Cancer Research Cancer Epigenetics Protein Translation and Cancer
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chang, H.-C.
Right arrow Articles by Hung, W.-C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chang, H.-C.
Right arrow Articles by Hung, W.-C.
[Cancer Research 66, 8413-8420, September 1, 2006]
© 2006 American Association for Cancer Research

Molecular Biology, Pathobiology, and Genetics

Silencing of the Metastasis Suppressor RECK by RAS Oncogene Is Mediated by DNA Methyltransferase 3b–Induced Promoter Methylation

Hui-Chiu Chang1,3, Chun-Yu Cho2 and Wen-Chun Hung2,3,4

1 Graduate Institute of Medicine, College of Medicine, Kaohsiung Medical University; 2 Institute of Biomedical Sciences, National Sun Yat-Sen University; 3 National Sun Yat-Sen University-Kaohsiung Medical University Joint Research Center, Kaohsiung, Taiwan and 4 Center for Gene Regulation and Signal Transduction Research, National Cheng Kung University, Tainan, Taiwan

Requests for reprints: Wen-Chun Hung, Institute of Biomedical Sciences, National Sun Yat-Sen University, No. 70, Lien-Hai Road, Kaohsiung 804, Taiwan. Phone: 886-7-5252000; Fax: 886-7-5250197; E-mail: hung1228{at}ms10.hinet.net.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
RECK is a membrane-anchored glycoprotein that may negatively regulate matrix metalloproteinase activity to suppress tumor invasion and metastasis. Our previous study indicated that oncogenic RAS inhibited RECK expression via a histone deacetylation mechanism. In this study, we address whether DNA methyltransferases (DNMT) participate in the inhibition of RECK by RAS. Induction of Ha-RASVal12 oncogene increased DNMT3b, but not DNMT1 and DNMT3a, expression in 2-12 cells. In addition, induction of DNMT3b by RAS was through the extracellular signal-regulated kinase signaling pathway. Oncogenic RAS increased the binding of DNMT3b to the promoter of RECK gene and this binding induced promoter methylation, which could be reversed by 5'-azacytidine and DNMT3b small interfering RNA (siRNA). The MEK inhibitor U0126 also reversed RAS-induced DNMT3b binding and RECK promoter methylation. Treatment of 5'-azacytidine and DNMT3b siRNA restored RECK expression in 2-12 cells and potently suppressed RAS-stimulated cell invasion. In addition, the inhibitory effect of 5'-azacytidine on RAS-induced cell invasion was attenuated after knockdown of RECK by siRNA. Interestingly, human lung cancer cells harboring constitutively activated RAS exhibited lower RECK expression and higher promoter methylation of RECK gene. 5'-Azacytidine and DNMT3b siRNA restored RECK expression in these cells and effectively suppressed invasiveness. Collectively, our results suggest that RAS oncogene induces RECK gene silencing through DNMT3b-mediated promoter methylation, and DNMT inhibitors may be useful for the treatment of RAS-induced metastasis. (Cancer Res 2006; 66(17): 8413-20)


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Mutations in the RAS proto-oncogene are frequently found in various types of human cancer. Aberrant activation of this oncogene has been implicated in many aspects of the malignant phenotypes, including proliferation, transformation, invasion, and metastasis (1, 2). Numerous studies have shown that oncogenic RAS induced the characteristics of the metastatic phenotype in transformed cells (3–5). However, the underlying mechanism is poorly characterized. Recently, several potential mediators for RAS-induced metastasis have been identified, including cytoskeleton proteins (6–8), integrin (9–12), cadherin (13–15), angiogenic factors (16, 17), and matrix metalloproteinase (MMP). The role of RAS in the regulation of MMPs is particularly interesting. MMPs are a large family of zinc-dependent endopeptidases that participate in various cellular processes. This family of proteins consists of at least 26 enzymes and can be grouped into different subtypes based on the substrate specificity and sequence characteristic (18, 19). Previous studies showed that RAS oncogene up-regulated MMP-2 and MMP-9, two MMPs that played critical roles in tumor metastasis, in transformed cells (20, 21). Therefore, MMPs are important mediators for RAS-induced metastasis.

It should be noted, however, that MMPs are synthesized as inactive precursors and must be activated by proteolytic cleavage, a process that can be blocked by MMP inhibitory proteins (22). Recently, several MMP inhibitory proteins, including tissue inhibitor of metalloproteinases and RECK, have been identified. The RECK gene was isolated as a transformation suppressor gene that induced flat reversion in v-Ki-RAS-transformed NIH3T3 cells (23). This gene encodes a membrane-anchored glycoprotein that can negatively regulate MMP-2 and MMP-9 activities and inhibit tumor angiogenesis and metastasis (24, 25). Although RECK mRNA is highly expressed in most of normal human tissues and untransformed cells, it is down-regulated or undetectable in many tumor cell lines or in cells ectopically expressed active oncogenes (24).

Our recent study showed that RAS oncogene suppressed RECK expression via inhibition of its transcription (26). Additionally, we found that oncogenic RAS inhibited RECK expression through a histone deacetylation mechanism. In this study, we further show the involvement of DNA methyltransferases (DNMT) in the inhibition of RECK by RAS oncogene.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Cell culture and experimental reagents. The 2-12 cells provided by Dr. H.S. Liu (National Cheng Kung University, Tainan, Taiwan) were established by introducing an inducible Ha-RASVal12 oncogene under the control of Escherichia coli lac operator/repressor system into NIH3T3 cells (27). H358 and CL1.0 human lung cancer cells were gifts of Dr. K. Fang (National Taiwan Normal University, Taiwan) and Dr. M.L. Kuo (National Taiwan University, Taiwan), respectively. Cells were routinely cultured in DMEM/F-12 containing 10% FCS and antibiotics. One-step reverse transcription-PCR (RT-PCR) kit was from Qiagen (Valencia, CA). Kinase inhibitor U0126 and SP600126 were from Tocris (Northpoint, United Kingdom). Anti-DNMT1, anti-DNMT3a, and anti-DNMT3b antibodies were obtained from Abcam (Cambridge, MA). Anti-RECK antibody was purchased from MBL (Nagoya, Japan). Anti-RAS antibody was obtained from BD Biosciences (San Jose, CA) and anti-actin antibody was purchased from Chemicon (Temecula, CA).

RNA extraction and RT-PCR. Total RNA was isolated from cells and RECK expression was investigated by using the one-step RT-PCR kit according to the manufacturer's protocol. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control to check the efficiency of cDNA synthesis and PCR amplification. cDNA synthesis was carried out at 50°C for 30 minutes and the condition for PCR was 30 cycles of denaturation (94°C for 1 minute), annealing (60°C for 1 minute), and extension (72°C for 1 minute) and 1 cycle of final extension (72°C for 10 minutes). The predicted sizes for PCR products for RECK and GAPDH were 477 and 512 bp, respectively. Sequences of the primers are RECK forward 5'-CCTCAGTGAGCACAGTTCAGA-3' and reverse 5'-GCAGCACACACACTGCTGTA-3' and GAPDH forward 5'-GAGTCAACGGATTTGGTCGT-3' and reverse 5'-TGTGGTCATGAGTCCTTCCA-3'. After reaction, PCR products were separated on a 2% of 0.5x Tris-borate EDTA agarose gel, stained with ethidium bromide, and visualized under UV light.

Western blot analysis. Cells were harvested and equal amount of cellular proteins was subjected to SDS-PAGE as described previously (28). Proteins were transferred to nitrocellulose membranes and the blots were probed with various primary antibodies. Enhanced chemiluminescence reagents were used to detect the proteins on the membranes.

DNA affinity precipitation assay. We used biotin-labeled DNA probe to interact with nuclear proteins and precipitated the DNA-protein complex by streptavidin-coated agarose beads. After centrifugation, the beads were collected and washed, and proteins were eluted by SDS-PAGE sample buffer. The binding proteins were separated by 7.5% or 12% polyacrylamide gels and analyzed by Western blot analysis. The sequence of DNA probe is 5'-GCGCGGGGGGCGGGGCCTGGTGCC-3' containing the Sp1 site localized at the –82/–71 bp region, originally designed as Sp1(B) in the study of Sasahara et al. (24), of mouse RECK promoter. DNA affinity precipitation assays (DAPA) were done according to the procedures described previously (26).

Chromatin immunoprecipitation assay. Cells were fixed with 1% formaldehyde at 37°C for 10 minutes. Cells were washed twice with ice-cold PBS containing protease inhibitors (1 mmol/L phenylmethylsulfonyl fluoride, 1 µg/mL aprotinin, 1 µg/mL pepstatin A), scraped, and pelleted by centrifugation at 4°C. Cells were resuspended in a lysis buffer [1% SDS, 10 mmol/L EDTA, 50 mmol/L Tris-HCl (pH 8.1)], incubated for 10 minutes on ice, and sonicated to shear DNA. After sonication, lysate was centrifuged for 10 minutes at 13,000 rpm at 4°C. The supernatant was diluted in chromatin immunoprecipitation dilution buffer [0.01% SDS, 1% Triton X-100, 2 mmol/L EDTA, 16.7 mmol/L Tris-HCl (pH 8.1), 167 mmol/L NaCl, protease inhibitors]. DNMT3b antibody was added to the supernatant and incubated overnight at 4°C with rotation. The immunocomplex was collected by protein A/G-agarose and washed sequentially with low-salt washing buffer [0.1% SDS, 1% Triton X-100, 2 mmol/L EDTA, 200 mmol/L Tris-HCl (pH 8.1), 150 mmol/L NaCl], high-salt buffer [0.1% SDS, 1% Triton X-100, 2 mmol/L EDTA, 200 mmol/L Tris-HCl (pH 8.1), 500 mmol/L NaCl], LiCl washing buffer [0.25 mol/L LiCl, 1% NP40, 1% deoxycholate, 1 mmol/L EDTA, 10 mmol/L Tris-HCl (pH 8.1)], and 1x TE buffer [10 mmol/L Tris-HCl, 1 mmol/L EDTA (pH 8.0)]. The immunocomplex was eluted by elution buffer (1% SDS, 0.1 mol/L NaHCO3, 200 mmol/L NaCl) and the cross-links were reversed by heating at 65°C for 4 hours. After reaction, the samples were adjusted to 10 mmol/L EDTA, 20 mmol/L Tris-HCl (pH 6.5), and 40 µg/mL proteinase K and incubated at 45°C for 1 hour. DNA was recovered and subjected to PCR amplification by using the primers specific for the detection of the –107/+52 region, which contained the Sp1(B) site of RECK promoter. The sequences for the primers are sense 5'-GAACCGAGAGACCACTAACC-3' and antisense 5'-GACGATGAAGACGACCGGC-3'. The predicted size for PCR product is 159 bp.

Methylation-specific PCR analysis. Genomic DNA was extracted from the control or isopropyl-L-thio-ß-D-galactopyranoside (IPTG)–treated cells by using the QIAamp DNA Mini kit (Qiagen). The DNA was modified with sodium bisulfite and analyzed according to the procedures of CpGenome DNA Modification kit (Chemicon). The sets of primers were methylated sense 5'-GTTTTTGATTTTTCGTTTGAAGATC-3' and antisense 5'-CTCTAATAATTAACTACGACTCGCT-3' for the methylated sequence of mouse RECK promoter and unmethylated sense 5'-TTTTGATTTTTTGTTTGAAGATTGT-3' and antisense 5'-TCCTAATAATTAACTACAACTCACT-3' for the unmethylated sequence of mouse RECK promoter. The predicted products for methylated and unmethylated DNA are 173 and 172 bp, respectively. We also used the following primers for the detection of methylation of human RECK promoter: methylated sense 5'-AATAAAGAGTTTTGGTACGGGGTAC-3' and antisense 5'-AAAACCGCGAAATACTCGAA-3' for the methylated sequence of human RECK promoter and unmethylated sense 5'-TAAAGAGTTTTGGTATGGGGTATGT-3' and antisense 5'-CTCCAAACCACAAAATACTCAAA-3' for the unmethylated sequence of human RECK promoter. The predicted products for methylated and unmethylated DNA are 195 and 199 bp, respectively. Modified DNA was amplified in 50 µL reaction mixtures containing 5 µL of 10x PCR buffer, 14 µL of 25 mmol/L MgCl2, 2.5 µL of 25 mmol/L deoxynucleotide triphosphate, 1 µL of each primer (300 ng/µL), and 0.5 unit AmpliTaq Gold DNA polymerase (Roche, Palo Alto, CA). PCR was carried out in a thermal cycler for 35 cycles (denaturation at 95°C for 1 minute, annealing at 56°C for 2 minutes, and extension at 72°C for 1 minute) followed by a final 5-minute extension at 72°C. PCR products were separated in 2% agarose gels, stained with ethidium bromide, and visualized under UV illumination.

In vitro invasion assay. In vitro invasion assay was done by using 24-well Transwell units with polycarbonate filters (pore size, 8 µm) coated on the upper side with Matrigel (Becton Dickinson Labware, Bedford, MA; ref. 29). 2-12 cells were treated without or with IPTG (1 mmol/L) for 24 hours for the induction of RAS protein. Cells were collected and 5 x 103 cells in 100 µL medium containing IPTG or 5'-azacytidine (2.5 µmol/L) were placed in the top part of the Transwell unit and allowed to invade for 24 hours. The bottom part of the Transwell unit was filled with 10% FCS. After incubation, noninvaded cells on the top part of the membrane were removed with a cotton swab. Invaded cells on the bottom surface of the membrane were fixed in formaldehyde, stained with Giemsa solution, and counted under a microscope. In some experiments, 2-12 or human lung cancer cells were transfected with nonspecific or DNMT3b small interfering RNA (siRNA; 30 nmol/L) and incubated with IPTG for 48 hours. Cells were harvested and seeded to the top part of the Transwell unit. After another 24 hours, invaded cell numbers were counted as described above.

siRNA treatment. RECK siRNA 5'-AAGACCCAGCCCUUGCCUCAA-3' (sense) and a nonspecific RNA 5'-AACGUUGCGAUAGCGUAGUAC-3' were synthesized (Dharmacon Research, Inc., Lafayette, CO). DNMT3b siRNA 5'-GGAUGCUAUUGUGAAUGUGTT-3' and 5'-GCACUUUAAUCUGGCUACCTT-3' were obtained from Ambion (Austin, TX) and used to target mouse (for 2-12 cells) and human (for H358 cells) DNMT3b, respectively. Cells were transfected with siRNA by using the LipofectAMINE reagent.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
RAS activation increases DNMT3b expression in 2-12 cells. We first studied the effect of RAS oncogene on the expression of DNMTs. We have shown previously that IPTG at the concentrations of 1 to 5 mmol/L effectively induced oncogenic RAS protein expression in 2-12 cells after a 24-hour stimulation (26). As shown in Fig. 1A , treatment of 1 mmol/L IPTG induced RAS protein expression and down-regulated RECK protein level in 2-12 cells. In addition, we found that IPTG up-regulated DNMT3b, but not DNMT1 and DNMT3a, in both protein (Fig. 1B) and RNA (Fig. 1C) levels in a concentration-dependent manner. We next investigated the signaling pathways by which RAS increased DNMT3b expression. Our data indicated that up-regulation of DNMT3b by RAS is mediated via the extracellular signal-regulated kinase (ERK) signaling pathway; we found that the MEK inhibitor U0126, but not c-Jun NH2-terminal kinase inhibitor SP600126, effectively suppressed this induction at both mRNA and protein levels (Fig. 1D).


Figure 1
View larger version (25K):
[in this window]
[in a new window]
  		Figure 1. RAS down-regulates RECK and up-regulates DNMT3b expression via the ERK signaling pathway in 2-12 cells. A, 2-12 cells were treated without (–) or with (+) 1 mmol/L IPTG for 24 hours. Western blot analysis was done to investigate the protein level of RAS and RECK. B, 2-12 cells were treated with various concentrations of IPTG for 24 hours. Cells were harvested and protein levels of DNMTs were examined by Western blot analysis. The signal intensities of DNMTs were normalized to the intensities of actin and the DNMT/actin ratio of untreated cells was defined as 1.0. C, total RNA was isolated and RT-PCR was done to study the expression of DNMTs. D, cells were preincubated with vehicle (–), SP600126 (SP; 10 µmol/L), or U0126 (U; 10 µmol/L) for 2 hours and stimulated without or with 1 mmol/L IPTG for 24 hours. Total RNA was isolated and DNMT3b mRNA level was examined by RT-PCR. Cellular proteins were also harvested and DNMT3b protein levels were studied by Western blot analysis.

 
Oncogenic RAS increases DNMT3b-mediated methylation of RECK promoter. The studies from our and other groups have shown that a Sp1-binding site localized at the –82/–71 bp region from translation start site of RECK gene is critical for the inhibition of RECK by RAS (24, 26). Thus, we investigated whether DNMT3b bound to this region after RAS activation. By DAPA assay, we found that Sp1 protein bound to the DNA probe containing the Sp1 site localized at the –82/–71 bp region of RECK gene (Fig. 2A ). Consistent with our previous data, the binding affinity of Sp1 was not altered after induction of RAS by IPTG. In contrast, RAS activation significantly increased the binding of DNMT3b to the DNA probe (Fig. 2A). As expected, DNMT1 did not bind to the DNA probe under the similar experimental conditions (data not shown). We also studied the binding of DNMT3b to RECK promoter in vivo. As shown in Fig. 2A, chromatin immunoprecipitation assays showed that RAS activation increased the binding of DNMT3b to RECK promoter in vivo. We next examined if DNMT3b binding induced methylation of RECK promoter. Our methylation-specific PCR (MSP) experiments showed that RAS activation significantly increased the methylation of RECK promoter, an effect that could be reversed by U0126 but not by SP600126 (Fig. 2B). Collectively, these results suggest that oncogenic RAS up-regulates DNMT3b expression via the ERK signaling pathway and enhances its binding and methylation of RECK promoter. To clarify the role of promoter methylation on the inhibition of RECK by RAS, we tested the effect of a specific DNMT inhibitor 5'-azacytidine on RECK expression. As shown in Fig. 2C, 5'-azacytidine effectively reversed RAS-induced RECK promoter methylation. In addition, 5'-azacytidine counteracted RAS-induced down-regulation of RECK mRNA and protein expression in 2-12 cells (Fig. 2D). These data suggested that promoter methylation by DNMT3b is critical for RAS-induced gene silencing on RECK. To further confirm the importance of DNMT3b, we investigated whether RAS-induced RECK promoter methylation could be reversed by DNMT3b siRNA. As shown in Fig. 3A , DNMT3b siRNA, but not nonspecific dsRNA, effectively reversed RAS-mediated promoter methylation of RECK gene. In addition, DNMT3b siRNA suppressed DNMT3b expression and reversed RAS-mediated inhibition of RECK at both mRNA and protein levels (Fig. 3B and C). Because our aforementioned results indicated that ERK was involved in the induction of DNMT3b, we also tested the effect of U0126 on RECK expression. As expected, U0126 counteracted the effect of RAS on down-regulation of RECK protein in 2-12 cells (Fig. 3D).


Figure 2
View larger version (24K):
[in this window]
[in a new window]
  		Figure 2. RAS induces DNMT3b binding and methylation of RECK promoter that can be reversed by 5'-azacytidine. A, 2-12 cells were treated without or with 1 mmol/L IPTG for 24 hours and nuclear proteins were extracted. DNA probe containing the Sp1 site localized at the –82/–71 bp region from translational start site of RECK gene was incubated with nuclear proteins and DAPA assays were done to study the binding of DNMT3b to the probe. In chromatin immunoprecipitation assay, cells treated without or with IPTG were fixed with 1% formaldehyde. Cells were harvested and sonicated to shear DNA. DNMT3b or anti-Myc (negative control) antibody was added and incubated overnight at 4°C with rotation. The immunocomplex was collected and the binding of DNMT3b to RECK promoter was examined by chromatin immunoprecipitation assay. B, cells were preincubated with vehicle (V), SP600126 (10 µmol/L), or U0126 (10 µmol/L) for 2 hours and stimulated with 1 mmol/L IPTG for 24 hours. Genomic DNA was isolated and modified with sodium bisulfite. The methylation status of RECK promoter was investigated by MSP analysis using the primers for the detection of unmethylated (U) or methylated (M) DNA. C, 2-12 cells were cotreated with IPTG and vehicle (H2O) or 5'-azacytidine (AZC; 2.5 µmol/L) for 24 hours. RECK promoter methylation was investigated by MSP assay. D, total RNA was harvested and RECK expression was studied by RT-PCR. Cellular proteins were also extracted and subjected to SDS-PAGE. Protein level of RECK was studied by Western blot analysis.

 

Figure 3
View larger version (18K):
[in this window]
[in a new window]
  		Figure 3. DNMT3b siRNA reversed RAS-induced promoter methylation and down-regulation of RECK. A, 2-12 cells were transfected with nonspecific (N) or DNMT3b (D) siRNA and incubated without (control) or with 1 mmol/L IPTG for 48 hours. Genomic DNA was isolated and modified with sodium bisulfite. The methylation status of RECK promoter was investigated by MSP analysis using the primers for the detection of unmethylated or methylated DNA. B, cells were treated as described in (A). Total RNA was isolated and RT-PCR was done to investigate the expression of DNMT3b and RECK. C, cellular proteins were also harvested and the protein levels of DNMT3b and RECK were examined by Western blot analysis. D, cells were preincubated with vehicle, SP600126 (10 µmol/L), or U0126 (10 µmol/L) for 2 hours and stimulated without or with 1 mmol/L IPTG for 24 hours. Protein level of RECK was studied by Western blot analysis.

 
RAS-induced cell invasion is reversed by 5'-azacytidine and DNMT3b siRNA. We next addressed the functional role of RECK on RAS-induced cell invasion. Induction of oncogenic RAS protein significantly increased the invasive ability of 2-12 cells (Fig. 4A ). We found that an average 2.5-fold increase of invaded cell number was observed in IPTG-treated cells (Fig. 4B). Treatment of 5'-azacytidine obviously suppressed cell invasion induced by RAS (Fig. 4B). To verify if this inhibitory effect was mediated by DNMT3b, we used siRNA to specifically knockdown DNMT3b expression. As shown in Fig. 4B, DNMT3b siRNA, but not nonspecific siRNA, effectively suppressed RAS-induced cell invasion. We next studied the importance of RECK in the inhibition of RAS-induced cell invasion by 5'-azacytidine. RECK expression was inhibited by siRNA and the inhibitory action of 5'-azacytidine on RAS-induced cell invasion was investigated. As shown in Fig. 4C, we found that RECK, but not nonspecific, siRNA effectively reduced RECK mRNA and protein levels in 2-12 cells. In agreement with our hypothesis, knockdown of RECK by siRNA reversed the inhibitory effect of 5'-azacytidine on RAS-induced cell invasion (Fig. 4D). These results indicated that RECK was a critical mediator for the anti-invasive activity from 5'-azacytidine.


Figure 4
View larger version (19K):
[in this window]
[in a new window]
  		Figure 4. 5'-Azacytidine and DNMT3b siRNA suppress RAS-induced cell invasion. A, in vitro invasion assay was done by using 24-well Transwell units with polycarbonate filters (pore size, 8 µm) coated on the upper side with Matrigel. 2-12 cells were treated without or with IPTG (1 mmol/L) for 24 hours for the induction of RAS protein. Cells were collected and 5 x 103 cells in 100 µL medium containing IPTG or 5'-azacytidine (2.5 µmol/L) were placed in the top part of the Transwell unit and allowed to invade for 24 hours. In some experiments, 2-12 cells were transfected with nonspecific or DNMT3b siRNA and incubated with IPTG (1 mmol/L) for 24 hours for the induction of RAS protein. Cells were then harvested and 5 x 103 cells in 100 µL medium containing IPTG were seeded to the top part of the Transwell unit. Representative figures for cell invasion of each treatment group. B, invaded cells on the bottom surface of the membrane were fixed in formaldehyde, stained with Giemsa solution, and counted under a microscope. Columns, mean of three independent experiments; bars, SD. C, 2-12 cells were transfected without or with nonspecific or RECK-specific siRNA for 48 hours. Total RNA was isolated and RECK expression was investigated by RT-PCR. Cellular proteins were also harvested to investigate the protein level of RECK. D, cells were transfected without or with nonspecific or RECK-specific siRNA for 24 hours and incubated with IPTG (1 mmol/L) and 5'-azacytidine (2.5 µmol/L) for another 24 hours. Cells were collected and placed in the top part of the Transwell unit and invasion assay was done as described above. Columns, mean of three independent experiments; bars, SD.

 
RAS activation is associated with the promoter methylation and the silencing of RECK gene in human lung cancer cells. Thus far, we showed a successful use of the inducible system to study the functional role of DNMT3b in RAS-induced down-regulation of RECK. Next, we sought to determine if this functional aberration indeed occurred in human cancer. We chose two human lung adenocarcinoma cell lines, H358 and CL1.0, to address this issue. H358 cells harbor a constitutively activated K-RAS gene (codon 12 mutation), whereas CL1.0 cells have wild-type K-RAS gene (30). We found that RECK expression in H358 cells was obviously lower than that of CL1.0 cells (Fig. 5A ). MSP analysis indicated that RECK promoter was highly methylated in H358 but not in CL1.0 cells (Fig. 5B) and the treatment of 5'-azacytidine obviously up-regulated RECK mRNA and protein levels in H358 cells, whereas no significant change was observed in CL1.0 cells (Fig. 5C). Because 5'-azacytidine restored RECK expression in H358 cells, we investigated the effect of 5'-azacytidine on cell invasion. As shown in Fig. 5D, 5'-azacytidine potently inhibited invasiveness of H358 cells. On the contrary, 5'-azacytidine did not significantly affect the invasive ability of CL1.0 cells. We next asked whether DNMT3b was involved in RAS-induced inhibition of RECK and up-regulation of cell invasion in H358 cells. We used siRNA to suppress DNMT3b expression and found that knockdown of DNMT3b increased RECK mRNA and protein expression in H358 cells (Fig. 6A ). In addition, up-regulation of RECK by DNMT3b siRNA was associated with reduced cell invasion of H358 cells (Fig. 6B). These results suggested that oncogenic RAS also acts via DNMT3b induction and promoter methylation to repress RECK expression in lung cancer cells.


Figure 5
View larger version (23K):
[in this window]
[in a new window]
  		Figure 5. RAS mutation is correlated with down-regulation of RECK, which can be reversed by 5'-azacytidine in human lung cancer cells. A, total RNA was isolated form CL1.0 (harbors wild-type K-RAS) and H358 (harbors constitutively activated mutant K-RAS) human lung cancer cell lines and RECK expression was investigated by RT-PCR and Western blot analysis. B, methylation status of RECK promoter in CL1.0 and H358 cells was studied by MSP. C, cells were treated without or with 5'-azacytidine (2.5 µmol/L) for 24 hours and RECK expression was examined by RT-PCR and Western blot analysis. D, CL1.0 or H358 cells were incubated without or with 5'-azacytidine (2.5 µmol/L) for 24 hours and cell invasion assay was done as described in Fig. 4.

 

Figure 6
View larger version (22K):
[in this window]
[in a new window]
  		Figure 6. DNMT3b siRNA restores RECK expression in H358 cells and RECK is critical for 5'-azacytidine-induced inhibition of H358 cell invasion. A, H358 cells were transfected without or with nonspecific or DNMT3b-specific siRNA for 48 hours. DNMT3b and RECK expression was investigated by RT-PCR and Western blot analysis. B, H358 cells were transfected without or with nonspecific or DNMT3b-specific siRNA for 48 hours. Cell invasion assays were done. Columns, mean of three independent experiments; bars, SD. C, H358 cells transfected without or with nonspecific or RECK-specific siRNA were incubated with 5'-azacytidine (2.5 µmol/L) for 48 hours. RECK expression was investigated by RT-PCR and Western blot analysis. D, cells were treated as described above and invasion assays were done. Columns, mean of three independent experiments; bars, SD.

 
5'-Azacytidine restores RECK expression and suppresses invasiveness of H358 cells. To verify the importance of RECK on the inhibition of invasiveness of H358 cells by 5'-azacytidine, we suppressed RECK expression by siRNA and tested the effect of 5'-azacytidine. As shown in Fig. 6C, RECK siRNA effectively repressed mRNA and protein levels of RECK in H358 cells. Furthermore, our results clearly indicated that knockdown of RECK almost completely abolished the anti-invasive activity of 5'-azacytidine (Fig. 6D). Thus, restoration of RECK expression by 5'-azacytidine is important for its inhibition of cell invasion of H358 lung cancer cells, which harbor constitutively activated RAS oncogene.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
In the current study, we clarify the functional role of DNMTs in RAS-induced down-regulation of the metastasis suppressor RECK by using a RAS-inducible system. We also study the methylation of RECK promoter and the regulation of RECK expression and test the effect of the DNMT inhibitor 5'-azacytidine in human lung cancer cells that either harbor wild-type or constitutively activated RAS gene. Results of both investigations provide similar conclusions that RAS-induced down-regulation of RECK is mediated via a DNMT3b/promoter methylation mechanism. A genome-wide survey of RAS transformation targets indicates that both gene activation and repression are critical for the transformation by oncogenic RAS (31). However, how oncogenic RAS suppresses gene transcription is largely unknown. Recent studies suggest that DNMTs may be involved in this process. For example, a study from Soejima et al. showed that DNMT3b was participated in RAS-induced inhibition of tumor suppressor genes, including FHIT, TLSC1, and RASSF1A (32). Similarly, a study from Ordway et al. showed that increased DNMT1 activity was critical for the transformation and gene repression by RAS oncogene (33). More recently, Pruitt et al. reported that RAS-induced loss of the proapoptotic response protein PAR-4 was mediated by DNA hypermethylation through RAF-independent and RAF-dependent signaling cascades in epithelial cells (34). These authors also indicated that expression of DNMT1 and DNMT3a were up-regulated by oncogenic RAS. In agreement with these results, we found that DNMT3b expression increased after induction of oncogenic RAS protein. Moreover, we found that the ERK signaling pathway was involved in the activation of DNMT3b transcription by RAS. However, two questions remain unanswered at present. First, why are the expressions of DNMTs induced by RAS found in these studies are different? One possible explanation is that oncogenic RAS induces DNMT expression in a cell type–specific manner. Thus, the DNMTs up-regulated by RAS are dependent on the experimental cell lines used in the studies. Another possible explanation are the differences of the methods for the induction of oncogenic RAS. In most studies, investigators transfected constitutively activated RAS vector into cells to study the alteration of DNMTs (32–34). It should be noted that DNMT1 expression is cell cycle dependent (35). Constitutively activated RAS may change cell cycle distribution that leads to alterations in DNMT expression. Conversely, an inducible system is used in this study to address the effect of oncogenic RAS on DNMTs. We found that cell cycle distribution of 2-12 cells was not significantly altered after treatment of IPTG for 24 hours (data not shown), indicating that the difference of DNMT expression observed is not caused by the change of cell cycle distribution. Second, how does RAS act via ERK to stimulate DNMT3b expression? Characterization of human DNMT3b promoter indicates that the –604/–67 bp region of the promoter possess the basal promoter activity (36). Prediction of transcription factor binding sites reveals that several Sp1-binding sites are localized in this region (36). A straightforward speculation is that RAS stimulates ERK activity to phosphorylate Sp1 to activate DNMT3b transcription. Indeed, recent studies showed that ERK might directly phosphorylate Sp1 and activate Sp1-controlled gene transcription (37, 38). More experiments are needed to verify whether RAS acts via Sp1 to stimulate DNMT3b.

Previous studies on oncogenic activity of RAS mainly focused on its effect on cell cycle regulators. RAS oncogene may activate cyclin D1 and E expression to promote cell cycle progression. In addition, RAS may negatively regulate p21 and p27 to antagonize the antiproliferative activity of these two cyclin-dependent kinase inhibitors. However, it is unclear if the inhibition of gene expression by oncogenic RAS plays a critical role in the induction of tumor metastasis. A recent genome-wide screening of RAS targets showed that RAS suppressed the expression of thrombospondin 1 and tissue inhibitor of metalloproteinase 1, two important antimetastatic genes, to enhance cell invasion (39). Pathologic analysis also shows that inactivation of RASSF1A, a tumor suppressor that exhibits antimetastatic activity, is closely linked with RAS mutation in cancer cells (40, 41). We now provide evidences that oncogenic RAS inhibits an important metastasis suppressor, RECK, through DNMT3b-mediated promoter methylation to enhance cell invasion. Most importantly, we found that reversion of RECK expression by DNMT inhibitor and DNMT3b siRNA effectively block RAS-induced cell invasion. These results are of clinical significance. RAS mutation is frequently found in human cancers. Therefore, targeting of RAS signaling pathway is an important issue for cancer therapy. Recently, several drugs have been developed for this purpose. The first class, which exhibits high target specificity, are antisense oligonucleotides. Several antisense oligonucleotides targeting RAS oncogene are now undergoing clinical trials (42). The second class, which possesses less specificity, are prenyltransferase inhibitors. The covalent attachment of the prenyl group to RAS proteins is the first step for their post-translational modifications and is essential for the oncogenic activity of RAS proteins. Three prenyltransferases, including farnesyltransferase and geranylgeranyltransferase I and II, have been shown to catalyze this reaction and the inhibitors of these enzymes exhibit potent anticancer effect in vitro and in vivo (43–45). However, recent studies indicated that prenyltransferase inhibitors may inhibit proliferation of RAS-mutated cancer cells via a RAS-independent pathway and may not be real anti-RAS drugs. Indeed, prenyltransferase inhibitors have been shown recently to target the phosphatidylinositol 3-kinase/AKT and mammalian target of rapamycin pathways to repress tumor growth (46–48). Results from our study suggest that DNMT inhibitors may be another novel class of drugs for the inhibition of RAS-induced tumorigenesis.

Because promoter methylation and gene silencing of tumor suppressor genes are frequently detected in various types of human cancer (49), it is rational to speculate that DNMT inhibitors may exhibit anticancer actions. Indeed, nucleoside or non-nucleoside DNMT inhibitors are progressing into clinical trials recently. For example, two nucleoside inhibitors 5'-azacytidine and decitabine have been tested in many phase I and II clinical trials for the treatment of different cancers (50). More recently, 5'-azacytidine (also named as Vidaza) has been approved by Food and Drug Administration for the treatment of myelodysplastic syndrome, a preleukemic bone marrow disorder (50). Although the antiproliferative activity of DNMT inhibitors has been well studied, the application of DNMT inhibitors for the therapy on tumor metastasis is not known. Our results suggest that DNMT inhibitors are a novel class of antimetastatic drugs and may be useful for the treatment of RAS-induced metastasis.


   Acknowledgments
 
Grant support: Center for Gene Regulation and Signal Transduction Research, National Cheng Kung University and National Sun Yat-Sen University-Kaohsiung Medical University Joint Research Center.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received 2/21/06. Revised 6/27/06. Accepted 7/ 3/06.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Grand RA, Owen D. The biochemistry of ras p21. Biochem J 1991;279:609–31.
  2. Bos JL. Ras oncogenes in human cancer: a review. Cancer Res 1989;49:4682–9.[Abstract/Free Full Text]
  3. Bondy GP, Wilson S, Chambers AF. Experimental metastatic ability of H-ras-transformed NIH3T3 cells. Cancer Res 1985;45:6005–9.[Abstract/Free Full Text]
  4. Muschel RJ, Williams JE, Lowy DR, Liotta LA. Harvey ras induction of metastatic potential depends upon oncogene activation and the type of recipient cell. Am J Pathol 1985;121:1–8.[Abstract]
  5. Ichikawa T, Kiprianou N, Isaacs JT. Genetic instability and the acquisition of metastatic ability by rat mammary cancer cells following v-H-ras oncogene transfection. Cancer Res 1990;50:6349–57.[Abstract/Free Full Text]
  6. Takai Y, Takuya S, Tanaka K, Nakanishi H. Rho as a regulator of the cytoskeleton. Trends Biochem Sci 1995;20:227–31.[CrossRef][Medline]
  7. Hughes DA. Control of signal transduction and morphogenesis by Ras. Semin Cell Biol 1995;6:89–94.[CrossRef][Medline]
  8. Khosravi-Far R, Der CJ. The Ras signal transduction pathway. Cancer Metastasis Rev 1994;13:67–89.[CrossRef][Medline]
  9. Hall A. Ras-related GTPases and cytoskeleton. Mol Biol Cell 1992;3:475–9.[Abstract]
  10. Zhang Z, Vuori K, Wang H, Reed JC, Ruoslahti E. Integrin activation by R-ras. Cell 1996;85:61–9.[CrossRef][Medline]
  11. Kinashi T, Katagiri K, Watanabe S, Vanhaesebroeck B, Downward J, Takatsu K. Distinct mechanisms of {alpha}5ß1 integrin activation by Ha-Ras and R-Ras. J Biol Chem 2000;275:22590–6.[Abstract/Free Full Text]
  12. Schramm K, Krause K, Bittroff-Leben A, Goldin-Lang P, Thiel E, Kreuser ED. Activated K-ras is involved in regulation of integrin expression in human colon carcinoma cells. Int J Cancer 2000;87:155–64.[CrossRef][Medline]
  13. Schmidt CR, Washington MK, Gi YJ, Coffey RJ, Beauchamp RD, Pearson AS. Dysregulation of E-cadherin by oncogenic Ras in intestinal epithelial cells is blocked by inhibiting MAP kinase. Am J Surg 2003;186:426–30.[CrossRef][Medline]
  14. Schmidt CR, Gi YJ, Coffey RJ, Beauchamp RD, Pearson AS. Oncogenic Ras dominates overexpression of E-cadherin in malignant transformation of intestinal epithelial cells. Surgery 2004;136:303–9.[CrossRef][Medline]
  15. Caulin C, Lopez-Barcons L, Gonzales-Garrigues M, et al. Suppression of the metastatic phenotype of a mouse skin carcinoma cell line independent of E-cadherin expression and correlated with reduced Ha-ras oncogene products. Mol Carcinog 1996;15:104–14.[CrossRef][Medline]
  16. Rak J, Mitsuhashi Y, Bayko L, et al. Mutant ras oncogenes upregulate VEGF/VPF expression: implications for induction and inhibition of tumor angiogenesis. Cancer Res 1995;55:4575–80.[Abstract/Free Full Text]
  17. Pironin M, Clement G, Benzakour O, Barritault D, Lawrence D, Vigier P. Growth in serum-free medium of NIH3T3 cells transformed by the EJ-H-ras oncogene: evidence for multiple autocrine growth factors. Int J Cancer 1992;51:980–8.[Medline]
  18. Birkedal-Hansen H. Proteolytic remodeling of extracellular matrix. Curr Opin Cell Biol 1995;7:728–35.[CrossRef][Medline]
  19. Sternlicht MD, Werb Z. How matrix metalloproteinases regulate cell behavior. Annu Rev Cell Dev Biol 2001;17:463–516.[CrossRef][Medline]
  20. Himelstein BP, Lee EJ, Sato H, Seiki M, Muschel RJ. Transcriptional activation of the matrix metalloproteinase-9 gene in an H-ras and v-myc transformed rat embryo cell line. Oncogene 1997;14:1995–8.[CrossRef][Medline]
  21. Meade-Tollin LC, Boukamp P, Fusenig NE, Bowen CP, Tsang TC, Bowden GT. Differential expression of matrix metalloproteinases in activated c-ras-Ha-transfected immortalized human keratinocytes. Br J Cancer 1998;77:724–30.[Medline]
  22. Overall CM, Lopez-Otin C. Strategies for MMP inhibition in cancer: innovations for the post-trial era. Nat Rev Cancer 2002;2:657–72.[CrossRef][Medline]
  23. Takahashi C, Sheng Z, Horan TP, et al. Regulation of matrix metalloproteinase-9 and inhibition of tumor invasion by the membrane-anchored glycoprotein RECK. Proc Natl Acad Sci U S A 1998;95:13221–6.[Abstract/Free Full Text]
  24. Sasahara RM, Takahashi C, Noda M. Involvement of the Sp1 site in ras-mediated downregulation of the RECK metastasis suppressor gene. Biochem Biophys Res Commun 1999;264:668–75.[CrossRef][Medline]
  25. Oh J, Takahashi R, Kondo S, et al. The membrane-anchored MMP inhibitor RECK is a key regulator of extracellular matrix integrity and angiogenesis. Cell 2001;107:789–800.[CrossRef][Medline]
  26. Chang HC, Liu LT, Hung WC. Involvement of histone deacetylation in ras-induced down-regulation of the metastasis suppressor RECK. Cell Signal 2004;16:675–9.[CrossRef][Medline]
  27. Liu HS, Scrable H, Villaret DB, Lieberman MA, Stambrook PJ. Control of Ha-ras-mediated mammalian cell transformation by Escherichia coli regulatory elements. Cancer Res 1992;52:983–8.[Abstract/Free Full Text]
  28. Pan MR, Hung WC. Nonsteroidal anti-inflammatory drugs inhibit matrix metalloproteinase-2 via suppression of the ERK/Sp1-mediated transcription. J Biol Chem 2002;277:32775–80.[Abstract/Free Full Text]
  29. Liu LT, Chang HC, Chiang LC, Hung WC. Histone deacetylase inhibitor up-regulates RECK to inhibit MMP-2 activation and cell invasion. Cancer Res 2003;63:3069–72.[Abstract/Free Full Text]
  30. Alemany R, Ruan S, Kataoka M, et al. Growth inhibitory effect of anti-K-ras adenovirus on lung cancer cells. Cancer Gene Ther 1996;3:296–301.[Medline]
  31. Zuber J, Tchernitsa OI, Hinzmann B, et al. A genome-wide survey of RAS transformation targets. Nat Genet 2002;24:144–52.
  32. Soejima K, Fang W, Rollins BJ. DNA methyltransferase 3b contributes to oncogenic transformation induced by SV40T antigen and activated Ras. Oncogene 2003;22:4723–33.[CrossRef][Medline]
  33. Ordway JM, Williams K, Curran T. Transcription repression in oncogenic transformation: common targets of epigenetic repression in cells transformed by Fos, Ras or Dnmt1. Oncogene 2004;23:3737–48.[CrossRef][Medline]
  34. Pruitt K, Ulku AS, Frantz K, et al. Ras-mediated loss of the pro-apoptotic response protein Par-4 is mediated by DNA hypermethylation through Raf-independent and Raf-dependent signaling cascades in epithelial cells. J Biol Chem 2005;280:23363–70.[Abstract/Free Full Text]
  35. Kimura H, Nakamura T, Ogawa T, Tanaka S, Shiota K. Transcription of mouse DNA methyltransferase 1 (DNMT1) is up-regulated by both E2F-Rb-HDAC-dependent and independent pathways. Nucleic Acids Res 2003;31:3101–13.[Abstract/Free Full Text]
  36. Yanagisawa Y, Ito E, Yuasa Y, Maruyama K. The human DNA methyltransferases DNMT3A and DNMT3B have two types of promoters with different CpG contents. Biochim Biophys Acta 2002;1577:457–65.[Medline]
  37. Milanini-Mongiat J, Pouyssegur J, Pages G. Identification of two Sp1 phosphorylation sites for p42/p44 mitogen-activated protein kinases: their implication in vascular endothelial growth factor gene transcription. J Biol Chem 2002;277:20631–39.[Abstract/Free Full Text]
  38. Merchant JL, Du M, Todisco A. Sp1 phosphorylation by Erk 2 stimulates DNA binding. Biochem Biophys Res Commun 1999;254:454–61.[CrossRef][Medline]
  39. Watnick RS, Cheng YN, Rangarajan A, Ince TA, Weinberg RA. Ras modulates Myc activity to repress thrombospondin-1 expression and increase tumor angiogenesis. Cancer Cell 2003;3:219–31.[CrossRef][Medline]
  40. Dammann R, Schagdarsurengin U, Liu L, et al. Frequent RASSF1A promoter hypermethylation and K-ras mutations in pancreatic carcinoma. Oncogene 2003;22:3806–12.[CrossRef][Medline]
  41. Kim DH, Kim JS, Park JH, et al. Relationship of Ras association domain family 1 methylation and K-ras mutation in primary non-small cell lung cancer. Cancer Res 2003;63:6206–11.[Abstract/Free Full Text]
  42. Cox AD, Der CJ. Ras family signaling: therapeutic targeting. Cancer Biol Ther 2002;1:599–606.[Medline]
  43. Zhu K, Hamilton AD, Sebti SM. Farnesyltransferase inhibitors as anticancer agents: current status. Curr Opin Investig Drugs 2003;4:1428–35.[Medline]
  44. Sebti SM, Adjei AA. Farnesyltransferase inhibitors. Semin Oncol 2004;31:28–39.[Medline]
  45. Gelb MH, Scholten JD, Sebolt-Leopold JS. Protein prenylation: from discovery to prospects for cancer treatment. Curr Opin Chem Biol 1998;2:40–8.[CrossRef][Medline]
  46. Jiang K, Coppola D, Crespo NC, et al. The phosphoinositide 3-OH kinase/AKT2 pathway as a critical target for farnesyltransferase inhibitor-induced apoptosis. Mol Cell Biol 2000;20:138–48.
  47. Law BK, Norgaard P, Moses HL. Farnesyltransferase inhibitor induces rapid growth arrest and blocks p70s6k activation by multiple stimuli. J Biol Chem 2000;275:10796–801.[Abstract/Free Full Text]
  48. Sebti SM. Protein farnesylation: implications for physiology, malignant transformation and cancer therapy. Cancer Cell 2005;7:297–300.[CrossRef][Medline]
  49. Esteller M. Aberrant DNA methylation as a cancer-inducing mechanism. Annu Rev Pharmacol Toxicol 2005;45:629–56.[CrossRef][Medline]
  50. Das PM, Singal R. DNA methylation and cancer. J Clin Oncol 2004;22:4632–42.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 J. Cell Sci.Home page
M.-R. Pan, T.-M. Chang, H.-C. Chang, J.-L. Su, H.-W. Wang, and W.-C. Hung
Sumoylation of Prox1 controls its ability to induce VEGFR3 expression and lymphatic phenotypes in endothelial cells
J. Cell Sci., September 15, 2009; 122(18): 3358 - 3364.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
M.-R. Pan, H.-C. Chang, Y.-C. Wu, C.-C. Huang, and W.-C. Hung
Tubocapsanolide A Inhibits Transforming Growth Factor-{beta}-activating Kinase 1 to Suppress NF-{kappa}B-induced CCR7
J. Biol. Chem., January 30, 2009; 284(5): 2746 - 2754.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Cancer Res.Home page
M.-C. Hsu, C.-C. Huang, H.-C. Chang, T.-H. Hu, and W.-C. Hung
Overexpression of Jab1 in Hepatocellular Carcinoma and Its Inhibition by Peroxisome Proliferator-Activated Receptor{gamma} Ligands In vitro and In vivo
Clin. Cancer Res., July 1, 2008; 14(13): 4045 - 4052.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
M.-R. Pan, M.-F. Hou, H.-C. Chang, and W.-C. Hung
Cyclooxygenase-2 Up-regulates CCR7 via EP2/EP4 Receptor Signaling Pathways to Enhance Lymphatic Invasion of Breast Cancer Cells
J. Biol. Chem., April 25, 2008; 283(17): 11155 - 11163.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chang, H.-C.
Right arrow Articles by Hung, W.-C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chang, H.-C.
Right arrow Articles by Hung, W.-C.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online