Rd was obtained from the Shanghai Research Center for Standardization of Chinese Medicines (Shanghai, China). Its structure was confirmed using ¹HNMR and ¹³C NMR spectral analysis, and its purity was >98% as determined by high pressure liquid chromatography (HPLC) analysis.

Phospho-p85 PI3K (Tyr458), PI3K, phospho-Akt (Thr308), Akt, phospho-mTOR (Ser2481), mTOR, cleaved caspase-3, Bax, Bcl-2, VEGFR2, GAPDH, goat anti-rabbit horseradish peroxidase (HRP)-conjugated, and goat anti-mouse HRP antibodies were obtained from Cell Signaling Technology, Inc. (Danvers, MA, USA). Other antibodies against Ki67, HIF-1α and CD31 were provided by Abcam (Cambridge, UK). The Pierce BCA protein assay kit was purchased from Thermo Fisher Scientific (Waltham, MA, USA). Human recombinant VEGF was supplied by PeproTech (Rocky Hill, NJ, USA). All of the other reagents were obtained from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise indicated.

HUVECs were cultured in endothelial cell medium (ECM; ScienCell, San Diego, CA, USA) supplemented with 5% fetal bovine serum (FBS), 100 U/ml penicillin and 100 µg/ml streptomycin (both from Gibco, Gaithersburg, MD, USA). Breast cancer cell line MDA-MB-231 obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA) was maintained in Dulbeccos modified Eagles medium (DMEM) containing 10% FBS, 100 U/ml of penicillin and 100 µg/ml streptomycin. All the cells were cultured at 37°C with 95% humidity and a 5% CO₂ gas environment.

HUVECs or MDA-MB-231 cells were treated with or without VEGF (10 ng/ml) and Rd for 48 h. The cell viability was determined using MTT assay (Sigma-Aldrich). The number of cells were counted after trypsinizing HUVECs. In addition, the final cell viability and the numbers of the treated cells were expressed as a percentage relative to that of the untreated control cells.

MDA-MB-231 cells were treated with Rd at 0, 25 and 50 µM for 24 h. The cells were collected by centrifugation at 400 × g, and stained with propidium iodide (PI) (50 µg/ml) and Annexin V-FITC (2 µg/ml) for 15 min in the dark. The staining was then immediately analyzed by flow cytometry using the FACScan and CellQuest program. The FCS Express program (BD Biosciences, San Jose, CA, USA) was used to determine the percentage of apoptotic cells.

The wound healing migration assay was performed as previously described (22). Briefly, HUVECs were treated with mitomycin C to inactivate cell proliferation. Scratches were drawn with sterile pipette tips. Fresh ECM was added with or without VEGF (10 ng/ml) and different concentrations of Rd. Images of the cells were captured using an inverted microscope (Olympus CKX41; Olympus, Tokyo, Japan) after incubation at 37°C for 10 h. The width of the scratches was evaluated and used as the indicator for the assessment of cell migration ability.

After incubation with ECM containing 1% FBS for 4 h, HUVECs were seeded at a density of 1×10⁴ cells/well into Matrigel (BD Biosciences, Bedford, MA, USA) coated 96-well plates followed by treatment with Rd at different concentrations for 4 h. Tubes forming intact networks were quantified by counting the number of branch points from 5 random fields/well in a blinded manner under an inverted microscope.

Rat aortic ring assay was performed as previously described (23). In brief, aortas isolated from Sprague-Dawley rats were cleaned of fibroadipose tissue and collateral vessels, and cut into rings of 1–1.5 mm of thickness. The aortic rings were randomly placed into growth factor reduced Matrigel-coated 48-well plates and further overlayed with 100 µl of Matrigel. Medium with or without VEGF (10 ng/ml) supplemented with different concentrations of Rd was added to the wells and incubated with the aortic rings for 6 days. At the end of the incubation period, the microvessel sprouts that had formed were fixed and photographed using an inverted microscope. After images were acquired, the outgrowth area was delineated and measured using Image-Pro Plus software (Media Cybernetics, Rockville, MD, USA), and used for the assessment of angiogenesis.

Matrigel plug assay is a widely used method to assess the in vivo anti-angiogenic effect of drugs (24). To examine the anti-angiogenic property of Rd, Matrigel (0.5 ml) containing 100 ng VEGF and 20 U of heparin with or without Rd (25 and 50 µM) were subcutaneously injected into the ventral area of female C57BL/6 mice (5 weeks old, n=6/group). After 7 days, the mice were sacrificed and the intact Matrigel plugs were isolated and photographed. The hemoglobin in the Matrigel plugs was quantified using Drabkin's reagent kit (Sigma-Aldrich) according to the manufacturer's instructions. The concentration of hemoglobin was calculated based on a set of hemoglobin standards. Blood vessels in the Matrigel were visualized with an antibody against CD31.

Healthy 5-week-old female athymic nude mice (BALB/c) were obtained from Shanghai Laboratory Animal Center. All studies were performed in accordance with the guidelines approved by the Animal Ethics Committee of Shanghai University of TCM (SHUTCM). MDA-MB-231 cells were subcutaneously injected (1×10⁷ cells/mouse) into the right flank of each mouse. Treatments were started 4 days after tumor cell implantation and lasted for 4 weeks. After the tumors grew to ~50 mm³, the tumor-bearing mice were randomly assigned into 5 groups (n=10/group): the vehicle control, the doxorubicin (DOC; 10 mg/kg, once a week for 4 weeks), and the Rd groups (1, 3 and 10 mg/kg). The vehicle control group received the vehicle solvent [0.1% v/v dimethyl sulfoxide (DMSO) in phosphate-buffered saline (PBS)]. The Rd groups were intraperitoneally administered with Rd diluted in vehicle solvent daily. The body weight of the mice was monitored once a week. The tumors were assessed every day using a digital caliper. The tumor volume was calculated using the formula: V (mm³) = [ab²] × 0.5, where a is the length, and b is the width of the tumor. At the end of treatment, the mice were sacrificed and the tumors of the mice from the different groups were collected for further analysis.

Solid tumors were fixed with 10% phosphate-buffered formalin, embedded in paraffin and longitudinally sectioned at 5-µm of thickness. The sections were incubated with 3% H₂O₂ for 10 min to deactivate the endogenous peroxidase. For antigen retrieval, the sections were soaked in 10 mM citrate buffer solution (pH 6.0), and heated twice in the microwave oven. The slides were then washed thoroughly with PBS (pH 7.4). After being blocked with 5% bovine serum albumin (BSA; Sigma-Aldrich) in TBS for 20 min, the sections were incubated with primary antibodies against CD31 and Ki67 at 4°C overnight followed by a thorough wash with PBS. Afterwards, the slides were sequentially incubated with a biotinylated secondary antibody for 20 min and streptavidin-HRP for another 20 min. The staining was visualized after incubation with a DAB-H₂O₂ solution. The slides were then counterstained with hematoxylin for 1 min, dehydrated with ethanol and sealed in resin for microscopic observation.

Cell and tissue homogenates were lysed in lysis buffer containing 50 mM Tris (pH 7.5), 1 mM EDTA, 150 mM NaCl, 20 mM NaF, 0.5% NP-40, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, 10 µg/ml aprotinin, 10 µg/ml leupeptin and 10 µg/ml pepstatin A on ice. After centrifugation at 12,000 × g for 15 min at 4°C, the supernatant was collected and the protein concentration was determined using the BCA method. Total proteins, 30 µg for each sample, were separated on 12% SDS-PAGE and transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore, Bedford, MA, USA). Blocking was performed in 5% BSA (Sigma-Aldrich) in 0.1% Tween-20 in PBS (PBST) for 1 h. The membranes were probed with respective primary antibodies overnight at 4°C. Binding of the primary antibody was detected using a peroxidase-conjugated secondary antibody (Pierce, Rockford, IL, USA) for 1 h at room temperature. The blots were developed using ECL detection reagents (GE Healthcare, Waukesha, WI, USA). The gray intensity of the protein bands was quantified using ImageJ and normalized to that of GAPDH in each sample.

To examine the difference among multiple groups, one-way ANOVA followed by Tukey's multiple comparison test were conducted with GraphPad Prism 5.0. The unpaired t-test was used to assess the difference between two groups. All data are presented as the mean ± SEM. A value of p<0.05 was considered as a significant difference.

As endothelial cell migration is one of the most important and early events during the process of angiogenesis (25), the wound-healing migration assay was performed to determine the effects of Rd on HUVEC migration. Upon stimulation with VEGF, HUVECs migrated much faster and the wounds healed faster compared with the control (Fig. 1B; p<0.01). Rd treatment at 25 and 50 µM significantly prevented VEGF-induced migration of HUVECs as the wound healing was delayed compared to the VEGF-treated cells (p<0.01 and p<0.001). Tube formation assay represents a simple, reliable and powerful model for studying inhibitors of angiogenesis. As shown in Fig. 1C, cells stimulated with VEGF formed robust tubular structures when seeded on growth factor-reduced two-dimensional Matrigel (p<0.05). The addition of Rd suppressed the formation of the capillary-like network (p<0.05 and p<0.01). The process of angiogenesis also requires the proliferation of endothelial cells. VEGF alone promoted the cell viability and increased the number of HUVECs (Fig. 1D; p<0.05, p<0.01). Rd treatment (5, 10, 25 and 50 µM) mitigated the VEGF-induced cell viability and number of HUVECs in a dose-dependent manner (p<0.05, p<0.01 and p<0.001). Overall, these findings clearly demonstrated that Rd exerted an anti-angiogenic effect through the inhibition of cell proliferation, migration and tube formation of endothelial cells.

To study whether Rd affected VEGF-induced angiogenesis ex vivo, an aortic ring assay was conducted. As shown in Fig. 2A, VEGF treatment significantly stimulated microvessel sprouting, leading to the formation of a network of vessels around the aortic rings (p<0.001). The addition of Rd at 25 and 50 µm significantly counteracted the VEGF-induced microvessel sprouting which appeared to be achieved in a dose-dependent manner (p<0.05 and p<0.01).

To further verify the inhibitory effect of Rd on angiogenesis in vivo, the Matrigel plug assay was carried out. As shown in Fig. 2B, Matrigel plugs containing VEGF alone appeared reddish-brown, inside of which increased hemoglobin was found (p<0.001), indicating the formation of functional vasculatures. Accordingly, more CD31 immunoreactive capillaries were found within the VEGF-treated Matrigel plugs (Fig. 2C) and the capillary density was significantly higher (p<0.001), compared with the vehicle-treated control. In contrast, Rd at 25 and 50 µM markedly inhibited VEGF-induced hemoglobin accumulation in the Matrigel plugs as the color of the Rd-treated Matrigel plugs became bleached (Fig. 2B; both p<0.001). Meanwhile, CD31 immunoreactive capillaries were decreased in Rd-treated Matrigel plugs (Fig. 2C; p<0.001). All of these results demonstrated that Rd effectively inhibited angiogenesis in vivo.

Interaction of VEGFR2 with VEGF leads to the activation of various downstream signaling molecules responsible for endothelial cell migration, proliferation and survival (26). HIF-1α is a key regulatory protein in hypoxic response, which is downstream of mTOR signaling and is an important modulator of VEGF (27). To further elucidate the underlying mechanism of the anti-angiogenic effect of Rd, the activation of the signaling molecules in HUVECs were examined under both normoxic and hypoxic conditions.

As shown in Fig. 3A, under normoxic conditions, VEGF induced the expression of VEGFR2, thereby, enhancing the phosphorylation of the PI3K/Akt/mTOR signaling molecules. As a result, downstream p70S6K and HIF-1α that are crucial to the regulation of protein synthesis and angiogenesis (28) were also phosphorylated. Conversely, VEGF-induced VEGFR2 was suppressed by Rd in a dose-dependent manner. Meanwhile, the PI3K/Akt/mTOR signaling pathway molecules as well as p70S6K and HIF-1α activated by VEGF were all inhibited with Rd treatment. We next examined the effect of Rd on the VEGF signaling cascade under hypoxic conditions using CoCl₂, a reagent used widely for the induction of hypoxia (29,30). Not surprisingly, CoCl₂ treatment enhanced the activation of the PI3K/Akt/mTOR/p70S6K signals and increased the expression of HIF-1α and VEGFR2 (Fig. 3B). However, similar to its effect under normoxic conditions, Rd treatment diminished the angiogenic signals induced by CoCl₂ on HUVECs. Therefore, Rd inhibited the VEGF-mediated PI3K/Akt/mTOR/p70S6K signaling cascade activation in both normoxic and hypoxic conditions.

To investigate the effect of Rd on tumor growth and tumor angiogenesis in vivo, a human breast tumor-bearing xenograft mouse model was employed. As shown in Fig. 4A, the administration of Rd at 3 and 10 mg/kg for 28 days substantially suppressed the tumor volume (Fig. 4A; p<0.01) and decreased the tumor weight (Fig. 4C; p<0.05 or p<0.01 in a dose-dependent manner. Notably, administration of Rd at all experimental doses exhibited no obvious toxicity on solid tumor model animals as no significant loss of body weight occurred during the course of the experiment (Fig. 4B). We next evaluated the effect of Rd on cell proliferation and angiogenesis in the solid tumors by immunohistochemical analysis. As shown in Fig. 4D, the number of Ki67 (a marker of cell proliferation) immunoreactive cells in tumor tissues of Rd-treated (3 mg/kg) mice was less than that in the control group of mice (p<0.01). Moreover, CD31 immunoreactive capillaries were decreased in the Rd-treated tumors (p<0.001). In addition, Rd treatment led to a decrease in the expression of HIF-1α and CD31 protein (Fig. 4E; p<0.05). All of these results revealed that Rd prevented angiogenesis and tumor growth in mice.

Since Rd effectively decreased cell proliferation in the xenografted breast tumors, we next examined whether it also had a direct influence on breast cancer cells. As shown in Fig. 5A-C, Rd treatment dose-dependently decreased the cell viability of cancer cell lines, such as MDA-MB-231, MCF-7 and MDA-MB-468. Flow cytometric analysis revealed that Rd significantly increased the percentage of the apoptotic cells (Fig. 5D). Furthermore, Rd treatment increased Bax and cleaved caspase-3 while it decreased Bcl-2 expression in MDA-MB-231 cells, implicating its pro-apoptotic effect. Rd treatment also inhibited the activation of PI3K, Akt, mTOR and p70S6K and mitigated the expression of HIF1-α in MDA-MB-231 cells (Fig. 5E). All of these results indicated that Rd induced direct apoptosis of breast cancer cells, which might be mediated through the inhibition of PI3K/Akt/mTOR signaling pathway.