Bone morphogenetic protein 2: a potential new player in the pathogenesis of diabetic retinopathy

Abstract

Diabetic retinopathy (DR) is one of the most common complications of diabetes mellitus. Vision loss in DR principally occurs due to breakdown of the blood-retinal barrier (BRB), leading to macular edema, retinal detachment and inner retinal and vitreous hemorrhage. Several growth factors have been shown to play crucial role in the development of these vascular changes; however, the cellular and molecular mechanisms of DR are not yet fully revealed. In the current study we investigated the role of bone morphogenetic protein-2 (BMP2) in DR. We examined the changes in the protein levels of BMP2 in human vitreous and retina in addition to the mouse retina of streptozotocin-induced diabetes. To detect the source of BMP2 during diabetes, human retinal endothelial cells (hRECs) were subjected to high glucose (HG) for 5 days and levels of BMP2 protein were analyzed in conditioned media of these cells relative to control. We also evaluated the effect of BMP2 on the levels of VEGF in cultured rat Müller cells (rMC1). In addition, we tested the pro-inflammatory effects of BMP2 by examining its effect on leukocyte adhesion to cultured hRECs, and levels of adhesion molecules and cytokines production. Finally, the effect of different concentrations of BMP2 on permeability of confluent monolayer of hRECs was evaluated using FITC-Dextran flux permeability assay and by measuring Transcellular Electrical Resistance (TER) using Electric Cell-substrate Impedance Sensing (ECIS). Our results show, for the first time, the up-regulation of BMP2 in diabetic human and mouse retinas in addition to its detection in vitreous of patients with proliferative DR (72 ± 7 pg/ml). In vitro, hRECs showed upregulation of BMP2 in HG conditions suggesting that these cells are a potential source of BMP2 in diabetic conditions. Furthermore, BMP2 induced VEGF secretion by Müller cells in-vitro; and showed a dose response in increasing permeability of cultured hRECs. Meanwhile, BMP2 pro-inflammatory effects were recognized by its ability to induce leukocyte adhesion to the hRECs, intercellular adhesion molecule-1 (ICAM-1) and upregulation of interleukin-6 and 8 (IL-6 and IL-8). These results show that BMP2 could be a contributing growth factor to the development of microvascular dysfunction during DR via enhancing both pro-angiogenic and inflammatory pathways. Our findings suggest BMP2 as a potential therapeutic target to prevent/treat DR.

Keywords: BMP 2; Müller cells; VEGF; blood–retinal barrier; diabetic retinopathy; inflammation; leukocyte adhesion; retinal endothelial cells.

Figure 1. Increased BMP2 Expression in Human Retina During Diabetic Retinopathy

A) Immunohistochemistry of BMP2 in normal and diabetic human retina. There is marked increase in the expression of BMP2 in DR compared to the normal retina (Control). BMP2 is increased different layers (brown stain) particularly in relation to retinal vessels (Arrow). B) Western blot analysis of BMP2 in normal and diabetic human retina. Showing up-regulation of BMP2 in retina of diabetic (D) subjects. (*: pValue < 0.05 versus control group, n=4)

Figure 2. Increased BMP2 expression in Retina of Diabetic mice

A) Immunofluorescence of BMP2 in normal and diabetic mouse retina. There is marked increase in the expression of BMP2 (Green) in diabetic mouse retina compared to the normal retina (Control). BMP2 is localized mainly in relation to retinal vasculature (Red) (Arrows). B) Western blot analysis of BMP2 in normal and diabetic mice retina showing upregulation of BMP2 in retina of diabetic mice compared to control group (n=5). (C) BMP2 ELISA of conditioned media of high glucose (30mM) treated hRECs for 5 days. Results are plotted as fold change/control and show a significant upregulation of BMP2 as a result of HG (2.9 ±0.4, pvalue: 0.02, n=4) (*:pValue < 0.05 versus control group).

Figure 3. BMP2 effect on leukocyte adhesion to hRECs and expression of adhesion molecule ICAM-1

(A) Bar graph representing fold change/control of myeloperoxidase activity (indicator of leukocyte adhesion) showing that the used BMP2 concentrations (50, 10 and the 5ng/ml) showed increased leukocyte adhesion to hRECs. These effects were significantly higher than both control and to LPS (10μg/ml) (n=4). (B) Western blot of ICAM-1 in hRECs treated with BMP2 showing up-regulation of ICAM-1 hRECs treated with BMP2 (50ng/ml) versus control group (n=3). (*: pvalue< 0.05 versus control, ***: pvalue< 0.001 versus control, ###: pvalue< 0.001 versus LPS, #: pvalue< 0.05 versus LPS).

Figure 4. Multiplex results of

(A) G-CSF, (B) IL-4, (C) IL-8, (D) INF-γ and (E) IL-6 measured in conditioned media of hRECs treated with 10 and 50 ng/ml BMP2. (*: pValue < 0.01 versus other groups, n=4).

Figure 5. ELISA of VEGF expression in cultured rat Müller cells

The levels of VEGF were evaluated after 12 and 24 hours from the beginning of BMP2 (10ng/ml) treatment. (*: pValue < 0.05 versus control group, n=5).

Figure 6. Effect of BMP2 on retinal endothelial cell barrier function

FITC-Dextran leakage relative to control at 1 hour (A), 3 hours (B) and 5 hours (C) time points. Both BMP2 and VEGF demonstrated 1.2 and 1.5 fold increase in the leakage of FITC-dextran after 3 and 5 hours respectively. (D) Normalized TER resistance measured using ECIS: showing a dose response effect of BMP2 on hRECs normalized resistance where the significant decrease in resistance versus control started at 2 Hours, 7 Hours and 10 Hours time points for the 50, 10 and 5ng/ml BMP2 respectively (*: pValue < 0.05 versus control group, #: pValue < 0.05 versus VEGF group, $: pValue < 0.05 versus all other groups; group n=3).