Antitumor effects of histone deacetylase inhibitor on Ewing's family tumors
Article first published online: 22 APR 2005
Copyright © 2005 Wiley-Liss, Inc.
International Journal of Cancer
Volume 116, Issue 5, pages 784–792, 20 September 2005
How to Cite
Sakimura, R., Tanaka, K., Nakatani, F., Matsunobu, T., Li, X., Hanada, M., Okada, T., Nakamura, T., Matsumoto, Y. and Iwamoto, Y. (2005), Antitumor effects of histone deacetylase inhibitor on Ewing's family tumors. Int. J. Cancer, 116: 784–792. doi: 10.1002/ijc.21069
- Issue published online: 20 JUL 2005
- Article first published online: 22 APR 2005
- Manuscript Accepted: 4 FEB 2005
- Manuscript Received: 6 OCT 2004
- Scientific Research, Japan Society for the Promotion of Science. Grant Numbers: 14207057, 15591586
- Cancer Research on Priority Areas, Ministry of Education, Science, Sports and Culture, Japan. Grant Number: 16023248
- Japan Society for the Promotion of Science Fujita Memorial Fund for Medical Research
- HDAC inhibitor;
A chimeric protein, EWS-Fli1, identified in most Ewing's family tumors (EFTs) has been shown to be associated with the tumorigenicity of EFTs. We have previously reported that p21Waf1/Cip1 expression was inhibited by EWS-Fli1 in EFTs. Histone deacetylase inhibitors (HDACIs) are known to up-regulate p21Waf1/Cip1 expression in various cells and show promise as a cancer therapy. Here, we demonstrate the possible involvement of EWS-Fli1 in the activities of both histone acetylation and deacetylation, as well as the potential use of HDACIs as an antitumor agent for EFTs. A novel HDACI, FK228, strongly induced p21Waf1/Cip1 expression, leading to the hypophosphorylation of retinoblastoma protein (Rb) in EFT cells. Results indicated that EWS-Fli1 deregulated histone acetylation through both the repression of histone acetyltransferase (HAT) and the enhancement of histone deacetylase (HDAC) activities in EFT cells. FK228 treatment blocked both of the abnormal functions of EWS-Fli1. Expressions of EWS-Fli1 protein and mRNA were also inhibited by HDACIs. We suggest that HDACIs might inhibit the expression of EWS-Fli1 via the suppression of the EWS promoter activity. FK228 demonstrated potent growth inhibitory effects on EFT cells at nanomolar concentrations, as well as an apparent distinction in the apoptotic effects between EFT and normal cells. Moreover, intraperitoneal administration of FK228 significantly inhibited tumor growth and induced apoptosis in EFTs in vivo. These results suggest that HDACI might be a promising reagent for use in molecular-based chemotherapy against EFTs. © 2005 Wiley-Liss, Inc.
Ewing's family tumors (EFTs), including Ewing's sarcoma and primitive neuroectodermal tumor, are bone and soft-tissue tumors common in children and adolescents. Although multimodal therapies utilizing chemotherapy, radiation, and surgery have been developed, treatment for patients presenting with metastases and/or early relapse has been rather unsuccessful to date. In the majority of EFTs, the chromosomal translocation t (11; 22) (q24: q12) results in the expression of a fusion protein, EWS-Fli1. Molecular analyses have shown that this rearrangement fuses the N-terminus of the EWS gene to the C-terminus of the Fli1 gene which contains an ets-like DNA-binding domain.1 It has also been shown that EWS-Fli1 has the biochemical characteristics of an aberrant transcription factor, and that EWS-Fli1 is associated with tumorigenecity.2 Consequently, an understanding of the mechanisms behind how EWS-Fli1 possesses transforming activity might lead to the identification of novel therapeutic targets and agents that will help toward more effective treatment.
We have previously reported that the inhibition of the expression of EWS-Fli1 by treatment with antisense oligonucleotides resulted in the growth arrest of various EFT cell lines both in vitro and in vivo.3 We have also demonstrated that the antisense-treated cells showed induced-expression of p21Waf1/Cip1, a cyclin dependent kinase (CDK) inhibitor, at both the mRNA and protein levels,4 whereas little p21Waf1/Cip1 expression was observed in the untreated EFT cells. Moreover, we have found that EWS-Fli1 directly downregulated the expression of p21Waf1/Cip1 by inhibiting the activity of its promoter.5 p21Waf1/Cip1 regulates cell-cycle progression and plays an important role in the control of cell senescence, apoptosis and carcinogenesis. The direct downregulation of p21Waf1/Cip1 by EWS-Fli1 might cause, at least in part, the uncontrolled proliferation of EFT cells. Furthermore, as a consequence of exploring the mechanism of p21Waf1/Cip1 downregulation by EWS-Fli1, we have identified that EWS-Fli1 suppressed the histone acetyltransferase (HAT) activity of p300 through interaction with p300.5 Histone acetylation plays a key role in the regulation of transcription by modulating the structure of chromatin, and it is precisely regulated by the balance between the activities of HAT and histone deacetylase (HDAC).6 The deacetylation of nucleosomal core histone tails by the HDACs leads to a chromatin conformation that inhibits the transcription, whereas histone acetylation results in transcriptional activation. Deregulation of HAT and/or HDAC has been shown to cause abnormal modulation of the transcriptional activity of target genes in several neoplasms.7, 8 In leukemia, for instance, different chromosomal translocation-generated fusion gene products, such as PML/RARα, PLZF/RARα and AML1/ETO, induce abnormal regulation of both HAT and HDAC activities, which plays a crucial role in the pathogenesis and maintenance of the transformed phenotype in leukemia cells.9, 10, 11 However, in EFTs, it is still unclear whether or not EWS-Fli1 modulates the activity of HDAC. The fact that EWS-Fli1 inhibits p21Waf1/Cip1 induction via the suppression of the HAT activity of p300 led us to explore the possible involvement of EWS-Fli1 in the activity of HDAC and the potential use of HDAC inhibitors (HDACIs) as antitumor agents for EFTs.
Recently, HDACIs have represented a novel class of antineoplastic agents and several HDACIs are currently under clinical trial. A number of studies have demonstrated that HDACIs cause a variety of phenotypic changes, such as cell-cycle arrest, morphological reversion of transformed cells, differentiation and apoptosis.12, 13, 14 It has been estimated that approximately 2–8% of genes are regulated by the induction of histone hyperacetylation in cancer cells.15, 16 Of these, the induced expression of p21Waf1/Cip1 has been found consistently in a number of cell lines studied. In this study, we report the antitumor activity of a novel HDACI, FK228, on EFTs both in vitro and in vivo. Treatment with FK228 induced p21Waf1/Cip1 expression in EFT cells. We demonstrate that EWS-Fli1 abnormally modulated not only HAT, but also HDAC activities, and that FK228 effectively suppressed the abnormal functions of EWS-Fli1 associated with both HAT and HDAC activities. The data also show that FK228 inhibited the expression of EWS-Fli1 both at the protein and mRNA levels. FK228 showed selective antiproliferation efficacy in EFTs due to the induction of apoptosis both in vitro and in vivo. These results indicate that HDACIs might be a promising novel tool for use in molecular-based chemotherapy against EFTs.
Material and methods
Cells and reagents
Human EFT cell lines, SK-N-MC, PN-KT-1, RDES and SK-ES-1, in addition to NIH3T3, a mouse fibroblast cell line, were cultured as described previously.4 WE68 and VH64, p53-positive EFT cell lines, were kindly provided by Dr. Valen (Department of Paediatrics, University of Munster, Munster, Germany), and Balb/c3T3 cells, a mouse fibroblastic cell line, were obtained from Dr. Kuwano (Department of Biochemistry, Kyushu University, Fukuoka City, Japan). Swiss3T3 (mouse embryo fibroblasts) were purchased from Clontech (Palo Alto, CA). Human primary fibroblasts were isolated from healthy human donors after receiving their informed consent. FK228 was generously provided by Fujisawa Pharmaceutical (Osaka, Japan). FK228 was dissolved in 100% ethanol at a concentration of 1 mg/ml and stored at –20°C.
Swiss3T3 and Balb/c3T3 cells were seeded onto 15-cm dishes (2 × 106 cells/dish). At 24 hr after seeding, transfection was done using FuGENE 6 (Roche Diagnostics, Indianapolis, IN) according to the manufacturer's protocol. Briefly, each dish was transfected with 8.0 μg of pCMV4-Flag-EWS-Fli1 or pCMV4 using 24 μl of FuGENE 6. At 48 hr after the transfection, the cells were subjected to nuclear extraction.
Nuclear extracts were prepared using the method of Dignam et al.17 with slight modification. In brief, 1 × 107 cells were harvested and washed in PBS, suspended in 200 μl of cold buffer A (10 mM HEPES, pH7.9, 50 mM NaCl, 1 mM dithiothreitol, 0.1 mM EDTA and 0.1 mM phenylmethylsulfonyl fluoride), and then incubated for 20 min on ice. The cells were added to cold buffer B (buffer A with 0.1% Nonidet P-40, Nakalai Tesque, Kyoto, Japan) and were gently pipetted and incubated on ice for another 20 min. The nuclei were pelleted (5,000g, 2 min) and washed in buffer A, and the nuclear proteins were extracted in 25 μl of buffer C (400 mM NaCl, 20 mM HEPES, pH 7.9, 1 mM EDTA, 1 mM dithiothreitol, 1 mM EGTA and 1 mM phenylmethylsulfonyl fluoride). The nuclear proteins were placed on ice for 30 min, followed by centrifugation at 4°C for 10 min. The supernatant was recovered, snap-frozen in liquid nitrogen and stored at –80°C until use.
RNA extraction and quantitative real-time RT-PCR (TaqMan PCR)
Total RNA was isolated from SK-N-MC cells treated by FK228 using an RNA-easy kit (Qiagen, Hilden, Germany). Of the total RNA, 1 μm was subjected to RT reaction using SuperScriptII reverse transcriptase (Invitrogen, Carlsbad, CA). Real-time PCR was performed as described by Heid et al.18 The amplification primers and fluorogenic hybridization probes of human p21Waf1/Cip1 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were designed as described previously.5 The primers used for the detection of EWS-Fli1 transcripts were the forward primer, 5′-GGCAGCAGCCTCCCACTAG-3′ and the reverse primer, 5′-CCATGCTCCTCTTCTGACTGAGT-3′. The sequence of the TaqMan probe (Applied Biosystems, Foster City, CA) used to quantify the RT-PCR products of EWS-Fli1 was 5′-(Fam) CCACCCCAAACTGGATCCTACAGCC (TAMRA)-3′. Quantitative real-time RT-PCR was carried out using a PE Applied Biosystems 7700 Sequence Detector. The relative p21Waf1/Cip1 and EWS-Fli1 expression data were calculated by dividing the concentrations of p21Waf1/Cip1 and EWS-Fli1 by those of GAPDH, respectively.
Western blot analyses were carried out as described previously.4 In brief, SK-N-MC cells treated with FK228 were solubilized. The samples were run on a 4–12% gradient precast MOPS-polyacrylamide gel (Novex, San Diego, CA) and blotted onto a nitrocellulose filter. The filter was incubated with antibody against human p21Waf1/Cip1, total Rb, actin (BD Pharmingen, San Diego, CA), phosphorylated Rb (Cell Signaling Technology, Beverly, MA), Fli1, p300, CBP, PCAF and TAFII p250 (Santa Cruz Biotechnology, Santa Cruz, CA), acetylated histone H3 and H4 (Upstate Biotechnology, Lake Placid, NY) or cleaved poly (ADP-ribose) polymerase PARP (Promega, Madison, WI). Mouse as well as human actin and PARP were detected by using anti-actin (BD Pharmingen) and anti-PARP (Roche Diagnostics) antibodies, respectively. After several washes, the horseradish peroxidase-conjugated secondary antibody (BioSource International, Inc., Chicago, IL) was added. The immunoreactivity of the blots was detected using an enhanced chemiluminescence system (Amersham Biosciences, Uppsala, Sweden).
Immunoprecipitation and immunoblotting
Nuclear extracts from SK-N-MC cells treated with or without 9.25 nM of FK228 for 24 hr were prepared as described above. After the final centrifugation, the salt concentration was reduced to 150 mM of NaCl. The nuclear extract was subjected to immunoprecipitation using 5 μl of rabbit polyclonal antibody against acetyl-Lysine (Upstate Biotechnology) or 4 μg of normal rabbit polyclonal immunoglobulin G (IgG; Santa Cruz Biotechnology). After incubation on ice for 2 hr, protein A-G-Sepharose beads (Santa Cruz Biotechnology) were added to the reactions, and were then shaken on a rotary shaker at 4°C for 8 hr. The beads were washed 5 times with PBS, boiled and subjected to electrophoresis. The procedures of Western blot were as described above. The antibody used was mouse monoclonal antibody against PCAF (Santa Cruz Biotechnology) at 1:100 dilution.
Cell growth analysis
Various cell lines were seeded at a density of 3 × 104 cells in 35-mm culture dishes (Falcon Labware, Franklin Lakes, NJ) in triplicate. At 24 hr after the cell preparation, various concentrations of FK228 were added to the media. After various time intervals (0–72 hr) cell numbers were determined using a Coulter Hematology Analyzer (Beckman Coulter, Fullerton, CA).
Trypan blue dye exclusion assay
Cells were plated at a density of 2 × 104 cells/ml in 12-well plates. Various concentrations of FK228 were added to the cells 48 hr after the cell seeding. Then, 72 hr after the treatment with FK228, the cells trypsinized and floating in the culture medium were collected and centrifuged. The viable cells were determined by means of the trypan blue dye-exclusion test and counted in a hemocytometer.
Cells treated with FK228 were harvested and fixed with 70% ethanol for 30 min at 4°C. Then, the samples were centrifuged and resuspended in PBTB (PBS containing 0.1% Tween 20 and 0.05% BSA) with 10 μg/ml RNase A and 50 μg/ml propidium iodide. Alterations in the cell-cycle distribution were analyzed using Epics-XL, a flow cytometer (Beckman Coulter). The cell proportions were analyzed using EXPO32 Software (Beckman Coulter). For each sample, 10,000 events were stored.
The siRNA duplexes used in this study were custom-synthesized by Dharmacon Research (Lafayette, CO). We designed 21 nt of siRNAs corresponding to the sequence of the breakpoint of EWS-Fli1 type I (siBPEFI) (Matsunobu et al., unpublished results). The scrambled siRNA (siScr) used as a control was commercially obtained from Dharmacon Research. Transfection of siRNA was carried out using Oligofectamine (Invitrogen) according to the manufacturer's protocol.
Nonradioactive histone acetyltransferase activity assay(HAT assay)
WE68 cells were treated with siBPEFI or siScr for 48 hr. FK228 treatment of the cells was carried out for 24 hr. Each nuclear extract was harvested and subjected to the nonradioactive HAT assays using a HAT assay kit (Upstate Biotechnology), according to the manufacturer's protocol. Briefly, each nuclear extract was mixed with 100 μM acetyl-coenzyme A (CoA) and 1× HAT assay buffer, and was incubated on an enzyme-linked immunosorbent assay plate precoated with histone H3 for 30 min. After several washes with PBS, acetylated histones were detected using an anti-acetyl-lysine rabbit polyclonal antibody followed by the horseradish peroxidase-based colorimetric assay.
Histone deacetylase (HDAC) assay
Histone deacetylase activity was measured using a Histone Deacetylase Assay Kit (Upstate Biotechnology), according to the manufacturer's protocol. Briefly, biotinylated histone H4 peptide corresponding to the amino acids 2–24 of histone H4 was acetylated in vitro by incubation for 4 hr with 12.5 μCi acetyl-CoA, HAT assay buffer and active PCAF recombinant GST fusion protein expressed in E. coli, corresponding to the amino acids 352–832. Then, the streptavidin agarose slurry was added to capture the labeled peptide, and the mixture was incubated for 15 min. The beads were washed with TBS and resuspended in a 50% slurry using TBS. For the deacetylase assays, 10,000 cpm of the acetylated peptide were incubated with 50 μg of nuclear extracts and HDAC assay buffer at room temperature for 20 hr. The reaction was stopped by the addition of acetic acid (0.04 M, final concentration) and HCl (250 mM, final concentration). The beads were pelleted by centrifugation and the supernatant from each sample was transferred to a separate scintillation fluid. The released acetate was quantified by scintillation counting. Differences among experiment groups were assessed by the unpaired Student's t-test.
The 1 μm of total RNA extracted from SK-N-MC cells treated with various concentrations of FK228 was subjected to RT reaction. The sequences of the primers were described previously.4, 19 PCRs were performed in a final volume of 50 μl for 27 cycles. Each PCR cycle consisted of a heat denaturation step at 94°C for 30 sec, a primer annealing step at 64°C for 1 min and an extension step at 72°C for 1 min. The PCR was performed within the linear range of amplification determined in a preliminary study. The PCR products were analyzed in 1.5% agarose gel (Sigma, St. Louis, MO).
In vivo antitumor activity of FK228
SK-N-MC cells were harvested from the monolayer cultures at 60% confluence, resuspended in DMEM at 4 × 107 viable cells/ml, and subcutaneously inoculated into athymic mice (1 × 107 viable cells/mouse). When the subcutaneous tumor had grown to a visible size, FK228 suspended in PBS was administered intraperitoneally at a total dose of 0, 1, 3 or 5 mg/kg in 3 evenly divided doses, 7, 11 and 15 days after the tumor inoculation. Control groups of mice were injected with PBS only. A total of 6 mice were used in each group and followed every 3–4 days by measuring the body mass. At 21 days after the tumor implantation, the length (A mm) and width (B mm) of each tumor were measured according to the method described by Ueda et al.20 The multivariate analysis was performed by ANOVA. Treatment-related toxicity was evaluated by the serial weight measurements and gross anatomy. All the animal experiments were done in accordance with the Institutional Guidelines for Animal Experiments of Kyushu University.
TdT-mediated dUTP-biotin nick end-labeling (TUNEL) staining
Tumors were fixed with 10% (v/v) formalin in 0.9% NaCl solution prior to paraffin embedding and routine sectioning. TUNEL staining was performed using an In Situ Apoptosis Detection Kit (Takara, Tokyo, Japan) according to the manufacturer's recommendations. In brief, after rehydration, the slides were incubated with proteinase K (20 μg/ml) at room temperature for 15 min. Endogenous peroxidase was inactivated by treatment with 3% hydrogen peroxide. Tissue sections were then subjected to a TUNEL reaction. The sections were then covered with anti-FITC HRP conjugate antibody for 30 min at 37°C, stained with 0.05% diaminobenzidine hydrochloride to detect the labeled nuclei, and counterstained with 3% methyl green.
Induced p21Waf1/Cip1 expression and reduced phosphorylation of Rb protein by FK228 in EFT cells
We have reported that EWS-Fli1 fusion protein directly inhibits p21Waf1/Cip1 gene expression in EFT cells.5 Recently, HDACIs, including FK228, were found to upregulate p21Waf1/Cip1 expression in various cancer cells.21 Herein, we investigated the effect of FK228 on the expression of p21Waf1/Cip1 in SK-N-MC cells. TaqMan PCR analysis showed that p21Waf1/Cip1 expression was dose-dependently induced by treatment with doses of 2.31–18.5 nM of FK228, whereas untreated cells expressed only a small amount of p21Waf1/Cip1 mRNA (Fig. 1a). In the time-course experiment, p21Waf1/Cip1 mRNA expression was upregulated 3 hr after treatment with FK228 (Fig. 1b). Western blot analysis showed that the expression of p21Waf1/Cip1 protein was also induced by FK228 in parallel with mRNA expression (Fig. 1a and b). We further examined Rb protein phosphorylation, which correlates with p21Waf1/Cip1 expression in SK-N-MC cells. Rb is hypophosphorylated indirectly by p21Waf1/Cip1, via the inhibition of CDK activity, resulting in the inactivation of the E2F transcription factor, which in turn controls the transcription of genes required for G1-S transition.22 Western blot analysis demonstrated that the level of Rb phosphorylationdose-dependently decreased anti-parallel with the increase in p21Waf1/Cip1 expression following FK228 treatment. On the other hand, total Rb protein level was not affected (Fig. 1c).
Effect of FK228 on HAT activity in EFT cells
Previously, we have shown that EWS-Fli1 negatively regulates p21Waf1/Cip1 expression by inhibiting p300-mediated HAT activity in EFT cells.5 In that study, the downregulation of EWS-Fli1 by the antisense oligonucleotides led to the enhancement of overall HAT activity in SK-N-MC cells. Sodium butylate also upregulated the HAT activity in SK-N-MC cells, meaning that it inhibited the abnormal function of EWS-Fli1 in EFT cells.5 To confirm whether the increased HAT activity was cell-type-specific and mediated by the inhibition of EWS-Fli1 expression, we assessed the HAT activity in WE68 cells expressing EWS-Fli1 type I by treatment with RNA interference to knock down EWS-Fli1. We designed 21 nt of siRNAs corresponding to the sequence of the breakpoint of EWS-Fli1 type I (siBPEFI) (Matsunobu et al., unpublished results). In agreement with our previous study, the treatment with siBPEFI for 48 hr inhibited the expression of EWS-Fli1 and significantly enhanced the HAT activity in WE68 cells, whereas the treatment with scrambled siRNA (siScr) did not alter either EWS-Fli1 expression or HAT activity (Fig. 2a). We next performed HAT assay using nuclear extracts of WE68 cells treated with FK228. As shown in Figure 2b, HAT activity in the FK228-treated cells was dose-dependently increased to up to 30-fold that in nontreated cells. To determine the mechanisms of the enhanced HAT activity in the FK228-treated cells, we investigated the expressions of HAT proteins. As shown in Figure 2c, the expression levels of HAT proteins, p300, CBP, PCAF, and TAFII p250 were not significantly altered by the treatment with FK228 on EFT cells. A recent study has shown that PCAF, which is a target of p300 acetylation, is autoacetylated in vivo, leading to an increment in its HAT activity.23 Therefore, we examined the possibility whether PCAF might be acetylated in EFT cells treated with FK228. To do that, we analyzed the acetylation status of the endogenous PCAF from FK228-treated or -nontreated SK-N-MC cells. Acetylated proteins were immunoprecipitated using the anti-acetyl-Lysine antibody and probed with the PCAF antibody. The results indicated that the endogenous PCAF was clearly acetylated in SK-N-MC cells treated with FK228, compared with that in nontreated cells (Fig. 2d). These data demonstrate that HDACIs might induce HAT enzymatic activity in EFT cells, due at least in part to the acetylation of endogenous PCAF.
HDAC activity enhanced by EWS-Fli1 and repressed by HDACIs
Dysfunctional events in histone acetylation and deacetylation play a role in the progression of cancers. However, it is still unclear whether or not EWS-Fli1 modulates the activity of HDAC. We herein investigated the potential involvement of EWS-Fli1 in histone deacetylation. The HDAC activity was examined using a radioactive assay that could measure the incorporation of a radiolabeled Histone H4 peptide. Overexpressed EWS-Fli1 significantly increased the activity of histone deacetylation in murine fibroblasts. HDAC activity in Swiss3T3 and Balb/c3T3 cells transfected with EWS-Fli1 exhibited 4.7-fold and 2.0-fold enhancement compared with that of the control, respectively (Fig. 3a). In order to confirm whether EWS-Fli1 might enhance HDAC activity in EFT cells, we used RNA interference to knock down EWS-Fli1 in WE68 and SK-N-MC cells, both of which express EWS-Fli1 type I fusion protein. Treatment with siBPEFI for 48 hr inhibited the expression of EWS-Fli1 and caused significant suppression of HDAC activity in both the cell lines used (Fig. 3b). We next examined HDAC activity in SK-N-MC cells treated with FK228. FK228 dose-dependently decreased HDAC activity in SK-N-MC cells (Fig. 3c). These data suggest that HDACIs inhibit the HDAC activity enhanced by EWS-Fli1 in EFTs. We also assessed histone H3 and H4 acetylation in order to determine the summation of HDAC and HAT activity. Western blot analysis showed that both histone H3 and H4 were strongly acetylated with FK228 treatment in SK-N-MC cells (Fig. 3d). Taken together, these findings seem to suggest that EWS-Fli1 might deregulate histone acetylation via the repression of HAT and the enhancement of HDAC activities. HDACIs inhibit the abnormal function of EWS-Fli1 by the induction of HAT activity and the suppression of HDAC activity in EFT cells.
Effect of HDACIs on the expression of EWS-Fli1 in EFT cells
Despite the fact that so few genes are modulated by the treatment with HDACIs, cluster analysis of gene expression patterns has demonstrated that the dependency of genes regulated by HDACIs varies among cell lines.24 Our previous studies have shown that the suppression of EWS-Fli1 expression by the antisense oligonucleotides is an effective way to inhibit the proliferation of EFT cells.3, 4, 5 To examine the expression of EWS-Fli1 in EFT cells treated with HDACIs, we performed Western blot and TaqMan PCR analyses. A dose-dependent decrease in EWS-Fli1 expression was observed in SK-N-MC cells treated with FK228 for 24 hr both at the protein and mRNA levels (Fig. 4a and b). Consistently, each Western blot analysis of 6 EFT cell lines, SK-N-MC, WE68, VH64, PN-KT-1, RDES and SK-ES-1, showed the suppressed expression of EWS-Fli1 following treatment with FK228 (data not shown). Other HDACIs, sodium phenylbutylate, trichostatin A and SAHA also decreased EWS-Fli1 protein expression in all of the 6 EFT cell lines (data not shown). In the time-course study, 10 ng/ml of FK228 inhibited the expression of EWS-Fli1 mRNA within 3 hr after the treatment, and the expression levels reached a plateau at around 24 hr (Fig. 4c). Since the EWS-Fli1 transcript is under the control of the EWS promoter in chromosome 22, HDACIs might decrease the intrinsic EWS mRNA transcription. Therefore, we also assessed changes in the normal EWS mRNA expression by HDACIs using RT-PCR. In accordance with the suppression of EWS-Fli1 expression, RT-PCR analyses using EWS-specific primers showed that FK228 decreased the intrinsic EWS mRNA in SK-N-MC cells in both dose- and time-dependent manners (Fig. 4d and e). The suppressed expression of normal EWS mRNA by FK228 treatment was detected not only in EFT cell lines, but also in Hela and Saos2 cells (data not shown). These results suggest that HDACIs might inhibit the activity of the EWS promoter.
Growth inhibitory effect of FK228 on EFT cells
Since HDACIs also inhibit the expressions of EWS-Fli1 in EFT cells, the results would underscore the potential of HDACIs for the treatment of EFTs. Then, the effects of HDACIs on the growth of EFT cells were examined. As shown in Figure 5a, the growth inhibitory effects of FK228 on 3 EFT cell lines, SK-N-MC, PNKT1 and VH64 were very potent. The growth of EFT cells were completely inhibited by the treatment with 4.62 nM of FK228 for 48 hr. On the other hand, the effect of FK228 on Balb/c3T3, a mouse fibroblast cell line, was weaker than that on EFT cells. However, FK228 inhibited the proliferation of Balb/c3T3 within the range of 2.31–9.25 nM (Fig. 5a). Then, to investigate the viability of the cells, we performed trypan blue dye exclusion assay. We used 4 EFT cell lines. Treatment with 18.5 nM of FK228 for 72 hr increased the number of nonviable cells to more than 70% of total cells. On the other hand, most of the normal cells, mouse fibroblastic cell lines and human normal primary fibroblasts, were still observed to be viable after the treatment with the same concentration of FK228 (Fig. 5b).
Selective induction of apoptosis by FK228 on EFT cells
To define alterations in the cell-cycle distribution by FK228 treatment in EFT cells, we performed flow cytometry analysis. We tested 6 EFT cell lines. Among them, SK-N-MC, PN-KT-1, RDES and SK-ES-1 have nonfunctional p53, whereas WE68 and VH64 are p53 wild-type. The treatment with FK228 at the concentrations of 1.85–185 nM for 48 hr caused the accumulation of populations within the sub-G1 fraction, indicating apoptosis in the EFT cells. p53 wild-type cell lines, as represented by WE68, showed induced apoptosis by FK228 via a predominance of the G1 arrest. On the other hand, p53-defective cell lines, such as the typical SK-N-MC, were led to apoptosis by FK228, without indicating either G1 or G2 cell-cycle arrest (Fig. 6a and b). FK228 had weak cytotoxic activities in response to normal cells. Mouse fibroblast cell lines, NIH3T3, Balb/c3T3 and Swiss3T3, and human primary fibroblasts were also examined by FACS analysis. The normal cells, represented by NIH3T3, were exclusively arrested in the G1 phase of the cell-cycle by FK228 treatment (Fig. 6a). As shown in the right panel of Figure 6b, few normal cells underwent apoptosis following FK228 treatment, even at the dose of 185 nM. We next examined the effect of FK228 on the proteolytic cleavage of the nuclear protein PARP as a marker of apoptosis in EFT cells. The 116-kDa PARP is specifically cleaved to produce an 85-kDa fragment during apoptosis.25 Western blot analysis using the antibody against the 85-kDa fragment of cleaved PARP demonstrated the dose-dependent induction of the cleaved PARP in SK-N-MC cells treated with FK228 (Fig. 6c). On the other hand, no cleavage of the PARP was observed following FK228 treatment in NIH3T3 cells (Fig. 6d). FK228 showed apparent selectivity between EFT and normal cells.
In vivo antitumor effects of FK228 on EFTs
The antitumor activity of FK228 was also examined in subcutaneous tumors of SK-N-MC developed in nude mice. At 7 days after cell inoculation, when the subcutaneous tumor had grown to a visible size, FK228 suspended in PBS was administered intraperitoneally at a total dose of 1, 3 or 5 mg/kg. At 13 days after the first treatment with FK228, the weight of each tumor was measured. As shown in Figure 7a, FK228 significantly inhibited tumor growth in a dose-dependent manner (p < 0.05). The toxicity of FK228 was assessed by monitoring body weight and survival of mice in each group. Intraperitoneal administration of FK228 did not cause any measurable toxicity in doses of up to 5 mg/kg (Fig. 7b). To study whether FK228 induces apoptosis in vivo, we examined TUNEL staining on the resected tumor sections. Although most of the tumor cells were TUNEL-negative in the untreated controls, approximately 40% of the EFT cells in the tumors treated with 5 mg/kg of FK228 were TUNEL-positive at day 13 (Fig. 7b). These results suggest that FK228 effectively killed the EFT cells due to the induction of apoptosis in vivo.
The development of therapeutic strategies that exploit tumor-specific biological mechanisms is at the frontier of cancer medicine. For EFTs, an understanding of the mechanisms through which EWS-Fli1 possesses transforming activity is so important that it might lead to the identification of novel therapeutic targets and agents. In this study, we demonstrated the possible involvement of EWS-Fli1 in the activity of HDAC, as well as the potential use of HDACIs as antitumor agents for EFTs. Our hypothesis was primarily based on our finding that EWS-Fli1 inhibited p21Waf1/Cip1 expression via the suppression of the HAT activity of p300.5 p21Waf1/Cip1 plays a critical role in both the Rb and p53 pathways of cell-cycle control.26, 27 Since EWS-Fli1 protein directly downregulates the expression of p21Waf1/Cip1 at the promoter level,5 the agents activating the p21Waf1/Cip1 promoter might be effective for molecular-targeting therapy against EFTs. We herein demonstrated that treatment with FK228, an HDACI, induced p21Waf1/Cip1 expression and the hypophosphorylation of Rb in EFT cells, subsequently causing tumor growth inhibition.
Moreover, we showed that HDACI enhanced the HAT activity that was suppressed by EWS-Fli1. The increased HAT activity might not correlate equally with the associated decrease in HDAC activity, because the HAT assay is used to determine the specific activity of HAT enzymes and the assay does not take into account the summation of HDAC and HAT activity. Our results demonstrate that HDACIs might induce HAT enzymatic activity in EFT cells. However, the data indicated that the enhanced HAT activity might not be caused by the alternations of the expressions of HAT proteins. A report has shown that PCAF, a CBP/p300-binding protein, is acetylated by itself and by p300 in vivo, leading to an increment in its HAT activity.23 In addition, Yamagoe et al.28 have demonstrated that the HAT proteins, PCAF and GCN5, are incorporated into the HDAC complex(es) at least in part by directly binding to HDAC proteins. In this study, we revealed that the endogenous PCAF was acetylated by the treatment with HDACI in EFT cells. Therefore, our data suggest that HDACIs might promote HAT activity due to the acetylation of PCAF, at least in part.
It had also been unclear as to whether or not EWS-Fli1 is involved in the modulation of histone deacetylation. Extensive studies have revealed that HDAC proteins play pivotal roles in carcinogenesis, including multiple steps such as cell cycle, apoptosis, differentiation, angiogenesis and metastasis.29 In this study, an HDAC assay was applied to determine HDAC activity specifically, and it was clearly revealed that EWS-Fli1 enhanced HDAC activity in EFT cells. Recent reports have shown that HDAC1 is overexpressed in gastric and prostate cancer compared to normal cells and benign prostatic hyperplasia cells, respectively.30, 31 However, in our study, the downregulation of EWS-Fli1 by the treatment with siRNA did not affect HDAC1 expression in SK-N-MC or WE68 cells (data not shown). In several types of leukemia, the translocation-generated oncoprotein recruits HDAC and inappropriately represses transcription.9, 10, 11 Therefore, it seems to be more likely that EWS-Fli1 might form complexes with HDAC and act as a corepressor, leading to the enhancement of HDAC activity. Thus, our data suggest that the inhibition of HDAC activity might be an effective strategy for EFT therapy.
It is noteworthy that HDACIs inhibited not only the abnormal functions of EWS-Fli1, but also the expression of EWS-Fli1 itself at the protein and mRNA levels. Considering the fact that the expression of normal EWS mRNA was also suppressed in parallel with EWS-Fli1 by the treatment with HDACIs and that the EWS-Fli1 gene is considered to be regulated by the EWS gene promoter, the promoter of the EWS gene might be targeted by HDACIs. Multiple transcription start sites associated with an absence of the TATA sequence and a high incidence of unmethylated CpG dinucleotides are recognized features of the EWS promoter region, suggesting that EWS has the characteristics of a housekeeping gene.32 The mechanism of the transcriptional control of the EWS gene has remained unknown and the mechanism of gene repression by HDACIs treatment is not well understood. However, it might result either from the direct or indirect effects of histone acetylation or from an increase in the acetylation of proteins other than histones. Our previous studies have indicated that the downregulation of EWS-Fli1 leads to the inhibition of tumor growth in EFTs.3, 4, 5 Since the treatment with HDACIs could suppress EWS-Fli1 expression, HDACIs have another beneficial effect in the therapy for EFTs.
Our results strongly suggested that HDACIs might directly and/or indirectly inhibit the transforming activity of EWS-Fli1, at least in part. FK228 greatly inhibited the proliferation of all the 6 EFT cell lines we tested. In addition, FK228 strongly induced the apoptosis of EFT cells, whereas normal cells exhibited G0/G1 cell cycle arrest following FK228 treatment. Accordingly, FK228 appears to be a novel-type of drug that has the potential to decrease side effects in normal cells. It has been reported that in normal cells, including MCF10A,13 Swiss3T3 and NIH3T3,33 HDACIs do not cause apoptosis, but they do cause G0/G1 cell-cycle arrest via p21Waf1/Cip1 induction. The selectivity of FK228 might depend upon whether the checkpoint mechanisms of the cell cycle are complete or disrupted in the cells. Although numerous studies have been explored, the mechanisms through which HDACIs exert apoptosis in cancer cells have not been fully delineated. Further investigations are needed to determine the underlying elements by which apoptosis is induced in EFT cells through HDACI treatment.
We confirmed the antitumor effects of FK228 on EFTs, not only in vitro but also in vivo. Intraperitoneal administration of FK228 significantly inhibited the growth of EFT cells in nude mice in a dose-dependent manner. However, TUNEL staining showed that approximately 50% of the total cells were still viable, although a number of cells underwent apoptosis in vitro. Thus, the cytotoxic effect of FK228 on EFTs in nude mice is only partial. This might be due to the short plasma half-life of FK228 which leads to difficulties in achieving adequate concentrations of FK228 in the tumor. Li and Chan34 showed that the plasma concentration of FK228 declined with a mean terminal half-life of 97 min following the intravenous administration of 10 mg/kg of FK228 into rats. Since no apparent side effects were observed in mice treated with 5 mg/kg of FK228, FK228 dose-escalation may be a possible way of achieving a higher serum concentration leading to a higher rate of apoptosis induction in EFT cells. It might be more effective when an administration-by-hours infusion is applied in clinical trials of FK228.35
In conclusion, we demonstrated the abnormal functions of EWS-Fli1 associated with both histone acetylation and deacetylation, and the effectiveness of HDACIs in the therapeutic application for EFTs. Accordingly, HDACIs might be a promising novel tool for use in molecular-based chemotherapy against EFTs.
We thank Dr. F. van Valen for VH64 and WE68. We also thank Dr. Yoshihiko Maehara for the ABI PRISM 7700 Sequence Detection System. FK228 was generously provided by Fujisawa Pharmaceutical (Osaka, Japan).