Calcium dobesilate inhibits the alterations in tight junction proteins and leukocyte adhesion to retinal endothelial cells induced by diabetes

Abstract

Objective: Calcium dobesilate (CaD) has been used in the treatment of diabetic retinopathy in the last decades, but its mechanisms of action are not elucidated. CaD is able to correct the excessive vascular permeability in the retina of diabetic patients and in experimental diabetes. We investigated the molecular and cellular mechanisms underlying the protective effects of CaD against the increase in blood-retinal barrier (BRB) permeability induced by diabetes.

Research design and methods: Wistar rats were divided into three groups: controls, streptozotocin-induced diabetic rats, and diabetic rats treated with CaD. The BRB breakdown was evaluated using Evans blue. The content or distribution of tight junction proteins (occludin, claudin-5, and zonula occluden-1 [ZO-1]), intercellular adhesion molecule-1 (ICAM-1), and p38 mitogen-activated protein kinase (p38 MAPK) was evaluated by Western blotting and immunohistochemistry. Leukocyte adhesion was evaluated in retinal vessels and in vitro. Oxidative stress was evaluated by the detection of oxidized carbonyls and tyrosine nitration. NF-κB activation was measured by enzyme-linked immunosorbent assay.

Results: Diabetes increased the BRB permeability and retinal thickness. Diabetes also decreased occludin and claudin-5 levels and altered the distribution of ZO-1 and occludin in retinal vessels. These changes were inhibited by CaD treatment. CaD also inhibited the increase in leukocyte adhesion to retinal vessels or endothelial cells and in ICAM-1 levels, induced by diabetes or elevated glucose. Moreover, CaD decreased oxidative stress and p38 MAPK and NF-κB activation caused by diabetes.

Conclusions: CaD prevents the BRB breakdown induced by diabetes, by restoring tight junction protein levels and organization and decreasing leukocyte adhesion to retinal vessels. The protective effects of CaD are likely to involve the inhibition of p38 MAPK and NF-κB activation, possibly through the inhibition of oxidative/nitrosative stress.

Figures

FIG. 1.

Diabetes increases BRB permeability: protective effect of CaD. A: Quantitative measure of BRB permeability by quantification of extravasated Evans blue. Data are presented as μg of Evans blue per retina wet wt (g) and represent the mean ± SEM of 7–10 animals. **P < 0.01, significantly different from control; ANOVA (one-way) followed by Dunnett post hoc test. #P < 0.05, significantly different from diabetic; ANOVA (one-way) followed by Bonferroni post hoc test. B: Representative images showing Evans blue fluorescence, allowing the detection of leaking sites (arrows) in retinal vessels. In the retina of control animals, Evans blue fluorescence is limited to the blood vessels, while in diabetic retinas, the dye leaks out of the vessels to the retinal tissue. CaD treatment prevents the leakage of Evans blue. Magnification: 100×; bar 200 μm. (A high-quality color representation of this figure is available in the online issue.)

FIG. 2.

CaD prevents the decrease in occludin and claudin-5 protein levels in the rat retinas induced by diabetes. ZO-1 levels were not significantly changed. Tight junction protein levels were assessed by Western blotting. A representative Western blot is shown above each graph. Data are presented as percentage of control and represent the mean ± SEM of 7–9 animals. *P < 0.05, **P < 0.01, significantly different from control; ANOVA (one-way) followed by Dunnett post hoc test.

FIG. 3.

CaD prevents the decrease in occludin and claudin-5 immunoreactivity and the changes in occludin and ZO-1 distribution (arrows), in rat retinal vessels induced by diabetes. Magnifications: 200× (bar 100 μm) and 400× (bar 50 μm). (A high-quality color representation of this figure is available in the online issue.)

FIG. 4.

Diabetes and elevated glucose increase the number of leukocytes adhering to retinal vessels and retinal endothelial cells and the content of ICAM-1: protective effect of CaD. A: Quantification of leukocytes adhering to retinal vessels. Data are presented as number of adherent leukocytes to retinal vessels per rat (two retinas) and represent the mean ± SEM of seven animals. B: Quantification of leukocyte adhesion to retinal endothelial cells (TR-iBRB2 cell line) using a fluorometric assay. Data are presented as percentage of control and represent the mean ± SEM of 7–10 independent experiments. C: The protein levels of ICAM-1 were evaluated in whole rat retinal extracts by Western blotting. Data are presented as percentage of control and represent the mean ± SEM of seven animals. D: The protein levels of ICAM-1 were evaluated in whole extracts of rat retinal endothelial cell cultures (TR-iBRB2 cell line) by Western blotting. Data are presented as percentage of control and represent the mean ± SEM of at least four independent experiments. *P < 0.05, **P < 0.01, significantly different from control; ANOVA (one-way) followed by Dunnett post hoc test. #P < 0.05, ##P < 0.01, ###P < 0.001, significantly different from diabetic rat or high glucose condition; ANOVA (one-way) followed by Bonferroni post hoc test. HG, high glucose (30 mmol/l, for 4 days); CaD12.5, CaD25, and CaD50, calcium dobesilate 12.5, 25, and 50 μg/ml, respectively (4 days).

FIG. 5.

CaD prevents oxidative and nitrosative stress induced by diabetes. A: The oxidized proteins were detected using an anti-DNP antibody by dot blot. A representative dot blot is shown above the graph. Data are presented as percentage of control and represent the mean ± SEM of 3–4 animals. *P < 0.05, significantly different from control; ANOVA (one-way) followed by Dunnett post hoc test. B: Representative images showing nitrotyrosine immunoreactivity (green), which allows the detection of nitrated tyrosine residues, and nuclear DAPI staining (blue). Magnification: 400×, bar 50 μm. GCL, ganglion cell layer; IPL, inner plexiform layer; OPL, outer plexiform layer; ONL, outer nuclear layer; CH, choroidal layer. (A high-quality digital representation of this figure is available in the online issue.)

FIG. 6.

CaD inhibits the activation of p38 MAPK and NF-κB in diabetic rat retinas. A: The activation of p38 MAPK was determined by Western blotting, analyzing the phospho-p38/p38 MAPK ratio. A representative Western blot is shown above the graph. B: NF-κB activation was determined in retinal homogenates by ELISA (kit from Active Motif) using an antibody specific for the p65 subunit of NF-κB. A secondary antibody conjugated to horseradish peroxidase was used to quantify spectrophotometrically the activated form. Data are presented as percentage of control and represent the mean ± SEM of 6–7 animals. *P < 0.05, significantly different from control; ANOVA (one-way) followed by Dunnett post hoc test. #P < 0.05, significantly different from diabetic animals; ANOVA (one-way) followed by Bonferroni post hoc test.