Primary HUVECs were isolated from the umbilical vein vascular wall based on a previous method (20). HUVECs were routinely passage in ECM (Life Technologies, Carlsbad, CA, USA) with 5% CO₂ at 37°C. Only those cells from passage 3 to 8 were used for experiments.

HUVECs (5×10³/well) were plated onto a 96-well culture plate and incubated for 24 h. Different dilutions of TSA were added and further incubated for 48 h. Viable cells were determined by MTT assay using the MTT Cell Proliferation Assay kit (Promega, Madison, WI, USA) according to the manufacturer’s instructions. The number of viable cells was presented relative to the untreated control. Meanwhile, HUVECs were seeded (5,000 cells/well) in a 96-well culture plate and incubated for 24 h. The medium was replaced with ECM, containing 1% fetal bovine serum (FBS), 10 ng/ml VEGF and various concentrations of TSA (4, 8 and 12 μM). After 48 h, relative cell proliferation was determined by the MTT assay. All assays were performed in triplicate wells.

Endothelial cell migration was assessed using a modified Boyden chamber assay. HUVECs (1×10⁵) were plated in Endothelial cell medium containing 1% FBS in the upper chamber of the Transwell (8-μm PET; Millipore, Germany). Cells were then treated for 30 min at 37°C. Endothelial cell medium containing 1% FBS, 10 ng/ml VEGF and various concentrations of TSA (4, 8 and 12 μM) were added to the lower chamber. Endothelial cell medium containing 1% FBS was a negative control. After 8 h, the non-migrated cells were removed by a cotton swap, and the migrated cells were stained with crystal violet and examined under a microscope. The number of migrated cells was quantified by counting the cells with ×40 objective. Migration was normalized to the percentage of migration.

The ECM gel-induced capillary tube formation assay was used as an in vitro measurement of angiogenesis. Briefly, a 24-well culture plate was coated with 50 μM/well ECM gel (Sigma, Japan) and allowed to stand for 30 min at 37°C to form a gel layer. After gel formation, 1×10⁵ HUVECs in 0.5 ml of growth medium were seeded to each well along with 10 ng/ml VEGF. Various dilutions of TSA were added into the wells and incubated for 6 h at 37°C in a humidified atmosphere with 5% CO₂, and the formation of capillary tubes was photographed using an inverted microscope.

The effect of TSA on ex vivo angiogenesis was determined by the CAM assay. Briefly, fertilized chick eggs which were incubated for 6 days were used for the experiment and were incubated at 70% humidity and 37°C. A window was opened on the top of each egg after 1 day of incubation. The windows were covered with sterile tape and the eggs were returned to the incubator. After an additional 1 day, the transparent tape was cut open and a drug carrier plate with a 15-mm diameter was placed in the chamber. Twenty microliters of each test reagent was added onto the plate. Test substances including TSA (the dose of TSA was 4, 8 and 12 μM, respectively). The eggs were incubated for 48 h and photographed. Blood vessel density was quantified by counting the number of branching blood vessels. Each experiment was performed three times and represented as a bar diagram.

The rat aortic ring assay was used as the ex vitro angiogenesis study model (21). Dorsal aortas from freshly sacrificed Sprague-Dawley rats were extracted in a sterile manner and rinsed in ice cold phosphate-buffered saline (PBS). They were then cut into 1-mm long rings using a surgical blade. Each ring was placed in a collagen pre-coated 96-well plate. Different dilutions of TSA were added to the wells. On day 7, the rings were analyzed by phase-contrast microscopy, and microvessel outgrowths were quantified (22) and photographed. All experimental protocols were approved by the Ethics Committee for Animal Experimentation of China Pharmaceutical University.

A human MMP-2 and MMP-9 ELISA kit (R&D Systems, Shanghai, China) was used to determine the levels of MMP-2 and MMP-9 in conditioned media collected from HUVECs with or without a 24-h treatment with TSA. The experimental steps were followed as described in the protocol provided by the manufacturer.

The VEGF165 expression plasmid pcVEGF165 was constructed in our laboratory. Cells were transfected with the pcVEGF165 using Lipofectamine 2000 (Life Technology) according to the manufacturer’s protocol. After 6 h, the medium was replaced with ECM containing 1% FBS, 100 ng/ml endothelial cell growth supplement (ECGS) and various concentrations of TSA (0, 4, 8 and 12 μM). After 24 and 48 h, total cellular RNA and protein were respectively extracted for the next detection.

Total cellular RNA was extracted using TRIzol reagent (Gibco-BRL, Gaithersburg, MD, USA), according to the manufacturer’s protocol. Total RNA was subjected to cDNA synthesis using M-MLV reverse transcriptase. The process was performed at 30°C for 10 min, 42°C for 1 h followed by denaturation at 95°C for 2 min. To determine mRNA expression levels, qRT-PCR was performed using the ABI Prism 7500 Sequence Detector (Applied Biosystems, Life Technologies). cDNA templates (2 μl) were amplified in a final volume of 20 μl containing the SYBR-Green PCR Master Mix and primer. Melt-curve analysis and agarose gel electrophoresis were used to confirm amplicon specificity. The length of the amplified product was confirmed using 2% agarose gel electrophoresis. Relative quantification was performed using the 2^−ΔΔCt method. GAPDH served as an appropriate reference gene in this experiment. qRT-PCR primers for KDR generated a 371-bp product and were forward, 5′-CTGGCATGGTCTTCTGTGAAGCA-3′ and reverse, 5′-CCGCATCACATCCACTGGTATT-3′; primers for CD146 were forward, 5′-TGGTCATCGTGGCTGTGATTGTG-3′ and reverse, 5′-CCTTTGGAGGCTTTGGCTGAGAGAA-3′. An alternative set of MMP-2 primers generated a 754-bp product and were forward, 5′-TGACGGTAAGGACGGACTC-3′ and reverse, 5′-AGTCCGCCAAATGAACCG-3′; primers for MMP-9 were forward, 5′-TGGGGGGCAACTCGGC-3′ and reverse, 5′-GGAATGATCTAAGCCCAG-3′; PCR primers for GAPDH were as described and generated a 450-bp product forward, 5′-AAGGTCGGAGTCACCGGATT-3′ and reverse, 5′-CTGGAAGATGGTGATGGGATT-3′.

Cells were washed twice with 1X PBS and then lysed in lysis buffer (20 mM Tris, 150 mM NaCl, 0.25% NP-40, 1 mM phenylmethylsulfonyl fluoride and 1X protease inhibitors). Fifty micrograms of protein was analyzed on 8% SDS-PAGE under denaturing conditions and electro-transferred to PVDF membranes (Millipore, Bedford, MA, USA). Non-specific protein binding was blocked by incubating the membranes with blocking solution (TBST and 10% non-fat dried milk) for 60 min at room temperature. Polyclonal antibodies specific for KDR (1:250 in TBST containing 5% BSA) and CD146 (1:500 in TBST containing 5% BSA) were applied to the membrane and incubated overnight at 4°C. Membranes were washed in TBST and then incubated in 1:5,000 diluted goat anti-rabbit IgG-HRP secondary antibody (Bio-Rad, Hercules, CA, USA) for 1 h at room temperature. The detection of specific signals was performed using the ECL detection system (Amersham Pharmacia Biotech, Piscataway, NJ, USA). Protein concentrations were measured using the method of Bradford (Bio-Rad).

All data were obtained from at least three independent experiments and expressed as mean ± SD unless stated otherwise. Statistical comparisons were made relative to the negative controls, and the significance was assessed using the Student’s t-test and was indicated as P<0.05 or P<0.01 using relevant symbols in the figures and legends. A P-value of <0.05 was considered to be statistically significant.

Cell viabilty was determined by the MTT assay. Effect of TSA on HUVEC viability in culture is shown in Fig. 1B. TSA was found to be non-toxic to HUVECs at concentrations of 4–12 μM and these concentrations were used for further in vitro study. The proliferation of endothelial cells is important for the formation of new blood vessels. To investigate the mechanism of the anti-angiogenic function of TSA, we performed an MTT assay to explore the effect of TSA on proliferation of endothelial cells. As shown in Fig. 1C, HUVECs showed a very high rate of proliferation when stimulated with VEGF. The endothelial cells were then exposed to TSA of different doses. At 48 h, treatment with TSA (12 μM) significantly inhibited the proliferation of HUVECs by ~60% compared to the untreated cells. This suggests that reduced proliferation contributes to the anti-angiogenic effect of TSA.

To further understand the anti-angiogenic function of TSA, we examined whether TSA could exert any effect on the migration of HUVECs by Transwell migration assay. The dose-dependent inhibitory effect of TSA on cell migration was demonstrated (Fig. 2A–F). The migration rates of the cells following treatment with 4, 8 and 12 μM TSA were inhibited by 28, 49 and 59%, respectively demonstrating that TSA also dose-dependently inhibits the migration of HUVECs.

In order to study the anti-angiogenic function of tanshinone IIA in vitro, we carried out an angiogenesis test by examining the VEGF-induced tube formation of HUVECs. Treatments with TSA (4, 8 and 12 μM) inhibited the VEGF-induced tube formation of HUVECs in a dose-dependent manner, which indicates that TSA inhibits the angiogenesis of HUVECs in vitro (Fig. 2G–L).

CAM is another ex vivo test for angiogenesis and it is more convenient to quantify vascular development with this method. We placed TSA-impregnated filter disks on blood vessels in avascular sections of CAM (day 10) for 48 h. The disks and underlying CAM tissues (day 12) were then harvested. We scored angiogenesis by counting the vessel branches present in the CAM tissue below the filter by digital images. Consistent with the data above, upregulation of TSA led to a significant reduction in angiogenesis, while the control CAM exhibited more neovascularization (Fig. 3A), which confirmed the anti-angiogenic potential of TSA through the ex vivo assay.

We futher explored the anti-angiogenic activity of TSA using ex vitro angiogenesis models. An isolated rat aortic ring was embedded in Matrigel in ECM containing different concentrations of TSA cultured for 7 days. TSA markedly suppressed the outgrowth of cells from the aortic arch in a dose-dependent manner, indicating that TSA inhibits angiogenesis in vitro (Fig. 3B).

VEGF is a critical factor in the process of angiogenesis, and activates the angiogenesis of endothelial cells via integration with its receptor (VEGFR) (23,24). Activation of the VEGF/VEGFR signaling pathway can trigger a network signal cascade, consequently contributing to proliferation, invasion and migration of vascular endothelial cells. Meanwhile, human soluble CD146, a coreceptor for KDR, represents both an attractive biomarker of placental vascular development and a therapeutical target in pregnancy complications associated with pathological angiogenesis (25). In order to further investigate the mechanism underlying the inhibitory effect of TSA on the invasion of human vascular endothelial cells, we analyzed the effect of TSA on the mRNA and protein expression of KDR and CD146 by RT-PCR and western blot analysis. HUVECs were transfected with the pcVEGF165 plasmid to induce the expression of KDR. Our results demonstrated the inhibitory effect of TSA on mRNA and protein levels of KDR and CD146 (Fig. 4) in a dose-dependent manner, which definitely indicates that TSA inhibits angiogenesis via repression of the VEGF/VEGFR2 pathway.

ECM degradation is also important for angiogenesis, and is mainly regulated by a balance between different groups of MMPs (26). MMPs and VEGF/KDR are two foremost factors in angiogenesis (27). To understand the potential mechanism of the inhibitory effect of TSA on the invasion of human vascular endothelial cells through matrix, we analyzed the effect of TSA on the production of MMP-2 and MMP-9 by ELISA method. Our results demonstrated a dose-dependent inhibitory effect of TSA on MMP-2 and MMP-9 (Fig. 5A and B). We further examined whether TSA affects the gene expression of MMP-2 and MMP-9 at the mRNA level. The mRNA levels of MMP-2 and MMP-9 remained unchanged after TSA treatment, which indicates that the anti-angiogenic function of TSA on MMP-2 and MMP-9 in HUVECs is likely mediated by a post-translational mechanism (Fig. 5C and D).