Endothelial von Willebrand factor regulates angiogenesis

Abstract

The regulation of blood vessel formation is of fundamental importance to many physiological processes, and angiogenesis is a major area for novel therapeutic approaches to diseases from ischemia to cancer. A poorly understood clinical manifestation of pathological angiogenesis is angiodysplasia, vascular malformations that cause severe gastrointestinal bleeding. Angiodysplasia can be associated with von Willebrand disease (VWD), the most common bleeding disorder in man. VWD is caused by a defect or deficiency in von Willebrand factor (VWF), a glycoprotein essential for normal hemostasis that is involved in inflammation. We hypothesized that VWF regulates angiogenesis. Inhibition of VWF expression by short interfering RNA (siRNA) in endothelial cells (ECs) caused increased in vitro angiogenesis and increased vascular endothelial growth factor (VEGF) receptor-2 (VEGFR-2)-dependent proliferation and migration, coupled to decreased integrin αvβ3 levels and increased angiopoietin (Ang)-2 release. ECs expanded from blood-derived endothelial progenitor cells of VWD patients confirmed these results. Finally, 2 different approaches, in situ and in vivo, showed increased vascularization in VWF-deficient mice. We therefore identify a new function of VWF in ECs, which confirms VWF as a protein with multiple vascular roles and defines a novel link between hemostasis and angiogenesis. These results may have important consequences for the management of VWD, with potential therapeutic implications for vascular diseases.

Figures

Figure 1

VWF-deficient cells display enhanced angiogenesis in vitro. VWF-specific (20nM; siVWF, closed bars), or control siRNA (20nM; siCTL, open bars) was transfected into HUVECs for 24 or 48 hours. VWF expression was measured in these cells by (A) RT-PCR, (B) Western blotting (representative blot shown), or (C) VWF ELISA. (D) HUVECs treated with control or VWF-specific siRNA for 24 hours were seeded onto Matrigel. Capillary network formation was observed in the presence (right panels) or absence (left panels) of 10 μg/mL VWF after 24 hours and quantified in panel E by measuring total tube length in micrometers. (Bar = 200 μm). Data for siVWF in panels A and C have been normalized to siCTL. siCTL, nonspecific siRNA-treated cells; siVWF, VWF siRNA-treated cells; UT, untreated cells (n = 3). Error bars = mean ± SEM. *P < .05; **P < .01; ***P < .001 (Student t test).

Figure 2

VWF-deficient cells display increased migration and proliferation in vitro due to increased VEGFR-2 signaling. (A) Representative trajectory plots of siCTL- or siVWF-treated HUVECs migrating into a wounded area for 16 hours, starting at 30 hours after siRNA treatment. The starting point of each cell trajectory is plotted at the center of the graph and the wounded area is to the left of zero on the x-axis. (B) Migration velocity (in micrometers per minute) with or without 50 ng/mL VEGF or with 4μM VEGFR-2 inhibitor SU4312. (C) Directionality of cell migration (Euclidean distance/accumulated distance). (D) Proliferation (in cells per square centimeter) of HUVEC 24 hours after control or siVWF treatment, cultured with or without 100 ng/mL VEGF for 96 hours. Data for siVWF in panel D has been normalized to siCTL-VEGF. siVWF was still effective at 120 hours after siRNA treatment (data not shown). siCTL, nonspecific siRNA-treated cells; siVWF, VWF siRNA-treated cells (n = 3). Error bars = mean ± SEM. *P < .05; **P < .01 (Student t test).

Figure 3

VWF-deficient cells display reduced β3 integrin expression and αvβ3-dependent adhesion, increased rate of β3 integrin internalization, and elevated release of Ang-2. (A) β3 mRNA expression was measured by RT-PCR. (B) Total β3 integrin protein expression was measured by Western blotting (representative blot shown) and quantified by densitometry relative to tubulin. (C) Surface levels of total αvβ3 (antibody: LM609) were measured by flow cytometry at 24 and 48 hours after siRNA treatment. (D) Control or siVWF-treated cells, 48 hours after transfection, were seeded onto different extracellular matrix substrates. After 40 minutes, nonadherent cells were removed by gentle washing and the number of bound cells was quantified relative to the total number seeded. (E) Internalization of β3 integrin from the cell surface after incubation at 37°C for 7.5 minutes. (F) Levels of Ang-2 in the supernatant of control or siVWF-treated cells were measured 48 hours after transfection by ELISA. Data (A-C, F) were normalized to control siRNA-treated cells at each time point. Open bars, siCTL; closed bars, siVWF. Error bars indicate mean ± SEM (n = 3). *P < .05; **P < .01 (Student t test).

Figure 4

VWF-deficient mice display increased angiogenesis and mature blood vessel density. (A) Matrigel was injected subcutaneously into littermate control (CTL; top panels) or VWF-deficient mice (KO; bottom panels). Seven-day-old plugs were excised, sectioned, and stained with hematoxylin and eosin (left panels), or for immunofluorescence microscopy, with anti-CD31 (middle panels) or VEGFR-2 (right panels). Size bar = 20μm. (B) Cellular infiltrate was quantified in hematoxylin and eosin sections from CTL (open bars) or KO (closed bars). n = 5 KO mice and n = 6 CTL mice. (C) CTL or KO mice were culled and whole ears were removed. Ears were stained with anti–α-SMA-Cy3 and a tile scan was performed to obtain a composite image of the whole ear (left panels). Size bar = 2 mm. Blood vessels were identified, segmented, and converted to binary images (right panels). Total ear area, blood vessel area (relative to total ear area), and fractal dimensions were calculated. n = 5 KO mice and n = 7 CTL mice, Error bars indicate mean ± SEM. **P < .01 (Student t test).

Figure 5

BOECs from patients with VWD show increased in vitro angiogenesis, proliferation, and migration. (A) Representative images of BOECs from a healthy control (top, far left) and a VWD patient (bottom, far left). Colonies appeared after 7-22 days in culture (size bar = 200 μm). BOECs were stained for VE-cadherin, ICAM-2, and CD45. Nuclei were stained with TO-PRO 3 (Invitrogen). Size bar = 20 μm. (B) Secretion of VWF (μg/mL) from cultured BOECs after 48 hours. (C) BOEC were seeded onto Matrigel for 24 hours and capillary network formation was quantified by measuring total tube length in micrometers. Size bar = 200 μm. (D) Proliferation of BOECs cultured with 100 ng/mL VEGF for 96 hours relative to unstimulated controls. (E) Migration velocity of cells at the periphery of endothelial colony prior to the first passage. Each patient is identified with a unique symbol (see supplemental Table 1 for code and information on VWD patients). Error bars indicate mean ± SEM. *P < .05; **P < .01 (Student t test).