole of the fundus at equal distances from the optic nerve head. Only mice with cavitation bubbles, which indicated the disruption of Bruch’s membrane, were included in the study. Immediately after CNV induction, the anesthetized mice received a single intravitreal injection of dasatinib (1 μg in 1 μl DMSO) or 1 μl DMSO (contralateral control). Two weeks later, the mice were euthanized using carbon dioxide inhalation for further analysis.

Quantification of laser-induced CNV

Anesthetized mice were perfused through the left ventricle with PBS followed by 5 mg/ml fluorescein isothiocyanate (FITC)-labeled dextran (MW=10,000 Da; Sigma, St. Louis, MO). The eyes were enucleated and fixed in a 4% paraformaldehyde solution. The anterior segment and retina were removed from the eyecup, and the remaining RPE/choroid/sclera complexes were flatmounted after four radial incisions. The dissected RPE/choroid/sclera were treated with blocking solution (5% BSA, 5% normal donkey serum, and 0.5% Triton X-100 in PBS) for 1 h at room temperature and incubated with Alexa Fluor^® 594-conjugated Griffonia simplicifolia isolectin B4 (1:100 dilution; Invitrogen) overnight at 4 °C. The samples were washed three times with PBS and mounted sclera side down on microscope slides. Images of CNV lesions were obtained using a fluorescence microscope (Olympus), and the CNV area in each image was calculated using ImageJ software (NIH). For the analysis of CNV volumes, Z-stack images of CNV lesions were acquired with an LSM780 confocal microscope (Zeiss, Jena, Germany). The sum of the CNV area in each Z-stack layer (multiplied by 5-μm thickness) was used as the volume of the CNV lesion [11].

Retinal toxicity

Normal mice received a single intravitreal injection of dasatinib (1 µg in 1 µl in DMSO) or 1 µl DMSO (contralateral control). Two weeks later, the mice were euthanized using carbon dioxide inhalation for histological and immunohistochemical examinations. For the analysis of the retinal toxicity of dasatinib in mice with OIR or CNV, ocular sections were obtained from the mice that were prepared for the OIR and CNV experiments reported in images.

Immunohistochemistry

The eyes were enucleated, fixed, embedded in optimum cutting temperature (OCT) compound (Sakura Finetek, Torrance, CA) or paraffin, sectioned, and mounted on slides. First, the endogenous peroxidase activity in the tissues was quenched. The sections were blocked with 10% normal goat serum and were then incubated with primary immunoglobulin G (IgG) against phospho-Y416 Src (p-Y416 Src; Invitrogen) or glial fibrillary acidic protein (GFAP; Dako), followed by the appropriate fluorescent secondary IgG (Dako). Blood vasculature was visualized by costaining with Alexa Fluor^® 594-conjugated Griffonia simplicifolia isolectin B4 (Invitrogen). Cell nuclei were stained with 4’,6-diamidino-2-phenylindole (DAPI; Vector Laboratories, Burlingame, CA). The images were acquired using a fluorescence microscope (Olympus), and six to eight sections were examined per group.

Histological analysis

Hematoxylin and eosin (H&E) staining was performed on paraffin-embedded sections. Retinal thickness was assessed as the ratio of A (the distance from the ganglion cell layer to the outer edge of the inner nuclear layer) to B (the distance from the ganglion cell layer to the outer edge of the outer nuclear layer).

TUNEL assay

Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assays were performed on paraffin-embedded sections using the In Situ Cell Death Detection Kit (Roche Applied Science, Mannheim, Germany) in accordance with the manufacturer’s instructions. Counterstaining for total nuclei was performed by mounting with antifading medium containing DAPI (Vector Laboratories). The positive controls were paraffin sections that were treated with DNase I (1,500 U/ml; Roche Applied Science) for 10 min at room temperature to induce DNA strand breaks. Images were obtained using a fluorescence microscope (Olympus); six to ten sections were examined per group.

Statistical analysis

All results are expressed as the mean ± standard error of the mean (SEM). Statistical significance was evaluated using an one-way ANOVA followed by Bonferroni’s post hoc multiple comparison test (Figure 1) or unpaired Student t test (Figure 2, Figure 3, and Figure 4). A p value of less than 0.05 was considered statistically significant.

Dasatinib inhibits VEGF-induced increase in endothelial tube formation, scratch wounding migration, and cell proliferation

We had previously determined that dasatinib (25 nM) completely blocked the VEGF-induced phosphorylation of Src on Y416 in HRMECs. Consequently, we treated the HRMECs with this concentration of dasatinib and evaluated its inhibitory effect on VEGF-induced angiogenesis [9]. As VEGF strongly enhances a variety of coordinated events required for angiogenesis, including the proliferation, migration, and morphogenesis of endothelial cells, we conducted three in vitro angiogenesis assays: Matrigel tube formation, scratch wounding migration, and CCK-8 cell proliferation assay. As shown in Figure 1, the VEGF-induced increase in tube formation, migration, and cell proliferation in HRMECs was significantly blocked by cotreatment with dasatinib. However, treatment with dasatinib alone did not affect the basal angiogenic activity of the HRMECs. These data support the hypothesis that VEGF-induced in vitro angiogenesis is effectively inhibited by dasatinib in HRMECs.

Dasatinib reduces retinal neovascularization in mice with OIR

To investigate the beneficial effect of dasatinib on pathological retinal neovascularization, we employed a mouse model of OIR, in which retinal ischemia increases the expression of various hypoxia-inducible angiogenic factors, including VEGF. Newborn pups were exposed to hyperoxia from P7 to P12. On P12, the pups were returned to the normoxic conditions and received a single intravitreal injection of dasatinib or DMSO vehicle. On P17, the extent of neovascularization was determined by quantifying the neovascular tufts, which represent pathological retinal neovascularization (Figure 2A). As illustrated in Figure 2B, the DMSO-treated control eyes had markedly high periretinal neovascularization (tuft formation). However, the periretinal neovascularization profoundly decreased in the retinas of the dasatinib-treated eyes. In the quantitative analysis, the percentage of neovascularization in the eyes treated with dasatinib and DMSO was 6.47±1.07 and 12.00±1.39 (n=6 mice per group) of the total retinal area, respectively. Thus, a single intravitreal injection of dasatinib induced a 46.08% reduction (*p<0.05) in the ischemia-induced pathological neovascularization. We also investigated whether intravitreal injection of dasatinib reduced Src phosphorylation in the retinal vasculature. In the immunohistochemical analysis of transverse retinal sections, the retinal tissues of DMSO-injected control eyes exhibited a strong staining of p-Y416 Src in isolectin B4–positive endothelial cells, whereas the retinal endothelial cells of dasatinib-injected eyes showed a substantial decrease in the p-Y416 Src signal (Figure 2C).

Dasatinib reduces CNV formation in mice with laser-induced CNV

To investigate the inhibitory effect of dasatinib on pathological CNV, we used a mouse model of laser injury-induced CNV, in which the rupture of Bruch’s membrane results in new vessel growth from the choroid into the subretinal space. Immediately after laser photocoagulation, the mice received a single intravitreal injection of dasatinib or DMSO. Two weeks later, CNV lesions were visualized by perfusing mice with FITC-labeled dextran and staining the choroidal vasculature with isolectin B4 (Figure 3A). Representative images of CNV lesions showed that the dasatinib-treated eyes had significantly less neovascularization than the DMSO-treated control eyes did (Figure 3B). Quantitatively, the area and volume of the CNV lesions in eyes treated with dasatinib were (35.28±5.13) × 10³ μm² and (36.74±4.43) × 10⁴ μm³ (n=10 mice), whereas those in the eyes treated with DMSO were (66.28±5.27) × 10³ μm² and (55.41±7.16) × 10⁴ μm³ (n=10 mice). Compared with that of the DMSO-treated control eyes, the dasatinib-treated eyes had a 46.77% (**p<0.01) and 33.69% reduction (*p<0.05) in the area and volume of the CNV lesions, respectively.

Intravitreal injection of dasatinib does not cause retinal toxicity in mice

To evaluate the retinal toxicity caused by the intravitreal injection of dasatinib in normal mice and in mice with OIR or CNV, we conducted histological examinations and immunohistochemical analyses on ocular tissue sections. The histological examinations of the H&E-stained tissue sections revealed that the dasatinib-injected retinal tissues showed no change in histological morphology or retinal thickness compared with that of the DMSO-injected contralateral eyes (Figure 4A). In addition, GFAP immunohistochemical staining and TUNEL assays revealed that the intravitreal injection of dasatinib into mice with OIR or CNV did not affect GFAP expression and apoptosis levels compared to those observed after DMSO injection (Figures 4B,C). The examination of retinal tissues from normal mice (i.e., without OIR or CNV) also revealed that there were no differences in TUNEL-positive cells and GFAP expression between the dasatinib-treated eyes and DMSO-treated controls (Figure 4B,C). These data suggest that a single intravitreal injection of dasatinib does not cause retinal toxicity in normal mice and in mice with OIR or CNV.