Abstract

Vascular endothelial growth factors (VEGFs) and their receptors (VEGFRs) are pivotal regulators of angiogenesis. The VEGF–VEGFR system is therefore an important target of anti-angiogenesis therapy. Based on the X-ray structure of VEGF-B/VEGFR-1 D2, we designed a cyclic peptide (known as VGB1) reproducing the α1 helix and its adjacent region to interfere with signaling through VEGFR-1. Unexpectedly, VGB1 bound VEGFR-2 in addition to VEGFR-1, leading to inhibition of VEGF-stimulated proliferation of human umbilical vein endothelial cells and 4T1 murine mammary carcinoma cells, which express VGEFR-1 and VEGFR-2, and U87 glioblastoma cells that mostly express VEGFR-2. VGB1 inhibited different aspects of angiogenesis, including proliferation, migration and tube formation of endothelial cells stimulated by VEGF-A through suppression of extracellular signal-regulated kinase 1/2 and AKT (Protein Kinase B) phosphorylation. In a murine 4T1 mammary carcinoma model, VGB1 caused regression of tumors without causing weight loss in association with impaired cell proliferation (decreased Ki67 expression) and angiogenesis (decreased CD31 and CD34 expression), and apoptosis induction (increased TUNEL staining and p53 expression, and decreased Bcl-2 expression). According to far-UV circular dichroism (CD) and molecular dynamic simulation data, VGB1 can adopt a helical structure. These results, for the first time, demonstrate that α1 helix region of VEGF-B recognizes both VEGFR-1 and VEGFR-2.

Introduction

Angiogenesis, the formation of blood vessels from the preexisting vasculature, occurs under the physiological as well as pathological conditions. Vascular endothelial growth factor (VEGF) family is of the most important pro-angiogenic factors, which have been most commonly implicated in pathological angiogenesis [1,2]. This family comprises seven subgroups of ligands: VEGF-A, -B, -C, -D, -E, -F and placenta growth factor (PlGF). VEGF-A, -B and PlGF are required for blood vessel formation, among which VEGF-A (also known as VEGF) is the pivotal regulator of angiogenesis that stimulates endothelial and tumor cells to proliferate and migrate as well as supports the invasion of stromal cells into the growing tumors [2,3]. VEGF-B that is structurally and functionally related to VEGF-A and PlGF controls the bioavailability of VEGF-A [4] and is implicated in several pathological conditions, such as tumor angiogenesis and metastasis [5].

The biological roles of VEGFs are mediated by three receptor tyrosine kinases (RTKs), including VEGF receptor (VEGFR)-1, -2 and -3 [also known as FLT-1, KDR (kinase insert domain receptor)/FLK-1 (fetal liver kinase 1) and FLT-4, respectively], and semaphorin co-receptor neuropilin (NRP) 1 and 2 [69]. RTKs are transmembrane proteins with multiple domains including seven extracellular immunoglobulin (Ig)-like domains, a transmembrane (TM) domain, a juxtamembrane domain, an intracellular tyrosine kinase domain and the kinase insert domain, followed by a sequence carrying several tyrosine residues. Receptor specificity in VEGF ligands has distinct patterns: VEGF-A binds to VEGFR-1 and -2, VEGF-B and PlGF exclusively bind to VEGFR-1 [10], and VEGF-C and -D recognize both VEGFR-2 and -3.

The binding of VEGFs to their receptors on the surface of endothelial cells activates the MAPK (mitogen-activated protein kinase)/ERK (extracellular signal-regulated kinase) 1/2 and phosphatidylinositol-3 kinase (PI3K)/AKT (Protein Kinase B) signaling pathways, leading to endothelial cell (EC) proliferation, survival, differentiation, migration and tube formation [1117]. Given that the angiogenesis is crucial for tumor growth and metastasis, the anti-angiogenic therapy can be efficacious for the treatment of the metastatic tumors [1821]. Peptides have a variety of advantages such as small size, few side effects, ease of synthesis and modification, penetrating ability into tumor environment and good biocompatibility, leading them to be considered as the safest and least toxic drugs [22]. Accordingly, many attempts have been made to identify peptide inhibitors of angiogenesis [23]. Several peptides derived from growth factors and their receptors are currently being evaluated in clinical trials for the treatment of human cancers [18]. Despite the clinical efficacy of these anti-cancer agents, they have not been shown capable of healing cancer, suggesting the requirement for more efficacious peptides that span a wide range of tumorigenic cellular processes [18].

In the present study, to interfere with the interaction of VEGF with VEGFR-1, an antagonistic peptide (known as VGB1) was designed according to the complex structure of VEGF-B/VEGFR-1 D2 complex. Interestingly, VGB1 mimicking the α1 helix region of VEGF-B, bound to VEGFR-2 in addition to VEGFR-1, as it blocked the binding of anti-VEGFR-1 and anti-VEGFR-2 monoclonal antibodies to human umbilical vein endothelial cells (HUVECs). VGB1 inhibited the VEGF-induced proliferation of HUVECs, 4T1 mammary carcinoma and U87 glioblastoma cells, as well as migration and tube formation of HUVECs through suppression of PI3K/AKT and MAPK/ERK1/2 signaling pathways. According to in vivo studies, VGB1 inhibited the growth of 4T1 mammary carcinoma tumors correlated with a significant decrease in microvessel density (MVD), tumor cell proliferation and an increase in apoptosis induction in tumor environment.

Experimental section

Antibodies and chemicals

Anti-AKT (Ab25893), anti-phospho-AKT (Ser473) (Ab81283), anti-VEGFR-2 (Ab9530), anti-VEGFR-1 (Ab212369), fluorescein isothiocyanate (FITC)-secondary anti-mouse (Ab6724), phycoerythrin (PE)-secondary anti-mouse (Ab97024), anti-CD31 (Ab32457), anti-CD34 (Ab81289), anti-Ki67 (Ab15580), anti-p53 (Ab131442), and anti-Bcl-2 (Ab59348) were obtained from Abcam, Cambridge, U.K. Anti-p44/p42 MAPK (ERK1/2) (9102S) and anti-phospho-p44/p42 MAPK (ERK1/2) (Thr202/Tyr204) (4377S) were purchased from Cell Signaling Technology, Danvers, MA, U.S.A. TUNEL assays were done utilizing an in situ Cell Death Detection Kit POD (Roche Diagnostic GmbH, Germany). The Bio-Rad protein assay kit (500-0006) was purchased from Bio-Rad, U.S.A. BSA (A7030) was purchased from Sigma, U.S.A.

Peptide synthesis

The peptide was synthesized and purified by high-performance liquid chromatography (HPLC) to a purity of 95%, analyzed by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF), and confirmed by electrospray ionization mass spectrometry (ESI-MS) analysis (Shine Gene Biotechnologies, Inc., Shanghai, China) (Supplementary Figures S1 and S2).

Ellman assay

Disulfide bond formation was confirmed by the Ellman assay [24,25]. Briefly, VGB1 (1.18 μM) was added to the 100 mM sodium phosphate reaction buffer at pH 8.0 containing 100 mM EDTA. Next, the Ellman reagent [4 mg/ml of DTNB (5,5'-dithiobis-(2-nitrobenzoic acid))] was added to the solution. The reaction was started by mixing the sample solution with an equal volume of the Ellman's reagent. After 15 min incubation at room temperature, the sulfhydryl concentration was calculated by reading the absorbance of reaction at 412 nm (ε = 1.36 × 104 M−1 cm−1) versus a sulfhydryl standard ‘cysteine’ treated in the same manner by using a UV-3000 spectrophotometer (Pharmacia Biotech Ultrospect, Cambridge, U.K.).

Cell culture

HUVECs were cultured in Dulbecco's Modified Eagle's Medium (DMEM; Gibco, Life Technologies, U.S.A.) supplemented with 10% fetal bovine serum (FBS; Sigma, St. Louis, Missouri, U.S.A.) and were preserved in 5% CO2 at 37°C and 13.8% O2 until 90% confluent. All protocols were carried out and confirmed by the Ethics Committee of the Kermanshah University of Medical Sciences. The 4T1 and U87 cell lines were acquired from the Iran Pasteur Institute and cultured and maintained.

Peptide conjugation to FITC

Peptide was labeled with FITC as described recently [23]. Briefly, FITC was dissolved in DMSO, getting 1 mg/ml solution, which was added into the reaction buffer (PBS, pH 8.5) containing 1 mg/ml as the peptide. The tube was covered with foil and was incubated at 37°C for 90 min. To eliminate unreacted FITC and peptide and exchange the peptide into the storage buffer (PBS, pH 7.5), the reaction mixture was loaded onto an equilibrated Sephadex G-10 column (1.5 × 150 cm) and was eluted with PBS, pH 7.5. The concentration of the collected sample was determined using the following equation: 
formula
where DF is the dilution factor, MW is the peptide molecular mass and ɛ is the molar extinction coefficient of each chromophore at 280 nm.

Peptide-binding assays

HUVECs (3 × 105) were incubated with various concentrations of the FITC-conjugated peptide or unlabeled peptide and the other indicated reagents (anti-VEGFR-1 primary antibody and PE-labeled anti-mouse secondary antibody or anti-VEGFR-2 primary antibody and FITC-labeled anti-mouse secondary antibody) overnight at 4°C in the dark. After washing with PBS, HUVECs were resuspended in PBS and were investigated using a BD FACSCalibur Flow Cytometer.

For further examination of VGB1 binding to VEGFR-1 and VEGFR-2, a competitive receptor-binding assay was performed by immunocytochemistry. HUVECs (5000 cells/well) were seeded into a 96-well culture plate, with DMEM and 5% FBS, and were cultured at 37°C overnight [26]. After replacing the media with DMEM, gradient concentrations (0, 0.58, 0.88 and 1.18 µM) of VGB1 were added and incubated at 37°C overnight. After fixing endothelial cells with 4% paraformaldehyde at room temperature for 5–10 min, ECs were washed with PBS and permeabilized with 0.3% Triton X-100 and PBS, and then blocked with 10% normal goat serum and 1% BSA/PBST. After another round of washing with PBS, anti-VEGFR-1 (5 µg/ml) antibody together with PE-secondary anti-mouse antibody was added. For nuclear staining, cells were mounted with 4′,6-diamidino-2-phenylindole (DAPI; Invitrogen, Carlsbad, CA, U.S.A.) (1 µg/ml). To assess VGB1 binding to VEGFR-2, ECs were treated with anti-VEGFR-2 (5 µg/ml) and FITC-secondary anti-mouse antibodies. To nuclear staining, these cells were treated with propidium iodide (PI; Sigma–Aldrich, U.S.A., P4170) (1 µg/ml). Finally, imaging was performed by fluorescence microscopy in red (PI; Sigma–Aldrich, U.S.A., P4170; 1 µg/ml, nuclear staining), blue (DAPI; Invitrogen, Carlsbad, CA, U.S.A.; 1 µg/ml, nuclear staining), red (VEGFR-1/PE-secondary anti-mouse antibody) and green (VEGFR-2/FITC-secondary anti-mouse antibody). All images were captured by a fluorescence microscope (Olympus BX51, Japan) equipped with a digital camera and merged before analysis by the NIH Image J plugin (http://imagej.nih.gov/ij/). Error bars denote the standard error of the mean (SEM) of measurements executed in triplicate, and scale bars are 20 µm.

Proliferation assay

The effects of VGB1 on the cell proliferation were examined by the MTT assay [27]. HUVE, 4T1 and U87 cells (2 × 103 cells/well) were cultured in DMEM containing 5% FBS in 96-well plates overnight. The supernatant was replaced with FCS-free culture medium containing 0.2 μg/ml VEGF-A (Sigma, St. Louis, MO, U.S.A.), and simultaneously, VGB1 was added at concentrations ranging from 0.15 to 1.18 µM. The plates were incubated for an additional 48 h at 37°C, and the supernatant was discarded. Then, 5 mg/ml 3-(4,5-dimethyl thiazolyl-2)-2, 5-diphenyltetrazolium bromide (10 μl) (MTT) (Sigma, St. Louis, MO, U.S.A.) and the culture medium (100 μl) was added to each well. The plates were incubated at 37°C for 4 h, then the medium was discarded, and DMSO (100 μl) was added to each well. The cells were incubated at 37°C overnight and then read on with background subtraction of 630 nm by an ELISA plate reader (Space Fax 2100, Awareness, U.S.A.) at 570 nm. VEGF-stimulated and unstimulated cells were used as positive and negative controls, respectively. The experiment was repeated three times.

Migration assay

The wound healing assay was used to investigate the effect of VGB1 on cell migration in vitro [28]. The same cell numbers of control and treatment groups were seeded into 24-well plates. Then, a cell-free area of confluent monolayer cells was created by scratching with a 100 µl pipette tip. After washing the detached cells, the medium was replaced with the one containing 0.2 μg/ml VEGF-A, and then cell monolayer was cultured in the presence or absence of various VGB1 concentrations (0.58–1.18 µM). After 24 h of the incubation, cells were washed with PBS and were fixed with 4% paraformaldehyde in PBS at the room temperature. The images were recorded by a microscope (scale bar: 20 µm; Olympus BX-51, Japan) at the suitable magnifications. For comparing the migration rate, scratches were photographed at five separate fields, and then the scratch area was measured by an online analysis of images on Wimasis image analysis software (http://www.wimasis.com/en/). The test was repeated three times.

Capillary-like tube formation assay

The tube formation of two-dimensional (2D) angiogenesis assay was carried out using the angiogenesis starter kit (Geltrex A1460901, Invitrogen, U.S.A.) and according to the manufacturer's instructions. In summary, the plate was coated with growth factor-reduced basal membrane extract (Geltrex, A1460901, Invitrogen, U.S.A.), thawed at 4°C overnight, and was allowed to solidify while being incubated for 30 min at 37°C. Then, HUVECs were suspended in LVES-supplemented Medium 200 or basal media free-serum containing 0.2 μg/ml VEGF-A and were seeded at 14 × 103 cells/200 µl per well. The cells were treated with the concentrations 0.3–1.18 μM of VGB1. At 14–18 h post-seeding, the images were recorded using a fluorescence microscope (Olympus BX-51, Japan). The experiment was repeated at least three times. Total tubes, tube length and branching points were estimated using Wimasis image analysis software (http://www.wimasis.com/en/).

In vitro sprout formation assay

The fibrin or collagen sprouting angiogenesis assay was performed as previously described [29]. Briefly, HUVECs were attached to cytodex microcarrier beads (Amersham Pharmacia Biotech, Sweden) by incubation in DMEM with 10% FBS at 37°C in 5% CO2 for 4 h. The HUVEC-coated beads embedded in a collagen matrix under sodium bicarbonate conditions (Sigma, St. Louis, MO, U.S.A.) were distributed in 96-well plates (100 µl/well), and were incubated in 37°C, 5% CO2 for 30 min. After the media were replaced by VEGF-A (0.2 μg/ml) without FBS, HUVE cells were treated with different concentrations of VGB1 (0–1.18 μM) and were incubated at 37°C, 5% CO2 for 72 h. Vascular sprout formation was photographed using an inverted microscope (Olympus BX-51, Japan). The experiment was repeated at least three times. The number of sprouting vessels was quantified by Wimsprout (http://www.wimasis.com/en/) software.

Western blot analysis

The expression and the phosphorylation levels of ERK1/2, AKT were measured by the western blot assay. HUVECs were seeded into six-well plates at 1 × 106/well in cell culture. After deprivation in serum-free media for 24 h, the cells were treated with (treatment group) or without (control group) VGB1 (0.58 and 1.18 μM) together with VEGF (0.2 μg/ml). HUVECs were washed twice with PBS and then were lysed with RIPA lysis buffer [50 mM Tris–HCl, pH 8.0, 150 mM NaCl, 1.0% NP-40, 0.5% sodium deoxycholate and 0.1% sodium dodecyl sulfate (SDS)] supplemented with a protease inhibitor cocktail and phosphatase inhibitor cocktail 3 (Sigma, St. Louis, MO, U.S.A.) [30,31]. Protein concentrations were estimated by the Bradford assay [32]. Briefly, five dilutions of BSA (A7030, Sigma, U.S.A.) as a protein standard were prepared. Each of the standard or sample solutions was pipetted into a clean, dry test tube. Then, a proper amount of Bradford solution (Bio-Rad protein assay kit, Bio-Rad, U.S.A.) was added to each tube, followed by absorbance measurement at 595 nm using a UV-3000 spectrophotometer (Pharmacia Biotech Ultrospect, Cambridge, U.K.). Finally, a comparison of the standard curve with the sample curve provided a relative measurement of protein concentration. Similar volumes of total protein were loaded into and parted by 10% SDS–polyacrylamide gel electrophoresis and afterward transported to Amersham™ Protran® Premium nitrocellulose western blotting membranes (Ge10600013, Sigma, St. Louis, MO, U.S.A.). The proteins were electrophoretically transferred to the nitrocellulose. After the transfer, the proteins were visualized by dyeing with Ponceau S solution for 5 min, were destained in the distilled water, and the marker was labeled with indelible ink and was decolored for 10 min. The nitrocellulose was later soaked in 5% dry milk in Tris-buffered saline and Tween 20 (TBST) (blocking buffer) for 1 h at room temperature to ‘block’ the nonspecific binding of proteins. The nitrocellulose membrane was rinsed again with washing buffer and was incubated with the primary antibodies anti-AKT, anti-AKT phospho-S473, anti-p44/p42 MAPK (ERK1/2) and phospho-p44/p42 MAPK (ERK1/2) (Thr202/Tyr204) at 4°C overnight. The next day, the nitrocellulose was washed with TBST and was incubated with the horseradish peroxidase-conjugated goat anti-mouse IgG secondary antibody for 2 h at room temperature. The nitrocellulose membrane was washed and was incubated with DAB (diaminobenzidine) solution (substrate) at room temperature and was observed for color development. The membrane was soaked with distilled water to stop the substrate reaction. ERK1/2, AKT proteins were recognized as a band in the nitrocellulose membrane, and β-actin was used as the control. Membranes were developed by exposure to a film and proteins were visualized using a chemiluminescence detection system (Amersham ECL™ Select western blotting detection kit; GE Healthcare, Little Chalfont, U.K.) according to the manufacturer's protocol (GERPN2236). Imaging was performed using a digital camera in the dark room. The experiment was repeated at least three times, and the protein bands were quantified using ImageJ software.

In vivo efficacy of VGB1

Animal studies were conducted in conformity with the institutional guide for the care and use of laboratory animals. All protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of the Tehran University of Medical Sciences. Female BALB/c mice (5–7 weeks old) were purchased from Iran Pasteur Institute. Tumor cells (4T1; 1 × 106 cells/500 µl or 1 × 105 cells/50 µl) were injected subcutaneously into the right flanks of mice (n = 5). To generate the metastatic model, mouse mammary tumors (derived from 4T1 as a stock model) were aseptically separated from the breast cancer-bearing BALB/c mice, were cut into pieces of <0.3 cm3, and subcutaneously were transplanted into the animals, right flank under ketamine (100 mg/kg, i.p.) and xylazine (10 mg/kg, i.p.) anesthesia [33]. After 14 days, mice bearing tumors ∼370–400 mm3 were selected and randomized into control and treatment groups (six mice per group). The treatment group received 1, 2.5 and 10 mg/kg (once a day, i.p.) of VGB1, whereas the control group received equivalent volumes of phosphate-buffered saline (PBS) for 2 weeks. Tumor length and width were determined every 2 days by a digital Vernier caliper (INSTAR, China) and tumor volume was calculated using the following formula: tumor volume = length × width2 × 0.52.

Immunohistochemistry analysis

After excision, the tumor tissues were immediately stored in 4% neutral-buffered formalin. The samples were embedded in paraffin and sliced into 3–5 μm sections. Then, based on the type of immunohistochemistry analysis, the sections were stained with appropriate compounds. The relative areas of necrosis in tumors were analyzed by staining of sections with hematoxylin–eosin (H&E) [34]. The incidence of apoptotic tumor cells was assessed by the TUNEL assay followed by enzymatic development in DAB (Invitrogen, Carlsbad, CA, U.S.A.) detection and counter-staining in hematoxylin. To investigate and quantify the effect of VGB1 on the apoptosis of tumor cells, the sections were also stained by mouse anti-Rat/Rabbit Bcl-2 polyclonal antibody and Rabbit anti-p53 monoclonal antibody. Images were acquired by a microscope (Olympus BX51, Japan) equipped with a digital camera. Apoptosis markers were evaluated based on the estimated proportion of positive cells as well as the average staining intensity of the positive cells for Bcl-2 and p53. Nuclear and cytoplasmic staining were considered positive for p53 and Bcl-2, respectively.

The amount of MVD was determined using mouse anti-Rat/Rabbit CD31 and CD34 polyclonal antibodies according to the manufacturer's instructions to detect CD31 and CD34 expression. Briefly, the sections were deparaffinized and rehydrated. The slides were placed in TBS-EDTA buffer and were microwaved for 15 min at 90°C. To inhibit endogenous peroxidase activity, the slides were incubated in 0.3% H2O2 buffer for 15 min. The biotinylated secondary antibody and avidin–biotin complex with horseradish peroxidase were applied followed by the addition of chromogen 3,3′-diaminobenzidine (Sigma Chemical). The representative images of tumor tissues were taken by a microscope (Olympus BX51, Japan) equipped with a digital camera at 400× magnifications (40× lens, 10× ocular). CD31- and CD34-positive microvessels were counted in 10 randomly chosen fields. The results were expressed as the number of microvessels per field. The staining intensity was estimated by ImageJ software.

Far-UV CD measurement

Far-UV CD spectra were recorded on a JASCO spectropolarimeter (J-715, Tokyo, Japan) at 25°C. In the absence or presence of TFE (0–40% (v/v)), the peptide was dissolved in Tris 20 mM, pH 7.5. The final concentration of the peptide was 0.5 mg/ml. Three scans on average were taken from 190 to 250 nm, and the buffer spectra were subtracted from their corresponding sample spectra using the J-717 software provided by the manufacturer. The cuvette path length was 0.1 cm. The results are reported as molar ellipticity, [θ] molar, λ = 100 × θ/m × d, in units of degrees cm2 dmol−1[35].

Building model and structural refinement

A three-dimensional (3D) model of VGB1 was constructed using homology modeling in MODELLER, version 9.16 [36]. The 3D structure of VEGF-B available from the PDB (protein data bank; PDB code: 2C7W) was considered as a template. Ten models were generated by MODELLER software, version 9.16, and the quality of the modeled structures was evaluated using PROCHECK [37] and QMEAN [38] with full model analysis via the SWISS-MODEL web server (http://swissmodel.expasy.org/). MD (molecular dynamics) simulation was used to refine the structure obtained from homology modeling. Structural refinement was performed using GROMACS package, version 5.1.4 [39]. The interaction time-steps were set to 2 fs during the whole simulation time. The long-range electrostatic interactions were assessed employing the Particle-Mesh Ewald method with an interpolation order of 4 and maximum Fast Fourier Transforms (FFT) grid spacing of 0.16 nm. A cutoff radius of 1.0 nm was applied for Coulomb interactions, and the short-range interactions were used as van der Waals and were set to a cutoff of 1.0 nm [40]. Periodic boundary conditions were used and the bonds were constrained using the LINCS algorithm to their equilibrium state. The initial structure was solvated with three-point model water molecules in a cubic box of 1.0 nm to fully cover the protein system in water following periodic boundary conditions (PBCs). The generated system was then neutralized by adding Na and Cl ions by ‘genion tool’ plugin of GROMACS package. The system was minimized for 50 000 iterations using the steepest descent (SD) algorithm. The studied system was simulated under isothermal–isobaric ensemble (NPT) using velocity rescaling (modified Berendsen) temperature coupling and Parrinello–Rahman pressure coupling procedures for 5 ns at constant temperature and pressure of 300 K and 1 atm, respectively. The equilibration (NVT and NPT) was conducted for 50 000 steps. After implementation of minimization and equilibration methods, one system of the peptide was subjected to 200 ns production run of MD simulations employing 2 fs as the time step. Co-ordinates of all atoms were recorded in the form of .trr GROMACS trajectory file for every 10 ps. MD simulation was repeated two times, and two independent trajectories were obtained. The analysis was done for both independent trajectories. Because there was no difference between the analysis results, the results were reported for one of two trajectories.

Statistical analysis

The Prism software (version 6.00 for Windows, GraphPad Software, La Jolla, CA, U.S.A.; www.graphpad.com) was used for data analysis, for the production of graphs, and for statistical analysis. Data were presented as mean ± SEM. One-way ANOVA supported by Tukey's post hoc test was used to statistical significance for multiple comparisons and two-way repeated-measures ANOVA supported by Tukey's post hoc test was used for therapeutic efficacy in affecting tumor growth. A value of P < 0.05 was considered significant.

Results

Peptide design

Five binding regions were identified in the complex between VEGF-B and VEGFR-1 D2 [41]: (i) α1 helix, including binding residues Trp17, Ile18, Tyr21, Thr25, Cys26 and Gln27; (ii) Val48 within loop1 (α2–β3 loop); (iii) loop region 60–67 (β3–β4 loop), including binding residues Pro62, Asp63, Gly65 and Leu66; (iv) the segment 79–93 within loop3 (β5–β6 loop), including binding residues Gln79, Leu81, Ile83, Ser88, Gln89 and Leu90 and (v) the C-terminal segment of β6, including Glu102, Cys103, Arg104 and Pro105. In the present study, we aimed to design a peptide that inhibits the interaction of VEGF-A with VEGFR-1. A recent study has demonstrated that peptides mimicking the α1 helix of VEGF-A or -B composed of amino acids 16–25 can bind to VEGFR-1 and inhibits angiogenesis [42]. In addition, it was found that the antagonist peptides derived from the α1 helix of VEGF-B bind to VEGFR-1 with higher affinity than those from VEGF-A. However, no information was provided on the in vivo activity of the peptide variant. Accordingly, we designed a peptide to mimic the α1 helix of VEGF-B. Notably, the segment next to the α1 helix of VEGF-B, including amino acids 26–35, was also shown to target VEGFR-1; a phage display study discovered a cyclic peptide CPQPRPLC composed of the short sequences Cys26-Gln27-Pro28-Arg29 and Pro34-Leu35 from VEGF-B that specifically targets VEGFR-1 and abrogates the angiogenesis in vitro and in vivo [4345]. Accordingly, VEGF-B segments 16–25, 26–29 and 34–35 were selected to be covalently linked into a single peptide as SWIDVYTRATCQPRPL. To increase peptide half-life in circulation, a disulfide bond was incorporated between a conveniently incorporated cysteine residue at the N-terminal and the cysteine residue available at the C-terminal side of the α-helix segment. Therefore, the sequence of the designed 17-amino acid peptide (referred to as VGB1) was 2HN-CSWIDVYTRATCQPRPL-COOH.

Peptide-binding assays

To evaluate the cell surface-binding capability of VGB1, HUVECs were incubated with various concentrations of FITC-conjugated VGB1 (0.58, 0.88 and 1.18 μM) and analyzed by flow cytometry. As shown in Figure 1A, fluorescence intensity was considerably increased compared with untreated HUVECs with increasing the concentration of FITC-VGB1. To establish whether VGB1 binding was mediated by VEGFR-1 and VEGFR-2, HUVECs were incubated either with anti-VEGFR-1 primary antibody/PE-labeled anti-mouse secondary antibody or with anti-VEGFR-2 primary antibody/FITC-labeled anti-mouse secondary antibody with (0.58, 0.88 and 1.18 μM) or without pretreatment with VGB1. As shown in Figure 1B,C, in HUVECs treated with VGB1 in the range of 0.58–1.18 μM, the fluorescent signals were decreased dose-dependently compared with those in the control group. These data show that VGB1 binds to both VEGFR-1 and VEGFR-2 on the surface of the endothelial cells and competes with antibodies which identify the ectodomains of the receptors. Next, fluorescence microscopy was used to confirm the ability of VGB1 to bind to VEGF receptors. To investigate the ability of VGB1 to bind VEGFR-1, we conducted a cell-binding assay using HUVECs by exposing them to different concentrations of VGB1 (0–1.18 μM), followed by the addition of anti-VEGFR-1 (5 µg/ml) together with PE-secondary antibodies to the medium. The binding of the anti-VEGFR-1/PE-secondary antibodies to HUVECs was significantly inhibited compared with untreated cells at 0.88 μM of VGB1, and the inhibition of binding was almost completed at 1.18 μM of VGB1 (P < 0.0001) (Figure 2A,C), confirming that VGB1 recognizes VEGFR-1.

Receptor-binding assays using flow cytometry.

Figure 1.
Receptor-binding assays using flow cytometry.

(A) HUVECs were incubated with various concentrations of FITC-VGB1 (0.58, 0.88 and 1.18 µM) (green curve). Autofluorescence of cells was colored in red. HUVECs were incubated either with (B) anti-VEGFR-1 primary antibody/PE-conjugated secondary antibody or with (C) anti-VEGFR-2 primary antibody/FITC-conjugated secondary antibody in the absence (green curve) or the presence of various concentrations of VGB1 (0.58, 0.88 and 1.18 µM) (blue curve) overnight in the dark. Next, flow cytometric analysis was executed using BD FACSCalibur Flow Cytometer.

Figure 1.
Receptor-binding assays using flow cytometry.

(A) HUVECs were incubated with various concentrations of FITC-VGB1 (0.58, 0.88 and 1.18 µM) (green curve). Autofluorescence of cells was colored in red. HUVECs were incubated either with (B) anti-VEGFR-1 primary antibody/PE-conjugated secondary antibody or with (C) anti-VEGFR-2 primary antibody/FITC-conjugated secondary antibody in the absence (green curve) or the presence of various concentrations of VGB1 (0.58, 0.88 and 1.18 µM) (blue curve) overnight in the dark. Next, flow cytometric analysis was executed using BD FACSCalibur Flow Cytometer.

Binding of VGB1 to VEGFR-1 and VEGFR-2 on HUVECs.

Figure 2.
Binding of VGB1 to VEGFR-1 and VEGFR-2 on HUVECs.

An immunocytochemical assay was performed as described in the Experimental section. HUVECs were treated with various VGB1 concentrations (0–1.18 µM). All images were displayed at the same scale. VGB1 inhibited binding of (A) anti-VEGFR-1 (5 μg/ml) together with PE-secondary anti-mouse antibody (red) to VEGFR-1 and (B) anti-VEGFR-2 (5 μg/ml) together with FITC-secondary anti-mouse antibody (green) to VEGFR-2. (C) Quantitative analysis was performed using ImageJ in VGB1-treated HUVECs and untreated cells. VGB1 significantly blocked the binding of VEGF at 1.18 µM. All images were captured by Olympus BX-51, Japan. Scale bar, 20 μm. Error bars signify ±SEM. n = 3, *P = 0.034, **P = 0.0024, ****P < 0.0001, ns = not significant; one-way ANOVA.

Figure 2.
Binding of VGB1 to VEGFR-1 and VEGFR-2 on HUVECs.

An immunocytochemical assay was performed as described in the Experimental section. HUVECs were treated with various VGB1 concentrations (0–1.18 µM). All images were displayed at the same scale. VGB1 inhibited binding of (A) anti-VEGFR-1 (5 μg/ml) together with PE-secondary anti-mouse antibody (red) to VEGFR-1 and (B) anti-VEGFR-2 (5 μg/ml) together with FITC-secondary anti-mouse antibody (green) to VEGFR-2. (C) Quantitative analysis was performed using ImageJ in VGB1-treated HUVECs and untreated cells. VGB1 significantly blocked the binding of VEGF at 1.18 µM. All images were captured by Olympus BX-51, Japan. Scale bar, 20 μm. Error bars signify ±SEM. n = 3, *P = 0.034, **P = 0.0024, ****P < 0.0001, ns = not significant; one-way ANOVA.

Given that VEGF-B has ∼80% sequence similarity with VEGF-A in the selected region, the ability of VGB1 for binding to VEGFR-2 was also examined in HUVECs (Figure 2B,C). Surprisingly, VGB1 significantly blocked the binding of the anti-VEGFR-2/FITC-secondary antibodies at 0.88 μM (P < 0.0001) and inhibition was almost completed at 1.18 μM (P < 0.0001) (Figure 2C), supporting that VGB1 also binds to VEGFR-2.

Proliferation assay

VEGF/VEGFR-2 system plays a critical role in endothelial and tumoral cell proliferation [46]. Hence, the VEGF-dependent proliferation of HUVECs and two tumor cell lines including 4T1 mammary carcinoma and U87 glioblastoma cells were assessed by the MTT assay. These cell lines express different levels of VEGFR-1 and VEGFR-2 on the cell surface: HUVEC expresses VEGFR-2 more than VEGFR-1, 4T1 mammary carcinoma cell line expresses VEGFR-1 more than VEGFR-2, and U87 glioblastoma cell line mostly expresses VEGFR-2. In the presence of a very high concentration of VEGF-A (0.2 µg/ml), the cells were treated with VGB1 (0.15–1.18 μM), and results were analyzed in comparison with untreated controls. However, no meaningful difference was observed in response to 0.58 μM of VGB1 (Figure 3B), a significant effect was obtained when 4T1 mammary carcinoma cells were treated with 0.88 μM of VGB1 when compared with control (P = 0.006). Even more strikingly, VGB1 inhibited the VEGF-induced proliferation of HUVECs and U87 glioblastoma cells at dosage of 0.58 μM (Figure 3A,C) (P ≤ 0.01 and P = 0.006, respectively), further confirming that VGB1 neutralizes VEGFR-2.

The effect of VGB1 on VEGF-stimulated cell proliferation and migration.

Figure 3.
The effect of VGB1 on VEGF-stimulated cell proliferation and migration.

The effect of VGB1 (0.15–1.18 µM) on proliferation of (A) HUVE, (B) 4T1 mammary carcinoma and (C) U87 glioblastoma cells in the presence (black columns) or absence (red columns) of VEGF (0.2 µg/ml). The results of three independent experiments expressed as inhibition percentage relative to the control. Error bars signify ±SEM. n = 3; **P ≤ 0.01, ***P = 0.006, ****P < 0.0001, ns = not significant; one-way ANOVA. (D) The effect of VGB1 (0.58–1.18 µM) on HUVEC wound closure migration after 24 h. Image analysis performed by Wimasis image analysis software is illustrated in the bottom panel (a′–d′). The green overlay and the gray area in the image display the cell-covered area and the scratch area, respectively. Scale bar, 20 µm. (E) Quantitative results of filling the scratched area by migrating cells. Results are expressed as the percentage of migration for the treatment group versus the corresponding control group. Error bars signify ±SEM. n = 3; *P < 0.01, ****P < 0.0001; one-way ANOVA.

Figure 3.
The effect of VGB1 on VEGF-stimulated cell proliferation and migration.

The effect of VGB1 (0.15–1.18 µM) on proliferation of (A) HUVE, (B) 4T1 mammary carcinoma and (C) U87 glioblastoma cells in the presence (black columns) or absence (red columns) of VEGF (0.2 µg/ml). The results of three independent experiments expressed as inhibition percentage relative to the control. Error bars signify ±SEM. n = 3; **P ≤ 0.01, ***P = 0.006, ****P < 0.0001, ns = not significant; one-way ANOVA. (D) The effect of VGB1 (0.58–1.18 µM) on HUVEC wound closure migration after 24 h. Image analysis performed by Wimasis image analysis software is illustrated in the bottom panel (a′–d′). The green overlay and the gray area in the image display the cell-covered area and the scratch area, respectively. Scale bar, 20 µm. (E) Quantitative results of filling the scratched area by migrating cells. Results are expressed as the percentage of migration for the treatment group versus the corresponding control group. Error bars signify ±SEM. n = 3; *P < 0.01, ****P < 0.0001; one-way ANOVA.

Migration assay

To assess if binding of VGB1 to VEGFR-1 and VEGFR-2 could also inhibit endothelial cell migration, a scratch wound healing assay was performed after stimulation of HUVECs with VEGF-A (0.2 µg/ml). After washing the detached cells, cell monolayer was treated with various concentrations of VGB1 (0–1.18 μM) for 24 h. VEGF-A-stimulated HUVECs filled the scratched area, whereas VGB1 inhibited the migration of ECs in a dose-dependent manner (P < 0.02) (Figure 3D,E), indicating that VGB1 could suppress the VEGF-induced migration of ECs.

Capillary-like tube formation assay

To further evaluate the anti-angiogenic property of VGB1, we determined the potential of VGB1 to inhibit the VEGF-induced capillary-like tube formation of ECs. The VEGF-A stimulated the formation of tubular structures in HUVECs, but a dose-dependent inhibition of two-dimensional tube formation was observed for VGB1-treated cells, and the suppression of tube formation was completed at 1.18 μM (Figure 4A). In addition, detailed analysis using Wimasis (http://www.wimasis.com/en/) software [47] revealed that different aspects of tubular structures, including total tubes, total tube length and total branching points, were inhibited by VGB1 compared with untreated control in a dose-dependent manner (P < 0.05) (Figure 4B).

Inhibition of tube formation and 3D sprouting in HUVECs with VGB1.

Figure 4.
Inhibition of tube formation and 3D sprouting in HUVECs with VGB1.

(A) Representative images of anti-tubologenesis activity of VGB1 in VEGF (0.2 µg/ml)-induced HUVECs. Geltrex matrix was treated with increasing concentrations of VGB1 (left panel; a–e: 0, 0.3, 0.58, 0.88 and 1.18 µM, respectively). Image analysis achieved by ImageJ is illustrated in the right panel (a′–e′). (B) Analysis of total tube, total tube length and total branching points. Error bars signify ±SEM. n = 3; *P = 0.04, **P = 0.001, ***P = 0.0001, ****P < 0.0001, ns = not significant; one-way ANOVA. (C) Representative images of anti-sprouting activity of VGB1 (left panel; a–d: 0, 0.58, 0.88 and 1.18 µM, respectively) in VEGF (0.2 µg/ml)-stimulated HUVECs. Image analysis performed by ImageJ is illustrated in the right panel (a′–d′). (D) The sprouts area, number of sprouts and mean sprout length were analyzed per field. Results were from three random fields. Error bars signify ±SEM. n = 3; **P = 0.002, ***P = 0.0005, ****P < 0.0001, ns = not significant; one-way ANOVA.

Figure 4.
Inhibition of tube formation and 3D sprouting in HUVECs with VGB1.

(A) Representative images of anti-tubologenesis activity of VGB1 in VEGF (0.2 µg/ml)-induced HUVECs. Geltrex matrix was treated with increasing concentrations of VGB1 (left panel; a–e: 0, 0.3, 0.58, 0.88 and 1.18 µM, respectively). Image analysis achieved by ImageJ is illustrated in the right panel (a′–e′). (B) Analysis of total tube, total tube length and total branching points. Error bars signify ±SEM. n = 3; *P = 0.04, **P = 0.001, ***P = 0.0001, ****P < 0.0001, ns = not significant; one-way ANOVA. (C) Representative images of anti-sprouting activity of VGB1 (left panel; a–d: 0, 0.58, 0.88 and 1.18 µM, respectively) in VEGF (0.2 µg/ml)-stimulated HUVECs. Image analysis performed by ImageJ is illustrated in the right panel (a′–d′). (D) The sprouts area, number of sprouts and mean sprout length were analyzed per field. Results were from three random fields. Error bars signify ±SEM. n = 3; **P = 0.002, ***P = 0.0005, ****P < 0.0001, ns = not significant; one-way ANOVA.

In vitro sprout formation assay

To further investigate the effect of VGB1 on angiogenesis of HUVECs in vitro, we performed a 3D sprout formation assay. In agreement with the results of the 2D tube formation assay, VGB1 inhibited the angiogenesis of ECs dose-dependently. The analysis of results by Wimsprout (http://www.wimasis.com/en/) software [47] revealed a significant decrease in the sprouts area, the number of sprouts and mean sprout length in the presence of 1.18 μM VGB1 against control (P < 0.0001) (Figure 4C,D).

Western blot analysis

VEGF binding to VEGFR-1 and VEGFR-2 on the endothelial cell surface activates MAPK/ERK1/2 and PI3K/AKT signaling pathways, leading to proliferation, migration and tube formation [12,14,48]. To address the mechanism by which VGB1 inhibited the hallmarks of angiogenesis, we evaluated the expression and phosphorylation of ERK1/2 and AKT in HUVECs. Quantitative western blot analysis demonstrated a marked inhibition of p-AKT and p-ERK1/2 levels in VEGF (0.2 µg/ml)-stimulated HUVECs after treatment with VGB1 (0.58 and 1.18 μM) compared with control (P < 0.0001), whereas the levels of ERK1/2 and AKT remained unaffected (Figure 5A,B). These results demonstrate that VGB1 suppressed different hallmarks of angiogenesis by neutralization of VEGFR-1 and -2 in ECs followed by inhibition of their downstream signaling pathways.

Abrogation of VEGF-driven signaling in EC and growth of mammary carcinoma tumors in BALB/c mice.

Figure 5.
Abrogation of VEGF-driven signaling in EC and growth of mammary carcinoma tumors in BALB/c mice.

(A) ERK1/2 and AKT phosphorylation were evaluated by western blot analysis. HUVECs were stimulated by 0.2 µg/ml VEGFA and were treated with PBS (control) and VGB1 (0.58 and 1.18 µM). (B) Quantification of ERK1/2 and AKT phosphorylation was done by densitometry using Image J. The data are illustrative of three independent experiments and are presented as the relative expression of p-ERK1/2 and p-AKT normalized to total ERK1/2 and AKT. β-Actin was used as the loading control. Error bars signify ±SEM. n = 3; ****P < 0.0001, ns = not significant; one-way ANOVA. (C) 5–7-week-old female BALB/c mice were used for the investigation of VGB1 efficacy in vivo. The treatment groups received 1, 2.5 and 10 mg/kg (once a day, i.p.) of VGB1, whereas the control group received equivalent volumes of PBS for a period of 2 weeks. The treatment with VGB1 resulted in significant inhibition of tumor growth versus control. Error bars signify ±SEM. n = 6; ***P < 0.001, ****P < 0.0001; two-way ANOVA. (D) Exploration of toxicity of VGB1. The weight of mice was monitored at the treatment period. In the treatment groups, as in the control group (which receive PBS), no decrease in weight of the mice was observed in the treatment period. Error bars signify ±SEM.

Figure 5.
Abrogation of VEGF-driven signaling in EC and growth of mammary carcinoma tumors in BALB/c mice.

(A) ERK1/2 and AKT phosphorylation were evaluated by western blot analysis. HUVECs were stimulated by 0.2 µg/ml VEGFA and were treated with PBS (control) and VGB1 (0.58 and 1.18 µM). (B) Quantification of ERK1/2 and AKT phosphorylation was done by densitometry using Image J. The data are illustrative of three independent experiments and are presented as the relative expression of p-ERK1/2 and p-AKT normalized to total ERK1/2 and AKT. β-Actin was used as the loading control. Error bars signify ±SEM. n = 3; ****P < 0.0001, ns = not significant; one-way ANOVA. (C) 5–7-week-old female BALB/c mice were used for the investigation of VGB1 efficacy in vivo. The treatment groups received 1, 2.5 and 10 mg/kg (once a day, i.p.) of VGB1, whereas the control group received equivalent volumes of PBS for a period of 2 weeks. The treatment with VGB1 resulted in significant inhibition of tumor growth versus control. Error bars signify ±SEM. n = 6; ***P < 0.001, ****P < 0.0001; two-way ANOVA. (D) Exploration of toxicity of VGB1. The weight of mice was monitored at the treatment period. In the treatment groups, as in the control group (which receive PBS), no decrease in weight of the mice was observed in the treatment period. Error bars signify ±SEM.

VGB1 inhibits tumor growth in vivo

The potential of VGB1 for the inhibition of tumor growth was investigated in 4T1 murine mammary carcinoma tumor model. Mice were daily treated with 1, 2.5 and 10 mg/kg of VGB1 for 2 weeks (n = 6/group, i.p.). On day 28, the average tumor volume in groups treated with 1, 2.5 and 10 mg/kg was 594.8, 533.2 and 586.5 mm3, respectively, indicating a marked tumor growth regression compared with the PBS-treated group (1385.0 mm3) (P < 0.0001, Figure 5C). Accordingly, inhibition of tumor growth was 57.1, 61.5 and 57.6% at dosages 1, 2.5 and 10 mg/kg, respectively, indicating that efficacy of VGB1 was not dose responsive between 1 and 10 mg/kg. During the experiment, VGB1-treated mice gained weight (Figure 5D), and no mortality and other adverse effects, such as skin ulcerations or toxic death, were observed (data not shown), which indicates that VGB1 is nontoxic at the dosages used in this work.

Immunohistochemical studies

To assess the effect of VGB1 on tumor cell proliferation, MVD and apoptosis, we stained 4T1 mammary carcinoma tumor sections at the end of treatment (28 days after implantation) and examined them for tumor changes.

The effect of VGB1 on the proliferation of tumor cells was measured by staining of tumor sections with the anti-Ki67 antibody. Tumor cell proliferation was significantly decreased in tumors treated with dosages 1 and 10 mg/kg compared with controls (P < 0.0001), and no significant differences were observed between VGB1-treated tumors (Figure 6A,D).

Immunohistochemical analysis of tumor proliferation and angiogenesis.

Figure 6.
Immunohistochemical analysis of tumor proliferation and angiogenesis.

Digital photographs were taken at 400× magnification with a microscope (Olympus BX51, Japan) equipped with a digital camera. 4T1 mammary carcinoma tumor tissues (were treated with 1 and 10 mg/kg of VGB1) were removed and cut into 4 µm thick slices. Representative images of (A) Ki67 staining as a cellular marker that correlated with cell proliferation; (B) CD31 and (C) CD34 staining as indicators of tumor microvessel density. Scale bar, 20 µm. (D) Quantification of Ki67, CD31 and CD34. Ki67-positive cells and number of microvessel per field for both markers. CD31 and CD34 were significantly decreased in treated tumors compared with PBS-treated tumors. Error bars signify ±SEM. n = 3; *P = 0.01, **P = 0.002, ***P = 0.0001, ****P < 0.0001 and ns = not significant; one-way ANOVA.

Figure 6.
Immunohistochemical analysis of tumor proliferation and angiogenesis.

Digital photographs were taken at 400× magnification with a microscope (Olympus BX51, Japan) equipped with a digital camera. 4T1 mammary carcinoma tumor tissues (were treated with 1 and 10 mg/kg of VGB1) were removed and cut into 4 µm thick slices. Representative images of (A) Ki67 staining as a cellular marker that correlated with cell proliferation; (B) CD31 and (C) CD34 staining as indicators of tumor microvessel density. Scale bar, 20 µm. (D) Quantification of Ki67, CD31 and CD34. Ki67-positive cells and number of microvessel per field for both markers. CD31 and CD34 were significantly decreased in treated tumors compared with PBS-treated tumors. Error bars signify ±SEM. n = 3; *P = 0.01, **P = 0.002, ***P = 0.0001, ****P < 0.0001 and ns = not significant; one-way ANOVA.

The effect of VGB1 treatment on MVD in tumor tissues was assessed by immunohistochemical examination of CD31 and CD34 levels. A notable reduction in CD31- and CD34-positive microvessels was observed in tumors treated with 1 and 10 mg/kg of VGB1 compared with untreated tumors (P < 0.0001) (Figure 6B–D). Also, immunohistochemical analysis showed a reduced number of microvessel per field in the tumors treated with 10 mg/kg compared with 1 mg/kg of VGB1 (P < 0.001). These results imply the VGB1-driven suppression of tumor angiogenesis.

Inhibition of angiogenesis results in apoptosis induction in tumors [49]. To assess whether the inhibition of VGB1-mediated tumor growth was also associated with the apoptosis induction, tumor tissues were analyzed by TUNEL, Bcl-2 and p53 staining. TUNEL staining revealed an increased number of apoptotic cells in VGB1-treated tumors compared with control tumors (P < 0.001 and P < 0.0001 for tumors treated with 1 and 10 mg/kg, respectively) (Figure 7A,E). Similarly, a significantly increased number of p53-positive cells were observed in peptide-treated tumors compared with those from control animals (P < 0.0001) (Figure 7B,E). Moreover, unlike in PBS-treated tumors, a limited number of cells were stained with anti-Bcl-2 antibody in VGB1-treated tumors (P = 0.0008) (Figure 7C,E). Finally, promotion of cell death in the VGB1-treated tumors was further confirmed by results of H&E staining, which indicated a considerable modification in the cell morphology as a result of VGB1 administration (Figure 7D). These results show that the anti-tumor property of VGB1 is also associated with apoptosis induction in tumor tissues.

Immunohistochemical analysis of apoptosis induction in tumor tissues.

Figure 7.
Immunohistochemical analysis of apoptosis induction in tumor tissues.

Digital photographs were taken at 100× or 400× magnification with a microscope (Olympus BX51, Japan) equipped with a digital camera. 4T1 mammary carcinoma tumor tissues (were treated with 1 and 10 mg/kg of VGB1) were removed and cut into 4 µm thick slices. Representative images of (A) TUNEL staining as a marker of late apoptosis; and (B) p53 and (C) Bcl-2 staining as markers of early apoptosis; (D) H&E staining showing dead cells. Scale bars, 100 and 20 µm. (E) Quantification of TUNEL, Bcl-2, and p53. Anti-apoptotic protein Bcl-2 notably reduced and TUNEL staining and apoptotic factor p53 markedly increased in VGB1-treated tumors compared with control. Error bars signify ±SEM. n = 3; *P = 0.01, **P < 0.002, ***P = 0.0004, ****P < 0.0001 and ns = not significant; one-way ANOVA.

Figure 7.
Immunohistochemical analysis of apoptosis induction in tumor tissues.

Digital photographs were taken at 100× or 400× magnification with a microscope (Olympus BX51, Japan) equipped with a digital camera. 4T1 mammary carcinoma tumor tissues (were treated with 1 and 10 mg/kg of VGB1) were removed and cut into 4 µm thick slices. Representative images of (A) TUNEL staining as a marker of late apoptosis; and (B) p53 and (C) Bcl-2 staining as markers of early apoptosis; (D) H&E staining showing dead cells. Scale bars, 100 and 20 µm. (E) Quantification of TUNEL, Bcl-2, and p53. Anti-apoptotic protein Bcl-2 notably reduced and TUNEL staining and apoptotic factor p53 markedly increased in VGB1-treated tumors compared with control. Error bars signify ±SEM. n = 3; *P = 0.01, **P < 0.002, ***P = 0.0004, ****P < 0.0001 and ns = not significant; one-way ANOVA.

Structural analysis by CD

The far-UV CD studies were performed to estimate this probability that whether the designed peptide can resemble the α1 helix region of VEGF-B. According to the molar ellipticity at 190, 208 and 222 nm at pH 7.5, the helical content of the peptide was 15.1%. To further assess the helix propensity of the peptide, far-UV CD spectra were recorded in the presence of increasing concentrations of 2,2,2-trifluoroethanol (TFE). Importantly, the helical content was increased with increasing the TFE concentration up to 30% (v/v), while no further increase in α-helix structure was observed when VGB1 dissolved in 40% (v/v) (Figure 8A). The estimated helical structure of VGB1 was 16.3, 30.2, 37.3 and 33.9% in the presence of 10, 20, 30 and 40% (v/v) TFE, respectively. These results demonstrate the helical propensity of VGB1.

Structural analysis of VGB1 and the structure obtained by modeling.

Figure 8.
Structural analysis of VGB1 and the structure obtained by modeling.

(A) The far-UV CD spectrum of the peptide in the absence or presence of different concentrations of TFE (0–40% (v/v)). As shown, VGB1 helicity increases with increasing the TFE concentration up to 30% (v/v). (B) A 3D view of VEGF-B (PDB code: 2C7W) (left) and a presentation of the peptide model obtained by the MODELLER program, version 9.16 (right). The selected region corresponding to the VGB1 sequence in VEGF-B is shown in orange, comprising α-helix, β-strand and random coil. (C) DSSP analysis suggests the probability of α-helix formation in VGB1.

Figure 8.
Structural analysis of VGB1 and the structure obtained by modeling.

(A) The far-UV CD spectrum of the peptide in the absence or presence of different concentrations of TFE (0–40% (v/v)). As shown, VGB1 helicity increases with increasing the TFE concentration up to 30% (v/v). (B) A 3D view of VEGF-B (PDB code: 2C7W) (left) and a presentation of the peptide model obtained by the MODELLER program, version 9.16 (right). The selected region corresponding to the VGB1 sequence in VEGF-B is shown in orange, comprising α-helix, β-strand and random coil. (C) DSSP analysis suggests the probability of α-helix formation in VGB1.

Building model and structural refinement

A 3D structure of the peptide was made using MODELLER program, version 9.16 [36]. Ten models were obtained by MODELLER program, and Ramachandran plot for all of the models was provided using the PROCHECK web tool. In the all of models, either 92.3% or 84.6% of residues presented in the most favored regions, and no residue was located in the disallowed regions of the Ramachandran plot (Supplementary Figure S3A,B). Ideally, a good quality model should have over 90% of the residues in the most favored regions [37]. Thus, 5 out of 10 models that had 92.3% of the residues in these regions were subjected to further analysis via QMEAN and the SWISS-MODEL web server (http://swissmodel.expasy.org/). Ramachandran analysis conducted by the SWISS-MODEL web server for every five models has shown over 90% of the residues in the favored regions that it was in agreement with the results from PROCHECK. QMEAN Z-scores for these models were −2.44, −2.3, −2.24, −2 and −0.96 (Supplementary Figure S4). Higher Z-scores correlate with more favorable models [50]. Based on this point, the model which has QMEAN Z-score equal to −0.96, 100% of its residues have been located in the most favored regions (according to the structure assessment by SWISS-MODEL) and were selected and subjected to the energy minimization. The bigger proteins tend to have lower energies and higher QMEAN scores [50], and thus, the negative value of Z-score is arising from the small size of the modeled peptide (17 residues). Molecular dynamics simulation was employed to allow conformational relaxation of the structure. The last generated structure in the trajectory was chosen as the final model (Figure 8B and Supplementary Material).

Dictionary of secondary structure of proteins (DSSPs) define that the secondary structure of peptides or proteins gave a set of 3D co-ordinates [51]. DSSP analysis for VGB1 throughout the simulation time showed that the segment Trp3-Ile4-Asp5-Val6-Tyr7-Thr8 folds into an α-helix structure (Figure 8C), suggesting that the α-helical structure can form in VGB1 at the approximately identical amino acids as VEGF-B.

Discussion

Both VEGFR-1 and -2 are implicated in physiological and pathological angiogenesis. Signaling through VEGFR-2 leads to proliferation, permeability, migration and tube formation of ECs [52], and VEGFR-1 triggers EC migration, survival and tube formation [53], emerging them as a therapeutic target to combat immoderate angiogenesis. Obviously, compared with the blockade of either VEGFR-1 or VEGFR-2, their simultaneous targeting is advantageous for anti-angiogenic therapy. Further evidence in agreement with this proposal is our recent observations that dual specific peptides, which bind to both VEGFR-1 and VEGFR-2, could effectively inhibit the angiogenesis even more potently than bevacizumab, which exclusively targets VEGF-A [23,54].

The receptor-binding properties of antagonist peptides issued from the α1 helix of VEGF-A, -B or placenta growth factor (PlGF) have been studied in detail [42,5559]. In the case of peptides derived from VEGF-B, previous studies deal with the binding to VEGFR-1 D2, and there are no data concerning their binding to VEGFR-2. Here, for the first time, we provided evidence that a peptide mimicking the α1 helix of VEGF-B and its adjacent C-terminal region can bind to both VEGFR-1 and VEGFR-2. First, our receptor-binding assays using flow cytometry and immunocytochemistry revealed that VGB1 behaves as a dual specific antagonist that binds to both receptors. The unexpected result that a peptide derived from VEGF-B could bind to VEGFR-2 was confirmed by its ability to inhibit the VEGF (0.2 µg/ml)-induced proliferation of U87 glioblastoma cell lines, which highly express VEGFR-2 on the cell surface, with an IC50 (half maximal inhibitory concentration) value of 0.7 µM. At the same concentration of VEGF, VGB1 also inhibited the proliferation of HUVECs and 4T1 mammary carcinoma cells, which express VEGFR-1 and VEGFR-2, with the IC50 values of 1.15 and 1.28 µM, respectively. Considering that signaling through VEGFR-2 but not VEGFR-1 triggers the HUVEC proliferation [60], the anti-endothelial property of VGB1 can be ascribed to the blockade of VEGFR-2. But, the convincing data support the antagonizing of VEGFR-1 and -2 provided by quantitative western blot analysis of AKT, p-AKT, ERK1/2 and p-ERK1/2 in HUVECs. VEGFR-2 signaling is mainly through the MAPK/ERK1/2 axis and, to a lesser extent, by the PI3K/AKT signaling pathway [52], whereas activation of VEGFR-1 is through the PI3K/AKT signaling pathway [61]. As a consequence, the dual blockade of VEGFR-1 and -2 is expected to substantially abrogate both pathways. As expected, the treatment of HUVECs with VGB1 (0.58 µM) in the presence of VEGF (0.2 µg/ml) led to the similar inhibition of AKT and ERK1/2 phosphorylation by 88 and 76%, respectively, compared with controls (P < 0.0001). Suppression of ERK1/2 phosphorylation is due to the blockade of VEGFR-2 but not VEGFR-1 [61], whereas the pronounced inhibition of AKT phosphorylation is attributable to the neutralization of both receptors [62,63]. This leads to the conclusion that VGB1 abrogated PI3K/AKT and MAPK/ERK1/2 pathways through neutralization of both VEGFR-1 and -2. VEGF-induced activation of ERK1/2 and/or AKT is required for several angiogenesis-related changes, including proliferation, migration and capillary-like vessel formation of ECs as well as tumor angiogenesis, growth and metastasis [61]. Our detailed in vitro studies demonstrated that different aspects of angiogenesis were inhibited by VGB1. As mentioned above, VGB1 inhibited VEGF-stimulated proliferation of ECs in a dose-dependent manner. In addition, VGB1 inhibited VEGF-stimulated migration of ECs. More data on the anti-angiogenic property of VGB1 were provided by two- and three-dimensional tube formation assays. VGB1 remarkably inhibited the formation of tubular structures grown on Geltrex matrix as well as three-dimensional microvessel formation by HUVECs in a dose-dependent manner. Remarkably, VGB1 led to the regression of murine 4T1 mammary carcinoma tumors by reduced CD31, CD34 and Ki67 expression in treated animals, suggesting that inhibition of angiogenesis was a mechanism of VGB1 efficacy. The anti-tumor effect of VGB1 was also associated with the induction of apoptosis in tumor tissues, as indicated by decreased expression of Bcl-2 and increased TUNEL staining and p53 expression. On the basis of these results, we indicated that through neutralization of VEGFR-1 and -2 followed by abrogation of PI3K/AKT and MAPK/ERK1/2 pathways in ECs, VGB1 suppressed critical steps of VEGF-induced angiogenesis, including proliferation, migration and tube formation of ECs as well as tumor angiogenesis and growth.

Basile et al. revealed that antagonist peptides derived from the α1 helix of VEGF-A and/or -B can inhibit angiogenesis in vitro. They showed that a peptide mimicking the α1 helix of VEGF-A, -B and PlGF (referred to as Peptide1) [55] can abolish the VEGF (25 ng/ml)-stimulated proliferation of HUVECs at 12.5 nM. By comparison, VGB1 completely inhibited the proliferative effect of a very high concentration of VEGF (200 ng/ml) on HUVECs at 0.88 µM (P < 0.0001). They also indicated a 36% reduction in A375 melanoma tumor growth after treatment with 200 nM Peptide1 compared with controls, whereas we observed that inhibition of 4T1 mammary carcinoma tumors was ∼57% at the VGB1 dosage of 1 mg/kg (∼120 nM) compared with PBS-treated controls (P < 0.001). Finally, they found that microvessel density, as indicated by CD31 expression in tumor tissues, inhibited by as 50% after 7-day treatment of Peptide1, while we indicated that VGB1 inhibited the microvessel density as ∼85% at the end of 14-day administration. Since no data are available for the inhibition of two- or three-dimensional tube formation in the previously reported antagonist peptides from the α1 helix, we are unable to compare the anti-angiogenic property of VGB1 with previously reported peptides. Wang et al. [42] have recently reported that antagonist peptides derived from the α1 helix, loop1 and loop3 of VEGF-A, VEGF-B and PlGF can bind to the VEGFR-1 with low, moderate and high affinity, respectively. They have not provided information on the anti-angiogenic activity and VEGFR-2-binding capability for peptides issued from the α1 helix. However, their best peptide variant, a peptide mimicking the loop1 of VEGF-B, could inhibit HUVEC capillary tube formation on Matrigel as potent as bevacizumab. Based on our tube formation assays, VGB1 inhibited the VEGF-A (0.2 µg/ml)-stimulated HUVEC growth on Geltrex™ matrix, so that the total number of tubes was reduced from 57 to 12 at 1.18 µM, whereas, in conditions quite similar to the present research, we have recently indicated that bevacizumab (2.14 µM) reduced the total number of tubes from 150 to 46 [54], suggesting that the VGB1 is more potent than the peptide reported by Wang et al. [42].

The anti-angiogenic and anti-tumor properties of VGB1 are also attributable to the segment CQPRPL at C-terminus. Of note, CPQPRPLC peptide was shown to interact with VEGFR-1 and NRP-1 but not with VEGFR-2 [4345], suggesting that α1 helix region is responsible for the VEGR-2-binding and segment CQPRPL does not participate in the VEGFR-2-binding property of VGB1. Giordano et al. [44] demonstrated that tripeptide RPL, as well as D(LPR), is a critical motif for receptor binding to NRP-1 and to VEGFR-1. They also indicated that D(LPR) acts as a potent inhibitor of angiogenesis in vivo. These findings suggest that the blockade of VEGFR-1 and, consequently, downstream signaling of VEGFR-1 by VGB1 resulted at least, in part, from its C-terminal segment CQPRPL.

Our structural analysis using far-UV CD suggested that VGB1 can adopt α-helix structure and high helical propensity of the peptide was also deduced from TFE-state structural estimation. In accordance to the data obtained by far-UV CD, theoretical analysis of the secondary structural elements (DSSP) proposed that VGB1 has helical structure (Figure 8C). The residues in α1 helix of VEGF-B accessible on the receptor-binding face, including Trp17, Ile18, Tyr21, Thr22 and Thr25, are highly similar to KDR-binding residues in α1 helix of VEGF-A, comprising of Phe17, Met18, Tyr21, Gln22 and Tyr25 [41,64], and aromatic residues Trp17 and Tyr21 in VEGF-B and Phe17, Tyr21 and Tyr25 in VEGF-A participate in the interaction with the receptors [64]. Considering that VGB1 was able to bind to VEGFR-1 and VEGFR-2, and results of far-UV CD showing the possibility of helix formation in the designed peptide, we speculate that aromatic-binding residues Trp3 and Tyr7, corresponding to the receptor-binding residues Trp17 and Tyr21 of VEGF-B, may be located in the α-helix.

The most surprising feature of VGB1 was its ability to bind to VEGFR-2. D´Andrea et al. [57] reported that a peptide reproducing α1 helix (residues 17–25) of VEGF-A was unstructured and binds to neither VEGFR-1 nor VEGFR-2. As a result of five mutations, including an amino acid substitution according to the corresponding residue in VEGF-B (F17W) and four substitutions based on the corresponding residues in PlGF (M18Q, D19E, Y21W and Q22G), the mutant peptide, referred to as QK, adopted helical structure and bound very strongly to both VEGFR-1 and VEGFR-2. Interestingly, although seven out of nine residues in the α-helix region of QK peptide were identical with those from PlGF and only three residues were identical with VEGF-A, this peptide exhibited a very high affinity for VEGFR-2. Moreover, it was notable that substitution of key binding residues (Phe17 and Tyr21) did not impair the VEGFR-2-binding ability of QK. They stated that the mutations increased helical propensity while retaining the three-dimensional arrangement of the interacting residues. Similarly, we suggest that the arrangement of the residues involved in receptor binding of VGB1 was probably fixed by preserving the helical structure.

Our results, for the first time, indicated that a peptide consisting of the α1 helix region and its C-terminal segment can bind to both VEGFR-1 and VEGFR-2 and, through abrogation of AKT and ERK1/2 phosphorylation, it inhibits different hallmarks of angiogenesis including proliferation, migration and tube formation of ECs in vitro. In vivo, VGB1 exhibited notable anti-angiogenic and anti-tumor properties together with inhibition of cell proliferation and apoptosis induction in tumors. The helical structure is critical for the receptor binding of the α1 helix region. Structural analyses suggested that VGB1 can assume the helical structure. Considering that VGB1 sequence was almost completely derived from VEGF-B, and its VEGFR-2 binding property is noteworthy. This unexpected feature of VGB1 can be because a very high (90%) similarity exists between α1 helix (residues 17–25) of VEGF-A and VEGF-B, especially in the binding residues (residues 17, 21 and 25). Simultaneous targeting of VEGFR-1 and -2 would be advantageous to combat immoderate angiogenesis than therapeutics targeting only one receptor. This peptide is a promising candidate for the anti-angiogenic treatment of breast as well as other cancer types.

Abbreviations

     
  • 2D

    two-dimensional

  •  
  • 3D

    three-dimensional

  •  
  • AKT

    Protein Kinase B

  •  
  • DAB

    diaminobenzidine

  •  
  • DAPI

    4′,6-diamidino-2-phenylindole

  •  
  • DSSPs

    dictionary of secondary structure of proteins

  •  
  • DTNB

    5,5'-dithiobis-(2-nitrobenzoic acid)

  •  
  • EC

    endothelial cell

  •  
  • ERK

    extracellular signal-regulated kinases

  •  
  • ESI-MS

    electrospray ionization mass spectrometry

  •  
  • FITC

    fluorescein isothiocyanate

  •  
  • FLK1

    fetal liver kinase 1

  •  
  • FLT1

    fms-like tyrosine kinase

  •  
  • H&E

    hematoxylin–eosin

  •  
  • HPLC

    high-performance liquid chromatography

  •  
  • HUVECs

    human umbilical vein endothelial cells

  •  
  • IC50

    half maximal inhibitory concentration

  •  
  • Ig

    immunoglobulin

  •  
  • KDR

    kinase insert domain receptor

  •  
  • MALDI-TOF mass spectrometry

    matrix-assisted laser desorption/ionization time-of-flight mass spectrometry

  •  
  • MAPK

    mitogen-activated protein kinases

  •  
  • MD

    molecular dynamics

  •  
  • microvessel density (MVD)

    microvessel density

  •  
  • NRP

    neuropilin

  •  
  • PDB

    protein data bank

  •  
  • PE

    phycoerythrin

  •  
  • PI

    propidium iodide

  •  
  • PI3K

    phosphatidylinositol 3-kinase

  •  
  • PlGF

    placental growth factor

  •  
  • RTK

    receptor tyrosine kinase

  •  
  • SDS

    sodium dodecyl sulfate

  •  
  • TBST

    Tris-buffered saline and Tween 20

  •  
  • TFE

    2,2,2-triflouroethanol

  •  
  • TM

    transmembrane

  •  
  • VEGF

    vascular endothelial growth factor

  •  
  • VEGFR

    vascular endothelial growth factor receptor

Author Contribution

S.M.A. directed this work. S.M.A. and E.A. performed peptide design. S.M.A., K.M. and F.M. carried out the development of methodology. S.M.A., E.A. and K.M. performed acquisition of in vitro data. A.R.E.R. and E.A. carried out animal studies. S.M.A., E.A., H.N. and F.M. performed acquisition of structural data. S.M.A., E.A., F.M., K.M. and H.N. performed analysis and interpretation of data. S.M.A., K.M., F.M., H.N. and A.R.E.R. supported administration, techniques or material of the present study. S.M.A., E.A. and F.M. wrote the present paper and have given approval to the final version of the manuscript.

Funding

This work was funded through the research council of the University of Guilan.

Acknowledgements

The authors express the gratitude to the research council of the University of Guilan, Rasht, Iran. They also thank the Pasteur Institute of Iran for animal husbandry and in vivo experiments, and the medical biology research center of the Kermanshah University of Medical Science for specialized in vitro laboratory.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Carmeliet
,
P.
(
2003
)
Angiogenesis in health and disease
.
Nat. Med.
9
,
653
660
2
Takahashi
,
H.
and
Shibuya
,
M.
(
2005
)
The vascular endothelial growth factor (VEGF)/VEGF receptor system and its role under physiological and pathological conditions
.
Clin. Sci.
109
,
227
241
3
Dvorak
,
H.F.
,
Brown
,
L.F.
,
Detmar
,
M.
and
Dvorak
,
A.M.
(
1995
)
Vascular permeability factor/vascular endothelial growth factor, microvascular hyperpermeability, and angiogenesis
.
Am. J. Pathol.
146
,
1029
1039
PMID:
[PubMed]
4
Olofsson
,
B.
,
Pajusola
,
K.
,
von Euler
,
G.
,
Chilov
,
D.
,
Alitalo
,
K.
and
Eriksson
,
U.
(
1996
)
Genomic organization of the mouse and human genes for vascular endothelial growth factor B (VEGF-B) and characterization of a second splice isoform
.
J. Biol. Chem.
271
,
19310
19317
5
Li
,
X.
,
Lee
,
C.
,
Tang
,
Z.
,
Zhang
,
F.
,
Arjunan
,
P.
,
Li
,
Y.
et al.  (
2009
)
VEGF-B: a survival, or an angiogenic factor?
Cell Adh. Migr.
3
,
322
327
6
Neufeld
,
G.
,
Cohen
,
T.
,
Shraga
,
N.
,
Lange
,
T.
,
Kessler
,
O.
and
Herzog
,
Y.
(
2002
)
The neuropilins: multifunctional semaphorin and VEGF receptors that modulate axon guidance and angiogenesis
.
Trends Cardiovasc. Med.
12
,
13
19
7
Pajusola
,
K.
,
Aprelikova
,
O.
,
Korhonen
,
J.
,
Kaipainen
,
A.
,
Pertovaara
,
L.
,
Alitalo
,
R.
et al.  (
1992
)
FLT4 receptor tyrosine kinase contains seven immunoglobulin-like loops and is expressed in multiple human tissues and cell lines
.
Cancer Res.
52
,
5738
5743
8
Shibuya
,
M.
,
Yamaguchi
,
S.
,
Yamane
,
A.
,
Ikeda
,
T.
,
Tojo
,
A.
,
Matsushime
,
H.
et al.  (
1990
)
Nucleotide sequence and expression of a novel human receptor-type tyrosine kinase gene (flt) closely related to the fms family
.
Oncogene
5
,
519
524
PMID:
[PubMed]
9
Terman
,
B.I.
,
Carrion
,
M.
,
Kovacs
,
E.
,
Rasmussen
,
B.
,
Eddy
,
R.
and
Shows
,
T.
(
1991
)
Identification of a new endothelial cell growth factor receptor tyrosine kinase
.
Oncogene
6
,
1677
1683
PMID:
[PubMed]
10
Olofsson
,
B.
,
Korpelainen
,
E.
,
Pepper
,
M.S.
,
Mandriota
,
S.J.
,
Aase
,
K.
,
Kumar
,
V.
et al.  (
1998
)
Vascular endothelial growth factor B (VEGF-B) binds to VEGF receptor-1 and regulates plasminogen activator activity in endothelial cells
.
Proc. Natl Acad. Sci. U.S.A.
95
,
11709
11714
11
Chang
,
L.
and
Karin
,
M.
(
2001
)
Mammalian MAP kinase signalling cascades
.
Nature
410
,
37
40
12
Kroll
,
J.
and
Waltenberger
,
J.
(
1997
)
The vascular endothelial growth factor receptor KDR activates multiple signal transduction pathways in porcine aortic endothelial cells
.
J. Biol. Chem.
272
,
32521
32527
13
Parenti
,
A.
,
Morbidelli
,
L.
,
Cui
,
X.-L.
,
Douglas
,
J.G.
,
Hood
,
J.D.
,
Granger
,
H.J.
et al.  (
1998
)
Nitric oxide is an upstream signal of vascular endothelial growth factor-induced extracellular signal-regulated kinase½ activation in postcapillary endothelium
.
J. Biol. Chem.
273
,
4220
4226
14
Pearson
,
G.
,
Robinson
,
F.
,
Beers Gibson
,
T.
,
Xu
,
B.-E.
,
Karandikar
,
M.
,
Berman
,
K.
et al.  (
2001
)
Mitogen-activated protein (MAP) kinase pathways: regulation and physiological functions
.
Endocr. Rev.
22
,
153
183
PMID:
[PubMed]
15
Coffer
,
P.J.
,
Jing
,
J.
and
Woodgett
,
J.R.
(
1998
)
Protein kinase B (c-Akt): a multifunctional mediator of phosphatidylinositol 3-kinase activation
.
Biochem. J.
335
,
1
13
16
Franke
,
T.F.
,
Kaplan
,
D.R.
and
Cantley
,
L.C.
(
1997
)
PI3K: downstream AKTion blocks apoptosis
.
Cell
88
,
435
437
17
Wymann
,
M.P.
and
Pirola
,
L.
(
1998
)
Structure and function of phosphoinositide 3-kinases
.
Biochim. Biophys. Acta, Mol. Cell Biol. Lipids
1436
,
127
150
18
Rosca
,
E.V.
,
Koskimaki
,
J.E.
,
Rivera
,
C.G.
,
Pandey
,
N.B.
,
Tamiz
,
A.P.
and
Popel
,
A.S.
(
2011
)
Anti-angiogenic peptides for cancer therapeutics
.
Curr. Pharm. Biotechnol.
12
,
1101
1116
19
Vasudev
,
N.S.
and
Reynolds
,
A.R.
(
2014
)
Anti-angiogenic therapy for cancer: current progress, unresolved questions and future directions
.
Angiogenesis
17
,
471
494
20
Lin
,
Z.
,
Zhang
,
Q.
and
Luo
,
W.
(
2016
)
Angiogenesis inhibitors as therapeutic agents in cancer: challenges and future directions
.
Eur. J. Pharmacol.
793
,
76
81
21
Zhang
,
Y.
,
Chen
,
Y.
,
Zhang
,
D.
,
Wang
,
L.
,
Lu
,
T.
and
Jiao
,
Y.
(
2018
)
Discovery of novel potent VEGFR-2 inhibitors exerting significant antiproliferative activity against cancer cell lines
.
J. Med. Chem.
61
,
140
157
22
Ebrahimi Samani
,
E.
,
Seraj
,
Z.
,
Naderimanesh
,
H.
,
Khajeh
,
K.
,
Esmaeili Rastaghi
,
A.R.
,
Droudi
,
T.
et al.  (
2017
)
Controlled release of an endostatin peptide using chitosan nanoparticles
.
Chem. Biol. Drug Des.
90
,
417
424
23
Behelgardi
,
M.F.
,
Zahri
,
S.
,
Mashayekhi
,
F.
,
Mansouri
,
K.
and
Asghari
,
S.M.
(
2018
)
A peptide mimicking the binding sites of VEGF-A and VEGF-B inhibits VEGFR-1/-2 driven angiogenesis, tumor growth and metastasis
.
Sci. Rep.
8
,
17924
24
Ellman
,
G.L.
(
1958
)
A colorimetric method for determining low concentrations of mercaptans
.
Arch. Biochem. Biophys.
74
,
443
450
25
Ellman
,
G.L.
(
1959
)
Tissue sulfhydryl groups
.
Arch. Biochem. Biophys.
82
,
70
77
26
Glynn
,
M.W.
and
McAllister
,
A.K.
(
2006
)
Immunocytochemistry and quantification of protein colocalization in cultured neurons
.
Nat. Protoc.
1
,
1287
1296
27
Denizot
,
F.
and
Lang
,
R.
(
1986
)
Rapid colorimetric assay for cell growth and survival: modifications to the tetrazolium dye procedure giving improved sensitivity and reliability
.
J. Immunol. Methods
89
,
271
277
28
Rodriguez
,
L.G.
,
Wu
,
X.
and
Guan
,
J.-L.
(
2005
)
Wound-healing assay
.
Methods Mol. Biol.
294
,
23
29
PMID:
[PubMed]
29
Mansouri
,
K.
,
Khodarahmi
,
R.
and
Ghadami
,
S.A.
(
2013
)
An in vitro model for spontaneous angiogenesis using rat mesenteric endothelial cells: possible therapeutic perspective for obesity and related disorders
.
Pharm. Biol.
51
,
974
980
30
Phung
,
T.L.
,
Ziv
,
K.
,
Dabydeen
,
D.
,
Eyiah-Mensah
,
G.
,
Riveros
,
M.
,
Perruzzi
,
C.
et al.  (
2006
)
Pathological angiogenesis is induced by sustained Akt signaling and inhibited by rapamycin
.
Cancer Cell
10
,
159
170
31
Kong
,
D.
,
Li
,
Y.
,
Wang
,
Z.
,
Banerjee
,
S.
and
Sarkar
,
F.H.
(
2007
)
Inhibition of angiogenesis and invasion by 3,3′-diindolylmethane is mediated by the nuclear factor–κB downstream target genes MMP-9 and uPA that regulated bioavailability of vascular endothelial growth factor in prostate cancer
.
Cancer Res.
67
,
3310
3319
32
Bradford
,
M.M.
(
1976
)
A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding
.
Anal. Biochem.
72
,
248
254
33
Chamani
,
R.
,
Asghari
,
S.M.
,
Alizadeh
,
A.M.
,
Mansouri
,
K.
,
Doroudi
,
T.
,
Kolivand
,
P.H.
et al.  (
2016
)
The antiangiogenic and antitumor activities of the N-terminal fragment of endostatin augmented by Ile/Arg substitution: the overall structure implicated the biological activity
.
Biochim. Biophys. Acta
1864
,
1765
1774
34
Shi
,
C.
,
Liu
,
D.
,
Xiao
,
Z.
,
Zhang
,
D.
,
Liu
,
G.
,
Liu
,
G.
et al.  (
2017
)
Monitoring tumor response to antivascular therapy using non-contrast intravoxel incoherent motion diffusion-weighted MRI
.
Cancer Res.
77
,
3491
3501
35
Ranjbar
,
B.
and
Gill
,
P.
(
2009
)
Circular dichroism techniques: biomolecular and nanostructural analyses — a review
.
Chem. Biol. Drug. Des.
74
,
101
120
36
Eswar
,
N.
,
Webb
,
B.
,
Marti-Renom
,
M.A.
,
Madhusudhan
,
M.S.
,
Eramian
,
D.
,
Shen
,
M.Y.
et al.  (
2006
)
Comparative protein structure modeling using modeller
.
Curr. Protoc. Bioinformatics
Chapter 5:Unit-5.6
37
Laskowski
,
R.A.
,
MacArthur
,
M.W.
,
Moss
,
D.S.
and
Thornton
,
J.M.
(
1993
)
PROCHECK: a program to check the stereochemical quality of protein structures
.
J. Appl. Crystallogr.
26
,
283
291
38
Benkert
,
P.
,
Tosatto
,
S.C.
and
Schomburg
,
D.
(
2008
)
QMEAN: a comprehensive scoring function for model quality assessment
.
Proteins
71
,
261
277
39
Abraham
,
M.J.
,
Murtola
,
T.
,
Schulz
,
R.
,
Páll
,
S.
,
Smith
,
J.C.
,
Hess
,
B.
et al.  (
2015
)
GROMACS: high performance molecular simulations through multi-level parallelism from laptops to supercomputers
.
SoftwareX
1–2
,
19
25
40
Tao
,
Y.
,
Rao
,
Z.-H.
and
Liu
,
S.-Q.
(
2010
)
Insight derived from molecular dynamics simulation into substrate-induced changes in protein motions of proteinase K
.
J. Biomol. Struct. Dyn.
28
,
143
157
41
Iyer
,
S.
,
Darley
,
P.I.
and
Acharya
,
K.R.
(
2010
)
Structural insights into the binding of vascular endothelial growth factor-B by VEGFR-1D2 recognition and specificity
.
J. Biol. Chem.
285
,
23779
23789
42
Wang
,
L.
,
Zhou
,
L.
,
Reille-Seroussi
,
M.
,
Gagey-Eilstein
,
N.
,
Broussy
,
S.
,
Zhang
,
T.
et al.  (
2017
)
Identification of peptidic antagonists of vascular endothelial growth factor receptor 1 by scanning the binding epitopes of its ligands
.
J. Med. Chem.
60
,
6598
6606
43
Giordano
,
R.J.
,
Cardó-Vila
,
M.
,
Lahdenranta
,
J.
,
Pasqualini
,
R.
and
Arap
,
W.
(
2001
)
Biopanning and rapid analysis of selective interactive ligands
.
Nat. Med.
7
,
1249
1253
44
Giordano
,
R.J.
,
Cardó-Vila
,
M.
,
Salameh
,
A.
,
Anobom
,
C.D.
,
Zeitlin
,
B.D.
,
Hawke
,
D.H.
et al.  (
2010
)
From combinatorial peptide selection to drug prototype (I): targeting the vascular endothelial growth factor receptor pathway
.
Proc. Natl Acad. Sci. U.S.A.
107
,
5112
5117
45
Giordano
,
R.J.
,
Anobom
,
C.D.
,
Cardó-Vila
,
M.
,
Kalil
,
J.
,
Valente
,
A.P.
,
Pasqualini
,
R.
et al.  (
2005
)
Structural basis for the interaction of a vascular endothelial growth factor mimic peptide motif and its corresponding receptors
.
Chem. Biol.
12
,
1075
1083
46
Shibuya
,
M.
(
2013
)
Vascular endothelial growth factor and its receptor system: physiological functions in angiogenesis and pathological roles in various diseases
.
J. Biochem.
153
,
13
19
47
Galvagni
,
F.
,
Nardi
,
F.
,
Spiga
,
O.
,
Trezza
,
A.
,
Tarticchio
,
G.
,
Pellicani
,
R.
et al.  (
2017
)
Dissecting the CD93-multimerin 2 interaction involved in cell adhesion and migration of the activated endothelium
.
Matrix Biol.
64
,
112
127
48
Shiojima
,
I.
and
Walsh
,
K.
(
2002
)
Role of Akt signaling in vascular homeostasis and angiogenesis
.
Circ. Res.
90
,
1243
1250
49
D'Andrea
,
L.D.
,
Gatto
,
A.
,
Pedone
,
C.
and
Benedetti
,
E.
(
2006
)
Peptide-based molecules in angiogenesis
.
Chem. Biol. Drug Des.
67
,
115
126
50
Benkert
,
P.
,
Biasini
,
M.
and
Schwede
,
T.
(
2010
)
Toward the estimation of the absolute quality of individual protein structure models
.
Bioinformatics
27
,
343
350
51
Kabsch
,
W.
and
Sander
,
C.
(
1983
)
Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features
.
Biopolymers
22
,
2577
2637
52
Simons
,
M.
,
Gordon
,
E.
and
Claesson-Welsh
,
L.
(
2016
)
Mechanisms and regulation of endothelial VEGF receptor signalling
.
Nat. Rev. Mol. Cell Biol.
17
,
611
625
53
Kowanetz
,
M.
and
Ferrara
,
N.
(
2006
)
Vascular endothelial growth factor signaling pathways: therapeutic perspective
.
Clin. Cancer Res.
12
,
5018
5022
54
Sadremomtaz
,
A.
,
Mansouri
,
K.
,
Alemzadeh
,
G.
,
Safa
,
M.
,
Rastaghi
,
A.E.
and
Asghari
,
S.M.
(
2018
)
Dual blockade of VEGFR1 and VEGFR2 by a novel peptide abrogates VEGF-driven angiogenesis, tumor growth, and metastasis through PI3K/AKT and MAPK/ERK1/2 pathway
.
Biochim. Biophys. Acta Gen. Subj.
1862
,
2688
2700
55
Basile
,
A.
,
Del Gatto
,
A.
,
Diana
,
D.
,
Di Stasi
,
R.
,
Falco
,
A.
,
Festa
,
M.
et al.  (
2011
)
Characterization of a designed vascular endothelial growth factor receptor antagonist helical peptide with antiangiogenic activity in vivo
.
J. Med. Chem.
54
,
1391
1400
56
Ziaco
,
B.
,
Diana
,
D.
,
Capasso
,
D.
,
Palumbo
,
R.
,
Celentano
,
V.
,
Di Stasi
,
R.
et al.  (
2012
)
C-terminal truncation of vascular endothelial growth factor mimetic helical peptide preserves structural and receptor binding properties
.
Biochem. Biophys. Res. Commun.
424
,
290
294
57
D'Andrea
,
L.D.
,
Iaccarino
,
G.
,
Fattorusso
,
R.
,
Sorriento
,
D.
,
Carannante
,
C.
,
Capasso
,
D.
et al.  (
2005
)
Targeting angiogenesis: structural characterization and biological properties of a de novo engineered VEGF mimicking peptide
.
Proc. Natl Acad. Sci. U.S.A.
102
,
14215
14220
58
García-Aranda
,
M.I.
,
González-López
,
S.
,
Santiveri
,
C.M.
,
Gagey-Eilstein
,
N.
,
Reille-Seroussi
,
M.
,
Martín-Martínez
,
M.
et al.  (
2013
)
Helical peptides from VEGF and vammin hotspots for modulating the VEGF–VEGFR interaction
.
Org. Biomol. Chem.
11
,
1896
1905
59
Goncalves
,
V.
,
Gautier
,
B.
,
Coric
,
P.
,
Bouaziz
,
S.
,
Lenoir
,
C.
,
Garbay
,
C.
et al.  (
2007
)
Rational design, structure, and biological evaluation of cyclic peptides mimicking the vascular endothelial growth factor
.
J. Med. Chem.
50
,
5135
5146
60
Cudmore
,
M.J.
,
Hewett
,
P.W.
,
Ahmad
,
S.
,
Wang
,
K.-Q.
,
Cai
,
M.
,
Al-Ani
,
B.
et al.  (
2012
)
The role of heterodimerization between VEGFR-1 and VEGFR-2 in the regulation of endothelial cell homeostasis
.
Nat. Commun.
3
,
972
61
Shibuya
,
M.
(
2011
)
Vascular endothelial growth factor (VEGF) and its receptor (VEGFR) signaling in angiogenesis: a crucial target for anti- and pro-angiogenic therapies
.
Genes Cancer
2
,
1097
1105
62
Tchaikovski
,
V.
,
Fellbrich
,
G.
and
Waltenberger
,
J.
(
2008
)
The molecular basis of VEGFR-1 signal transduction pathways in primary human monocytes
.
Arterioscler. Thromb. Vasc. Biol.
28
,
322
328
63
Guo
,
S.
,
Colbert
,
L.S.
,
Fuller
,
M.
,
Zhang
,
Y.
and
Gonzalez-Perez
,
R.R.
(
2010
)
Vascular endothelial growth factor receptor-2 in breast cancer
.
Biochim. Biophys. Acta, Rev. Cancer
1806
,
108
121
64
Muller
,
Y.A.
,
Christinger
,
H.W.
,
Keyt
,
B.A.
and
de Vos
,
A.M.
(
1997
)
The crystal structure of vascular endothelial growth factor (VEGF) refined to 1.93 Å resolution: multiple copy flexibility and receptor binding
.
Structure
5
,
1325
1338

Supplementary data