Cheiradone was found to specifically inhibit the binding of VEGF₁₆₅ to VEGFR-1 and VEGFR-2 in a dose dependent manner with IC₅₀ values of 2.9 ± 0.31 μM and 0.61 ± 0.14 μM respectively (Figure 1; Table 1). No significant inhibition of FGFR-1 and -2 was observed even at the highest concentration (328.20 μM) tested (data not included).

Cheiradone was tested to evaluate its effect on cell proliferation in the presence of VEGF, EGF and FGF-2. A concentration-dependent inhibition of VEGF-stimulated BAEC and HDMEC proliferation with IC₅₀ values of 7.4 ± 0.74 and 7.8 ± 1.2 μM respectively (p < 0.005) (Figure 2; Table 1) was observed. However, no significant inhibition of FGF-2- and EGF-triggered cell proliferation was observed (data not included).

The effect of cheiradone on the migration of ECs was analysed using both two- and three-dimensional cell migration assays. In the two-dimensional assay, a wound-healing model was used to assess the migratory behaviour of BAECs and HDMECs (Figure 3). In the VEGF treated control group, significant wound healing was found 24 h after the cell monolayer was wounded with a sterile razor blade (Figure 3Biv; results shown for HDMEC). No significant inhibition was observed on non-stimulated wound recovery (Figure 3Biii) However, VEGF- stimulated HDMEC and BAEC migration was inhibited after 24 h in a dose dependent manner by cheiradone with IC₅₀ values of 6.5 ± 0.97 and 7.1 ± 0.77 μM respectively (p < 0.05; Figure 3A and 3Bv). In a three-dimensional cell migration assay, BAECs and HDMECs treated with VEGF showed 2.8 and 2.5 fold increase in migration to the lower chamber compared to non-treated cells respectively (Figure 3C). Cheiradone was found to significantly inhibit (p = 0.001 in each case) VEGF-induced BAEC and HDMEC migration with IC₅₀ values of 7.5 ± 0.92 and 5.2 ± 0.38 μM respectively.

Cheiradone had no effect on FGF-2 or EGF stimulated migration in the concentration range used (results not shown).

To examine the role of cheiradone on EC differentiation into vascular structures in vitro, tube formation of BAECs and HDMCs on Matrigel was assessed. When these cells were stimulated with VEGF, elongated tube-like structures were formed and the process was inhibited in a dose dependent manner by cheiradone (Figure 4A). Cheiradone reduced the width and length of VEGF-induced HDMEC and BAEC tubes with IC₅₀ values of 6.0 ± 0.38 and 7.7 ± 1.5 μM respectively (p < 0.005; Fig 4Bi–iv, results are shown for HDMEC). No significant effect of cheiradone was seen on non-stimulated tube formation (Figure 4Bii).

The effect of cheiradone on cell invasion was analysed using a Transwell Boyden chamber system coated with reconstituted growth factor-reduced Matrigel. BAECs or HDMECs were allowed to invade the lower chamber in the presence and absence of VEGF and cheiradone. A statistically significant increased in cell invasion was observed in VEGF treated HDMECs (2.2 fold, p = 0.002) and BAECs (3.3 fold, p = 0.001) (Figure 5). Cheiradone showed dose-dependent inhibition of VEGF stimulated cell invasion of HDMEC and BAEC with IC₅₀ values of 8.3 ± 1.0 and 6.3 ± 0.31 μM respectively (p < 0.05).

The above in vitro data suggests that cheiradone inhibits several steps of angiogenesis in vitro. Therefore, we analyzed its effect on in vivo angiogenesis using the CAM assay. After exogenous stimulation of angiogenesis with VEGF₁₆₅, significant new blood vessel growth was observed towards the stimulus after 6 days (Figure 6C, m = 3, n = 5). There was no significant angiogenic response to cheiradone alone (Fig 6B, m = 0.5, p = 0.0875, n = 5). VEGF₁₆₅-induced blood vessels formation was completely abolished in the presence of cheiradone (m = 0, n = 5, p = 0.0001) (Figure 6D). There was no evidence of an inflammatory response with cheiradone alone (Fig 6) or in the control (Figure 6A).

No significant cytotoxicity was found at the tested concentrations of cheiradone (Figure 7A), whereas staurosporine induced a noticeable cytotoxic effect in the MTT (Figure 7B) and immunofluorescence (Figure 7D) assays.

Cheiradone (Fig 8; RMM 613) was extracted and purified from Euphorbia cheiradenia Boiss. Full chemical characterisation and purification is detailed elsewhere 24.

Matrigel (Becton Dickinson, UK), FGF-2, goat anti-active caspase-3 antibody (R&D System), VEGF₁₆₅ (Apollo Cytokine Research, Cambridge, UK), EGF, VEGFR-1 and -2 and FGFR-1 and -2, anti-FGF-2 and anti-VEGF antibodies (Santa Cruz Biotechnology, Heidelberg, GDR), ABTS peroxidase substrate kit (Vector, UK), Transwell chamber system, culture plates and flasks (SLS, UK), anti-goat Alexa flour 488 conjugated green fluorescence dye and other chemical of commercial grade were purchased from Sigma (Poole, UK).

Human dermal microvascular endothelial cells (HDMECs) and the appropriate medium were purchased from TCS Cellworks (Buckingham, UK) and were cultured and maintained according to the supplier's instructions. Bovine aortic endothelial cells (BAECs) were from an established large vessel primary cell culture obtained and characterised as described previously 34. They were routinely cultured in Dulbecco's modified Eagles medium (DMEM) containing varying concentrations of foetal calf serum (FCS). All cells were used between passages six to nine.

Cells were seeded in triplicate at a concentration of 2.5 × 10⁴/ml, in 2 ml of complete medium in 6-well plates. After attachment (24 h), medium was replaced with serum poor medium (SPM), containing 2.5% foetal calf serum in which the cells grew at a significantly reduced rate. Growth factors FGF-2 and EGF (25 ng/ml), and VEGF₁₆₅ (10 ng/ml) with and without the test compound at different concentrations was added and cells incubated for a further 72 h. Control wells were treated with 5 μl DMSO. Concentration ranges of test compounds and pre-incubation times were based on pilot studies. MTT and immunofluorescence studies using active caspase-3 as a measure of apoptosis confirmed that test compounds were not cytotoxic at the concentrations used (see below). After 72 h, cells were washed in PBS, detached in 0.05%w/v trypsin in PBS, and counted on a Coulter counter (Coulter Electronics, Hialeah, FL) set to a threshold of 30 μm. Experiments were performed at least twice in triplicate wells and significance was determined by the Student t test. A representative example is shown.

Cell migration was examined in vitro using a Transwell chamber system with 8.0 μm pore polycarbonate filter inserts (TSL, UK). The Transwell insert was coated overnight with 0.1%w/v gelatine, and air-dried. Cells (1 × 10⁵) were placed in the upper part of the filter and test compounds at different concentrations with and without growth factors were added to the lower part in SPM. Cells were incubated at 37°C for 16 h. After fixation (4% paraformaldehyde), and staining (Geimsa), cell migration in duplicate wells was determined by counting cell numbers on the lower surface. Experiments were performed at least twice and a representative example is shown. Significance was determined by the Student t test.

Cells (6 × 10⁴/ml) were added to a Thrermanox cover slip in a 24-well plate in complete medium and incubated for 24–48 h. When confluent, the medium was replaced with DMEM containing 0.1% FCS and incubated for a further 48 h. Cover slips were washed (PBS, ×3), wounded with a sterile razor to produce a straight edged cut and washed in PBS to remove dislodged cells. Cover slips were added to a fresh 24-well plate in 0.1% FCS and incubated with VEGF₁₆₅ or the other growth factors and a range of concentrations of test compound for 24 h. Under these conditions there was negligible proliferation but measurable migration. Slides were fixed in ethanol (100%), stained with methylene blue and photographed. For each slide, 10 fields of view (2 mm × 1.45 mm) were counted at random. Each experiment was performed at least twice and significance was determined by the Student t test.

Cells (1.0 × 10⁶/ml) were mixed with an equal volume of Matrigel at 4°C. Aliquots (45 μl) were added to the wells of a 48-well plate and allowed to polymerise (1 h) when 500 μl of microvascular endothelial cell medium containing VEGF₁₆₅ or the other growth factors, with or without the test compound was added. The cells were incubated for 24 h at 37°C then fixed in 4% paraformaldehyde (5 min), washed in cold ethanol and air dried. Cells were stained with Geimsa (30 sec), air dried and photographed. Ten random fields were selected and the number of closed tubes counted.

All experiments were performed in triplicate and repeated at least twice and significance was determined by the Student t test.

A Transwell cell culture chamber with 6.5-mm-diameter polycarbonate filters (8-μm pore size) were coated with 30 μg/ml Matrigel. Cells (1 × 10⁵) were added to the upper chamber suspended in an appropriate medium. The medium containing a range of concentrations of test compound was added to the lower chamber in the presence and absence of VEGF₁₆₅ or the other growth factors. After 24 h incubation at 37°C, the medium from the lower chamber was removed, cell fixed (4% paraformaldehyde) and stained. The number of cells invaded to the lower chamber through the Matrigel was counted under phase-contrast microscopy. Each invasion experiment was performed in triplicate and repeated at least twice. Statistical significance was determined using the Student t test.

Competition between growth factors, their cell surface receptors and test compounds was assessed as described previously 35. A 96-well plate was coated overnight with growth factors (2 μg/ml of VEGF₁₆₅ or FGF-2) and blocked with 1% BSA in PBS containing 0.05% Tween-20. The test compound was separately pre-mixed with soluble receptors VEGFR-1, VEGFR-2, FGFR-1 or FGFR-2 (2 μg/ml in each case) for 2 h and added to the plate and incubated for a further 2 h. The plate was washed (×3 with PBS-Tween-20) and incubated with anti-VEGF or anti-FGF-2 antibody (1:500 in PBS-Tween-20) for 45 min, washed and incubated with horseradish peroxidase conjugated goat anti-IgG (Santa Cruz, 1:1000) for a further 45 min. After washing (×3), ABTS peroxidase substrate (Vector, UK) was added and the absorbance read at 405 nm. IC₅₀ values were calculated from the data using the EZ-Fit enzyme kinetic software (Perella Scientific Inc., Amherst, USA).

The angiogenic activity of cheiradone was determined using the semi-quantitative chick chorioallantoic assay (CAM) as described previously 36. To expose the CAM a window was created in the shells of 4 day-old chicken eggs. On day 8, a 2 mm³ methycellulose pellet (5 μl of 1% sterile methylcellulose; 400 centipoise, Sigma UK) containing no additions (control), the test compound (10 μg) with and without VEGF (100 ng) were applied to the membrane. The resultant angiogenesis scored on day 14 as 0- negative; 0.5- change in vessel architecture; 1- partial spokewheel (1/3 of circumference exhibits directional angiogenesis); 2- spokewheel; 3 or greater-strong and fully spokewheel. This approach enabled calculation of an accumulated response in each group. To photograph the membrane, 2 cm³ of a 50% emulsion of aqueous paraffin oil containing 2% Tween-80 was injected at the site of application and photographed using a Leitz dissecting microscope. Each experiment was performed five times and statistical significance was determined by the Mann-Whitney U test and the data is expressed as a median value (m).

Cheiradone toxicity was determined using MTT and active caspase-3 assays. BAECs or HDMECs (7.5 × 10³) were seeded in a 96 well plate and incubated for 4 h to allow cell adhesion. Cheiradone or staurosporine, an inducer of active caspase-3 and therefore, of apoptosis (1 μM) was added to the wells. Control cells were treated with PBS and the plate was incubated at room temperature for 72 hours. MTT reagent (10 μl) was added followed by incubation at room temperature for 2–4 h. When a purple precipitate was visible, detergent reagent (100 μl) was added to the plates and incubated at room temperature for 2 h in the dark. Absorbance was measured at 570 nm using a microplate reader.

In the apoptosis assay, HDMECs or BAEC (4 × 10⁴/ml) in complete medium were added to the chambers slide and allowed to adhere for 24 h. Cheiradone (8.2 μM) or staurosporine (1 μM) were added to all wells except control (PBS) and incubated for 24 hours. Wells treated with staurosporine were immediately washed (PBS) and fixed (4% paraformaldehyde) when cell morphology became round (2–4 h). After washing and fixing, cells were permeablized (0.1% Triton X-100; 10 min), washed (×5 ~ 5 min each), air dried and blocked with 1% BSA in 1:50 TBS Tween for 1 h at room temperature. Cells were incubated with goat anti-active caspase-3 (R&D system, UK; 1% BSA in TBS Tween) for 1 h. The plates were incubated with anti-goat Alexa-Flour 488 conjugated green fluorescent dye for 1.5 h at room temperature. Ten random homogeneous fields were viewed, and photographed.