Introduction
Osteosarcoma (OS) is one of the most common types of non-hematologic malignant bone cancer in humans, particularly in children and adolescents (1,2). Although other types of cancer can eventually penetrate into parts of the skeleton, OS is one of the few that actually initiate in bones and sometimes metastasize elsewhere, usually to the lungs (1–4). Approximately 80% of OS patients have metastatic disease at the time of diagnosis, and yet 10–15% of these lesions are detectable with current radiographic imaging modalities (4). The main clinical approach for OS patient is comprised of surgery with wide excision of the primary tumor and chemotherapy, a treatment which often proves difficult for patients due to the systemic toxicities of chemotherapy agents (5,6). Chemotherapy drugs may cause hearing loss, anemia, abnormal bleeding, kidney and liver damage. Nevertheless, chemotherapy plays a major role in the treatment of OS currently (7–9). The effectiveness of chemotherapy in the prognosis of OS patients cannot be further improved simply by increasing the dose of chemotherapeutic drugs. Therefore, there is a great clinical need to explore new agents for the treatment of OS.
It is currently accepted that herbal and natural products are valuable resources for anticancer drugs (10,11). Artemisinin (ART) is a natural product originally isolated from the plant Artemisia annua L., an herbal drug that has been used in traditional Chinese medicine for centuries (12,13). Dihydroartemisinin (DHA) is a derivative of ART and is a potent anti-malarial drug with only mild adverse effects. Recent studies have shown that ART derivatives including DHA also have profound effects against human tumors including breast, pancreas, prostate and liver cancer with fewer toxic effects (14–18). However, the exact molecular mechanisms by which DHA exerts its anticancer effects remain to be fully investigated.
Here, we investigated the antitumor effect of DHA on human OS cells, as well as the possible molecular basis underlying the function. Our results strongly indicate that DHA can inhibit tumor growth of OS cells both in vitro and in vivo. The inhibitory effect of DHA on OS cells can be mediated by inactivating Wnt/β-catenin signaling, which may be processed in part through increasing catalytic activity of glycogen synthase kinase 3β (GSK3β).
Materials and methods
Cell culture and drug preparations
OS cell lines 143B, U2OS, SaoS2 and MG63 were obtained from the American Type Culture Collection, and maintained in Dulbecco’s Modified Eagle’s Medium (DMEM) with 10% fetal bovine serum (FBS), 100 U/ml of penicillin and 100 μg/ml of streptomycin at 37°C in 5% CO2. DHA was purchased from Sigma-Aldrich, dissolved in dimethyl sulfoxide (DMSO) for in vitro test, or prepared with 0.5% carboxymethyl cellulose sodium (CMC-Na) as suspension for in vivo experiments. Antibodies were from Santa Cruz Biotechnology, Inc., and Cell Signaling Technology. All other reagents were purchased form Sigma-Aldrich, unless otherwise indicated.
Crystal violet assay
OS cells were seeded in 24-well plates and treated with different concentrations of DHA or DMSO as control. At the scheduled time points, cells were stained with crystal violet to visualize the cell viability. For quantification, crystal violet in the stained cells was extracted with 1 ml 20% acetic acid at room temperature for 20 min with shaking. A total of 100 μl were removed and added to 1 ml ddH2O. Absorbance at 570 nm was measured (19). Each assay was performed in triplicate.
MTT proliferation assay
143B cells were seeded in 96-well plates and treated with different concentrations of DHA. Cell proliferation was assessed by MTT assay at the scheduled time points. Optical absorbance was measured at 570 nm using a 96-well microplate reader. All experiments were performed in triplicate.
Colony formation assays
One day prior to DHA treatments, 200 cells were seeded into 6-well plates in complete DMEM. The cells were then treated with variable concentrations of DHA and cultured in complete DMEM for up to 14 days. Colonies were visualized by crystal violet staining and counted. The colony forming rate was obtained by the following calculation: (colony number/seeded cell number) ×100%. All experiments were performed in triplicate.
Transwell migration and invasion assay
The migration and invasion assay was performed as previously described (20,21). For the migration assay, 143B cells were seeded at 2×105/well in the upper chamber of type I-collagen-coated 24-well culture inserts. For the invasion assay, the upper side of the insert was coated with extracellular matrix (ECM) gel. DHA was added to the lower chamber. After 24 h, the cells were dried for 5 min, fixed with dehydrated alcohol, and stained with hematoxylin-eosin. The transmembrane cells were counted. All experiments were performed in triplicate.
Hoechst apoptosis staining
143B cells were plated in 24-well plates and treated with different concentrations of DHA or DMSO. After maintaining at 37°C in 5% CO2 for 48 h, cells were collected and subjected to Hoechst 33258 staining. The apoptotic cells were visualized under fluorescence microscopy as previously reported (22). Each assay was carried out in triplicate.
Western blot assay
143B cells were plated in 24-well plates and treated with different concentrations of DHA or DMSO. The cells were then collected and lysed in Laemmli buffer. Cleared total cell lysate was denatured by boiling and loaded onto a 4–20% gradient SDS-PAGE. Following electrophoretic separation, proteins were transferred to an Immobilon-P membrane. The membrane was blocked with SuperBlock blocking buffer and probed with the primary antibody, followed by incubation with a secondary antibody-conjugated with horseradish peroxidase. The proteins of interest were detected by using SuperSignal West Pico Chemiluminescent Substrate kit. Each assay was carried out in triplicate.
Luciferase reporter assay
143B cells were seeded in T-25 flask and transfected with 3 μg of Top-luc luciferase reporter plasmids. Twelve hours later, trypsinized cells were seeded to 24-well plates and were then treated with DHA or DMSO. At the scheduled time points, cells were lysed and subjected to luciferase assays using luciferase assay kit. Luciferase activity was normalized by total cellular protein concentrations of the samples. Each assay was performed in triplicate.
Establishment of orthotopic OS tumor animal model
All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Chongqing Medical University. Athymic mice (female, 4–6 weeks old, 5/group) were ordered from the Animal Centre of Chongqing Medical University (Chongqing, China). 143B cells were collected and resuspended in cold (4°C) phosphate buffer solution (PBS) to a final density of 2×107 cells/ml. Cells in 50 μl of cold PBS were injected into the proximal tibia of athymic mice. Animals were then treated with either different doses of DHA (5, 10 and 20 mg/kg) or solvent by intragastric administration, once a day until the 37th day. The tumor size was measured every day from the 1st week. Animals were sacrificed at 37 days after injection. The tumor samples were retrieved for histological evaluation.
Statistical analysis
Data presented are the results of at least 2 independent experiments performed in triplicate. Microsoft Excel was employed to calculate the standard deviations. The differences were analyzed using the Student’s t-test. p<0.05 was considered to indicate a statistically significant difference.
Results
DHA inhibits the proliferation of OS cells
We first employed crystal violet staining to test the proliferation inhibitory effect of DHA on human OS cells. The results showed that DHA inhibited the proliferation of 143B OS cells effectively in a concentration-dependent manner, even at the minimum concentration of 2.5 μM (Fig. 1A and B). Similar results were obtained in other commonly-used OS cell lines, including U2OS, SaoS2 and MG63 (Fig. 1C and D). Moreover, the inhibitory effect of DHA on 143B cell proliferation was further confirmed by MTT assay (Fig. 1E) and colony formation assay (Fig. 1F and G). Proliferating cell nuclear antigen (PCNA) was a simple but potent marker for the assessment of OS growth (23). Therefore, we also tested the level of PCNA with western blotting. The results showed that the level of PCNA in 143B cells was decreased by DHA as well (Fig. 1H). Collectively, these results suggest that DHA effectively inhibits the proliferation of OS cells.
DHA inhibits the migration and invasion of OS cells
Cell migration and invasion play a crucial role in the process of tumor metastasis. Herein, transwell assay without/with ECM was used to assess the alteration in cellular migration and invasion induced by DHA in OS cells. As shown in Fig. 2A and B, DHA inhibited the number of migrated OS cells in a dose-dependent manner. We also found that when 143B cells were treated with DHA, the number of invaded cells across the ECM-coated membranes decreased accordingly (Fig. 2C and D). Moreover, as illustrated in Fig. 2E, DHA reduced the protein level of matrix metalloproteinase 9 (MMP9), vascular endothelial growth factor (VEGF) and cyclooxygenases 2 (COX2), which have been proved to highly associate with migration and invasion in cancer metastasis (24–26). These results demonstrate that DHA effectively attenuates the migrating and invasiveness phenotype of OS cell in vitro. This function may at least partly result from the reduction of MMP9, VEGF and COX2 proteins in OS cells.
DHA induces apoptosis in OS cells
We next investigated whether DHA induces OS cells to undergo apoptosis with Hoechst 33258 staining assay. The result showed that the apoptotic cells in the DHA-treated group were more pronounced than those of the blank or DMSO group (Fig. 3). Furthermore, we found that the level of cleaved caspase 3, a well-established central executor in cell apoptosis, was increased by DHA in a concentration-dependent manner. DHA also disturbed the balance between proapoptotic protein Bad and anti-apoptotic protein Bcl2, led to an increase in Bad but a reduction in Bcl2 protein (Fig. 2C). The data indicate that DHA is capable of inducing and promoting apoptosis, a characteristic shared by most of the currently used chemotherapy agents (27,28).
DHA inhibits tumor growth of OS cells in vivo
The above data demonstrate that DHA inhibited tumor growth of OS cells in vitro. We sought to confirm these findings in an orthotopic OS tumor animal model in vivo. 143B cells were implanted into the proximal tibia of athymic mice under DHA treatment, and tumor size was monitored. As shown in Fig. 4A, the DHA treatment group exhibited significantly decreased tumor size, compared with the control group. At 37 days, the animals were euthanized and the tumor masses were retrieved (Fig. 4B). The H&E staining results showed that sample in the control group was more aggressive than that of the DHA-treated groups, the OS cells invaded and replaced almost all the bone at primary site (Fig. 4C). However, the samples retrieved from the DHA group displayed relative intact bone structure and less tumor cells, particularly in the high dose group (Fig. 4C). These results confirm that DHA inhibits OS tumor growth in vivo.
DHA suppresses Wnt/β-catenin signaling in OS cells
We found that DHA inhibits tumor growth of OS cells both in vitro and in vivo. Next, we sought to examine the possible mechanism behind the anticancer activity of DHA in OS cells. Western blot results showed that p-P53, P53 and MDM2 were not altered under DHA treatment, indicating that the level of these molecules is possibly not related to the anticancer effects of DHA in OS cells (Fig. 5A). However, total β-catenin protein was effectively reduced by DHA, as were the β-catenin targets C-myc and cyclin D1 (Fig. 5A). Furthermore, we found that DHA was able to inhibit reporter activity of TOP-Luc, which contains TCF/LEF-responsive elements and reflects β-catenin transcriptional activity (Fig. 5B). These results indicate that DHA suppresses Wnt/β-catenin signaling in OS cells. To investigate how DHA reduces the β-catenin protein in OS cells, we employed western blot assay to detect the activation status of GSK3β in OS cells. We found that DHA increased GSK3β activity by suppressing phosphorylation of GSK3β at serine 9 (Ser 9), which has been proved to be an inactive form of GSK3β (Fig. 5A). Notably, we observed that the level of DVL proteins, a main switchboard of Wnt/β-catenin signaling, is decreased by treatment of DHA as well (Fig. 5A). These findings suggest that downregulation of β-catenin protein in OS by DHA resulted from upregulation of GSK3β activity and destabilize the β-catenin.
To validate the role of β-catenin in the inhibitory effect of DHA on OS cells, the overexpression of exogenous β-catenin or knockdown of β-catenin in OS cells was achieved respectively (Fig. 6A). The proliferation of 143B cells was suppressed by knockdown of β-catenin, although it was not altered by exogenous β-catenin (Fig. 6B). Markedly, we found that exogenous expression of β-catenin attenuated the growth inhibitory function of DHA, whereas knockdown of β-catenin enhanced this function in 143B cells, suggesting that decrease of β-catenin protein plays an important role in the anti-proliferation function of DHA in OS cells (Fig. 6B). In the transwell invasion assay, we observed that the invasion suppressive effect of DHA was neutralized by exogenous expression, but augmented by β-catenin knockdown (Fig. 6C and D). Collectively, the above data indicate that the inhibitory effect of DHA on OS cells may result from the inactivation of Wnt/β-catenin signaling.
Discussion
OS is a common primary bone cancer with considerable morbidity and mortality. The current clinical strategies for OS include surgical resection, radiotherapy and chemotherapy. However, OS patients still have a poor prognosis under these aggressive treatments (29). Thus, there is an urgent need to develop more effective therapies for OS treatment. Herbal and natural products have been proved to be highly suitable sources for anticancer drugs (10,11,14–18). DHA, a derivative of ART, was initially used as an anti-malarial. However, recent studies indicated that DHA is capable of improving immunity, preventing inflammation, regulating cell differentiation and inhibiting cancer growth with low host toxicity (30–34). The anticancer effect of DHA is partly associated with iron homeostasis (33). Moreover, DHA is shown to activate p38 and ERK1/2, indicating these signal molecules are relevant to the anticancer effects of DHA (34). Despite these findings, the molecular mechanism through which DHA inhibits tumor growth remains largely unclear and warrants extensive studies. In this study, we investigated the anticancer effect of DHA on human OS cells. We found that DHA prevents cell proliferation in OS cells effectively. Of note, however, the inhibitory effect of DHA on OS cells is possibly mediated by suppressing the Wnt/β-catenin signaling through activation of GSK3β and subsequent reduction of β-catenin protein.
We found that DHA can reduce metastatic phenotype, as well as expression of MMP9, VEGF and COX2 of OS cells in vitro. MMP9, member of the MMP family, is capable of degrading extracellular matrix and is involved in the metastasis of cancer (35,36). VEGF is a key angiogenic factor that can stimulate vasculogenesis and angiogenesis (37–39). Cancer that expresses VEGF is prone to invade and metastasize (38,39), and downregulation of VEGF can attenuate the invasiveness phenotype of malignant tumor (40,41). COX2 is known to be closely associated with tumor growth and metastasis in several types of human tumors (42,43,25). COX2 is unexpressed under normal conditions in most cells, but is upregulated in several types of cancer. The overexpression of COX2 is associated with increased angiogenesis and aggressive invasion in cancer (43,25). In human OS, COX2 is directly associated with OS cell migration and invasion (25,44). Therefore, inhibition of MMP9, VEGF and COX2 expression by DHA may be an important mechanism responsible for the decreased metastatic phenotype in OS cells.
Induction of apoptosis is a common mechanism underlying the anticancer effects of synthetic and natural drugs. Hoechst 33258 staining assay shows that DHA can induce apoptosis in OS cells. Cleaved caspase 3, the activated form of caspase 3, is regarded as a critical effector in the sequential activation of caspases, and plays a central role in the execution-phase of mitochondrial apoptosis pathway. Our results indicated that DHA can increase the level of cleaved caspase 3 in OS cells, suggesting that the mitochondrial apoptosis pathway is likely involved in DHA-induced apoptosis. Bad and Bcl2 belong to the Bcl2 family members that are critical regulators of apoptosis. However, the role of Bad and Bcl2 in cell apoptosis is different. Bad is a well-established proapoptotic factor, whereas Bcl2 is an antiapoptotic molecule known for its ability to inhibit the release of cytochrome c from mitochondria (45). DHA can reduce the levels of Bcl2 in OS cell lines, whereas it increases Bad protein levels. Therefore, disturbance of the balance between proapoptotic protein (Bad) and antiapoptotic protein (Bcl2) by DHA provides relative evidence to demonstrate the DHA-induced apoptosis in OS cells. Induction of apoptosis of human OS cells by DHA may be an important mechanism responsible for the inhibitory effect on OS cell growth.
It has been reported that the abnormal activation of Wnt/β-catenin signaling accounts for tumor initiation of human OS (45–49). Our results showed that DHA can reduce the total protein level of β-catenin. Overexpression of exogenous β-catenin alone in OS cells cannot further enhance proliferation and invasion of OS cells. In the present study, we hypothesized that the β-catenin protein in OS cells is already high enough, therefore exogenous β-catenin is not necessary to further increase malignant phenotype of OS cells. However, overexpression of exogenous β-catenin can neutralize the function of DHA, and knockdown of β-catenin can enhance the growth inhibitory effect of DHA. Therefore, we propose that the effect of DHA on OS cells may, in part, result from targeting Wnt/β-catenin signaling.
The Wnt/β-catenin signaling is a canonical Wnt pathway that causes an accumulation of β-catenin and its eventual translocation into the nucleus. The β-catenin in the nucleus interacts with TCF/LEF family to regulate downstream gene expression (50). Wnt/β-catenin pathway plays important roles in embryogenesis and tumorigenesis. Without Wnt ligands, free β-catenin is rapidly degraded through proteosome machinery promoted by a destruction complex assembled with APC, Axin and GSK3β. In the presence of Wnt ligands, this destruction complex function becomes disrupted and the β-catenin protein becomes stable in the cytoplasm and translocates to the nucleus (50). Following translocation to the nucleus, β-catenin binds to T-cell factors (TCF) and activates the transcription of specific target genes, including C-myc and cyclin D1 (50). GSK3β is a master protein in the Wnt/β-catenin pathway by facilitating phosphorylation and promoting degradation of β-catenin (50). Catalytic kinase activity of GSK3β is controlled through differential phosphorylation of serine/threonine residues. Phosphorylation at tyrosine 216 in GSK3β enhances the enzymatic activity of GSK3β, while phosphorylation of serine 9 significantly decreases catalytic activity (51). We found that phosphorylation of serine 9 residue in GSK3β is decreased by DHA treatment, which lead to an increased catalytic activity of GSK3β. This result indicates that the decrease of β-catenin protein in OS cells may result from the degradation initialized by GSK3β. Our results show that DHA can downregulate DVL proteins in OS cells. DVL proteins are upstream controllers of APC/Axin/GSK3β complex, and are positive mediators of the Wnt/GSK3β pathway. It has been shown that DVL proteins are frequently upregulated and are significantly associated with abnormal elevated β-catenin protein in several types of cancer (52–54). Moreover, targeted inhibition of DVL proteins decreased β-catenin protein, reduced TCF-dependent transcription and inhibited cell growth in human cancer (55,56). Although further analysis should be conducted, the decreased DVL proteins by DHA are possibly related to the anticancer effects of DHA by disturbing Wnt/β-catenin signaling in OS cells as well.
In conclusion, the present study suggests that DHA can be used as an effective chemotherapy agent for human OS. The proliferation and invasion inhibitory effects of DHA on OS cells may result from inactivating Wnt/β-catenin signaling through increasing GSK3β activity and thereby promoting the degradation of β-catenin. Future studies will be directed at elucidating the exact mechanism of DHA on the degradation of the β-catenin protein, as well as at combining DHA with other traditional chemotherapy drugs to develop a novel therapy strategy for human OS cancer.