High performance liquid chromatography (HPLC) grade acetonitrile and ethanol were purchased from Fisher Scientific (Thermo Fisher Scientific, Inc., Leicester, UK). The iScript one-step RT-PCR kit and SYBR Green reagents used were purchased from Bio-Rad Laboratories, Inc. (Hercules, CA, USA). HUVECs were provided by the American Type Culture Collection (Manassas, VA, USA). The authentic standards (purity, >95%) of ginsenoside Ro, chikusetsusaponin IVA, β-ecdysterone and cyasterone were from Shanghai Tauto Biotech Co., Ltd. (Shanghai, China). Trypsin, fetal bovine serum (FBS), penicillin and streptomycin were supplied by Gibco (Thermo Fisher Scientific, Inc., Grand Island, NY, USA). Matrigel™ Matrix Growth Factor Reduced was bought from BD Biosciences (Franklin Lakes, NJ, USA). F12 medium, heparin, endothelial cell growth supplement (ECGS) and all other unspecified chemicals were supplied by Sigma-Aldrich (Merck KGaA, Darmstadt, Germany).

The dried raw herbs of RAB (harvested in Henan, China) and RC (harvested in Sichuan, China) were purchased from Guangzhou Zhixin Pharmaceutical Co., Ltd. (Guangzhou, China). Their voucher specimens (no. 2013-3413 for RAB and no. 2013-3414 for RC) were kept in the museum of the Institute of Chinese Medicine of the Chinese University of Hong Kong (Hong Kong, China) and were authenticated by thin layer chromatography (HPTLC) according to the Chinese Pharmacopoeia 2010. The presence of oleanolic acid and cyasterone were determined in RAB and RC, using HPTLC silica gel F₂₅₄ plates (Merck KGaA).

The raw materials (60 g) of RAB and RC were extracted by 1-h heat water reflux extraction with 2 L of water. Following filtration, the supernatant was collected, the residue extracted by heat ethanol reflux extraction with 2 L of 95% ethanol for 1 h, then filtered again to collect the ethanol. The ethanol was removed by a vacuum rotary evaporator under reduced pressure, and then mixed with the supernatant. Following overnight precipitation, the supernatant was subjected to freeze-drying in order to collect the whole extract. The dried extract was then stored in the desiccator before use.

RAB and RC whole extracts were completely dissolved in water to a concentration of 5 mg/ml and filtered with a 0.22-µm filter. Next, they were subjected to chemical analysis by an Agilent 1290 Infinity HPLC system (Agilent Technologies, Inc., Santa Clara, CA, USA). HPLC was equipped with an online degasser, a binary-pump, an autosampler and a diode array detector, and an Alltima HPLC C18 column (250×4.6 mm, 5 µm; Fisher Scientific) protected by a C18 guard column was used for HPLC analysis according to previous reports, with minor modifications (10,11). Moreover, the column was maintained at 40°C. The mobile phase for the RAB extract consisted of (A) 0.1% acetate acid and (B) acetonitrile with the following gradient: 10–20% B from 0 to 10 min; 20–40% B from 20 to 30 min; 40–60% B from 30 to 40 min; 60–90% B from 40 to 60 min and 90% B from 60 to 70 min. The mobile phase for the RC extract consisted of (A) 0.1% acetate acid and (B) acetonitrile in the following gradient: 10–20% B from 0 to 10 min; 20–28% B from 10 to 20 min; 28% B from 20 to 40 min; 28–60% B from 40 to 50 min; 60–90% B from 50 to 60 min and 90% B from 60 to 70 min. Moreover, the flow rate was set to 0.7 ml/min and the injection volume was 10 µl.

Mass spectrometry was performed using an Agilent 6530 Accurate-Mass Quadrupole Time-of-Flight mass spectrometer (Agilent Technologies, Inc.) equipped with a Jet Stream electrospray ionization (ESI; Agilent Technologies, Inc., Santa Clara, CA, USA) source. Parameters for the ESI source were set as follows: Positive ion mode; gas temperature, 350°C; drying gas, 11 l/min; nebulizer, 55 psi; and capillary, 4,200 V. Moreover, the mass range was set to 100–1,700. Data were qualitatively analyzed using MassHunter Workstation software, version B.05.00 (Agilent Technologies, Inc.).

Cell viability was measured using an MTT assay. DMEM/F12 medium was mixed with heparin (100 mg/l), ECGS (30 mg/l) and 0.5% antibiotics (penicillin-streptomycin-100X, 50 mg⁄ml; Gibco; Thermo Fisher Scientific, Inc.). Next, the mixed DMEM/F12 medium was supplemented with 10% FBS as culture medium or 0.5% FBS as test medium. In brief, HUVECs (1×10⁴ cells per well) were seeded onto the 96-well plates with culture medium overnight, and the medium was then replenished with test medium for another 12 h. RAB and RC whole extract were redissolved into the stock solution (800 µg/ml) with test medium, filtered by a 0.22-µm filter, and diluted with test medium to the working concentrations (0–800 µg/ml) for a 24-h incubation. Following drug treatment, the medium was replaced with MTT (0.5 mg/ml in test medium) for another 3-h incubation at 37°C, and replenished with dimethyl sulfoxide (100 µl) in order to dissolve the formed formazan. The absorbance was detected at 540 nm using a microplate reader.

A wound healing assay (scratch assay) was performed as described previously (12). HUVECs (1.5×10⁵ cells per well) were seeded in culture medium onto 24-well plates overnight. Two crosses were scratched on the cells with a p200 pipette tip, then medium was removed and the cells were washed with pre-warmed PBS. Different concentrations (0–200 µg/ml) of RAB and RC whole extracts in test medium were prepared as mentioned above and were added for an 8-h treatment. Images were then captured before and after drug treatment at a ×40 magnification by an inverted microscope (Eclipse TS100; Nikon Corporation, Tokyo, Japan). Data were analyzed using TScratch software (13). Four replicates were performed for each individual experiment, and four experiments were performed.

RAB and RC whole extracts were redissolved into the stock solution (800 µg/ml) with culture medium and filtered. HUVECs (1×10⁴ cells per well) were mixed with various concentrations (0–200 µg/ml) of RAB and RC whole extracts in culture medium, and seeded onto matrigel-pre-coated 96-well plates for a 2-h incubation. Next, the formed tube networks were monitored at a ×40 magnification using an inverted microscope. The total length of tubes in each image was measured using Image-Pro Plus 6.0 software (Media Cybernetics, Inc., Rockville, MD, USA). Duplicates or triplicates were performed for each experiment, and three experiments were performed in total.

The zebrafish TG (fli1:EGFP) ^y1/+(AB) line was transgenic with ECs expressing enhanced green fluorescent protein, which was supplied by the Zebrafish International Resource Centre of the University of Oregon (Eugene, OR, USA). Zebrafish experiments were approved by the Department of Health, Hong Kong SAR according to the guidelines in the Care and Use of Animals. Zebrafish were sacrificed by TRIzol (Qiagen Sciences, Inc., Gaithersburg, MD, USA Briefly, zebrafish embryos at 1–4 cell stage were collected for disinfection with methyl-blue solution (2 µg/ml) as reported previously (12). Disinfected embryos were washed twice with embryo medium (0.06 g/l Instant Ocean Sea Salt, Blacksburg, VA, USA) and RAB and RC whole extracts were redissolved into the stock solution (800 µg/ml) with embryo medium, filtered by a 0.22-µm filter. Next, they were diluted with embryo medium to the working concentrations. Embryos were incubated with different concentrations (0–200 µg/ml) of RAB and RC whole extracts. Moreover, embryos receiving embryo medium alone served as a control. After 72-h post-fertilization, the morphology of the sub-intestinal vessel (SIV) region in zebrafish larvae was monitored using an IX71S8F-2 inverted fluorescent microscope (Olympus Corporation, Tokyo, Japan). The mean sprout number in the SIV region was calculated by dividing the sum of the sprout number by the total number of embryos in each group, which was used for indicating the pro-angiogenic effect.

After counting sprout numbers, zebrafish were stored in TRIzol (Qiagen Sciences, Inc.). The total RNA was extracted from whole embryos using an RNeasy Mini kit (Qiagen Sciences, Inc.) according to manufacturer's instructions. qPCR was performed with Qiagen iScript one-step RT-PCR kit and SYBR Green using a CFX96™ Real-Time System (Bio-Rad Laboratories, Inc.). HotStarTaq DNA Polymerase (Qiagen Sciences, Inc.). The cycling conditions were set as follows: 50°C for 10 min, 95°C for 5 min, then 50 cycles of 95°C for 10 sec and 60°C for 30 sec. The RT-PCR primers for migration genes were synthesized by Tech Dragon Limited (Hong Kong, China) and are listed in Table I according to our previous study (12). β-actin served as a housekeeping gene. All detections were performed in triplicate, and gene expression analysis was performed to calculate their normalized expression using the 2^−ΔΔCq method (14) with the CFX Manager™ software (Bio-Rad Laboratories, Inc.).

Data are presented as the mean ± standard error of the mean. Statistical analysis was performed by Student t-test or one-way analysis of variance followed by Dunnett's post-test and analyzed using GraphPad Prism software, version 4 (GraphPad Software, Inc., La Jolla, CA, USA). P<0.05 was used to indicate a statistically significant difference.

As shown in Fig. 1, the chemical markers in the RAB extract were identified with their respective molecular weights and retention times when compared to their authentic standards, respectively (Fig. 1A). β-ecdysterone, which is the chemical marker indicated in the Chinese pharmacopoeia for RAB, was found with [M+H]⁺ m/z at 481.3157 (error, 17 ppm). The other two chemical markers, namely ginsenoside Ro and chikusetsusaponin IVA, were detected with [M+Na]⁺ m/z at 979.4871 (error, 0.8 ppm) and [M+NH₄]⁺m/z at 812.4779 (error, 2.2 ppm), respectively.

Cyasterone is the chemical marker indicated in the Chinese pharmacopoeia for RC, and it was identified in the RC extract with [M+H]⁺m/z at 521.3095 (error, 3.6 ppm) when compared with its authentic standard.

As detected by the MTT assay, 24-h treatment of RAB whole extract did not significantly alter the cell viability from 3.125 to 200 µg/ml but displayed cytotoxicity at 400 and 800 µg/ml when compared to the control (P<0.001) (Fig. 2A). Meanwhile, the RC whole extract also did not significantly increase the cell viability from 3.125 to 400 µg/ml but only demonstrated a significant cytotoxicity at 800 µg/ml (Fig. 2B).

As shown in Fig. 3A, the RAB whole extract significantly increased cell migration at 50 µg/ml after an 8-h treatment (P<0.001). Moreover, the open wound area at 50 µg/ml was smaller than that of the control group by ~21.7%. In Fig. 3B, the RC whole extract evidently enhanced cell migration with a smaller open wound area by ~14.9%.

Following a 2-h treatment, RAB and RC whole extracts at 100 µg/ml both individually increased tube formation in HUVECs with 5% enhancement, but the increase was not significant compared to the control (Fig. 4).

When compared to the control group, RAB whole extract at 100 and 200 µg/ml significantly increased the sprout numbers in the SIV region of zebrafish (P<0.001) (Fig. 5A). Meanwhile, RC whole extract at 200 µg/ml markedly enhanced SIV sprout numbers (Fig. 5B). These indicate that RAB and RC whole extracts can both promote angiogenesis in zebrafish.

In Fig. 6, RAB and RC whole extracts regulated migration-related genes. In detail, RAB whole extract (200 µg/ml) increased the gene expression of β-catenin (2.4-fold) but decreased the expression of matrix metallopeptidase 2 (MMP2) by 45%. Moreover, the expression of MMP9, plasminogen (PLG) or tissue inhibitor of metalloproteinase 2 (TIMP2) was not significantly reduced by the RAB whole extract.

In addition, the RC whole extract (200 µg/ml) significantly enhanced the gene expression of β-catenin (1.7-fold; P<0.05), but reduced the expression of MMP2 by 47%. However, the RC whole extract did not significantly regulate the expression of PLG and TIMP2.