Calcium dobesilate reduces VEGF signaling by interfering with heparan sulfate binding site and protects from vascular complications in diabetic mice

Abstract

Inhibiting vascular endothelial growth factor (VEGF) is a therapeutic option in diabetic microangiopathy. However, VEGF is needed at physiological concentrations to maintain glomerular integrity; complete VEGF blockade has deleterious effects on glomerular structure and function. Anti-VEGF therapy in diabetes raises the challenge of reducing VEGF-induced pathology without accelerating endothelial cell injury. Heparan sulfate (HS) act as a co-receptor for VEGF. Calcium dobesilate (CaD) is a small molecule with vasoprotective properties that has been used for the treatment of diabetic microangiopathy. Preliminary evidence suggests that CaD interferes with HS binding sites of fibroblast growth factor. We therefore tested the hypotheses that (1) CaD inhibits VEGF signaling in endothelial cells, (2) that this effect is mediated via interference between CaD and HS, and (3) that CaD ameliorates diabetic nephropathy in a streptozotocin-induced diabetic mouse model by VEGF inhibition. We found that CaD significantly inhibited VEGF165-induced endothelial cell migration, proliferation, and permeability. CaD significantly inhibited VEGF165-induced phosphorylation of VEGFR-2 and suppressed the activity of VEGFR-2 mediated signaling cascades. The effects of CaD in vitro were abrogated by heparin, suggesting the involvement of heparin-like domain in the interaction with CaD. In addition, VEGF121, an isoform which does not bind to heparin, was not inhibited by CaD. Using the proximity ligation approach, we detected inhibition of interaction in situ between HS and VEGF and between VEGF and VEGFR-2. Moreover, CaD reduced VEGF signaling in mice diabetic kidneys and ameliorated diabetic nephropathy and neuropathy, suggesting CaD as a VEGF inhibitor without the negative effects of complete VEGF blockade and therefore could be useful as a strategy in treating diabetic nephropathy.

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Fig 1. Calcium dobesilate (CaD) inhibits VEGF-induced VEGFR-2 activation.

VEGF165 (25 ng/mL) was premixed and incubated with: (A) various CaD concentrations (6–100 μM) or (B) 50–100 μM CaD for 1 h, before exposure to HUVECs for 2 min. (C) HUVECs were incubated with 50 and 100μM CaD for 60 min with three subsequent washing steps with warm medium before stimulation with VEGF165 for 2 min. Western blot analysis was performed using anti-phospho-VEGFR-2 antibody and total VEGFR2 was used as a loading control after membrane stripping. Each bar represents the mean ± SD (n = 3). *P<0.05, ***P<0.001 versus VEGF-treated HUVECs.

Fig 2. Calcium dobesilate (CaD) inhibits VEGF-induced MEK/ERK MAP kinase activation.

CaD at 50, 100 and 200μM was incubated with VEGF (25 ng/mL) for 60 min before exposure to HUVECs for 15 min. Phosphorylated ERK and MEK1/2 was determined by Western blot as described in Figure 2. Data are expressed as mean ± SD (n = 3). * p

Fig 3. Inhibition of endothelial cell proliferation, invasion and migration by CaD.

HUVECs were plated in 96-well plates, allowed to attach overnight, and then cultured for 24 h-48 h with CaD in the presence of VEGF165 (25 ng/ml)-CaD mixture (A) at the indicated concentration. The proliferation was measured as described in the Materials and Methods section. (B) Serum-starved HUVECs were allowed to migrate through trans-well membranes towards a vehicle, VEGF165 (25 ng/ml) and/or CaD at the indicated concentration for 24 h. Cells that had migrated to the underside of the membrane were processed for calcein-AM staining as described in the materials and methods section. (C) A monolayer of HUVECs was scratched and fresh medium containing vehicle, VEGF and/or CaD was then added. After 14 h, migration distance of HUVECs was quantified. Original magnification, 40x. (D) HUVECs were incubated with vehicle, VEGF165 (25 ng/ml) and CaD for 15 min. Cells were stained with phalloidin-Alexa Fluor 488 (left images) and DAPI (middle images). Merged view is represented in the right images. Original magnification, 40x. Scale bars represent 5 μm. The results shown are the means ± SD of four independent experiments conducted in triplicate. *P<0.05, **P<0.01, versus VEGF165-treated HUVECs.

Fig 4. CaD prevents the decrease in tight junction protein levels and the increase in HUVECs permeability induced by VEGF165.

CaD at 50 and 100 μM was incubated with VEGF (25 ng/mL) for 60 min before exposure to HUVECs for 2 h. Tight junction protein levels were determined by Western blotting (A), immunofluorescence (B) and endothelial cells permeability was by FITC-Dextran assay (C) as described in material and methods section. Representative images are shown and each bar indicates the relative expression to that of control or VEGF165 alone. Scale bars represent 25 μm and the average of three independent experiments ± SD. *P<0.05, ***P<0.001.

Fig 5. Effect of heparin and heparanase (HPSE) treatment on the inhibitory effect of CaD.

(A) VEGF165 (25 ng/mL), CaD (100 μM) and heparin (10 μg/ml) were pre-mixed and incubated at 37°C for 60 min prior to HUVECs stimulation for 2 min. (B) HUVECs were stimulated with VEGF121 (25 ng/ml)-CaD mixture as described in Fig 1 for 2 min. (C) HUVECs were digested with HPSE (0.5 μg/ml). After 30 min, each dish was washed and then stimulated with 25 ng/ml VEGF165 alone or VEGF-CaD (100 μM) mixture for 2 min. Samples were subjected to gel electrophoresis, and Western blotting was performed as described. Each bar indicates the relative phosphorylation to that of VEGF alone or VEGF-HPSE and the average of at least three independent experiments ± SD. *P<0.05, **P<0.01, ***P<0.001.

Fig 6. Effect of CaD on VEGF165 binding to VEGFR-1/2, heparin and HS.

(A) bt-VEGF165 binding assay to immobilized VEGR-1/2 in the presence of increasing concentrations of CaD. Data are expressed as percentage of control (no CaD) ± SD in binding from four independent experiments; *p < 0.05 vs. no CaD. (B) bt-VEGF165 binding in the absence or presence of 1 μg/ml heparin or CaD (100 μM) to VEGFR-1 and VEGFR-2. Data are expressed as mean OD at 450 nm ± SD from four independent experiments; *p < 0.05 vs. no heparin or CaD (Ctrl). (C) Representative images were obtained by proximity ligation assay. Cells were fixed and incubated with antibodies to human VEGF (goat IgG) and to human VEGFR-2 (rabbit IgG) (upper panel) or to VEGF and heparin sulphate (mAb 10E4) (middle panel) or to VEGFR-2 and HS (lower panel) followed by proximity ligation assay reagents. Each red dot indicates a protein interaction. Nuclei are shown in blue. The antibody (Ab) control panel represents cells incubated with the corresponding IgG. (D) Quantified data presented as red dots (signal intensity)/cell. The error bar represents SD. n = 3 separate experiments *P<0.05, ***P<0.001 relative to that of VEGF alone. All images were taken with a Leica DMI3000 B microscopy, scale bar 100 μm and analyzed with NIH ImageJ software.

Fig 7. Effect of CaD treatment on diabetic complications in vivo.

Type I diabetes was induced in Sv129 mice and thereafter the diabetic mice were daily treated with vehicle or CaD (100mg/kg) or enalapril (30mg/kg). CaD treatment had no effect on hyperglycemia (A and B) and body weight (C). CaD treatment protects kidney function in diabetic mice as reflected by serum creatinine levels (D) and albuminuria (E). CaD treatment reduces diabetic neuropathy as reflected by sensory nerve conduction velocity measurements (F). The results are expressed as the mean+SEM (n = 20 mice for each group). *p

Fig 8. CaD inhibits VEGFR-2 phosphorylation and signaling in STZ diabetic mice.

STZ mice were daily treated with vehicle, CaD or Enalapril and sacrificed at week 12 for the analysis. Non diabetic mice were used as control. Kidneys were isolated and homogenized as described in materials and methods. Phosphorylated VEGFR-2 (A) in the kidney lysates was determined by ELISA mean ± SD of 5–6 animals. Phosphorylated ERK1/2 (B) was analyzed by Western blot as described in Fig 1. Vinculin was used as a loading control. Phosphorylated P38 (C) was analyzed by immunohistochemistry as described in the materials and methods section. Autofluorescence is shown in green, scale bar 100 μM. Representative image for n = 6/condition is depicted. *P

Fig 9. CaD prevents the increase in diabetes-induced renal inflammation.

The mRNA levels were assessed by real time PCR and normalized to Actin (A). VEGF production was determined by VEGF164-188 ELISA according to the manufacturer’s instructions (B). Data represent the mean ± SD of 4–8 animals. * p < 0.05. Significantly different from vehicle treated STZ mice. Immunohistochemistry for F4/80 (red) in kidney cross sections of non-diabetic (control), diabetic vehicle treated (STZ), enalapril and CaD treated STZ mice (C). Autofluorescence is shown in green, scale bar 100 μm. Images are exemplary for n = 5/condition. Infiltrating cells were quantitatively analyzed by scoring five areas in each kidney section for F4/80-positive cells. Each bar represents the mean ± SD.* p < 0.05, **P<0.01, *** p < 0.001. Significantly different from vehicle treated STZ mice.

Fig 10. Postulated model of interactions between VEGF165, VEGFR-2, and CaD.

VEGF165 binds to its co-receptor heparin sulfates (HS) of the endothelial glycocalyx with a specific binding site (left box) which stabilizes the VEGF-VEGF-R binding leading to phosphorylation of VEGFR-2 receptor, intracellular signaling and cell activation (middle box). CaD interacts with the heparin-binding domain of the VEGF165 (right box), thereby displacing HS from its binding site, and decreases VEGF-induced intracellular signaling. CaD also regulates VEGF165 activity by participating in the formation of unstable VEGF-VEGFR-2 complex.