Chordin-like 1, a bone morphogenetic protein-4 antagonist, is upregulated by hypoxia in human retinal pericytes and plays a role in regulating angiogenesis

Abstract

Purpose: Pericytes play a specialized role in regulating angiogenesis and vascular function by providing vascular stability and controlling endothelial cell proliferation. Disorders in pericyte function and pericyte-endothelial interaction have been observed in several disease states including tumor angiogenesis and diabetic microangiopathy. In ischemic retinal disease, hypoxia is a potent driver of retinal angiogenesis. This study investigated the effects of hypoxia on retinal pericyte gene expression, and demonstrates a role in angiogenesis regulation for the hypoxia driven gene, chordin-like 1 (CHL-1).

Methods: In the current studies, we investigated hypoxia-induced gene expression in human retinal pericytes and found that expression of CHL-1, a member of the bone morphogenetic protein (BMP) superfamily, is upregulated by hypoxia. We investigated regulation of CHL-1 expression and the ability of CHL-1 to antagonize the antiangiogenic properties of BMP-4 using a human cell-based angiogenesis assay.

Results: We report that hypoxia induced hypoxia inducible factor-1alpha-driven expression of CHL-1. Both CHL-1 and BMP-4 were secreted from human retinal pericytes. We found that CHL-1 complexes with BMP-4 to antagonize the antiangiogenic effects of BMP-4, and that BMP-4 and vascular endothelial growth factor (VEGF) co-regulate angiogenesis.

Conclusions: We propose that hypoxia-induced upregulation of CHL-1 alters the homeostatic balance between BMP-4 and VEGF to synergize with VEGF in driving retinal angiogenesis.

Figure 1

Validation of three genes differentially expressed in human retinal pericytes (hRPC) in response to hypoxia. The upregulation of a selection of genes was validated with real time PCR, using a PerkinElmer 7700 analyzer, on cDNA generated from human retinal pericytes exposed to increasing periods of hypoxia (0, 6, 24, and 48 h). All results were normalized to 18S rRNA, using a pre-developed assay reagent. Data are expressed as mean relative quantity of mRNA, relative to control, for three independent experiments ±standard error of measurement for (A) CHL-1 mRNA, (B) VEGF mRNA, and (C) Cox 2 mRNA. Data are expressed as mean±SEM values. The asterisk indicates a significance at p<0.05 and the double asterisk indicates a significance at p<0.001.

Figure 2

HIF-1α drives expression of chordin-like 1 in retinal pericytes exposed to hypoxia. A: western blot analysis of nuclear extracts generated from human retinal pericytes exposed to increasing periods of hypoxia (0, 6, 24, and 48 h) for HIF-1α shows upregulation of the protein by 6 h. B: Transfection of human retinal pericytes maintained in normoxia with expression vectors for HIF-1 α, C is control/empty vector, WT is wild type vector, WT-HIF-1 α, DM is double mutant vector, DM-HIF-1 α, using the transfection reagent Fugene6, induces expression of many of the genes upregulated in response to hypoxia, as measured by RT–PCR. 18S PCR is shown as a loading control and western blot analysis confirmed expression from each of the HIF-1α expressing plasmids. C: Induction of CHL-1 mRNA in response to HIF-1α overexpression was quantitated by real time PCR. CHL-1 levels were normalized to 18S rRNA, using a pre-developed assay reagent. Data are expressed as mean relative quantity of mRNA, to control, for three independent experiments ±standard error of measurement. D: HeLa cells were transfected, using the transfection reagent Fugene6, with the CHL-1 promoter or a luciferase reporter construct containing four HIF-1α responsive elements (HRE), alone (control), cotransfected with the HIF-1α expression vector DM-HIF-1α, or alone and subsequent exposure to hypoxia for 24 h (1% O₂). Cotransfection with DM-HIF-1α as well as exposure to hypoxia induced activation of the CHL-1 promoter and the HRE construct. Data are expressed as mean±SEM values. The asterisk indicates a significance at p<0.05 and the double asterisk indicates a significance at p<0.001.

Figure 3

Chordin-like 1 expressed in human retinal pericytes is secreted and binds to bone morphogenetic protein-4. A: Conditioned media from HRPC exposed to 1% O₂ for 24 and 48 h was examined by western blot analysis for secretion of CHL-1, using an anti-CHL-1 antibody. B and C: BMP-2 and BMP-4 expression in HRPC was examined in cells cultured in normoxia (N) and hypoxia (H) by RT–PCR (B) and secreted BMP2 and BMP-4 were detected in conditioned media from HRPC exposed to 1% O₂ for 24 and 48 h (C). D: Transfection of the expression vector pcDNA6/CHL-1 V5-His into Cos7 cells, using the transfection reagent Fugene 6, resulted in expression of an approximately 60 kDa protein, which was detectable using an anti-V5 antibody. Cells were transfected with either an empty vector (E), pcDNA6/V5-His C, or a V5 tagged CHL-1 expressing vector (CHL-1), pcDNA6/CHL-1 V5-His. E: Whole cell extracts from Cos7 cells were transfected, using the transfection reagent Fugene 6, with empty pcDNA6/V5His (E), or with the expression vector pcDNA6 CHL-1/V5His (CHL-1) expressing V5His tagged CHL-1, were incubated with 250 ng rhBMP-4 and 100 ml NiNTA magnetic beads at 4 °C overnight. The complexes were washed, the beads and examined by western blotting for the presence of CHL-1, using anti-V5 antibody, and BMP-4, using an anti-BMP-4 antibody.

Figure 4

Chordin-like 1 modulates the antiangiogenic effect of bone morphogenetic protein-4. Human umbilical vascular endothelial cells(HUVECs) and human diploid fibroblasts were obtained (day 1) as cocultures in 24 well plates. Medium, with treatments or vehicle, was replenished on days 1, 4, 7, and 9. The assay was treated with VEGF, Suramin, recombinant human BMP-2, BMP-4, and CHL-1. Tubule formation was examined at day 11. Cells were fixed, quantitated, and visualized using a combined ELISA and histology kit. A: VEGF (2 ng/ml) and Suramin (20 mM) were used as positive and negative angiogenesis controls, respectively. B: Cells were treated with rhBMP-2 and rhBMP-4. BMP-4 significantly inhibited angiogenesis at 10 ng/ml. C: CHL-1 inhibited BMP-4; CHL-1 alone had no significant effect on angiogenesis, however it inhibited BMP-4s anti-angiogenic effects. Images A-C are shown at magnification 10X. Representative images are shown in A-C. D: Angiogenesis was quantitated by using anti-CD31 antibody coupled to a soluble substrate, ρ-nitrophenol phosphate (ρ-NPP), which permits quantitation by an optical density measurement. The asterisk indicates a significance at p<0.05 and the double asterisk indicates a significance at p<0.001.

Figure 5

Vascular endothelial growth factor and bone morphogenetic protein-4 co-regulate angiogenesis. Human umbilical vascular endothelial cells (HUVECs) and human diploid fibroblasts were obtained (day 1) as cocultures in 24 well plates. Medium, with treatments or vehicle, was replenished on days 1, 4, 7, and 9. The assay was treated with vascular endothelial growth factor (VEGF), Suramin, and recombinant human BMP-4. Tubule formation was examined at day 11. Cells were fixed, quantitated, and visualized using a combined ELISA and histology kit. A: Angiogenesis assay demonstrating the combined effects of VEGF (pro-angiogenic) and BMP-4 (anti-angiogenic) on angiogenesis. Suramin was used as a negative control. (magnification 10X). Representative images are shown. B: Angiogenesis was quantitated by using anti-CD31 antibody coupled to a soluble substrate, ρ-nitrophenol phosphate (ρ-NPP), which permits quantitation by an optical density measurement. Data are expressed as mean±SEM values. The asterisk indicates a significance at p<0.01 and the double asterisk indicates a significance at p<0.001.