UA, with a purity of 98.6%, was obtained from Xi'an Hao-Xuan Bio-Tech Co., Ltd. (Xi'an, China). The human OS cell line 143B was obtained from the American Type Culture Collection (Manassas, VA, USA). Pifithrin-α (PFT-α) was purchased from Selleck Chemicals (Houston, TX, USA). UA and PFT-α were dissolved with dimethyl sulfoxide (DMSO) for experiments in vitro. For in vivo experiments, UA was suspended in 0.4% carboxymethylcellulose sodium. The primary antibodies rabbit anti-human STAT3 and p-STAT3 were obtained from Abcam (Cambridge, MA, USA), and other antibodies were obtained from Santa Cruz Biotechnology, Inc. (Dallas, TX, USA). Cells were cultured with DMEM (containing 10% FBS, 100 U/ml of penicillin and 100 μg/ml of streptomycin). Cells were incubated in 5% CO₂ and 37°C.

Cell viability was determined with Cell Counting Kit-8 (CCK-8). In brief, 143B cells were seeded in 96-well plates with a final density of 3×10³ cells/well and incubated for 24 h. The cells were treated with different concentrations of UA, recombinant adenovirus or DMSO for 24, 48 and 72 h. Thereafter, 10 μl of CCK-8 (Dojindo Laboratories, Kumamoto, Japan) were added into each well and incubated for another 4 h. The absorbance was determined at 450 nm with a microplate reader. Each test was conducted in triplicate.

The clonogenic assay was employed to determine the ability of cells in a given population to undergo unlimited division and form colonies. This assay was carried out as described (25). Briefly, cells were treated with different concentrations of UA for 24 h and then replated with 2,000 cells/well into 6-well plates. Then cells were maintained up to 14 days until colonies were formed. Plates were washed gently with PBS and incubated with 0.25% crystal violet formalin solution at room temperature for 20 min. Each test was conducted in triplicate.

The 143B cells were plated into a 6-well plate. For cell cycle assay, cells were treated with different concentrations of UA or DMSO for 24 h. Then cells were harvested, washed with cold (4°C) PBS, fixed with cold (4°C) 70% ethanol. Finally, cells were suspended in 300 μl PBS, and incubated with propidium iodide (PI) (20 mg/ml) and RNase (1 mg/ml) for 30 min. The cells were detected with fluorescence-activated cell sorting (FACS) subsequently. The DNA contents were analyzed with ModFit LT software. For apoptosis analysis, cells were treated with UA for 24 h. Then the cells were collected and washed with cold (4°C) PBS, incubated with Annexin V-FITC/PI following the instruction of the kits (KeyGen, Nanjing, China). Finally, the processed cells were sorted with FACS and the data were analyzed with FlowJo. Each test was conducted in triplicate.

Recombinant adenoviruses expressing β-catenin (AdBC) and small interfering RNA fragments targeting β-catenin (AdsiBC) were constructed with AdEasy system (26), respectively. AdBC was tagged with green fluorescence protein and AdsiBC was tagged with red fluorescence protein. The adenovirus-expressing green fluorescence protein (AdGFP) only was used as vector control.

Subconfluent 143B cells were plated in a 6-well plate and treated with pre-designated concentrations of UA or DMSO. For total cellular protein or tissue protein, cells and tissues were harvested and lysed using ice-cold lysis buffer at pre-designated time-points. For subcellular fractionation, the protein was extracted with NE-PER™ Nuclear and Cytoplasmic Extraction Reagents (Pierce Biotechnology, Inc., Rockford, IL, USA) based on the manufacturer's instructions. The lysates were boiled for 10 min, subjected to SDS-PAGE separation and transferred to polyvinylidene difluoride (PVDF) membranes. Then the membranes were blotted with corresponding primary antibodies, followed by incubation with HRP-labelled second antibodies. Finally, the bands of target proteins were developed with the SuperSignal West Pico Substrate (Pierce Biotechnology, Inc.). All assays were performed in triplicate.

Cells were treated with indicated concentrations of UA in T-25 culture flasks. Total RNA was extracted with TRIzol reagents (Invitrogen, Carlsbad, CA, USA) and transcribed to cDNA templates with RT reaction at pre-designated time-points. Then, the cDNA templates were used to detect the expression levels of target genes by PCR. The primer sequences are available upon request. All assays were performed in triplicate.

Cells were seeded in T-25 culture flasks and transfected with β-catenin/TCF-4 luciferase reporter (pTOP-luc) 3 μg per flask with Lipofectamine 2000 (Invitrogen) (25,27). The cells were replated into a 24-well plate 16 h after transfection, and then treated with indicated concentrations of UA or DMSO. The cell lysates were subjected to luciferase assays with luciferase assay kit (Promega Corp., Madison, WI, USA) 24 h after treatment. All assays were performed in triplicate.

The animal experiment was approved by the Institutional Animal Care and Use Committee (IACUC) of Chongqing Medical University. Athymic nude mice (female, 4–6 weeks old, 5/group) were from the Animal Center of Chongqing Medical University (Chongqing, China). The 143B cells were collected and re-suspended in cold PBS (4°C) to 2×10⁷ cells/ml. Then cells in 100 μl of PBS were injected subcutaneously into the right flanks of the nude mice. Three days after injection, the athymic nude mice were given UA (100 and 200 mg/kg) or solvent by intragastric administration once a day for 4 weeks. The mice were sacrificed and the tumor samples were photographed and harvested for histological evaluation.

Retrieved tumor masses were fixed with 4% paraformaldehyde and embedded with paraffin, respectively. Serial sections were deparaffinized and rehydrated in a gradient fashion. Then the slides were stained with hematoxylin and eosin (H&E) (25). For immunohistochemical staining, the slides were further processed for antigen retrieval, and incubated with proliferating cell nuclear antigen (PCNA) antibody (1:100 dilution), or Wnt/β-catenin antibody (1:50 dilution) or isotype IgG as control. Finally, the slides were incubated with streptavidin-labelled secondary antibodies and visualized with 3,3′-diaminobenzidine (DAB) tetrahydro-chloride reagent (25,27).

All quantitative tests were performed in triplicate. Statistical analyses were performed with GraphPad Prism 5 (GraphPad Software, Inc., La Jolla, CA, USA). All measurement results were expressed as mean ± SD. Statistical significances between the two groups were determined with Student's t-test. p<0.05 was considered statistically significant.

To identify whether UA may serve as an effective chemotherapeutic reagent for human OS, the CCK-8 assay was employed to validate the anti-proliferative effect of UA in 143B cells. We found that the proliferation of 143B cells can be inhibited markedly by UA in a time- and concentration-dependent manner (Fig. 1A). Cell cycle analyses indicated that UA induces cell cycle arrest at G1 phase in 143B cells (Fig. 1B). We further checked the biomarkers of G1 arrest. The results indicated that UA inhibited the expression of cyclin-dependent kinase 2 (CDK2), CDK4 and CDK6 (Fig. 1C). Moreover, UA effectively suppresses the protein level of PCNA (Fig. 1C), an indicator for the status of proliferation (28). We further checked whether UA can affect the long-term colony formation ability in human OS cells. Our results illustrated that UA concentration-dependently inhibits the colony formation in 143B cells (Fig. 1D). The above results showed that UA is capable of inhibiting cell proliferation in 143B cells.

Next, we determined whether apoptosis occurs in human OS cells with the treatment of UA. 143B cells were treated with different concentrations of UA for 24 or 48 h. Then cells were analyzed with flow cytometric assay or lysed for western blotting. The results showed that UA can increase the apoptotic cell rate (Fig. 2A), enhance the protein level of Bad and cleaved caspase-3, and reduces the level of Bcl-2 concentration-dependently (Fig. 2B). According to the above results, UA can induce apoptosis in OS cells.

We next assessed the antitumor activity of UA in vivo with a well-established xenograft OS model (27). The results showed that tumor masses in UA-treated group are smaller than those in control group, and UA inhibits the tumor growth significantly compared with control group (Fig. 3A and B). Subsequently, histologic assay was conducted to evaluate the xenograft samples. H&E staining results revealed that more necrotic cells occur in UA-treated groups than that of the control group (Fig. 3C). Furthermore, the expression of PCNA was markedly decreased in UA-treated groups (Fig. 3D), which was consistent with our data in vitro. In addition, we evaluated tumor tissue at molecular level and found that p53 was strongly elevated by UA, while β-catenin, NF-κB and the phosphorylation of STAT3 were decreased (Fig. 3E). These data implied that UA may suppress the growth of OS via β-catenin and inflammatory signaling. Collectively, these in vivo results supported that UA may be a potential antitumor reagent for human OS.

Cell proliferation is well regulated by multiple signaling pathways. With luciferase reporter assay, we found that the transcriptional activity of β-catenin/TCF-4 reporter was effectively reduced by UA (Fig. 4A). Given that the stabilization and nuclear translocation of β-catenin are critical events in the activation of Wnt/β-catenin signaling (27), we employed western blotting assay to check whether UA can decrease the level of β-catenin in the whole cell, cytoplasm, and nucleus. The results indicated that UA decreases the protein level of β-catenin not only in the nucleus, but also in the cytoplasm and the whole cells (Fig. 4B). Moreover, we checked the level of downstream targets in Wnt/β-catenin signaling. The results showed that the expression of c-Myc and cyclin D1 were both decreased by UA concentration-dependently (Fig. 4C). The immunohistochemical results showed that β-catenin positive cells were reduced with UA treatment dose-dependently (Fig. 4D). These results suggested that the anti-proliferative effects of UA in OS cells may be associated with the suppression of Wnt/β-catenin signaling.

To investigate the role of Wnt/β-catenin signaling in the anti-proliferative effect of UA in 143B cells, we employed recombinant adenovirus to mediate the exogenous expression or knockdown for β-catenin. With CCK-8 assay, we found that exogenous expression of β-catenin attenuated the anti-proliferative effects of UA, while knockdown of β-catenin enhanced this function of UA in 143B cells (Fig. 5A). FACS analysis results indicated that overexpression of β-catenin attenuated the G1 phase arrest induced by UA in 143B cells. On the contrary, β-catenin knockdown augmented UA-induced G1 phase arrest (Fig. 5B). Thus, our data indicated that UA may exert its antitumor effects in OS cells by partly inactivating Wnt/β-catenin signaling.

Although inactivation of Wnt/β-catenin signaling partly mediates the anti-proliferative effects of UA in 143B cells, the mechanism on how UA regulates Wnt/β-catenin signaling remains unknown. With further research, we discovered that UA upregulated the mRNA level of p53 (Fig. 6A), increased the protein expression level of p53 and reduced the expression of MDM2 time- and concentration-dependently (Fig. 6B). A previous study demonstrated that overexpression of p53 downregulates β-catenin in human and mouse cells (29). Therefore, we hypothesized that UA-induced inactivation of Wnt/β-catenin signaling may be mediated through the activation of p53. With western blotting assay, we found that the effects of UA on β-catenin, c-Myc and cyclin D1 were partly reversed by p53 inhibitor (PFT-α) (Fig. 6C). Furthermore, the results of CCK-8 assay also showed that PFT-α can partly attenuate the anti-proliferative effects of UA in 143B cells (Fig. 6D), which is similar with the effects of exogenous expression of β-catenin on anti-proliferative effects of UA (Fig. 5A). Our data suggested that the inactivation of Wnt/β-catenin signaling induced by UA may be mediated by upregulating p53 in OS cells.