All the animal experiment procedures were approved by the Committee of Animal Experimentation and the Ethics Committee of Capital Medical University (Permit No. 2013-X-83). All the operations related with animals were in strict accordance with the requirements and guidelines of experimental animal ministry.

Quercetin, dimethyl sulfoxide (DMSO), 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) and hydroxypropyl-β-cyclodextrin (HPβCD) were purchased from Sigma (St. Louis, MO, USA). RPMI-1640 medium, 0.25% trypsin-EDTA, phosphate-buffered saline (PBS) and fetal bovine serum (FBS) were purchased from HyClone (Logan, UT, USA). VascuLife VEGF medium complete kit, 0.05% trypsin-EDTA and trypsin neutralizing solution were purchased from Cascade Biologics (Invitrogen, USA). Matrigel was from BD Biosciences (Franklin Lakes, NJ, USA).

Human prostate cancer androgen-independent PC-3 cells were obtained from Peking Union Medical College and cultured in RPMI-1640 medium supplemented with 10% FBS. Human umbilical vein endothelial cells (HUVECs) were purchased from Peking Union Medical College and grown in VascuLife VEGF medium. PC-3 cells and HUVECs were maintained at 37°C in the incubator with 95% air and 5% CO₂.

Quercetin was dissolved in DMSO to firstly form a concentration of 10 mM. It was then diluted using cell culture medium to reach the required concentrations (25, 50 and 100 µM) and equal volume of variable concentrations was added to the media. Following culture for the designated time, cell proliferation rate was determined by MTT assay. Comparable amount of DMSO without quercetin acted as the vehicle control during the research process.

MTT assay was generally used to evaluate the cell viability in the culture after drug treatment (10). PC-3 cells at a density of 3×10⁴ cells/ml and HUVECs at a density of 5×10⁴ cells/ml were seeded into 96-well plates in triplicate. After cultured for 24 h allowing both PC-3 cells and HUVECs to attach, cells were treated with variable concentrations of quercetin (25, 50 and 100 µM) or DMSO for further 24, 48 and 72 h. Four hours before the termination of each time points, 20 µl MTT with the concentration of 5 mg/ml was added to each well and co-cultured with PC-3 cells and HUVECs for another 4 h in the humidified environment containing 5% CO₂ at 37°C. Then culture medium was carefully removed leaving the formazan crystals preserved completely, which was dissolved with 150 µl me were calculated based on the following formula: Proliferation inhibition rate = [1 − (A₅₇₀treated/A₅₇₀control)] × 100% (10).

Transwell assay (BD Biosciences) was employed to analyze the cell migration capability. Cells (1×10⁴) in serum-free medium with different concentrations of quercetin or DMSO were seeded in the upper chamber of a 24-well Transwell cluster plates in triplicate. Cell culture medium containing 10% FBS was added into the lower chamber that was separated from the upper one by an 8-µm pore size filter membrane. After incubated in a 5% CO₂ incubator at 37°C for 24 h, cells in the the upper surface of the filter were discarded, and the migrated cells on the lower surface were fixed in ethanol and stained with crystal violet. Finally, migrated cells were photographed with inverted microscope (magnification, ×40) and counted from 5 random high powered fields (magnification, ×200). The data are normalized to control as means ± SD.

Transwell assay was also performed to analyze the cell invasion capability. The Matrigel (1 mg/ml) (BD Biosciences) was well-distributed inside the upper surface of Transwell filter membrane overnight at 4°C. After it was well formed and washed two times with medium before use, cells (1×10⁴) in serum-free medium with different concentrations of quercetin or DMSO were seeded in the upper compartment of the filter insert that was already coated with Matrigel. The remaining procedures were carried out as described in the above migration assay.

Capillary-like tube formation assay was used to evaluate the inhibitory effect of quercetin on angiogenesis in vitro. Experimental procedure was performed as previously described (10). Briefly, 100 µl Matrigel solution was added to the 96-well plates and placed in the incubator with controlled conditions to form a Matrigel gel. HUVECs (1×10⁴/well) were seeded in the Matrigel layer and cultured in VascuLife VEGF medium with quercetin (25, 50 and 100 µM) or DMSO for 24 h at 37°C in 5% CO₂ environment. Tube formation was detected under microscopy and photographed (magnification, ×40). The number was calculated from 5 random fields and inhibition rate was determined according to the formula: Tube formation inhibition rate = [1 − (Tubes_treated/tubes_control)] × 100%.

Male BALB/c nude mice 4–6 weeks old were provided by the experimental animal ministry of Capital Medical University and raised in pathogen-free environment with controlled constant 12-h light and 12-h dark cycle, suitable humidity and temperature. Mice were allowed one week to get accustomed to the new environment before the experiments.

PC-3 cells ~5×10⁵ were suspended in 100 µl PBS and inoculated subcutaneously into the right back. When the xenograft tumor volume grew to ~100 mm³, mice were randomly assigned to vehicle control and three quercetin treated groups (n=8 each group) and treated intraperitoneally. The therapeutic dose and regimen, administration manner and dissolution method were determined according to our preliminary experiments, our previous research results and numerous other research studies (8,11–15). The concrete implementation was as follows: i) vehicle control group, vehicle of quercetin, namely 25% HPβCD (w/v in ddH₂O) on day 1; ii) quercetin treated groups, quercetin 25, 50 and 75 mg/kg dissolved in 25% HPβCD (w/v in ddH₂O) were injected in the three separate groups on day 1 (8,16); iii) three days was a treatment cycle during which mice were weighed and xenograft tumors were measured by slide caliper, and the treatment process was terminated 4 weeks later. Tumor volume was obtained on the basis of the formula: L × S² × 0.5, where L was the longest diameter and S was the shortest diameter of xenograft tumor (8,4). Mental state, food and water intake, health condition and mortality rate were recorded as well (5). At the end of the experiments the mice were sacrificed by cervical dislocation following anesthesia by chloral hydrate. Xenograft tumors were taken out promptly, measured, weighed and photographed. Except for the portion preserved for immunohistochemical examination being kept in 10% neutral buffered formalin, the remaining was rapidly placed into liquid nitrogen for other biomarker analysis.

Western blot experiments were performed as previously described (8,11). Briefly, after HUVECs, PC-3 cells and mice were treated with different concentrations of quercetin, proteins from cells and xenograft tumor tissues were extracted and quantified. Equivalent amount (80 µg) of proteins were loaded, separated by 6–10% SDS-PAGE gels and transferred to polyvinylidene fluoride (PVDF) membrane (Pall, New York, NY, USA) which was blocked by 5% non-fat milk, incubated by primary antibody TSP-1 (1:500, Abcam) overnight at 4°C and GAPDH (1:10,000; Sigma) for 1 h at room temperature, and incubated with horseradish peroxidase-conjugated secondary antibodies (1:2,000; Cell Signaling) for another 1 h at room temperature in turn. Enhanced chemiluminescence kit (ECL Plus; Amersham Pharmacia Biotech, Piscataway, NJ, USA) was used to detect the antigen-antibody complex. GAPDH was used as the loading control.

RT-qPCR was performed as previously described (8). Briefly, Total RNA of HUVECs, PC-3 cells and xenograft tumor tissues were isolated using TRIzol reagent (Invitrogen). The first-strand cDNA was synthesized with oligo(dT) primers by Transcriptor First Strand cDNA Synthesis Kit (Roche, Germany). Sequences of the PCR primers were as follows: TSP-1 forward, 5′-CACGCTGCAG GACAGCAT-3′ and reverse, 5′-GGCCGCCTCAGCTCATT-3′. β-actin forward, 5′-CGGGAAATCGTGCGTGAC-3′ and reverse, 5′-GTGGCCATCTCTTGCTCGAA-3′. SYBR-Green PCR master mix was used to quantify the cDNA under certain controlled conditions (ViiA 7 Real-Time PCR System) (both from Applied Biosystems, Foster City, CA, USA) and the samples were in triplicate. TSP-1 mRNA level was calculated based on the 2^−ΔΔCt method and then normalized to β-actin expression.

The procedures for immunohistochemistry were carried out as previously described (8). Briefly, paraffin was removed and endogenous peroxidase activity was eliminated. Following antigen retrieval, sections were incubated with primary antibody CD31 (1:50) and CD34 (1:100) (both from Abcam) for 24 h at 4°C. The slides were then interacted with horseradish peroxidase-conjugated secondary antibody (1:100; Cell Signaling) at room temperature for another 30 min. Diaminobenzidine reaction was employed to visualize the signals which were counterstained with hematoxylin. CD31- and CD34-positive vessels were counted from three random high power fields in each slide. In addition, at least three sections were analyzed independently in each group.

All the results are represented as means ± SD. Statistical analysis and plots were performed by SPSS 17.0 and SigmaPlot 10.0, respectively. Differences between treated groups and vehicle control were compared using independent-samples t-test. P-value <0.05 denotes significant difference.

After treatment with varied doses of quercetin for the certain time, proliferation of HUVECs and PC-3 cells were inhibited in a dose- and time-dependent manner. Following culture with quercetin at 25, 50 and 100 µM for 24 h, the inhibition rate was 9.2, 13.6 and 18.8%, respectively, for HUVECs, and 13.2, 15.57 and 25.2%, respectively, for PC-3 cells. When treated with above doses for 48 h, HUVECs were inhibited by 12.02, 19.91 and 31.22%, respectively, and PC-3 cells were inhibited by 14.5, 22.8 and 39.3%, respectively. When exposed to the three doses for 72 h, the inhibition rate was 16.74, 31.28 and 55.36%, respectively, for HUVECs, and 21.5, 32.5 and 50.9%, respectively, for PC-3 cells (Fig. 1).

Transwell migration, invasion and tube formation of HUVECs were performed to evaluate the inhibitory effect of quercetin on angiogenesis in vitro. After treated with quercetin at 25, 50 and 100 µM for 24 h, the migration, invasion and tube formation were also inhibited and the level was in proportion to the concentration increase. Migration inhibition rate was 13, 25 and 55%, invasion inhibition rate was 21, 36 and 61% (Fig. 2), and tube formation inhibition rate was 11, 29 and 51%, respectively (Fig. 4).

Effect of quercetin on TSP-1 protein and mRNA expression in HUVECs was investigated by western blotting and Q-PCR. Fig. 5A and C, show that quercetin 25 µM greatly increased TSP-1 protein expression compared with vehicle control, and this phenomenon was more obvious when drug dose was increased. As for TSP-1 mRNA (Fig. 5E), it showed the same pattern as protein with larger dose of quercetin resulting in more mRNA expression than the lower dose.

Inhibitory effect of quercetin on PC-3 cells was assessed by Transwell migration and invasion. After PC-3 cells were cultured with quercetin at 25, 50 and 100 µM for 24 h, the migration inhibition rate was 17, 32 and 57%, and invasion inhibition was 24, 36 and 69%, respectively. The inhibitory degree had a positive association with the dose increase as well (Fig. 3).

TSP-1 protein and mRNA expression were detected to explore the underlying mechanisms of quercetin inhibiting PC-3 cell growth. After PC-3 cells were treated with quercetin for 24 h, 25 µM greatly increased TSP-1 protein and mRNA expression compared with vehicle control group. When quercetin reached 50 and 100 µM, TSP-1 protein and mRNA expression was more than that of the lower dose (Fig. 5A, B and D).

When PC-3 cell xenografts reached ~100 mm³, male BALB/c nude mice were randomly assigned and treated with vehicle or quercetin at the dose of 25, 50 and 75 mg/kg. At the end of the intervention procedure lasting 4 weeks, xenografts of the four groups were harvested and it was found that PC-3 tumors were smaller in quercetin 25 mg/kg treated groups than in the control group. The xenografts become much smaller when the drug dose increased (Fig. 6A). The inhibition rate of 25, 50 and 75 mg/kg groups were 22.85, 29.6 and 37.5%, respectively (Fig. 6B and C). In the entire treatment process, even large dose of 75 mg/kg was well tolerated by the nude mice. Drug-related toxic reaction or side-effects such as poor mental state, convulsion and hematuria were not observed. Weight loss in the treated groups did not show significant difference from the control group (Fig. 6D).

Microvessel density examination was carried out to explore the inhibitory effect of quercetin on angiogenesis in vivo. Immunohistochemistry showed that relative CD31- and CD34-positive vessels in quercetin 25 mg/kg group were 73.5 and 77.8%, respectively, as compared to 100% in the vehicle control group. The positive rates for CD31 and CD34 were 60.08 and 65.87%, respectively, in quercetin 50 mg/kg group, and 44.68 and 46.11%, respectively, in 75 mg/kg group (Fig. 7). The above verified the ability of quercetin to inhibit angiogenesis in prostate cancer cell xenografts in a dose-dependent manner.

Western blotting and Q-PCR detected TSP-1 protein and mRNA expression in xenograft tumor tissues to investigate the reason for quercetin inhibition of prostate cancer PC-3 cell growth in vivo. Fig. 8, shows that quercetin 25 mg/kg led to increased TSP-1 protein and mRNA expression in tumor tissues compared with control, whereas larger doses resulted in greater amount of TSP-1 protein and mRNA. These results indicated that quercetin increased TSP-1 protein and mRNA expression to inhibit angiogenesis thereby leading to antagonize human prostate cancer growth in vivo.