Zebularine enhances apoptosis of human osteosarcoma cells by suppressing methylation of ARHI.

ARHI is an imprinted tumor suppressor gene and its methylation suppresses ARHI transcription levels to cause the development and progression of malignant tumors. Zebularine exerts a demethylation function for tumor suppressor genes. Our study aims to investigate the effect and mechanism of action of zebularine on the epigenetic modification of the ARHI gene, and whether this effect may modulate the viability and apoptosis of human osteosarcoma cells. We found that zebularine inhibited the viability and promoted apoptosis in osteosarcoma cells. Zebularine potentiated the expression of ARHI at both the protein and mRNA level. This was related to the downregulation of methylation of ARHI caused by zebularine. Zebularine suppressed the interaction of DNA methyltransferase 1 (DNMT1) with histone methyltransferase G9a, but had no effect on G9a alone. Knockdown of DNMT1 or G9a can induce a reduction of ARHI methylation. Therefore, we inferred that zebularine was likely to directly repress DNMT1 alone, but G9a was necessary to regulate the function of DNMT1 on ARHI methylation. Moreover, knockdown of ARHI rescued cell viability and apoptosis under the zebularine-treated condition. We showed that zebularine inhibited viability and promoted apoptosis by disturbing the interaction between DNMT1 and G9a, thereby resulting in lower ARHI methylation and elevated ARHI expression in osteosarcoma cells.

Osteosarcoma is a common primary malignant bone cancer in children and adolescents.1 Epidemiologic data showed that the annual incidence of osteosarcoma is approximately three cases/million population, accounting for 0.2% of all malignant tumors.2 The current optimal treatment for osteosarcoma includes neoadjuvant chemotherapy and surgical resection of resectable osteosarcoma. Nonetheless, surgical resection has great limitations for patients with relapsed or metastatic disease, and the effectiveness of postoperative chemotherapy does not satisfy all patients. Moreover, the frequent acquisition of drug‐resistant phenotypes and the occurrence of “secondary malignancies” are often associated with chemotherapy.1 It is difficult to elect appropriate and effective chemotherapeutic drugs for the treatment of osteosarcoma.

Zebularine (1‐[β‐D‐ribofuranosyl]‐1,2‐dihydropyrimidin‐2‐one) is a cytidine analogue that may form a covalent complex with DNA methyltransferase to inhibit DNA methylation.3 In contrast to other DNA methylation inhibitors, such as 5‐aza‐2′‐deoxycytidine, zebularine has higher stability and lower toxicity detected both in vitro and in vivo.4 In several studies to date, zebularine has been shown to upregulate tumor suppressor genes by demethylation in carcinoma cells.5 Furthermore, methylation of tumor suppressor genes causes subsequent interruption of pro‐apoptotic pathways, which are deemed to contribute to the improvement of proliferation and/or drug resistance.6 ARHI, also called DIRAS3 (GTP‐binding protein Di‐Ras3), is an imprinted tumor suppressor gene; its methylation suppresses ARHI activity.7 As ARHI is frequently downregulated by methylation, the loss of its expression may contribute to the pathogenesis of the majority of cancers.8 Therefore, methylation of ARHI may participate in the pathogenesis of malignant tumors. Thus, there may be an association between zebularine and ARHI methylation, which may be applied in tumor therapy.

In this study, we examined the effects of zebularine on viability and apoptosis in human osteosarcoma cells, and investigated the impact of zebularine on ARHI expression. Additionally, we explored the mechanism of zebularine on modulating ARHI methylation in human osteosarcoma cells.

Materials and Methods

Cell culture

Human osteosarcoma cell lines, including those derived from fibroblastic (HOS, MG‐63) or osteoblastic (U2OS, Saos‐2) high‐grade osteosarcoma, and normal human osteoblasts (hFOB 1.19), were obtained from ATCC (Manassas, VA, USA). All cells were cultured in DMEM (Invitrogen, Carlsbad, CA, USA) supplemented with 10% FBS (Invitrogen), 100 U/mL penicillin, and 100 mg/mL streptomycin (Invitrogen), and maintained at 37°C in a humidified incubator with 5% CO2.

siRNA transfection

The single‐stranded DNA methyltransferase 1 (DNMT1) siRNAs, histone methyltransferase G9a siRNAs, ARH1 siRNA, and related negative control siRNAs were respectively transfected into U2OS cells using Lipofectamine 2000 (Invitrogen). The siRNA sequences were designed by Invitrogen Block‐iT RNAi Designer (http://rnaidesigner.thermofisher.com/rnaiexpress/). G9a siRNA1, 5′‐GCCUCUAUGCCAACUGGUU‐3′; G9a siRNA2, 5′‐CCAUGCUGUCAACUACCAUGG‐3′; G9a siRNA3, 5′‐UCACGCACUCAGGAGCGCAC‐3′. DNMT1 siRNA1, 5′‐GGAGCUGUUCUUGGUGGAU‐3′; DNMT1 siRNA2, 5′‐UUCAUGUCAGCCAAGGCCAC‐3′; DNMT1 siRNA3, 5′‐ ACCATGAGCACCGTTCTCC‐3′; control siRNA, 5′‐UUUAGCGCCGAAAAGAAUCC‐3′. ARH1 siRNA, 5′‐GCCAACAAUGUAUACGCGGAU‐3′; control siRNA: 5′‐UUCUCCGAACGUGUCACGU‐3′.

Cell viability analysis

The hFOB 1.19, U2OS, and MG‐63 cells were treated with 50, 100, 200, and 300 μM zebularine for 72 h, or the cells were treated with 200 μM for different times. Cell viability was analyzed by purchased cell counting kits (Sigma‐Aldrich, St. Louis, MO, USA). Assays were repeated four times for each sample.

Cell apoptosis assay

The apoptotic cells were measured by flow cytometry using an annexin V–FITC/propidium iodide apoptosis detection kit (Abcam, Cambridge, UK) in U2OS cells. The fluorescence intensity was detected at 488 nm using flow cytometry. Cells were sorted by the FACSCalibur flow cytometer (Becton Dickinson, San Diego,CA), and analyzed using CellQuest software (Becton Dickinson).

Western blot analysis

Total proteins were extracted using the Tissue or Cell Total Protein Extraction Kit (Amresco, Solon, OH USA) from HOS, MG‐63, U2OS, Saos‐2, and hFOB 1.19 cell lines. All primary antibodies were purchased from Abcam. The proteins were separated by SDS‐PAGE followed by electrotransfer to nitrocellulose membranes. The membranes were probed using antibodies against ARHI (1:1000), DNMT1 (1:2000), and G9a (1:1000) followed by an HRP‐conjugated secondary antibody (Abcam). Bands were revealed with ECL reagent (Millipore, Boston, MA, USA) and recorded on X‐ray films (Kodak, Rochester, NY, USA). The densitometry of each band was quantified by a gel imaging system and Quantity One 4.62 software (Bio‐Rad, Hercules, CA, USA).

Reverse transcription–PCR

Total RNA was extracted using TRIzol reagents (Invitrogen) from HOS, MG‐63, U2OS, Saos‐2, and hFOB 1.19 cell lines. Isolated RNA was electrophoresed on 1% agarose gel to detect the purity of total RNA. The first‐strand cDNA was synthesized using 1 μg total RNA and SuperScript III Reverse Transcriptase (Invitrogen). Polymerase chain reaction amplification was carried out using a PCR amplification kit (Takara Biotechnology, Dalian, China). The specific primers were designed using Primer Premier 6.0 software and synthesized by Sangon Biotech (Shanghai, China). The primers for ARHI were 5′‐TCTCTCCGAGCAGCGCA‐3′ (forward) and 5′‐ATCTTCCTGTGGGGCTTGAAGG‐3′ (reverse). The primers for GAPDH as an internal control were 5′‐ACCACAGTCCATGCCATCAC‐3′ (forward) and 5′‐TCCACCACCCTGTTGCTGTA‐3′ (reverse). The PCR production was electrophoresed on 1% agarose gel and visualized by the Gel Imaging System of Bio‐Rad. Each band was analyzed by Quantity One 4.62 software (Bio‐Rad).

Combined bisulfite restriction analysis

Combined bisulfite restriction analysis (COBRA) was carried out by referring to Xiong and Laird's study.9 Genomic DNA was extracted from cell lines using QIAamp DNA Mini Kit (Qiagen, Dusseldorf, Germany) and treated with bisulfite (Sigma‐Aldrich). The modified DNA was amplified using PCR. The specific primers were synthesized by Sangon Biotech. The primers for CpG I (187 bp) were 5′‐GTAAGGGAGAAAGAAGTTAGA‐3′ (forward) and 5′‐TACTATCCTAACAAAACCCTC‐3′ (reverse). The primers for CpG II (207 bp) were 5′‐GTTGGGTTAGTTTTTTATAGTTGGTT‐3′ (forward) and 5′‐AACCAAACAACCTAAAAAACAAATAC‐3′ (reverse). The primers for CpG III (184 bp) were 5′‐GTTTTTAAGTTTTATAGGAAGATT‐3′ (forward) and 5′‐ATAATATACAAAAAAAACACACC‐3′ (reverse). After amplification, PCR products were digested with the restriction enzyme TaqI (New England Biolabs) for CpG I and III or BstUI (New England Biolabs, Ipswich, MA, USA) for CpG II. DNA was then electrophoresed on 2% polyacrylamide gel. The gels were stained with ethidium bromide and visualized by Bio‐Rad's Gel Imaging System. Each band was analyzed by Quantity One 4.62 software (Bio‐Rad).

Co‐immunoprecipitation

For co‐immunoprecipitation, cells were washed with ice‐cold PBS and lysed in lysis buffer (25 mM HEPES [pH 7.5], 100 mM NaCl, 1 mM EDTA, 10% glycerol, 1% NP‐40, 0.5 mM MgCl2, 10 mM NaF, 1 mM PMSF, 1 mM Na orthovanadate, and 1 mM aprotinine and leupeptin) for 30 min. Immunoprecipitation was carried out on cleared lysates (13 000 g for 15 min at 4°C) with the DNMT1 or G9a antibodies and protein A/G agarose beads (Merck, Darmstadt, Germany) at 4°C overnight. The precipitates were washed in lysis buffer (3×) and boiled in SDS sample buffer containing 100 mM DTT. Samples were subjected to Western blotting analysis using anti‐DNMT1and anti‐G9a antibodies and stained using the ECL (Millipore, Boston, MA, USA) system.

Statistical analysis

Data are reported as mean ± SD in at least four replicates per group. Data were analyzed by spss 13.0 software (IBM, Almon, NY, USA). Statistical differences between means were calculated using anova, followed by least significant difference multiple comparison tests. Differences were considered significant at P < 0.05.

Results

Expression and methylation status of ARHI in human osteosarcoma cells

Expression of the ARHI protein in human osteosarcoma cells showed a marked reduction compared to normal osteoblast cells (P < 0.01; Fig. 1a,b). Meanwhile, the expression of ARHI mRNA in osteosarcoma cells was also significantly reduced compared with normal osteoblast cells (Fig. 1a,c). To explore the cause of reduction of ARHI expression, we further assessed the methylation status of CpG islands associated with the ARHI gene by COBRA. The sites of CpG I and II, and CpG III were located in the proximate promoters Exon 1 and Exon 2, respectively (Fig. 1d). CpG I and II were hypomethylated, and CpG III was partially methylated in normal human osteoblast cells (Fig. 1e). Conversely, in human osteosarcoma cells, CpG I, II, and III were partially methylated, especially in Saos‐2 and U2OS cell lines that were derived from osteoblastic osteosarcoma.

Figure 1

expression and methylation status. (a) Western blot and RT‐PCR analyses of expression in protein and mRNA levels in human osteosarcoma cells and normal osteoblast cells. (b,c) Semiquantitative histograms of Western blot (b) and RT‐PCR (c) analyses. (d) The distribution of CpG islands on the promoter region of ARHI gene. (e) Combined bisulfite restriction analysis was used to analyze the methylation status of CpG islands associated with the gene. DNA was amplified and digested with restriction enzymes that can distinguish methylated (c) and unmethylated (a or b) fragments. The methylation percentage, show below each lane, was calculated using the formula: %methylation = 100 × (c/(b + c)). %methylation >85, hypermethylation; %methylation 15–85, partial methylation; %methylation <15, hypomethylation. The values are presented as mean ± SD of four independent experiments. **P < 0.01 versus hFOB 1.19 normal osteoblast cells.

Zebularine suppresses cellular viability in U2OS and MG‐63 cells

To discuss the effect of zebularine on the viability of osteosarcoma and normal osteoblast cells, 50, 100, 200, and 300 μM zebularine was used to incubate the cells for 72 h. The Cell Counting Kit‐8 (CCK‐8) assay showed that the viability of osteosarcoma cells was greatly decreased in a dose‐dependent manner, whereas normal osteoblast cells were less sensitive to zebularine, compared with the osteosarcoma cells (Fig. 2). The cells were then treated with 200 μM zebularine for 0, 12, 24, 48, and 72 h. The viability of osteosarcoma cells and normal osteoblast cells were both suppressed by zebularine treatment, especially following incubation 200 μM zebularine for more than 48 h (Fig. 2). Therefore, in the following experiments, 200 μM and 48 h were the concentration and time applied in the incubation.

Figure 2

Effect of zebularine on cell viability in U2OS and MG‐63 osteosarcoma cells. (a) Normal osteoblast hFOB 1.19 cells and U2OS and MG‐63 cells were treated with 0, 50, 100, 200, and 300 μM zebularine for 72 h, and Cell Counting Kit‐8 (CCK‐8) assay was used to detect cell viability. (b) hFOB 1.19, U2OS, and MG‐63 cells were incubated with 200 μM zebularine for 0, 12, 24, 48, and 72 h, and CCK‐8 assay was used to detect cell viability. The values are presented as mean ± SD of four independent experiments. *P < 0.05, **P < 0.01 versus 0.

Zebularine promotes apoptosis in human osteosarcoma cells

To determine the effect of zebularine on apoptosis in U2OS cells, normal and U2OS cells were respectively treated with 200 μM zebularine for 48 h, apoptosis was monitored by annexin V–FITC/PI double‐labeled staining. As shown in Figure 3, the apoptosis rate in normal osteoblast cells did not vary significantly after treatment with 200 μM zebularine for 48 h, but it was in U2OS cells that the apoptosis rate clearly elevated (P < 0.01; Fig. 3e).

Figure 3

Effect of zebularine on apoptosis in U2OS osteosarcoma cells. (a) Normal, untreated cells. (b) Normal cells treated with 200 μM zebularine for 48 h. (c) Osteosarcoma cells, untreated. (d) Osteosarcoma cells treated with 200 μM zebularine for 48 h. (e) Sum of quadrant (Q)2 and Q4. Values are presented as mean ± SD of four independent experiments. **P < 0.01. PE, phycoerythrin.

Zebularine enhances ARHI expression

We then elucidated the effect of zebularine on ARHI expression. The expression of ARHI protein had no significant difference after treatment with zebularine in normal osteoblast cells, but was dramatically enhanced in osteosarcoma cells after treatment with zebularine (P < 0.01; Fig. 4a,b). Similar results were obtained with the effect of zebularine on ARHI mRNA expression. Zebularine markedly increased expression of ARHI mRNA in osteosarcoma cells (P < 0.01), but not in normal osteoblast cells (Fig. 4a,c). To further identify the effect of zebularine on ARHI methylation, COBRA was used to analyze the methylation status of CpG islands associated with the ARHI gene after treatment with zebularine. In U2OS and MG63 cells, zebularine suppressed the methylation of ARHI gene in a dose‐dependent manner (Fig. S1). Moreover, as shown in Figure 4(d), the methylation status of CpG I was reduced by treatment with 200 μM zebularine in osteosarcoma cells, specifically in HOS and MG‐63 cell lines. The methylation status of CpG II was also downregulated by treatment with zebularine, specifically in U2OS and Saos‐2 cell lines. Moreover, zebularine triggered weak demethylation of CpG III in all osteosarcoma cell lines.

Figure 4

Effect of zebularine on expression and methylation. U2OS osteosarcoma cells were treated with nothing or 200 μM zebularine for 48 h. (a) Analysis of the effect of zebularine on expression in protein and mRNA levels in both normal and U2OS cells by Western blot and RT‐PCR. (b,c) Semiquantitative histograms of Western blot (b) and RT‐PCR (c) analyses. (d) Combined bisulfite restriction analysis was used to analyze the effect of zebularine on the methylation status of CpG islands associated with the gene in osteosarcoma cells. Values are presented as mean ± SD of four independent experiments. **P < 0.01.

Zebularine prevents interaction DNMT1 with G9a to inhibit ARHI methylation

Zebularine is a DNMT inhibitor, and there was direct cooperation between DNMT1 and G9a.10 Thus, we intended to estimate the effect of zebularine on interaction between DNMT1 and G9a. As shown in Figure 5(a,b), zebularine markedly downregulated DNMT1 expression (P < 0.01), but not G9a. We undertook co‐immunoprecipitation of DNMT1 and G9a using anti‐DNMT1 or anti‐G9a antibodies. Compared with the untreated group, zebularine treatment sharply reduced the binding of DNMT1 and G9a, and the results were consistent when the G9a antibody was used in the immunoprecipitation assay (Fig. 5c). G9a is a novel lysine‐preferring mammalian histone methyltransferase with hyperactivity and specific selectivity to lysines 9 and 27 of histone H3.11 Because the modification of histone methylation frequently has an intimate connection with DNA methylation, we intended to evaluate the effect of silenced DNMT1 or G9a on ARHI methylation. DNMT1 siRNA or G9a siRNA were transfected into U2OS cells using Lipofectamine 2000; the ARHI methylation status was detected by COBRA. Results showed that DNMT1 silencing obviously suppressed methylation of CpG I and CpG II of the ARHI gene from partial methylation to hypomethylation, and also significantly inhibited the methylation of CpG III from hypermethylation to partial methylation (Figs. 5d,S2). Similar results were observed when G9a was silenced; CpG II of the ARHI gene was markedly demethylated, and CpG I and CpG III were also feebly demethylated, in U2OS cells (Figs. 5d,S2). These results indicated that ARHI methylation is not only related to DNA methylation, but also related to histone methylation.

Figure 5

Effect of zebularine on interaction between DNA methyltransferase 1 (DNMT1) and histone methyltransferase G9a in U2OS osteosarcoma cells. (a) U2OS cells were treated with nothing or 200 μM zebularine for 48 h. Zebularine treatment increased the expression of DNMT1 and G9a in U2OS cells measured by Western blot analysis. (b) Semiquantitative histograms of Western blot analysis. Values are presented as mean ± SD of four independent experiments. **P < 0.01. (c) Interaction of DNMT1 and G9a was tested by co‐immunoprecipitation in U2OS cells. The cells were successfully transfected with DNMT1 siRNA or G9a siRNA using Lipofectamine 2000. IB, immunoblotting; IP, immunoprecipitation; T, treated; Un, untreated. (d) Combined bisulfite restriction analysis was used to analyze the methylation status of CpG islands associated with the gene in U2OS cells (d). +, Gene is present (control); –, gene is absent (silenced with siRNAs).

Silencing of ARHI rescues cell viability and apoptosis by zebularine treatment

To further validate that zebularine suppressed cell viability and increased cell apoptosis through upregulation of ARHI, negative control siRNA or ARHI siRNA was transfected into U2OS cells with or without zebularine treatment. The CCK‐8 analysis showed that knockdown of ARHI rescues the decrease of cell viability (Fig. 6a) and resisted the increase of cell apoptosis (Fig. 6b) under zebularine treatment.

Figure 6

Knockdown of rescues cell viability and apoptosis under zebularine (Zeb) treatment. Negative control siRNA (Ctrl‐siRNA) or siRNA was transfected into U2OS osteosarcoma cells with or without zebularine treatment. After incubation for 48 h, cell viability and apoptosis were tested. (a) Knockdown of rescued the decrease in cell viability under zebularine treatment. (b) Knockdown of resisted the increase in cell apoptosis under zebularine treatment. *P < 0.05, **P < 0.01 versus Ctrl‐siRNA; #P < 0.05 versus both Ctrl‐siRNA and Zeb + Ctrl‐siRNA. n.s., Not significant.

Discussion

Drug resistance almost invariably occurs, limiting the treatment effectiveness for osteosarcoma. Chemotherapy shrinks the tumor mass, but may also exert a selective pressure on tumor cells leading to the outgrowth of the fittest surviving clones. Therefore, it is necessary to develop novel drugs with high efficiency to improve the survival of osteosarcoma patients. Our findings indicated that zebularine not only suppressed the viability of osteosarcoma cells in a time‐ and dose‐dependent manner, but also potentiated the dramatic apoptosis of osteosarcoma cells. Moreover, zebularine inhibited ARHI methylation by preventing the interaction of DNMT1 with histone methyltransferase G9a, leading to improvements in ARHI protein and mRNA. As a methylation transferase inhibitor, zebularine has been proven to downregulate the expression level of DNMT1.12, 13 During this process, zebularine was metabolized to deoxyadenosine triphosphate zebularine (dtZeb) in vivo. The dtZeb and guanine formed base pairs, reducing the required energy of G>C substitutions, so that the DNA and DNMT1 bound more easily. The dtZeb replaced cytosine to bind with DNMT1 to form a dense covalent complex, and DNMT1 was locked and degraded.14 Therefore, zebularine should be considered a promising anticancer drug for epigenetic therapy of osteosarcoma.

As a maternally imprinted tumor suppressor gene, ARHI maps to a location on chromosome 1p31 and encodes a small GTPase with 60% homology to Ras and Rap.7, 15 Despite its homology to Ras, ARHI, when re‐expressed, induces apoptosis in ovarian and breast cancer cells by activating the calpain‐dependent pathway, but not the caspase‐dependent pathway,16 and inhibits tumor angiogenesis in hepatocellular carcinoma by downregulating phosphorylation of two mTOR substrates, S6K1 and 4E‐BP1.17 Moreover, the overexpression of ARHI blocks the cell cycle at G1 phase in pancreatic cancer,18 but the ARHI protein is downregulated in 60% of ovarian cancers.8 A study reported that the methylation of CpG II alone abolished the promoter activity of the ARHI gene, whereas the methylation of CpG I could not completely abolish it.7 Hypermethylation was found in 3 of 9 breast cancer cell lines (33%) at CpG island II, and in 6 of 9 (67%) at CpG island I.7 Therefore, low expression of ARHI protein caused by hypermethylation of the ARHI gene is considered to contribute to the development and progression of malignant tumor. In the current study, we found that CpG islands of ARHI promotor region (CpG I and CpG II) in four human osteosarcoma cell lines all showed higher methylation, whereas the CpG island of the encoding region (CpG III) had a lower methylation status. Moreover, ARHI expression in the four cell lines was markedly reduced compared with normal osteoblasts. Therefore, we considered that the downregulation of ARHI expression is mainly caused by hypermethylation of the ARHI promotor region, leading to loss of the ability to suppress cell survival in osteosarcoma cells. A recent study clarified the inhibitory role of ARHI in growth and its role in promoting apoptosis in osteosarcoma cells.19 We did not repeat showing the role of ARHI in cell viability and apoptosis in this study. However, we showed that knockdown of ARHI rescues cell viability and apoptosis following zebularine treatment, supporting the proposition that ARHI could inhibit osteosarcoma growth and induce apoptosis.

Recently, the DNA methylation inhibitor zebularine has received increasing attention; studies have reported that continuous treatment with zebularine effectively sustains demethylation in human bladder cancer cells12 and prevents cell growth of gastric cancer.20 Besides being an effective inhibitor of DNA methylation, zebularine possesses many properties desirable for a therapeutic agent, such as high stability and low toxicity.4 Compared with the classic DNA demethylation agent 5‐aza‐2′‐deoxycytidine, zebularine may more effectively suppress the expression of tumor suppressors, such as RASSF1A, ARHI, and BLU, which may be due to a lack of the 4′‐amino group and a 5′‐nitrogen in the structure.6 ARHI was upregulated 2.5‐fold by zebularine in ovarian cancer.6 However, the mechanism by which zebularine regulates ARHI expression has many different viewpoints. Zebularine not only inhibits DNMT1, but also possibly causes chromatin remodeling and post‐transcriptional modification in vitro.20 We found that zebularine reduced DNMT1 in combination with histone methyltransferase G9a. G9a is a lysine‐preferring histone methyltransferase. During carcinogenesis, it usually appears in the form of a Snail/G9a/Dnmt1 triple complex in carcinoma cells. It interacts with DNMT1 and then transfers methyl groups to lysine 9 in histone H3 in vitro with hyperactivity and specific selectivity.21, 22, 23 Moreover, a recent study showed that DNMT1 could directly bind to the ARHI gene.24 Therefore, zebularine suppresses G9a and DNMT1 to regulate ARHI methylation.

In summary, we showed that zebularine inhibited the growth and promoted the apoptosis of osteosarcoma cells by suppressing G9a/DNMT1‐mediated ARHI methylation. Therefore, zebularine has the prospect of becoming an epigenetic drug in the treatment of osteosarcoma.

Disclosure Statement

The authors have no conflict of interest.

Fig. S1. Zebularine suppresses ARHI methylation in a dose‐dependent manner.

Click here for additional data file.

Fig. S2. Silencing of DNA methyltransferase 1 (DNMT1) or histone methyltransferase G9a suppresses ARHI methylation.

Click here for additional data file.