A novel antiangiogenic peptide derived from hepatocyte growth factor inhibits neovascularization in vitro and in vivo

Abstract

Purpose: To study the antiangiogenic activity of two small peptides (H-RN and H-FT) derived from the hepatocyte growth factor kringle 1 domain (HGF K1) using in vitro and in vivo assays.

Methods: RF/6A rhesus macaque choroid-retina endothelial cells were used for in vitro studies. The inhibiting effect of two peptides on a vascular endothelial growth factor (VEGF)-stimulated cell proliferation, cell migration, and endothelial cell tube formation were investigated. For in vivo assays, the antiangiogenic activity of H-RN and H-FT in the chick chorioallantoic membrane model (CAM) and a mice oxygen-induced retinopathy model (OIR) were studied. A recombinant mouse VEGF-neutralizing antibody, bevacizumab, and a scrambled peptide were used as two control groups in separate studies.

Results: H-RN effectively inhibited VEGF-stimulated RF/6A cell proliferation, migration, and tube formation on Matrigel™, while H-FT did not. H-RN was also able to inhibit angiogenesis when applied to the CAM, and had antineovascularization activity in the retinal neovascularization of a mouse OIR model when administrated as an intravitreous injection. The antiangiogenic activity of H-RN was not as strong as that of VEGF antibodies. The H-FT and scrambled peptide had no such activity.

Conclusions: H-RN, a new peptide derived from the HGF K1 domain, was shown to have antiangiogenic activity in vitro and in vivo. It may lead to new potential drug discoveries and the development of new treatments for pathological retinal angiogenesis.

Figure 1

H-RN inhibits endothelial cells (EC) proliferation while H-FT does not. A: After 24 h of starvation, rhesus macaque choroid-retina endothelial (RF/6A) cells were incubated in vascular endothelial growth factor (VEGF) and various concentrations of bevacizumab or peptides for 24 h; RF/6A cell proliferation was inhibited by bevacizumab and H-RN. An H-RN concentration of 1 mM inhibited the VEGF-induced response significantly. However, the inhibitory effect was not as strong as bevacizumab’s. H-FT and scrambled peptide had no inhibitory activity on RF/6A cells at various concentrations (**p<0.01, each condition versus control). B: Compared to the culture medium group, cell viability was not affected by H-RN alone at any concentration tested (p>0.05, each condition versus control.)

Figure 2

Effects of peptides on the migration of RF/6A cells stimulated by vascular endothelial growth factor. Bevacizumab and peptides were added in the upper chamber containing RF/6A cells 30 min before the stimulation of vascular endothelial growth factor (VEGF) in the lower chamber (100 ng/ml). After 24 h of incubation, migratory activity of the cells was estimated. Bevacizumab and H-RN inhibited migration of RF/6A cells. H-FT and scrambled peptide did not inhibit migration up to 1 µM concentration, whereas at 10 µM there was a mild inhibition (*p<0.05; **p<0.01, each condition versus control).

Figure 3

Inhibition effects of H-RN and H-FT on endothelial cell differentiation into capillary structures. Starved RF/6A cells were pre-incubated with different concentrations of bevacizumab, H-RN, and H-FT for 30 min before 100 ng/ml vascular endothelial growth factor (VEGF) was added in peptides treated and VEGF groups, and then cells were seeded into Matrigel-coated wells. A-F: Representative photographs of anti-tube formation activity of bevacizumab, H-RN, H-FT, and scrambled peptide: A, no-VEGF; B, 100 ng/ml VEGF; C, 100 ng/ml VEGF +2.5 mg/ml bevacizumab; D, 100 ng/ml VEGF +1 mM H-RN; E, 100 ng/ml VEGF +100 nM H-FT; F, 100 ng/ml VEGF +1 mM scrambled peptide. VEGF at 100 ng/ml strongly stimulated endothelium tube formation. Bevacizumab and H-RN potently inhibited VEGF-stimulated RF/6A cell tube formation on growth factor-reduced Matrigel™. H-FT and scrambled peptide did not inhibit RF/6A cell tube formation effectively (magnification ×200). G: Quantitative analysis of tube formation under the different experimental conditions using Image Program Plus software. The values are mean tube lengths from three repeated experiments (** p<0.01, each condition versus control).

Figure 4

H-RN inhibits angiogenesis of chorioallantoic membrane model. A-D: Representative photographs of antiangiogenesis in chick chorioallantoic membrane (CAM) of H-RN, H-FT, and scrambled peptide. Filter paper disks saturated with H-RN (10 µg, 50 µg), H-FT (10 µg, 50 µg), scrambled peptide (10 µg, 50 µg) or PBS were placed on the CAMs for 72 h of incubation. Vessels around the filter disks were photographed: A, Control (PBS); B, 50 µg H-RN per egg; C, 50 µg H-FT per egg; D, 50 µg scrambled peptide per egg. (magnification was ×11.5). E: Quantitative results showed different effects of two peptides on CAM angiogenesis. H-RN inhibited CAM angiogenesis significantly, compared to the PBS group. No such antiangiogenesis activity of H-FT or scrambled peptide was found (** p<0.01, each condition versus control).

Figure 5

H-RN inhibits retinal neovascularization. A: Retinal vascular distribution of a normal pup raised in room air. B: This figure shows a representative retina from 17-day-old pups that were subjected to 5 days of 75% oxygen tension and then maintained in room air for an additional 5 days. Circles indicate the neovascular tufts. C: A retina from a mouse pup injected with vascular endothelial growth factor ab (VEGFab; 50 ng/µl). D: A retina from a mouse pup injected with H-RN (50 mM). E: A retina from a mouse pup injected with H-FT (50 mM). F: A retina from a mouse pup injected with scrambled peptide (50 mM, magnification, ×40). G: A hematoxylin- and eosin-stained section of P17 control raised in room air. H: P17 retina exposed to hyperoxia from P7 to P12, and returned to room air from P12 to P17. Arrows indicate neovascular tufts extending into the vitreous. I: A section from 50 ng/µl VEGFab-treated hyperoxia group. J: A section from 50 mM H-RN-treated hyperoxia group. The retina of I and J presented obvious reduced abnormal neovascularization tufts (arrows). K: A section from 50 mM H-FT-treated hyperoxia group. L: A section from 50 mM scrambled peptide treated hyperoxia group (magnification, ×400). M: Quantification of neovascular lumens response to hyperoxia. Lumens extending from the internal limiting membrane into the vitreous were counted. Data in each column present means±SD values of total number of vascular lumens per retinal cross-section, from 7 to 9 eyes of 7–9 mice. Note that the lumens in VEGFab and H-RN-treated group are reduced significantly more than the control and PBS groups (p<0.001), while the number of lumens in the H-FT and scrambled peptide-treated group had no difference from the control and PBS treated group (p>0.05). These experiments were repeated three times with similar results (** p<0.01, each condition versus control).