Cancer Research Meeting Calendar EMT and Cancer Progression and Treatment
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 Ushijima, T.
Right arrow Articles by Kaneda, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ushijima, T.
Right arrow Articles by Kaneda, A.
[Cancer Research 65, 11-17, January 1, 2005]
© 2005 American Association for Cancer Research

Molecular Biology, Pathobiology and Genetics

Decreased Fidelity in Replicating CpG Methylation Patterns in Cancer Cells

Toshikazu Ushijima1, Naoko Watanabe1, Kimiko Shimizu2, Kazuaki Miyamoto1, Takashi Sugimura1 and Atsushi Kaneda1

1 Carcinogenesis Division and 2 Cancer Genomics Division, National Cancer Center Research Institute, Tokyo, Japan

Requests for reprints: Toshikazu Ushijima, Carcinogenesis Division, National Cancer Center Research Institute, 1-1 Tsukiji 5-chome, Chuo-ku, Tokyo 104-0045, Japan. Phone: 81-3-3547-5240; Fax: 81-3-5565-1753; E-mail: tushijim{at}ncc.go.jp.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
The unmethylated or methylated status of individual CpG sites is faithfully copied into daughter cells. Here, we analyzed the fidelity in replicating their methylation statuses in cancer cells. A single cell was clonally expanded, and methylation statuses of individual CpG sites were determined for an average of 12.5 DNA molecules obtained from the expanded population. By counting the deviation from the original methylation patterns inferred, the number of errors was measured. The analysis was done in four gastric cancer cell lines for five CpG islands (CGI), and repeated six times (total 1,495 clones sequenced). HSC39 and HSC57 showed error rates <1.0 x 10–3 errors per site per generation (99.90-100% fidelity) for all the five CGIs. In contrast, AGS showed significantly elevated error rates, mainly due to increased de novo methylation, in three CGIs (1.6- to 3.2-fold), and KATOIII showed a significantly elevated error rate in one CGI (2.2-fold). By selective amplification of fully methylated DNA molecules by methylation-specific PCR, those were stochastically detected in KATOIII and AGS but never in HSC39 and HSC57. When methylation of entire CGIs was examined for eight additional CGIs, KATOIII and AGS had frequent methylation, whereas HSC39 and HSC57 had few. KATOIII and AGS had four and eight times, respectively, as high expression levels of DNMT3B as HSC39. These data showed that some cancer cells have decreased fidelity in replicating methylation patterns in some CGIs, and that the decrease could lead to methylation of the entire CGIs.

Key Words: methylation pattern • maintenance methylation • de novo methylation • fidelity • CpG island • 02-12-03 DNA methylation/epigenetics • 02-12-03 DNA methylation/epigenetics • 02-12-03 DNA methylation/epigenetics • 02-10-03 Genetic instability: multistep progression • 02-12-00 Gene Expression, Chromatin Regulation, and Oncogenomics


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
CpG methylation serves as a long-term cellular memory (1). The methylation pattern of each cell is formed during embryonic development and is maintained in adult somatic cells (1, 2). Methylation of CpG islands (CGI) in gene promoters causes silencing of the downstream genes in normal and cancer cells (3–6). In order to maintain the methylated or unmethylated status of CpG sites, hemi-methylated CpG sites are methylated at DNA replication into fully methylated CpG sites, mainly by DNA methyltransferase 1 (DNMT1) that has a high affinity to hemi-methylated CpG sites (1, 5). It is also important for the maintenance of methylation patterns that unmethylated CpG sites are not methylated during DNA replication and in other phases of the cell cycle. The presence of Sp1 binding sites and gene transcription has been implicated as a mechanism to protect CGIs from de novo methylation (1, 7–10) .

As for the fidelity in replicating methylation patterns, including both methylated and unmethylated statuses of CpG sites, there had been limited information. The methylated status of exogenous and endogenous sequences was inherited with fidelities of ~94% and 98.8% to 99.9% per site per generation, respectively (11, 12) . Recently, taking advantage of methylation analysis by sequencing after bisulfite modification (13), we measured the fidelity in replicating the methylation status of each CpG site in normal cells (14). A single human mammary epithelial cell was expanded to 106 cells, methylation patterns in the expanded cell population were analyzed, and deviation from the inferred original methylation patterns of the two alleles in the starting single cell was analyzed. The fidelity in replicating methylation patterns was in the range of 99.85% to 99.92% per site per generation in promoter CGIs and 99.56% to 99.83% in CGIs outside promoter regions. Fidelities of 99.90% and 99.60% are expected to yield 0.020 and 0.077 methylation errors per site, respectively, in 20 generations.

When a CGI is methylated, almost all CpG sites in the CGI are methylated. The methylation of a normally unmethylated CGI is known to be triggered by methylation of multiple CpG sites in the CGI (15, 16)). With the high fidelity observed in normal cells, the probability that multiple CpG sites are methylated is very low, assuming that de novo methylation of proximate CpG sites is not cooperative. However, methylation of CGIs is frequently observed in cancers (6). Moreover, some cancers are associated with frequent aberrant methylation of multiple CGIs, called the CGI methylator phenotype (17, 18)). Although there remains a dispute over the presence of the CGI methylator phenotype (19), one possible mechanism for the CGI methylator phenotype is a general decrease in the fidelity in replicating methylation patterns. Alternatively, methylation of CGIs could cluster by a possible growth advantage conferred by methylation of promoter CGIs of tumor suppressor genes. Cancer cells with a long history would be expected to have accumulated more aberrant methylation than cancer cells with a short history. Also, "aberrant" methylation observed in cancer cells could be simply representing methylation present in cancer precursor cells, which is too small a fraction to be detected in normal tissues.

In this study, we measured fidelity in replicating methylation patterns in four gastric cancer cells, HSC39, HSC57, KATOIII and AGS, using promoter CGIs of five genes, bA305P22.2.3, FLJ32130, a homologue of RIKEN2210016 (RIKEN2210016), E-cadherin and Cyclophilin A. The first four genes were selected because they can be potentially silenced in human gastric cancers (18, 20)).


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Cell Culture and DNA/RNA Extraction. HSC39 and HSC57 were kind gifts from Dr. K. Yanagihara, National Cancer Center Research Institute, and KATOIII and AGS were obtained from the Japanese Collection of Research Bioresources (Tokyo, Japan) and the American Type Culture Collection (Manassas, VA), respectively. Human mammary epithelial cells were purchased from Cambrex (East Rutherford, NJ).

For analysis of fidelity in replicating methylation patterns, a single cell in log-phase growth was plated in a well of a 96-well plate. Cells were serially transferred to a well of a 12-well plate and to a 10-cm dish. When the cells grew to the target cell number (5 x 106), they were collected after measuring the actual number of cells. The number of cell generations observed was calculated from the plating efficiencies and the final cell count. The culture was repeated six times for each cell line.

High molecular weight genomic DNA was extracted by serial extraction with phenol/chloroform and ethanol precipitation, and RNA was isolated using ISOGEN (Nippon Gene, Tokyo, Japan).

Fluorescence In situ Hybridization. Chromosome spreads were obtained by using standard protocols (21). BAC clones (RP11-305P22 for bA305P22.2.3, CIT987SK-A-635H12 for FLJ32130, RP11-575L7 for RIKEN2210016, RP11-354M1 for E-cadherin, and RP11-105B9 for Cyclophilin A) were obtained from the mapping core group at the Sanger Institute or from Dr. H. Shizuya at the California Institute of Technology. A BAC was labeled with biotin 16-dUTP, using a nick translation kit (Roche, Basel, Switzerland). Probes were mixed after denaturation and then hybridized to the chromosome preparations. After washing, the signals were visualized after incubation with avidin-FITC. The washed slides were counterstained with diamidino-2-phenylindole (VYSIS, Downers Grove, IL).

Bisulfite Modification and Sequencing. Sodium bisulfite modificationwas done according to previous reports (13, 22). After restriction of genomic DNA with BamHI, 500 ng of the DNA were denatured in 0.3 N NaOH. The denatured DNA was dissolved in a solution containing 3.1 mol/L NaHSO3 (pH 5.3) and 0.5 mmol/L hydroquinone, and the solution underwent 15 cycles of denaturation at 95°C for 30 seconds and incubation at 50°C for 15 minutes. After desalting the sample with the Wizard DNA clean-up system (Promega, Madison, WI), the sample was desulfonated in 0.3 N NaOH. The DNA sample was ethanol-precipitated with ammonium acetate and dissolved in 20 µL of TE buffer.

Each region of a specific CGI was amplified by PCR using 1 µL of the solution and primers common to the methylated and unmethylated DNA sequences (Supplementary Table; refs. 14, 18, ). PCR products were cloned into pGEM-T Easy vector (Promega) and were cycle-sequenced using an ABI automated DNA sequencer (Applied Biosystems, Foster City, CA). At least 10 clones and more than thrice the gene copy number obtained by fluorescence in situ hybridization (FISH) were analyzed. To monitor the efficiency of bisulfite modification, the PCR product of the CGI of the E-cadherin gene mixed with rat genomic DNA was simultaneously processed. The unconversion rate of cytosines at CpG sites was <0.002. The methylation-specific PCR (MSP) product of the RIKEN2210016 CGI was similarly cloned and three to four clones were analyzed.

Methylation-specific PCR. MSP (23) was done using 1 µL of the bisulfite-modified DNA solution and primers specific for methylated (M) or unmethylated (U) sequences (Supplementary Table; refs. 18, 23, 24). The optimal annealing temperature was determined by amplifying DNA of normal human mammary epithelial cells and DNA treated by SssI-methylase (NEB, Beverly, MA). PCR was done for 33 to 35 cycles for analysis of CGI methylation, and a sample was regarded to have aberrant methylation when the PCR product was obtained with the M set primers. To examine the presence of DNA molecules fully methylated in the RIKEN2210016 CGI, PCR was done for 40 cycles in four replicates.

Methylation Pattern Error Rate, Fidelity, and Error Rate of a CpG Site. All these three parameters were calculated as reported (14). To infer the original methylation patterns in the single starting cell, the bisulfite-sequenced clones were classified according to their methylation patterns. The most common patterns were used to assign starting methylation status for each allele of a particular CGI, and deviations in other bisulfite-derived clones were assigned to a particular allele based on the minimum number of changes required to match the inferred original patterns. Bias in the selection of the original patterns was confirmed to be <20% of the total number of errors by selecting different clones as the original patterns. The total number of deviations in each culture was divided by the number of CpG sites observed, and an average number of errors per CpG site during the culture (21.6-23.1 generations), which we termed methylation pattern error rate (MPER), was calculated. The average MPER for a CGI in a cell line was calculated from the MPERs obtained for the six cultures. Fidelity in replicating methylation patterns (F: % per site per generation) was calculated from MPERs (M: errors per site per observed generations) by an equation: M = 1 – Fg (g = number of generations). The error rate of a CpG site (E: errors per site per generation) was calculated from the equation F + E = 1.

Quantitative Reverse Transcription-PCR. From 3 µg of DNase-treatedtotal RNA, cDNA was synthesized in 20 µL with a random hexamer primer and Superscript II reverse transcriptase (Life Technologies, Inc., Rockville, MD). One microliter of the cDNA solution was amplified in a solution that contained SYBR Green PCR Core Reagents (Applied Biosystems) and 200 nmol/L of primers. The sequences of the primers are listed in the Supplementary Table. PCR was done using an iCycler iQ detection system (Bio-Rad Laboratories, Hercules, CA). The absence of non-specific amplification was confirmed by observing melting curves of the product. The number of cDNA molecules in a sample was quantified by normalizing to the amplification curves of standard samples that contained 101 to 107 copies of the gene. The amount was normalized to that of PCNA.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Analysis of Fidelities in Replicating Methylation Pattern. A single cell of HSC39, HSC57, KATOIII, and AGS was expanded to 5.0 ± 0.2, 7.1 ± 0.2, 2.9 ± 0.0, and 5.6 ± 0.1 (mean ± SE) x 106, respectively. Plating efficiencies during the two transfers were 96 ± 2%, 86 ± 3%, 92± 2%, and 95 ± 1%, respectively. Based on these values, the number of cells that should have been produced at the time of harvest was calculated as 5.4, 9.5, 3.4, and 6.2 x 107 (final cell count per plating efficiency per plating efficiency), and they were estimated to have undergone 22.3, 23.1, 21.6 and 22.5 generations, respectively. The copy numbers of the five CGIs in the four cancer cell lines were analyzed by FISH (representative results in Fig. 1; summarized in Table 1).



View larger version (32K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Representative of FISH analysis of the four gastric cancer cell lines (bA305P22.2.3). At least five metaphases were observed, and the numbers of each CGI in a cell was measured. A, HSC39; B, HSC57; C, KATOIII; and D, AGS. HSC39 shows hybridization to two separate chromosomal loci, consistent with the normal location of this CGI on chromosome 20; HSC57 shows five loci on one normal chromosome 20 and two isochromosome 20s; KATOIII shows four loci on two normal chromosome 20s and one isochromosome 20; AGS shows three loci on three normal chromosome 20s.

 

View this table:
[in this window]
[in a new window]

  		Table 1. Fidelities in replicating methylation patterns in cancer cell lines

 
The methylation status of each CpG site was determined by bisulfite sequencing of promoter CGIs of bA305P22.2.3, FLJ32130, RIKEN2210016, E-cadherin, and Cyclophilin A (representative results in Fig. 2). Based on the observed errors and copy numbers obtained by FISH, MPERs, fidelities, and error rates were calculated (Table 1). Two cell lines, HSC39 and HSC57, showed error rates smaller than 1.0 x 10–3 errors per site per generation for all the five CGIs, which were in the same range as our previous data in normal cells (14). In contrast, AGS had significantly elevated error rates forbA305P22.2.3 (3.2-fold compared with HSC57, P < 0.05), RIKEN2210016 (5.7-fold, P < 0.01) and E-cadherin (1.6-fold, P < 0.01). KATOIII showed an elevated error rate for RIKEN2210016 (1.9-fold compared with HSC57, P < 0.05) and tended to show an elevated error rate for bA305P22.2.3 (2.2-fold, P = 0.06).



View larger version (81K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Representative of bisulfite sequencing. Unmethylated sites ({circ}) and methylated CpG sites ({bullet}), and nucleotide positions of the CpG sites in the PCR product are shown above. Brackets, original methylation patterns inferred; right, numbers of deviation from them. A, RIKEN2210016. In HSC39 and HSC57, methylation patterns were stable, and original patterns were easily inferred in some cultures (culture #2 of HSC39 and #1 of HSC57). In KATOIII and AGS, methylation patterns were unstable, and it was difficult to infer the original methylation patterns. Especially, in AGS, a DNA molecule methylated in 9 of 19 CpG sites was present (arrowhead). B, E-cadherin. Methylation patterns were stable in all the four cell lines.

 
The errors observed were mainly due to methylation of unmethylated CpG sites, but demethylation of methylated sites was also observed. In RIKEN2210016, a DNA molecule methylated at 9 of 19 CpG sites was observed in AGS (arrowhead, Fig. 2A). This showed that AGS, and possibly KATOIII, had decreased fidelities in replicating methylation patterns, that the decrease was prominent in CGIs of bA305P22.2.3 and RIKEN2210016, and that the decrease could lead to methylation of nearly half of all the CpG sites.

Emergence of Fully Methylated DNA Molecules in Cell Lines with Decreased Fidelity. The decreased fidelity produced DNA molecules methylated at nearly half of all the CpG sites. Considering that "seeds of methylation" could trigger methylation of the entire CGI (15, 16), we further analyzed the presence of fully methylated DNA molecules (Fig. 3A). Since the number of fully methylated DNA molecules was expected to be extremely small, we adopted a strategy of selective amplification of methylated DNA molecules by MSP (25). MSP primers used recognized three CpG sites located 5' to the region analyzed by bisulfite sequencing and three 3' CpG sites within the region (Fig. 3B). Stochastic amplification of methylated DNA molecules was specifically observed for KATOIII (10 positive reactions out of 24 reactions) and AGS (9 positive reactions out of 24 reactions), showing that zero to a few DNA molecules were contained in these reactions. All the PCR products were sequenced, and 16 CpG sites between MSP primers were fully methylated for all the clones sequenced (representative results in Fig. 3B).



View larger version (50K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Emergence of fully methylated DNA molecules and the presence of CIMP in cell lines with decreased fidelity. A, selective amplification of methylated DNA molecules by MSP (40 cycles of PCR). For each sample of bisulfite modified DNA, four reactions (A-D) were performed. Although unmethylated DNA molecules were consistently detected, methylated DNA molecules were stochastically detected only in KATOIII and AGS (arrowheads). This indicated that zero to a few fully methylated DNA molecules were present in 250 to 380 molecules. B, methylation statuses of CpG sites between the MSP primers. From all the 19 PCR products obtained with M set primers, three to four clones were sequenced. All the clones were fully methylated. C, Methylation statuses of eight additional promoter CGIs and five CGIs whose fidelity was analyzed. Presence of only the M band (closed boxes), that of both M and U bands (hatched boxes) and that of the U band only (open boxes) by MSP (33-35 cycles). Whereas HSC39 and HSC57 had accumulation of no and only one, respectively, aberrantly methylated CGIs, KATOIII and AGS had that of five and seven, respectively, aberrantly methylated CGIs. Data already reported (18) and data in this study were combined.

 
Accumulation of aberrant methylation of the entire CGIs was examined, using eight additional promoter CGIs that can be methylated in human gastric cancers (Fig. 3C). None and one CGI were aberrantly methylated in HSC39 and HSC57, respectively, whereas five and seven CGIs were methylated in KATOIII and AGS, respectively. The frequency of aberrant methylation was in accordance with the decreased fidelity in replicating methylation patterns.

Gene Expression Levels of the Analyzed Genes and DNMTs. Considering that gene transcription has been implicated as one of the mechanisms to protect a CGI from de novo methylation (1, 15, 16, 26), expression levels of the five genes were measured by quantitative reverse transcription-PCR (Fig. 4). In KATOIII and AGS, in which decreased fidelity was observed for bA305P22.2.3, but not for FLJ32130, the former gene had higher expression levels than the latter. KATOIII and AGS had RIKEN2210016 expression levels comparable to those of HSC39 and HSC57. Therefore, the decreased fidelity was unlikely to be due to low expression levels.



View larger version (47K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. mRNA expression levels of the five genes whose CGIs were analyzed and three DNMTs. The expression levels were measured quantitatively by real-time reverse transcription-PCR and normalized to PCNA. Expression levels of bA305P22.2.3 and RIKEN2210016 among the cancer cell lines did not correlate with the presence of decreased fidelities. The expression levels of DNMT3B were high in the two cell lines with decreased fidelities, KATOIII, and AGS. DNMT3A was not expressed in KATOIII at all, and the meaning of this is unclear.

 
Most of the errors in the methylation patterns in CGIs were considered to be due to de novo methylation. Therefore, expression levels of two de novo methyltransferases, DNMT3A and DNMT3B, and the maintenance methyltransferase, DNMT1, were also measured. It was shown that DNMT3B was expressed at 4- to 8-fold higher levels in the two cell lines with low fidelities, KATOIII and AGS, and that DNMT3A was not expressed at all in KATOIII.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
It was shown here that KATOIII and AGS gastric cancer cell lines had decreased fidelities in replicating methylation patterns, which was prominent in specific CGIs, such as those of bA305P22.2.3 and RIKEN2210016. The decreased fidelity was mainly due to methylation of unmethylated CpG sites, and led to emergence of fully methylated DNA molecules. This is the first experimental evidence that some cancer cells have decreased fidelity in replicating methylation patterns compared with other cancer cell lines and normal cells (14). Sequencing an average of 12.5 clones for each CGI in each cell line and repeating six experiments corresponded to analysis of 30,398 CpG sites in 1,495 clones,and this enabled to detect statistically significant increases at 1.6- to 3.2-folds.

The number of fully methylated DNA molecules was very small, in the range of zero to a few per reaction tube. For their MSP, 500 ng of genomic DNA, which corresponds to 1.5 x 105 gene copies, were treated with bisulfite, and 1 of 20 of it was used. Our protocol of bisulfite treatment is known to reduce copy numbers to 1 of 20 to 30 due to DNA degradation (data not shown). This calculates as 250 to 375 copies of template DNA having been present in one reaction of MSP, and that the fraction of fully methylated DNA molecules was 0-1/250-375 (0-0.4%). Although the fraction was very small, a cell could gain a growth advantage and expand clonally if the full methylation was induced in a promoter CGI of a tumor suppressor gene. It can be noted that mutations, although their frequency is very low, have a significant impact on carcinogenesis. The decreased fidelity was also considered to be related to the frequent methylation of entire CGIs in KATOIII and AGS.

Decreased fidelity was prominent in the CGIs of bA305P22.2.3 andRIKEN2210016. In AGS, Cyclophilin A was most abundantly expressed, followed by bA305P22.2.3, FLJ32130, RIKEN2210016, and E-cadherin. Although absence of transcription is considered to be important for gene silencing (1, 15, 16, 26), the decreased fidelity in cancer cells was shown to affect even actively transcribed genes like bA305P22.2.3 and FLJ32130. Sp1 binding sites are also known to be important to protect CGIs from being methylated (7–9). There are five putative Sp1 binding sites for both bA305P22.2.3 and RIKEN2210016, whereas there are one, three and four sites for FLJ32130, E-cadherin, and Cyclophilin A, respectively. Therefore, no direct correlation was observed between the number of Sp1 binding sites and the resistance to decreased fidelity. Accordingly, promoter CGIs affected by decreased fidelity in cancers seemed to be determined by some other factors than gene transcription and Sp1 binding sites. The number of chromosomes obtained by FISH did not show any clear correlation with the fidelity, and thus histone modifications or other chromatin structures were likely to be involved.

As for the role of overexpression of DNA methyltransferases in induction of CGI methylation, reports that support the role (27–29) and those that do not (15, 30) are available. Since the decreased fidelity observed in this study was mainly due to high susceptibility to de novo methylation, we examined expression levels of DNA methyltransferases, DNMT3A, DNMT3B, and DNMT1. Although DNMT1 expression was not different in cell lines with and without the decreased fidelity, DNMT3B expression was 4- to 8-fold higher in the two cell lines with the decreased fidelity. Therefore, we could not rule out the overexpression of DNMT3B as a mechanism for the high susceptibility to de novo methylation in KATOIII and AGS, but that of DNMT1 was unlikely.

The error rates obtained here were examined in six independent cultures for each CGI in each cell line, and all the differences obtained were statistically significant and led to appearance of fullymethylated DNA molecules. Possible selection bias of the original methylation patterns was confirmed to be <20% of the total number of errors, as in our previous study (14). However, ourassay system neglects interaction among neighboring CpG sitesfor the purpose of simple calculation, and calculates the number of errors in one replication from the number of errors accumulated during a culture. The number of errors in one replication can be underestimated by the presence of a mechanism that corrects methylation errors, and can be overestimated by the presence of an error in the early phase of culture. A method recently developed by Laird et al. (31) neatly measures methylation errors in one replication, and reports larger amounts of errors than thisreport.

In summary, this study showed that some cancers have decreased fidelity in replicating methylation patterns, mainly due to de novo methylation, that the decreased fidelity was prominent in some specific CGIs, and that the decrease could lead to full methylation of a CGI.


   Acknowledgments
 
Grant support: Ministry of Health, Labour and Welfare, Japan Grant-in-Aid for the Third Term Comprehensive 10-year Strategy for Cancer Control.

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.

We thank Dr. H. Shizuya for providing BAC clones and Drs. E. Okochi-Takada and M. Mihara for their critical discussion.

Received 8/13/03. Revised 9/22/04. Accepted 10/22/04.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Bird A. DNA methylation patterns and epigenetic memory. Genes Dev 2002;16:6–21.[Free Full Text]
  2. Li E. Chromatin modification and epigenetic reprogramming in mammalian development. Nat Rev Genet 2002;3:662–73.[CrossRef][Medline]
  3. Futscher BW, Oshiro MM, Wozniak RJ, et al. Role for DNA methylation in the control of cell type specific maspin expression. Nat Genet 2002;31:175–9.[CrossRef][Medline]
  4. Ohgane J, Hattori N, Oda M, Tanaka S, Shiota K. Differentiation of trophoblast lineage is associated with DNA methylation and demethylation. Biochem Biophys Res Commun 2002;290:701–6.[CrossRef][Medline]
  5. Jaenisch R, Bird A. Epigenetic regulation of gene expression: how the genome integrates intrinsic and environmental signals. Nat Genet 2003;33 Suppl:245–54.
  6. Jones PA, Baylin SB. The fundamental role of epigenetic events in cancer. Nat Rev Genet 2002;3:415–28.[Medline]
  7. Brandeis M, Frank D, Keshet I, et al. Sp1 elements protect a CpG island from de novo methylation. Nature 1994;371:435–8.[CrossRef][Medline]
  8. Macleod D, Charlton J, Mullins J, Bird AP. Sp1 sites in the mouse aprt gene promoter are required to prevent methylation of the CpG island. Genes Dev 1994;8:2282–92.[Abstract/Free Full Text]
  9. Machon O, Strmen V, Hejnar J, Geryk J, Svoboda J. Sp1 binding sites inserted into the rous sarcoma virus long terminal repeat enhance LTR-driven gene expression. Gene 1998;208:73–82.[CrossRef][Medline]
  10. Han L, Lin IG, Hsieh CL. Protein binding protects sites on stable episomes and in the chromosome from de novo methylation. Mol Cell Biol 2001;21:3416–24.[Abstract/Free Full Text]
  11. Wigler M, Levy D, Perucho M. The somatic replication of DNA methylation. Cell 1981;24:33–40.[CrossRef][Medline]
  12. Pfeifer GP, Steigerwald SD, Hansen RS, Gartler SM, Riggs AD. Polymerase chain reaction-aided genomic sequencing of an X chromosome-linked CpG island: methylation patterns suggest clonal inheritance, CpG site autonomy, and an explanation of activity state stability. Proc Natl Acad Sci U S A 1990;87:8252–6.[Abstract/Free Full Text]
  13. Clark SJ, Harrison J, Paul CL, Frommer M. High sensitivity mapping of methylated cytosines. Nucl Acids Res 1994;22:2990–7.[Abstract/Free Full Text]
  14. Ushijima T, Watanabe N, Okochi E, Kaneda A, Sugimura T, Miyamoto K. Fidelity of the methylation pattern and its variation in the genome. Genome Res 2003;13:868–74.[Abstract/Free Full Text]
  15. Song JZ, Stirzaker C, Harrison J, Melki JR, Clark SJ. Hypermethylation trigger of the glutathione-S-transferase gene (GSTP1) in prostate cancer cells. Oncogene 2002;21:1048–61.[CrossRef][Medline]
  16. Stirzaker C, Song JZ, Davidson B, Clark SJ. Transcriptional gene silencing promotes DNA hypermethylation through a sequential change in chromatin modifications in cancer cells. Cancer Res 2004;64:3871–7.[Abstract/Free Full Text]
  17. Toyota M, Ahuja N, Ohe-Toyota M, Herman JG, Baylin SB, Issa JP. CpG island methylator phenotype in colorectal cancer. Proc Natl Acad Sci U S A 1999;96:8681–6.
  18. Kaneda A, Kaminishi M, Yanagihara K, SugimuraT, Ushijima T. Identification of silencing of nine genes in human gastric cancers. Cancer Res 2002;62:6645–50.[Abstract/Free Full Text]
  19. Yamashita K, Dai T, Dai Y, Yamamoto F, Perucho M. Genetics supersedes epigenetics in colon cancer phenotype. Cancer Cell 2003;4:121–31.[CrossRef][Medline]
  20. Tamura G, Yin J, Wang S, et al. E-Cadherin gene promoter hypermethylation in primary human gastric carcinomas. J Natl Cancer Inst 2000;92:569–73.[Abstract/Free Full Text]
  21. Harnden DG, Klinger HP. An international system for human cytogenetic nomenclature. Basel: Karger; 1985.
  22. Rein T, Zorbas H, DePamphilis ML. Active mammalian replication origins are associated with a high-density cluster of mCpG dinucleotides. Mol Cell Biol 1997;17:416–26.[Abstract]
  23. Herman JG, Graff JR, Myohanen S, Nelkin BD, BaylinSB. Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands. Proc Natl Acad SciU S A 1996;93:9821–6.[Abstract/Free Full Text]
  24. Fleisher AS, Esteller M, Wang S, et al. Hypermethylation of the hMLH1 gene promoter in human gastric cancers with microsatellite instability. Cancer Res 1999;59:1090–5.[Abstract/Free Full Text]
  25. Suter CM, Martin DI, Ward RL. Germline epimutation of MLH1 in individuals with multiple cancers. Nat Genet 2004;36:497–501.[CrossRef][Medline]
  26. De Smet C, Loriot A, Boon T. Promoter-dependent mechanism leading to selective hypomethylation within the 5' region of gene MAGE-A1 in tumor cells. Mol Cell Biol 2004;24:4781–90.[Abstract/Free Full Text]
  27. Ahluwalia A, Hurteau JA, Bigsby RM, Nephew KP. DNA methylation in ovarian cancer. II. Expression of DNA methyltransferases in ovarian cancer cell lines and normal ovarian epithelial cells. Gynecol Oncol 2001;82:299–304.[CrossRef][Medline]
  28. Vertino PM, Yen RW, Gao J, Baylin SB. De novo methylation of CpG island sequences in human fibroblasts overexpressing DNA (cytosine-5-)-methyltransferase. Mol Cell Biol 1996;16:4555–65.[Abstract]
  29. Biniszkiewicz D, Gribnau J, Ramsahoye B, et al. Dnmt1 overexpression causes genomic hypermethylation, loss of imprinting, and embryonic lethality. Mol Cell Biol 2002;22:2124–35.[Abstract/Free Full Text]
  30. Eads CA, Danenberg KD, Kawakami K, Saltz LB, Danenberg PV, Laird PW. CpG island hypermethylation in human colorectal tumors is not associated with DNAmethyltransferase overexpression. Cancer Res 1999;59:2302–6.[Abstract/Free Full Text]
  31. Laird CD, Pleasant ND, Clark AD, et al. Hairpin-bisulfite PCR: assessing epigenetic methylation patterns on complementary strands of individual DNA molecules. Proc Natl Acad Sci U S A 2004;101:204–9.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Mol Cancer ResHome page
R. Nishiyama, L. Qi, M. Lacey, and M. Ehrlich
Both Hypomethylation and Hypermethylation in a 0.2-kb Region of a DNA Repeat in Cancer
Mol. Cancer Res., November 1, 2005; 3(11): 617 - 626.
[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 Ushijima, T.
Right arrow Articles by Kaneda, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ushijima, T.
Right arrow Articles by Kaneda, A.

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