March 2016
Volume 57, Issue 3
Open Access
Retinal Cell Biology  |   March 2016
The cKit Inhibitor, Masitinib, Prevents Diabetes-Induced Retinal Vascular Leakage
Author Affiliations & Notes
  • So Ra Kim
    College of Pharmacy Ajou University, Suwon, Korea
  • Ji-Eun Im
    College of Pharmacy, Chung-Ang University, Seoul, Korea
  • Ji Hoon Jeong
    School of Pharmacy, Sungkyunkwan University, Suwon, Korea
  • Ji Yeon Kim
    College of Pharmacy, Chung-Ang University, Seoul, Korea
  • Jee Taek Kim
    Department of Ophthalmology, College of Medicine, Chung-Ang University, Seoul, Korea
  • Se Joon Woo
    Department of Ophthalmology, College of Medicine, Seoul National University, Seongnam, Korea
  • Jong-Hyuk Sung
    College of Pharmacy, Yonsei University, Incheon, Korea
    STEMORE Co., Ltd., Incheon, Korea
  • Sang Gyu Park
    College of Pharmacy Ajou University, Suwon, Korea
  • Wonhee Suh
    College of Pharmacy, Chung-Ang University, Seoul, Korea
  • Correspondence: Wonhee Suh, College of Pharmacy, Chung-Ang University, Seoul 156-756, Korea; wsuh@cau.ac.kr
  • Footnotes
     SRK and J-EI contributed equally to the work presented here and should therefore be regarded as equivalent authors.
Investigative Ophthalmology & Visual Science March 2016, Vol.57, 1201-1206. doi:10.1167/iovs.15-18065
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      So Ra Kim, Ji-Eun Im, Ji Hoon Jeong, Ji Yeon Kim, Jee Taek Kim, Se Joon Woo, Jong-Hyuk Sung, Sang Gyu Park, Wonhee Suh; The cKit Inhibitor, Masitinib, Prevents Diabetes-Induced Retinal Vascular Leakage. Invest. Ophthalmol. Vis. Sci. 2016;57(3):1201-1206. doi: 10.1167/iovs.15-18065.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: Stem cell factor (SCF) has recently demonstrated activity as a novel endothelial permeability factor that contributes to the development of diabetes-induced hyperpermeable retinal vasculature. This study investigated the therapeutic potential of masitinib, a pharmacologic inhibitor of the SCF receptor cKit, for prevention of diabetes-induced breakdown of blood retinal barrier (BRB).

Methods: Permeability assays were performed with human retinal microvascular endothelial cells (HRMECs) and murine retinal vasculature. Localization of vascular endothelial (VE)-cadherin and activation of SCF signaling pathway was determined by immunofluorescence and Western blotting assays. Mice and rats with streptozotocin (STZ)-induced diabetes were used to investigate the role of cKit and masitinib in diabetes-induced retinal vascular hyperpermeability.

Results: Masitinib substantially blocked SCF-induced phosphorylation of cKit in HRMECs. In vitro and in vivo vascular permeability assays showed that masitinib significantly inhibited SCF-induced endothelial hyperpermeability and junctional loss of VE-cadherin. Streptozotocin-induced diabetes was induced in cKit-mutant mice with low cKit expression in their endothelial cells. Although diabetic wild-type mice exhibited enhanced retinal vascular leakage, diabetic cKit-mutant mice showed no increase in retinal vascular leakage or alteration in the distribution of VE-cadherin; this indicates the crucial role of cKit in diabetes-induced breakdown of BRB. Moreover, in vivo prevention experiments showed that an intravitreal injection of masitinib substantially inhibited the development of hyperpermeable retinal vasculature.

Conclusions: These results provide the first demonstration that cKit inhibitors, such as masitinib, might be promising therapeutics for prevention of diabetes-induced breakdown of the BRB.

In the normal retina, blood retinal barrier (BRB) regulates the entry of molecules into the inner retina and protects vulnerable retinal neural tissue from potentially harmful substances in the blood. Complex intercellular junctions that restrict paracellular permeability and induce high transendothelial electrical resistance (TEER) tightly seal the endothelium in the inner BRB. However, these BRB properties are disrupted and retinal vascular permeability is enhanced during the early pathogenesis of many ocular diseases including diabetic retinopathy, AMD, and retinal vascular occlusions. Several studies have reported that the breakdown of BRB and retinal vascular leakage in these ocular diseases stems from an altered distribution and expression level of junction proteins in retinal endothelial cells.13 
We recently identified stem cell factor (SCF) as a novel endothelial permeability factor.4 In human umbilical vein endothelial cells (HUVECs), SCF induced activation of its own receptor (cKit), and increased endothelial permeability and internalization of vascular endothelial (VE)-cadherin. In addition, expression levels of both SCF and cKit were significantly upregulated in the retina of streptozotocin (STZ)-induced diabetic mice. Other investigators have similarly reported that SCF and cKit are highly upregulated in the epiretinal membranes of patients with diabetic retinopathy as compared with their expression in nondiabetic patients.5 Quantitative analysis of protein levels in human vitreous fluid also revealed a marked increase in SCF levels in patients with diabetic retinopathy.6 These findings suggest that SCF might be a major contributor to diabetes-induced breakdown of the BRB. 
In this regard, we propose that masitinib, a potent and selective inhibitor of cKit, may have beneficial actions for the prevention of diabetes-induced retinal vascular leakage. Masitinib was developed for the treatment of diseases associated with cKit activation, such as mastocytosis and gastrointestinal stromal tumors.7 Masitinib exhibited high affinity and improved selectivity for cKit when compared with other cKit inhibitors.7 We now show that masitinib can prevent the diabetes-induced retinal vascular leakage when administered to STZ-induced diabetic rats. 
Methods
Measurement of Endothelial Paracellular Permeability and TEER
Human retinal microvascular endothelial cells (HRMECs; Cell Systems, Kirkland, WA, USA) were cultured in endothelial growth medium (Lonza, Walkersville, MD, USA). Confluent HRMEC monolayers were established on transwell membranes (Corning, Tewksbury, MA, USA). Recombinant human SCF (rhSCF; R&D Systems, Minneapolis, MN, USA) or PBS was added to the culture medium in the upper chamber of the transwell. Masitinib (Selleckchem, Houston, TX, USA) was added 30 minutes before SCF stimulation. For the paracellular permeability assay, the medium in the upper chamber was supplemented with fluorescein isothiocyanate (FITC)-conjugated dextrans (molecular weight = 40 kDa; Molecular Probes, Carlsbad, CA, USA). After incubation for 30 minutes, the fluorometric signals derived from FITC-dextran in the upper and lower chamber were measured using a fluorometer. For the TEER assay, electrical resistance across a monolayer of HRMECs established on a transwell membrane was measured using a Millicell ERS-2 Volt ohmmeter (Millipore, Billerica, MA, USA). Specified reagents were added to the medium in upper chamber at time zero and serial changes in electrical resistance were measured thereafter. The TEER values were calculated by subtracting the background TEER from the experimental TEER, then multiplying the result by the surface area of the filter. 
Western Blotting
Cell lysates were separated using SDS-PAGE. The blots were hybridized with the appropriate primary IgGs: phospho-cKit (p-cKit; R&D Systems), cKit (R&D Systems), phospho-endothelial nitric oxide synthase (p-eNOS; Cell Signaling Technology, Danvers MA, USA), eNOS (BD Biosciences, San Jose, CA, USA), and β-actin (Santa Cruz Biotechnology, Santa Cruz, CA, USA), followed by horseradish peroxidase-conjugated secondary IgGs. Then, immunoreactive bands were visualized with a chemiluminescent reagent (Amersham Biosciences, Piscataway, NJ, USA). 
Vascular Endothelial Cadherin Internalization and Co-Immunofluorescence Assays
The VE-cadherin internalization assay was performed according to a previously described protocol.4 Briefly, HRMECs were incubated with anti-human VE-cadherin IgG (R&D Systems) at 4°C for 1 hour and washed with cold basal medium to remove unbound IgG. Next, the cells were treated with rhSCF (50 ng/mL) or PBS at 37°C for 30 minutes. The cells were washed with acidic PBS to remove membrane-bound IgG, permitting the detection and quantification of internalized VE-cadherin. Cells with or without the acid washes were fixed, blocked, and stained with FITC-conjugated secondary IgG (Molecular Probes). For the co-immunofluorescence assay, the cells were further stained with primary IgG against early endosome antigen 1 (EEA1; BD Biosciences) or zona occludens 1 (ZO-1; BD Pharmingen, Palo Alto, CA, USA), followed by a rhodamine-conjugated secondary IgG (Molecular Probes). The nuclei were counterstained with 4,6-diamidino-2-phenylindole (DAPI). The amount of internalized VE-cadherin was assessed by counting the number of acid-resistant VE-cadherin–positive vesicles per cell. Cells containing acid-resistant VE-cadherin–positive vesicles were counted. 
Animals
Animal experiments were conducted with 6- to 8-week-old C57BL6 mice (Charles River Laboratories, Yokohama, Japan), KitW-sh/W-sh mice (Jackson Laboratory, Bar Harbor, ME, USA), and 8-week-old male Sprague-Dawley (SD) rats (Orient, Seoul, Korea). All animal protocols were reviewed and approved by the Institutional Animal Care and Use Committee (Approval Reference Number: AMC 116) in accordance with the ARVO statement for the use of animals in ophthalmic and vision research. For the intravitreal injection, the mice and rats were anesthetized with an intraperitoneal injection of ketamine/xylazine (79.5/9.1 mg/kg for mice, 80/5 mg/kg for rats). The adequacy of the anesthesia was assessed by monitoring the pedal withdrawal reflex response. At the end of the experiments, the animals were killed using carbon dioxide inhalation and tissues were collected for further analysis. The number of animals in most groups ranged from four to six. 
Retinal Vascular Permeability Assay
The FITC-dextran (molecular weight = 70 kDa; Molecular Probes) was intravenously injected into mice and rats, then the animals were killed by an anesthetic overdose 4 to 5 hours later. The retinas were isolated and prepared as flat mounts. Digital images of randomly selected areas were generated using identical fluorescence microscopic settings for each experiment. The fluorescence intensity was analyzed in each image by using Image J software (http://imagej.nih.gov/ij/; provided in the public domain by the National Institutes of Health, Bethesda, MD, USA). The average retinal fluorescent intensity was then normalized to the plasma fluorescence intensity. 
Rodent Models of STZ-Induced Diabetic Mellitus (DM)
Experimental diabetes was induced by intraperitoneal injections of STZ (Sigma-Aldrich Corp., St. Louis, MO, USA). For DM induction, animals received three (rats) or five (mice) consecutive daily injections of STZ (50 mg/kg) after an overnight fast. Rats and mice injected with an equal volume of PBS served as the nondiabetic (non-DM) controls. Serum glucose levels were measured using an Accu-Check Advantage glucose meter (Roche, Indianapolis, IN, USA) under nonfasting conditions. Blood glucose levels were measured at 24 hours after the final STZ injection and animals with consistently elevated glucose levels (>300 mg/dL) were considered diabetic. On the same day as DM onset, diabetic SD rats received a single intravitreal injection of the indicated amount of masitinib in 5 μL of dimethyl sulfoxide (DMSO) or the equivalent volume of DMSO. Two weeks after intravitreal injection, retinal vascular permeability was assessed. Body weight and blood glucose levels were recorded. 
Reverse Transcriptase PCR Analysis
Total RNA was extracted from cells or tissues by using the Trizol reagent (Invitrogen, Carlsbad, CA, USA). Complementary DNA was synthesized using the Superscript first-strand synthesis kit (Invitrogen) and PCR amplified (30–35 cycles) using specific primers for SCF, cKit, or glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Real-time PCR was performed with the SYBR-Green PCR master mix (Applied Biosystems, Carlsbad, CA, USA) by using the StepOnePlus Real-Time PCR System (Applied Biosystems). Data analysis was performed based on the ΔΔ Ct method and raw data were normalized to GAPDH. All reactions were performed in triplicate. The specific primer sequences were as follows: rat SCF, forward 5′-caaaactggtggcgaatctt-3′, reverse 5′-ttcttccatacatgccacga-3′; rat cKit, forward 5′-cgcctttgtcaaatggactt-3′, reverse 5′-tgcgtcattgtcttctttgc-3′; rat GAPDH, forward 5′- ctcatgaccacagtccatgc-3, reverse 5′-ttcagctctgggatgacctt-3′. 
Immunohistochemistry
Retinas were fixed in 10% neutral buffered formalin (Sigma), embedded in paraffin, and sectioned. After quenching endogenous peroxidase activity and blocking with 10% normal horse serum, the sections were incubated with primary IgG against phospho-cKit (p-cKit; Cell Signaling Technology), p-eNOS (Cell Signaling Technology), or CD31 (DakoCytomation, Carpinteria, CA, USA), followed by the appropriate fluorescence-conjugated secondary IgG. For whole-mount retina staining, samples were blocked with 5% bovine serum albumin and 5% normal donkey serum in 0.5% Triton X-100, then stained with a primary IgG against VE-cadherin (BD Biosciences). The whole mounts were incubated with the appropriate Alexa Fluor 488-conjugated secondary IgG (Invitrogen) and further stained with Alexa Fluor 594-conjugated isolectin GS-IB4 (Invitrogen) overnight. The number of sections examined ranged from 6 to 10 per group. 
Statistical Analysis
All data are presented as mean ± SEM. Statistical significance was evaluated using one-way ANOVA followed by Bonferroni's post hoc multiple comparison test. A P value less than 0.05 was considered statistically significant in each case. The number of samples is indicated by n
Results
Masitinib Inhibits SCF-Induced Disruption of the Retinal Endothelial Barrier In Vitro
We first carried out in vitro endothelial permeability assays to investigate whether masitinib would be able to prevent SCF-mediated endothelial hyperpermeability in HRMECs. Figures 1A and 1B reveal that SCF significantly increases the paracellular passage of FITC-dextran and reduces the electric resistance across the HRMEC monolayer. However, masitinib completely abrogated the SCF-induced increase in retinal endothelial permeability at 1 μM, a concentration that was previously reported to completely inhibit the kinase activity of recombinant human cKit protein.7 We then performed Western blotting experiments to confirm that SCF-induced activation of cKit and its downstream molecule, eNOS, is blocked by masitinib in HRMECs. In our previous study, SCF substantially increased the internalization of VE-cadherin, a major adherens junction protein controlling endothelial barrier function, in HUVECs.3 We next examined whether masitinib blocked the SCF-mediated endocytosis of VE-cadherin at retinal endothelial junctions. Immunofluorescence analysis revealed that SCF treatment substantially decreases immunoreactivity for VE-cadherin at retinal endothelial borders that were positively stained for ZO-1, a tight junction molecule, and increases staining for a punctuate form of intracellular VE-cadherin that is colocalized with the endosomal marker, EEA1 (Fig. 1E). This SCF-induced internalization of VE-cadherin is completely blocked by pretreatment with masitinib. These data indicate that masitinib effectively inhibits SCF-induced retinal endothelial hyperpermeability and disassembly of VE-cadherin in primary HRMECs. 
Figure 1
 
Masitinib blocks SCF-induced increase in paracellular permeability of human retinal endothelial cells in vitro. (A, B) Masitinib inhibits SCF-induced hyperpermeability of HRMECs. Endothelial permeability was determined by measuring the passage of FITC-dextran (A) and TEER (B) in an HRMEC monolayer that was pretreated with masitinib (0.1 ∼ 10 μM in [A], 1 μM in [B]) for 30 minutes before stimulation with rhSCF (50 ng/mL). Fluorescein isothiocyanate-dextran permeability is expressed as the fold increase with respect to PBS control (n = 4). Transendothelial electrical resistance values were calculated by subtracting the background electrical resistance and multiplying by the surface area of the filter (n = 5). (C, D) Masitinib inhibits SCF-induced activation of cKit (C) and eNOS (D) in HRMECs. The HRMECs treated with masitinib (1 μM) or untreated cells were stimulated with rhSCF (50 ng/mL) or PBS for 20 minutes. Phosphorylation of cKit or eNOS was assessed by Western blotting. (E) Masitinib inhibits SCF-induced internalization of VE-cadherin. The HRMECs were pretreated with masitinib (1 μM) for 30 minutes before stimulation with rhSCF (50 ng/mL). The arrows in the left image indicate the disappearance of VE-cadherin at endothelial junctions that were positively stained for anti-ZO-1 IgGs (red). The arrows in the right image indicate VE-cadherin (green) internalization in endosomes that were positively stained for EEA1 (red). Internalization of endogenous VE-cadherin was quantified based on the percentage of HRMECs with intracellular acid-resistant vesicles (three independent experiments). Scale bar: 20 μm. All data are presented as mean ± SEM (*P < 0.05 vs. PBS, #P < 0.05 vs. SCF only).
Figure 1
 
Masitinib blocks SCF-induced increase in paracellular permeability of human retinal endothelial cells in vitro. (A, B) Masitinib inhibits SCF-induced hyperpermeability of HRMECs. Endothelial permeability was determined by measuring the passage of FITC-dextran (A) and TEER (B) in an HRMEC monolayer that was pretreated with masitinib (0.1 ∼ 10 μM in [A], 1 μM in [B]) for 30 minutes before stimulation with rhSCF (50 ng/mL). Fluorescein isothiocyanate-dextran permeability is expressed as the fold increase with respect to PBS control (n = 4). Transendothelial electrical resistance values were calculated by subtracting the background electrical resistance and multiplying by the surface area of the filter (n = 5). (C, D) Masitinib inhibits SCF-induced activation of cKit (C) and eNOS (D) in HRMECs. The HRMECs treated with masitinib (1 μM) or untreated cells were stimulated with rhSCF (50 ng/mL) or PBS for 20 minutes. Phosphorylation of cKit or eNOS was assessed by Western blotting. (E) Masitinib inhibits SCF-induced internalization of VE-cadherin. The HRMECs were pretreated with masitinib (1 μM) for 30 minutes before stimulation with rhSCF (50 ng/mL). The arrows in the left image indicate the disappearance of VE-cadherin at endothelial junctions that were positively stained for anti-ZO-1 IgGs (red). The arrows in the right image indicate VE-cadherin (green) internalization in endosomes that were positively stained for EEA1 (red). Internalization of endogenous VE-cadherin was quantified based on the percentage of HRMECs with intracellular acid-resistant vesicles (three independent experiments). Scale bar: 20 μm. All data are presented as mean ± SEM (*P < 0.05 vs. PBS, #P < 0.05 vs. SCF only).
Masitinib Blocks SCF-Induced Increase in Retinal Vascular Leakage In Vivo
To determine whether masitinib can prevent the SCF-induced breakdown of retinal endothelial barrier function in vivo, a retinal permeability assay was performed in mice where masitinib was intraperitoneally injected before the intravitreal injection of SCF or PBS. In SCF-injected retinal tissues, substantial leakage of FITC-dextran with consequent FITC-dextran diffusion throughout the surrounding tissues and a pronounced increase in background fluorescence were observed (Figs. 2A, 2B). However, pretreatment with masitinib largely blocks the SCF-induced extravasation of FITC-dextran. Immunohistochemical data also reveal that intravitreal injection of SCF strongly induces the phosphorylation of cKit (Fig. 2C) and the focal loss of VE-cadherin in the retinal vasculature (Fig. 2D). However, masitinib abolishes SCF-mediated cKit phosphorylation and the disappearance of VE-cadherin from endothelial junctions. These data support the hypothesis that masitinib can attenuate BRB disruption in vivo in response to SCF. 
Figure 2
 
Masitinib blocks SCF-induced retinal vascular leakage in vivo. Masitinib (5 mg/kg) or DMSO was intraperitoneally injected into mice at 1 hour before intravitreal injection of recombinant mouse SCF (50 ng in 1 μL of PBS) or an equivalent volume of PBS. (A) Representative images of FITC-dextran–perfused retina flat mounts. The bright green fluorescent background indicates vascular leakage of FITC-dextran in the SCF-injected eyes. (B) Vascular leakage of FITC-dextran was quantified by assessing the fluorescence intensity of digital images taken from retinal whole mounts perfused with FITC-dextran. Fluorescence intensity is expressed as the fold increase ± SEM with respect to PBS control (*P < 0.05 vs. PBS, #P < 0.05 vs. SCF only, n = 4). (C) Immunofluorescence staining was performed using anti-p-cKit (green) and anti-CD31 IgGs (red) in retina sections. Arrowheads indicate double labeling with anti-CD31 IgGs with anti-p-cKit IgG. DAPI-stained nuclei are shown in blue. (D) Immunofluorescence staining in retinal whole mounts was performed using anti–VE-cadherin IgG (green) and Alexa Fluor 594-conjugated isolectin GS-IB4 (red). The arrowhead in the enlarged inset shows loss of VE-cadherin in the retinal vasculature. Scale bars in (A) to (D) are 100 μm.
Figure 2
 
Masitinib blocks SCF-induced retinal vascular leakage in vivo. Masitinib (5 mg/kg) or DMSO was intraperitoneally injected into mice at 1 hour before intravitreal injection of recombinant mouse SCF (50 ng in 1 μL of PBS) or an equivalent volume of PBS. (A) Representative images of FITC-dextran–perfused retina flat mounts. The bright green fluorescent background indicates vascular leakage of FITC-dextran in the SCF-injected eyes. (B) Vascular leakage of FITC-dextran was quantified by assessing the fluorescence intensity of digital images taken from retinal whole mounts perfused with FITC-dextran. Fluorescence intensity is expressed as the fold increase ± SEM with respect to PBS control (*P < 0.05 vs. PBS, #P < 0.05 vs. SCF only, n = 4). (C) Immunofluorescence staining was performed using anti-p-cKit (green) and anti-CD31 IgGs (red) in retina sections. Arrowheads indicate double labeling with anti-CD31 IgGs with anti-p-cKit IgG. DAPI-stained nuclei are shown in blue. (D) Immunofluorescence staining in retinal whole mounts was performed using anti–VE-cadherin IgG (green) and Alexa Fluor 594-conjugated isolectin GS-IB4 (red). The arrowhead in the enlarged inset shows loss of VE-cadherin in the retinal vasculature. Scale bars in (A) to (D) are 100 μm.
cKit-Mutant Mice Are Resistant to the Diabetes-Induced Development of Retinal Vascular Hyperpermeability
To provide functional evidence for the critical role of cKit in diabetes-induced retinal vascular leakage, we induced DM in wild-type (WT) mice and cKit-mutant mice (Kitw-sh/w-sh) via intraperitoneal administration of STZ. Kitw-sh/w-sh mice carry an inversion mutation in the 5′ transcriptional regulatory elements of cKit and exhibit a marked reduction in cKit signaling and gene expression in endothelial cells.8,9 In the current study, STZ injection markedly elevated blood glucose levels in both WT and Kitw-sh/w-sh mice, with similar increases observed for both WT and Kitw-sh/w-sh mice (Fig. 3A). Two weeks after DM onset, WT DM mice show a significant increase in retinal vascular leakage and a loss of junctional VE-cadherin staining in the retinal vasculature (Figs. 3B, 3C). In contrast, Kitw-sh/w-sh DM mice failed to demonstrate enhanced retinal vascular leakage or altered expression of VE-cadherin. These observations further confirm the critical contribution of cKit to the progression of diabetes-induced retinal vascular hyperpermeability in the STZ-induced diabetic animal model. 
Figure 3
 
cKit-mutant mice are resistant to the development of diabetes-induced retinal vascular hyperpermeability. (A) Blood glucose levels in WT and Kitw-sh/w-sh mice were measured at 2 weeks after intraperitoneal injection of STZ (DM) or PBS (non-DM). (B) Representative images of FITC-dextran–perfused retinas of Kitw-sh/w-sh and WT mice provide an indication of the extent of retinal vascular hyperpermeability in non-DM and DM mice. Fluorescence intensity is expressed as the fold increase ± SEM with respect to the non-DM WT control (*P < 0.05, n = 4). Scale bar: 200 μm. (C) Representative images of retinal whole mounts stained with anti-VE-cadherin IgG (green) and Alexa Fluor 594-conjugated isolectin GS-IB4 (red). Arrowheads in the enlarged inset show the focal loss of VE-cadherin in the retinal vasculature. Scale bar: 100 μm.
Figure 3
 
cKit-mutant mice are resistant to the development of diabetes-induced retinal vascular hyperpermeability. (A) Blood glucose levels in WT and Kitw-sh/w-sh mice were measured at 2 weeks after intraperitoneal injection of STZ (DM) or PBS (non-DM). (B) Representative images of FITC-dextran–perfused retinas of Kitw-sh/w-sh and WT mice provide an indication of the extent of retinal vascular hyperpermeability in non-DM and DM mice. Fluorescence intensity is expressed as the fold increase ± SEM with respect to the non-DM WT control (*P < 0.05, n = 4). Scale bar: 200 μm. (C) Representative images of retinal whole mounts stained with anti-VE-cadherin IgG (green) and Alexa Fluor 594-conjugated isolectin GS-IB4 (red). Arrowheads in the enlarged inset show the focal loss of VE-cadherin in the retinal vasculature. Scale bar: 100 μm.
Intravitreal Injection of Masitinib Prevents the Development of Retinal Vascular Hyperpermeability in STZ-Induced Diabetic Rats
Next, we investigated whether masitinib could prevent the development of diabetes-induced retinal vascular hyperpermeability. Immediately after STZ-induced DM onset, the rats received a single intravitreal injection of either masitinib or DMSO. Two weeks after the STZ injection, a time point when SCF and cKit expression is elevated in the retina (Supplementary Fig. S1), substantial perivascular leakage of FITC-dextran and a diffuse background were observed in the retinas of DMSO-treated DM rats (Fig. 4A). However, the retinas of masitinib-treated DM rats demonstrate a substantial decrease in vascular permeability. The retinal vascular permeability of DM rats treated with masitinib is low with a level similar to the non-DM controls (Fig. 4B). Intravitreal injection of masitinib does not affect blood glucose levels or body weights (Supplementary Fig. S2). Immunohistochemical analysis revealed that p-eNOS signal increased in the retinal vasculature of STZ-induced DM rats was abolished by administration of masitinib (Fig. 4C). 
Figure 4
 
Intravitreal injection of masitinib prevents the development of retinal vascular hyperpermeability in STZ-induced diabetic rats. (A) Representative images of FITC-dextran–perfused retina whole mounts from non-DM and DM rats. The DM rats received a single intravitreal injection of the indicated amount of masitinib or DMSO on the same day as DM onset. Vascular leakage of FITC-dextran was measured 2 weeks later. Scale bar: 200 μm. (B) Fluorescence intensity in the retina is expressed as the fold change ± SEM with respect to the non-DM control (*P < 0.05 vs. non-DM, #P < 0.05 vs. DMSO-treated DM, n = 6). (C) Immunofluorescence staining was performed using anti-p-eNOS (green) and anti-CD31 IgGs (red) in retina sections. The DM rats received a single intravitreal injection of masitinib (50 μg) or DMSO on the same day as DM onset. The arrowheads indicate the double labeling of CD31 with p-eNOS. The nuclei are shown in blue (DAPI). Scale bar: 50 μm.
Figure 4
 
Intravitreal injection of masitinib prevents the development of retinal vascular hyperpermeability in STZ-induced diabetic rats. (A) Representative images of FITC-dextran–perfused retina whole mounts from non-DM and DM rats. The DM rats received a single intravitreal injection of the indicated amount of masitinib or DMSO on the same day as DM onset. Vascular leakage of FITC-dextran was measured 2 weeks later. Scale bar: 200 μm. (B) Fluorescence intensity in the retina is expressed as the fold change ± SEM with respect to the non-DM control (*P < 0.05 vs. non-DM, #P < 0.05 vs. DMSO-treated DM, n = 6). (C) Immunofluorescence staining was performed using anti-p-eNOS (green) and anti-CD31 IgGs (red) in retina sections. The DM rats received a single intravitreal injection of masitinib (50 μg) or DMSO on the same day as DM onset. The arrowheads indicate the double labeling of CD31 with p-eNOS. The nuclei are shown in blue (DAPI). Scale bar: 50 μm.
Discussion
The present study demonstrated that SCF and cKit might be highly implicated in the diabetes-induced BRB breakdown. In primary human retinal endothelial cells, SCF augmented paracellular permeability and induced the disassembly of endothelial adherens junctions. In addition, mutant mice with a genetic deficit in cKit expression did not exhibit the diabetes-induced increase in retinal vascular leakage. Because Kitw-sh/w-sh mice used in the present study were mast cell deficient, the possibility that mast cells may influence retinal vascular hyperpermeability cannot be ruled out. However, our previous work showed that most cKit-positive cells in the diabetic retinas were CD31-positive vascular endothelial cells.4 Other reports also showed that CD45-positive leukocytes present in the retinal tissues of diabetic mice were mainly CD11b-positive monocytes and macrophages.10 Furthermore, no mast cells were observed in retinal specimens of human patients with diabetic retinopathy.11 These findings indicate that mast cells are unlikely to participate in the development of diabetes-induced hyperpermeable retinal vasculature. 
Considering the critical actions of SCF and cKit in BRB breakdown during the early stages of STZ-induced diabetes in rodent models, we hypothesized that cKit inhibitors might therefore have a therapeutic impact on diabetes-induced retinal vascular leakage. In support of this idea, the present study demonstrated that a single intravitreal injection of masitinib, a potent and selective cKit inhibitor, on the same day as DM onset was sufficient to prevent the development of retinal hyperpermeable vasculature. Because masitinib attenuates inflammatory pathogenesis and tumor proliferation through inhibition of platelet-derived growth factor receptor (PDGFR) signaling and inflammatory action of cKit-positive mast cells, potential off-target effect of masitinib should be considered to account for the reduced retinal permeability in masitinib-treated eyes. However, PDGF/PDGFR signaling plays a crucial role in recruiting pericytes into retinal vascular endothelial cells and maintaining the BRB function. Several reports have shown that loss of PDGFR signaling markedly increases retinal vascular permeability.12,13 Given the protective effect of PDGFR signaling on the BRB and minor contribution of mast cells to DM-induced breakdown of BRB, off-target effects of masitinib would be negligible. It is reasonable that masitinib predominantly acts to prevent the DM-induced development of retinal hyperpermeable vasculature through inhibition of endothelial SCF/cKIT signaling. 
In the past decade, new pharmacologic strategies have been proposed to inhibit retinal vascular leakage in many ocular diseases. For example, VEGF blocking agents have demonstrated promising efficacy in halting disease progression and improving the vision of patients, which is at least partly due to the inhibition of VEGF-induced increases in vascular permeability.1416 In clinical studies for the treatment of neovascular age-related macular degeneration, intravitreal anti-VEGF agents have shown great safety profiles with no systemic adverse effects attributable to VEGF inhibition.17 On the other hand, there has been increasing concern that patients with diabetes might have a higher incidence of myocardial infarction, stroke, or vascular death after intravitreal injection of anti-VEGF agents.18 Because patients with diabetes are at increased risk for developing cardiovascular diseases due to diabetes-mediated endothelial dysfunction and impaired angiogenic activity, systemic entry of anti-VEGF agents after intravitreal injection could compromise the critical vascular regeneration response to ischemic events in such patients, whereas it would not significantly affect vascular function in nondiabetic patients.19 In this regard, co-administration of anti-VEGF agents with cKit inhibitors such as masitinib might lead to reduction in the dose of anti-VEGF agents and a reduction in their adverse effects, which might be beneficial in the case of patients with high cardiovascular risk. Moreover, the mesylate salt of masitinib is orally bioavailable. Unlike anti-VEGF agents whose target, VEGF, is present in the vitreous humor, cKit is expressed on retinal vascular endothelium that is easily accessible to drugs circulating in the blood. These suggest that masitinib mesylate might be systemically delivered to inhibit diabetes-induced retinal vascular leakage. Although oral administration brings about additional concerns such as drug-drug interactions or possible systemic toxicity of the drug, the oral administration route would be advantageous in circumventing complications of intravitreal injection, such as endophthalmitis, rhegmatogenous retinal detachment, hemorrhage, or IOP elevation. 
In conclusion, the present study demonstrated the substantial protective effect of masitinib against diabetes-induced BRB breakdown. Although the experimental evidence in animal models cannot be directly extrapolated to the mechanism of BRB breakdown in humans, recent reports demonstrating increased expression of SCF and cKit in patients with diabetic retinopathy suggest that inhibition of SCF signaling using cKit inhibitors such as masitinib might provide a novel therapeutic opportunity for the prevention of diabetes-induced retinal vascular leakage. 
Acknowledgments
The authors thank Joon Hee Cho (Department of Ophthalmology, Seoul National University Bundang Hospital) for his assistance with analyzing images of optical coherence tomography. 
Supported by National Research Foundation of Korea grants funded by the Korea government (MSIP) (2013R1A2A2A04016796, 2015R1A2A1A15052509, 2015R1D1A1A02061724). 
Disclosure: S.R. Kim, None; J.-E. Im, None; J.H. Jeong, None; J.Y. Kim, None; J.T. Kim, None; S.J. Woo, None; J.-H. Sung, None; S.G. Park, None; W. Suh, None 
References
Antonetti DA, Barber AJ, Khin S, Lieth E, Tarbell JM, Gardner TW. Vascular permeability in experimental diabetes is associated with reduced endothelial occludin content: vascular endothelial growth factor decreases occludin in retinal endothelial cells. Penn State Retina Research Group. Diabetes. 1998; 47: 1953–1959.
Barber AJ, Antonetti DA, Gardner TW. Altered expression of retinal occludin and glial fibrillary acidic protein in experimental diabetes. The Penn State Retina Research Group. Invest Ophthalmol Vis Sci. 2000; 41: 3561–3568.
Navaratna D, McGuire PG, Menicucci G, Das A. Proteolytic degradation of VE-cadherin alters the blood-retinal barrier in diabetes. Diabetes. 2007; 56: 2380–2387.
Kim JY, Choi JS, Song SH, et al. Stem cell factor is a potent endothelial permeability factor. Arterioscler Thromb Vasc Biol. 2014; 34: 1459–1467.
Abu El-Asrar AM, Struyf S, Opdenakker G, Van Damme J, Geboes K. Expression of stem cell factor/c-kit signaling pathway components in diabetic fibrovascular epiretinal membranes. Mol Vis. 2010; 16: 1098–1107.
Abu El-Asrar AM, Nawaz MI, Kangave D, Mairaj Siddiquei M, Geboes K. Angiogenic and vasculogenic factors in the vitreous from patients with proliferative diabetic retinopathy. J Diabetes Res. 2013; 2013: 539658.
Dubreuil P, Letard S, Ciufolini M, et al. Masitinib (AB1010), a potent and selective tyrosine kinase inhibitor targeting KIT. PLoS One. 2009; 4: e7258.
Duttlinger R, Manova K, Berrozpe G, et al. The Wsh and Ph mutations affect the c-kit expression profile: c-kit misexpression in embryogenesis impairs melanogenesis in Wsh and Ph mutant mice. Proc Natl Acad Sci U S A. 1995; 92: 3754–3758.
Fang S, Wei J, Pentinmikko N, Leinonen H, Salven P. Generation of functional blood vessels from a single c-kit+ adult vascular endothelial stem cell. PLoS Biol. 2012; 10: e1001407.
Serra AM, Waddell J, Manivannan A, Xu H, Cotter M, Forrester JV. CD11b+ bone marrow-derived monocytes are the major leukocyte subset responsible for retinal capillary leukostasis in experimental diabetes in mouse and express high levels of CCR5 in the circulation. Am J Pathol. 2012; 181: 719–727.
Lopez R, Rand LI, Zetter BR. Absence of mast cells in diabetic retinopathy. Microvasc Res. 1982; 24: 87–93.
Enge M, Bjarnegard M, Gerhardt H, et al. Endothelium-specific platelet-derived growth factor-B ablation mimics diabetic retinopathy. EMBO J. 2002; 21: 4307–4316.
Geraldes P, Hiraoka-Yamamoto J, Matsumoto M, et al. Activation of PKC-delta and SHP-1 by hyperglycemia causes vascular cell apoptosis and diabetic retinopathy. Nat Med. 2009; 15: 1298–1306.
Aiello LP, Avery RL, Arrigg PG, et al. Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N Engl J Med. 1994; 331: 1480–1487.
Mitchell P, Bandello F, Schmidt-Erfurth U, et al. The RESTORE study: ranibizumab monotherapy or combined with laser versus laser monotherapy for diabetic macular edema. Ophthalmology. 2011; 118: 615–625.
Nguyen QD, Brown DM, Marcus DM, et al. Ranibizumab for diabetic macular edema: results from 2 phase III randomized trials: RISE and RIDE. Ophthalmology. 2012; 119: 789–801.
Falavarjani KG, Nguyen QD. Adverse events and complications associated with intravitreal injection of anti-VEGF agents: a review of literature. Eye (Lond). 2013; 27: 787–794.
Simo R, Hernandez C. Intravitreous anti-VEGF for diabetic retinopathy: hopes and fears for a new therapeutic strategy. Diabetologia. 2008; 51: 1574–1580.
Grundy SM, Benjamin IJ, Burke GL, et al. Diabetes and cardiovascular disease: a statement for healthcare professionals from the American Heart Association. Circulation. 1999; 100: 1134–1146.
Figure 1
 
Masitinib blocks SCF-induced increase in paracellular permeability of human retinal endothelial cells in vitro. (A, B) Masitinib inhibits SCF-induced hyperpermeability of HRMECs. Endothelial permeability was determined by measuring the passage of FITC-dextran (A) and TEER (B) in an HRMEC monolayer that was pretreated with masitinib (0.1 ∼ 10 μM in [A], 1 μM in [B]) for 30 minutes before stimulation with rhSCF (50 ng/mL). Fluorescein isothiocyanate-dextran permeability is expressed as the fold increase with respect to PBS control (n = 4). Transendothelial electrical resistance values were calculated by subtracting the background electrical resistance and multiplying by the surface area of the filter (n = 5). (C, D) Masitinib inhibits SCF-induced activation of cKit (C) and eNOS (D) in HRMECs. The HRMECs treated with masitinib (1 μM) or untreated cells were stimulated with rhSCF (50 ng/mL) or PBS for 20 minutes. Phosphorylation of cKit or eNOS was assessed by Western blotting. (E) Masitinib inhibits SCF-induced internalization of VE-cadherin. The HRMECs were pretreated with masitinib (1 μM) for 30 minutes before stimulation with rhSCF (50 ng/mL). The arrows in the left image indicate the disappearance of VE-cadherin at endothelial junctions that were positively stained for anti-ZO-1 IgGs (red). The arrows in the right image indicate VE-cadherin (green) internalization in endosomes that were positively stained for EEA1 (red). Internalization of endogenous VE-cadherin was quantified based on the percentage of HRMECs with intracellular acid-resistant vesicles (three independent experiments). Scale bar: 20 μm. All data are presented as mean ± SEM (*P < 0.05 vs. PBS, #P < 0.05 vs. SCF only).
Figure 1
 
Masitinib blocks SCF-induced increase in paracellular permeability of human retinal endothelial cells in vitro. (A, B) Masitinib inhibits SCF-induced hyperpermeability of HRMECs. Endothelial permeability was determined by measuring the passage of FITC-dextran (A) and TEER (B) in an HRMEC monolayer that was pretreated with masitinib (0.1 ∼ 10 μM in [A], 1 μM in [B]) for 30 minutes before stimulation with rhSCF (50 ng/mL). Fluorescein isothiocyanate-dextran permeability is expressed as the fold increase with respect to PBS control (n = 4). Transendothelial electrical resistance values were calculated by subtracting the background electrical resistance and multiplying by the surface area of the filter (n = 5). (C, D) Masitinib inhibits SCF-induced activation of cKit (C) and eNOS (D) in HRMECs. The HRMECs treated with masitinib (1 μM) or untreated cells were stimulated with rhSCF (50 ng/mL) or PBS for 20 minutes. Phosphorylation of cKit or eNOS was assessed by Western blotting. (E) Masitinib inhibits SCF-induced internalization of VE-cadherin. The HRMECs were pretreated with masitinib (1 μM) for 30 minutes before stimulation with rhSCF (50 ng/mL). The arrows in the left image indicate the disappearance of VE-cadherin at endothelial junctions that were positively stained for anti-ZO-1 IgGs (red). The arrows in the right image indicate VE-cadherin (green) internalization in endosomes that were positively stained for EEA1 (red). Internalization of endogenous VE-cadherin was quantified based on the percentage of HRMECs with intracellular acid-resistant vesicles (three independent experiments). Scale bar: 20 μm. All data are presented as mean ± SEM (*P < 0.05 vs. PBS, #P < 0.05 vs. SCF only).
Figure 2
 
Masitinib blocks SCF-induced retinal vascular leakage in vivo. Masitinib (5 mg/kg) or DMSO was intraperitoneally injected into mice at 1 hour before intravitreal injection of recombinant mouse SCF (50 ng in 1 μL of PBS) or an equivalent volume of PBS. (A) Representative images of FITC-dextran–perfused retina flat mounts. The bright green fluorescent background indicates vascular leakage of FITC-dextran in the SCF-injected eyes. (B) Vascular leakage of FITC-dextran was quantified by assessing the fluorescence intensity of digital images taken from retinal whole mounts perfused with FITC-dextran. Fluorescence intensity is expressed as the fold increase ± SEM with respect to PBS control (*P < 0.05 vs. PBS, #P < 0.05 vs. SCF only, n = 4). (C) Immunofluorescence staining was performed using anti-p-cKit (green) and anti-CD31 IgGs (red) in retina sections. Arrowheads indicate double labeling with anti-CD31 IgGs with anti-p-cKit IgG. DAPI-stained nuclei are shown in blue. (D) Immunofluorescence staining in retinal whole mounts was performed using anti–VE-cadherin IgG (green) and Alexa Fluor 594-conjugated isolectin GS-IB4 (red). The arrowhead in the enlarged inset shows loss of VE-cadherin in the retinal vasculature. Scale bars in (A) to (D) are 100 μm.
Figure 2
 
Masitinib blocks SCF-induced retinal vascular leakage in vivo. Masitinib (5 mg/kg) or DMSO was intraperitoneally injected into mice at 1 hour before intravitreal injection of recombinant mouse SCF (50 ng in 1 μL of PBS) or an equivalent volume of PBS. (A) Representative images of FITC-dextran–perfused retina flat mounts. The bright green fluorescent background indicates vascular leakage of FITC-dextran in the SCF-injected eyes. (B) Vascular leakage of FITC-dextran was quantified by assessing the fluorescence intensity of digital images taken from retinal whole mounts perfused with FITC-dextran. Fluorescence intensity is expressed as the fold increase ± SEM with respect to PBS control (*P < 0.05 vs. PBS, #P < 0.05 vs. SCF only, n = 4). (C) Immunofluorescence staining was performed using anti-p-cKit (green) and anti-CD31 IgGs (red) in retina sections. Arrowheads indicate double labeling with anti-CD31 IgGs with anti-p-cKit IgG. DAPI-stained nuclei are shown in blue. (D) Immunofluorescence staining in retinal whole mounts was performed using anti–VE-cadherin IgG (green) and Alexa Fluor 594-conjugated isolectin GS-IB4 (red). The arrowhead in the enlarged inset shows loss of VE-cadherin in the retinal vasculature. Scale bars in (A) to (D) are 100 μm.
Figure 3
 
cKit-mutant mice are resistant to the development of diabetes-induced retinal vascular hyperpermeability. (A) Blood glucose levels in WT and Kitw-sh/w-sh mice were measured at 2 weeks after intraperitoneal injection of STZ (DM) or PBS (non-DM). (B) Representative images of FITC-dextran–perfused retinas of Kitw-sh/w-sh and WT mice provide an indication of the extent of retinal vascular hyperpermeability in non-DM and DM mice. Fluorescence intensity is expressed as the fold increase ± SEM with respect to the non-DM WT control (*P < 0.05, n = 4). Scale bar: 200 μm. (C) Representative images of retinal whole mounts stained with anti-VE-cadherin IgG (green) and Alexa Fluor 594-conjugated isolectin GS-IB4 (red). Arrowheads in the enlarged inset show the focal loss of VE-cadherin in the retinal vasculature. Scale bar: 100 μm.
Figure 3
 
cKit-mutant mice are resistant to the development of diabetes-induced retinal vascular hyperpermeability. (A) Blood glucose levels in WT and Kitw-sh/w-sh mice were measured at 2 weeks after intraperitoneal injection of STZ (DM) or PBS (non-DM). (B) Representative images of FITC-dextran–perfused retinas of Kitw-sh/w-sh and WT mice provide an indication of the extent of retinal vascular hyperpermeability in non-DM and DM mice. Fluorescence intensity is expressed as the fold increase ± SEM with respect to the non-DM WT control (*P < 0.05, n = 4). Scale bar: 200 μm. (C) Representative images of retinal whole mounts stained with anti-VE-cadherin IgG (green) and Alexa Fluor 594-conjugated isolectin GS-IB4 (red). Arrowheads in the enlarged inset show the focal loss of VE-cadherin in the retinal vasculature. Scale bar: 100 μm.
Figure 4
 
Intravitreal injection of masitinib prevents the development of retinal vascular hyperpermeability in STZ-induced diabetic rats. (A) Representative images of FITC-dextran–perfused retina whole mounts from non-DM and DM rats. The DM rats received a single intravitreal injection of the indicated amount of masitinib or DMSO on the same day as DM onset. Vascular leakage of FITC-dextran was measured 2 weeks later. Scale bar: 200 μm. (B) Fluorescence intensity in the retina is expressed as the fold change ± SEM with respect to the non-DM control (*P < 0.05 vs. non-DM, #P < 0.05 vs. DMSO-treated DM, n = 6). (C) Immunofluorescence staining was performed using anti-p-eNOS (green) and anti-CD31 IgGs (red) in retina sections. The DM rats received a single intravitreal injection of masitinib (50 μg) or DMSO on the same day as DM onset. The arrowheads indicate the double labeling of CD31 with p-eNOS. The nuclei are shown in blue (DAPI). Scale bar: 50 μm.
Figure 4
 
Intravitreal injection of masitinib prevents the development of retinal vascular hyperpermeability in STZ-induced diabetic rats. (A) Representative images of FITC-dextran–perfused retina whole mounts from non-DM and DM rats. The DM rats received a single intravitreal injection of the indicated amount of masitinib or DMSO on the same day as DM onset. Vascular leakage of FITC-dextran was measured 2 weeks later. Scale bar: 200 μm. (B) Fluorescence intensity in the retina is expressed as the fold change ± SEM with respect to the non-DM control (*P < 0.05 vs. non-DM, #P < 0.05 vs. DMSO-treated DM, n = 6). (C) Immunofluorescence staining was performed using anti-p-eNOS (green) and anti-CD31 IgGs (red) in retina sections. The DM rats received a single intravitreal injection of masitinib (50 μg) or DMSO on the same day as DM onset. The arrowheads indicate the double labeling of CD31 with p-eNOS. The nuclei are shown in blue (DAPI). Scale bar: 50 μm.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×