| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Biology and Genetics |
Departments of Pathology [N. S., A. M., N. F., N. T. v. H., H. M., C. A. I-D., C. R., M. G.], Oncology [M. G.], and Medicine [M. G.], The Johns Hopkins Medical Institutions, Baltimore, Maryland 21205
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, trefoil factor2, S100A4, mesothelin, and prostate stem cell antigen) that were overexpressed in the neoplastic cells of pancreatic cancers and not expressed in normal pancreatic duct displayed a high prevalence of hypomethylation in pancreatic cancer cell lines and primary pancreatic carcinomas. By contrast, only 1 of 12 genes not overexpressed in pancreatic cancer demonstrated hypomethylation (P = 0.0002). In pancreatic cancer cell lines that retained methylation of 1 or more of the 7 aforementioned overexpressed and hypomethylated genes, treatment with 5-aza-2'-deoxycytidine or with trichostatin A, either alone or in combination, almost invariably reactivated the transcription of each of these 7 genes. These results indicate that gene hypomethylation is a frequent epigenetic event in pancreatic cancer and is commonly associated with the overexpression of affected genes. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The importance of DNA methylation in the transcriptional silencing of cancer-associated genes is increasingly recognized (10)
. For example, a variety of tumor suppressor, growth regulatory, and mismatch repair genes, including those encoding p16, preproenkephalin, and hMLH1 are inactivated by promoter region hypermethylation in benign and malignant pancreatic neoplasms (11, 12, 13, 14)
. Genome-wide hypomethylation has also been described in various human cancers (15, 16, 17)
and has been associated with genetic instability characterized by chromosomal aberrations or elevated mutation rates (18
, 19)
. The genome-wide DNA hypomethylation found in cancers is prevalent in normally methylated repeat sequences (reviewed in Ref. 20
). DNA hypomethylation at unique sequences could also increase expression of cancer-promoting genes (20)
. In fact, a correlation between site-specific hypomethylation and transcriptional activation has been observed for several genes such as MAGE (21
, 22)
, S100A4 (23)
, synuclein 
(24)
, and other genes (20)
. In addition, the extent of hypomethylation appears to correlate with tumor grade and with prognosis in certain cancers (25)
. These findings suggest an important role of DNA hypomethylation in the overexpression of genes in human cancers, but an understanding of the global pattern of hypomethylation at multiple loci has been limited by the few genes analyzed to date. To address this issue, we analyzed the methylation status of a panel of genes that are expressed at higher levels in ductal adenocarcinoma of the pancreas and not known to be expressed in normal pancreatic tissue and compared these methylation patterns to those of a group of genes that are not overexpressed in pancreatic cancer.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
cDNA Microarray and Data Analysis.
A comprehensive analysis of global gene expression profiles of pancreatic cancer using cDNA microarrays and significance analysis of microarrays was performed as described elsewhere (9)
. This study involved the analysis of 31 pancreatic cancers, including 14 pancreatic cancer cell lines and 17 primary pancreatic cancers, as well as 5 samples of normal pancreas tissue.
Sequencing Analysis of Bisulfite-treated Genomic DNA.
Genomic DNA was isolated from the cell lines and frozen tissues using a DNA isolation kit (Qiagen, Valencia, CA) and subjected to sodium bisulfite treatment. Sequencing of bisulfite-treated DNA samples was performed as described previously (12)
.
MSP.
The methylation status of each gene was also determined by MSP as described previously (12)
. Primers were designed to detect the sequence differences between methylated and unmethylated DNA as a result of bisulfite modification, and each primer pair contained at least three CpG sites to provide optimal specificity. Because knowledge of the regulatory regions of many of the genes overexpressed in pancreatic cancer is not known, we decided to characterize the methylation status of CpGs within a few hundred bp of the transcriptional start site, where CpG methylation has been implicated in transcriptional silencing (27)
. The primer sequences for bisulfite sequencing and MSP are available upon request.
Oligonucleotide Array Hybridization and Data Analysis.
Total RNA was isolated from cultured cells using Trizol reagent (Invitrogen, Carlsbad, CA). First- and second-stranded cDNA were synthesized from 10 µg of total RNA using T7-(dT)24 primer (Genset Corp., South La Jolla, CA) and SuperScript Choice system (Invitrogen). Labeled cRNA was synthesized from the purified cDNA by in vitro transcription reaction using the RNA Transcript Labeling kit (Enzo Diagnostics, Inc., Farmingdale, NY) at 37°C for 6 h and was purified using RNeasy Mini Kit (Qiagen). The cRNA was fragmented at 94°C for 35 min in a fragmentation buffer [40 mM Tris-acetate (pH 8.1), 100 mM potassium acetate, and 30 mM magnesium acetate]. The fragmented cRNA was then hybridized to the Human Genome U133A chips (Affymetrix, Santa Clara, CA) with 18,462 unique gene/expressed sequence tag transcripts at 45°C for 16 h. The probes were then scanned using a laser scanner, and signal intensity for each transcript (background-subtracted and adjusted for noise) and detection call (present, absent, or marginal) were determined using Microarray Suite Software 5.0 (Affymetrix).
Semiquantitative RT-PCR.
RT-PCR was performed with primers for the specific genes (the primer sequences are available upon request) and for GAPDH in duplex reactions. The range of linear amplification for each gene and the GAPDH gene was examined with serial PCR cycles, and the optimal PCR cycles were determined. The relative intensity of mRNA expression for each sample was then corrected for variable RNA recovery using the corresponding GAPDH mRNA measurement as a surrogate for total mRNA.
Inhibition of DNA Methylation and Histone Deacetylase Inhibition.
We treated two pancreatic cancer cell lines (MiaPaCa2 and Hs766T) with 5Aza-dC (Sigma, St. Louis, MO) and TSA (Sigma), either alone or in combination. Cells were exposed continuously to 5Aza-dC (1 µM) for 4 days or TSA (1 µM) for 24 h. For combined treatment, these cells were cultured in the presence of 5Aza-dC (1 µM) for 3 days and were then treated for another 24 h with TSA (0.5 µM).
Statistical Analysis.
Statistical analysis was performed using Mann-Whitney U nonparametric test, Student’s t test, or Fisher’s exact probability test. Differences were considered significant at P < 0.05.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. We selected a previously identified a panel of 18 genes for which one or more of evidence [including oligonucleotide array analysis (8)
, cDNA microarray analysis (9)
, and SAGE analysis (7)
] suggested that they were significantly overexpressed in pancreatic cancer. The 18 genes were claudin4, LCN2, mesothelin, PSCA, S100A4, 14-3-3, TFF2, BarH-like homeobox1, CDH11, fer-1-like3 (FER1L3, myoferlin), amphiregulin, caveolin2, cornichon-like, inhibin, mal, met proto-oncogene (HGF receptor), thyroid hormone receptor interactor 13, and tumor growth factor ß-induced (Table 1)
. In choosing these 18 genes, we aimed to identify genes that were not only overexpressed in pancreatic cancers but were also not normally expressed in the pancreas. We used this criterion because we hypothesized that genes that were normally methylated at 5' CpGs would be more likely to be transcriptionally silent, and the transcriptional activation of these genes would be more likely to be associated with hypomethylation.
|
|
|
. We have previously reported immunohistochemical analyses demonstrating overexpression of mesothelin (28)
, PSCA (29)
, S100A4 (23)
, and 14-3-3 (9)
in neoplastic cells of primary pancreatic carcinomas, in contrast to little or no expression in normal ductal epithelium. We next performed RT-PCR analyses on 12 genes (claudin4, LCN2, mesothelin, PSCA, S100A4, 14-3-3, TFF2, BarH-like homeobox1, CDH11, FER1L3, caveolin2, and tumor growth factor ß-induced) in a panel of 20 pancreatic cancer cell lines as compared with a nonneoplastic epithelial cell line, HPDE. This analysis revealed overexpression of claudin4, LCN2, S100A4, and TFF2 (Ref. 9
and data not shown). Several genes such as mesothelin, PSCA, and 14-3-3 were expressed in most pancreatic cancer cell lines and also expressed in HPDE by RT-PCR but were not expressed in nonneoplastic pancreatic epithelium by immunohistochemistry (9
, 28
, 29)
. Previous reports from other groups have shown that claudin4 (30)
, LCN2 (31)
, TFF2 (32)
, amphiregulin (33)
, and met proto-oncogene (34
, 35)
are overexpressed in pancreatic cancer. On the basis of these findings, we were able to confirm that 9 of 18 genes demonstrated overexpression in pancreatic cancer cells and had little or no expression in normal pancreatic ductal epithelium (Table 1)
.
For comparison, we also studied the methylation pattern of 14 additional genes not known to be overexpressed in pancreatic cancer (Table 1)
. These genes include CD8
, CD8ß, ITGAL (CD11a), CD21, oxytocin receptor gene, proopiomelanocortin, monocarboxylate transporter 3, paired box gene 5, lactate dehydrogenase B, LEP503, cytochrome P450 (CYP19), uncoupling protein homologue, uroplakin 1B, and thyroid transcription factor-1. These 14 genes were chosen for methylation analysis because of their limited tissue distribution and cell lineage-specific expression, and many of them have been previously described as harboring methylation in normal tissues (36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48)
. Gene expression data from oligonucleotide or cDNA microarray and SAGE databases showed no evidence for overexpression of any of these 14 genes in pancreatic cancer (data not shown).
Methylation Analysis by Bisulfite Sequencing of Genes Overexpressed in Pancreatic Cancer.
To determine the methylation pattern of genes overexpressed in pancreatic cancer, bisulfite genomic sequencing was performed on 4 selected genes (claudin4, mesothelin, PSCA, and 14-3-3) that were confirmed as being overexpressed in pancreatic cancer. For each gene, we analyzed a region regarded as important for gene regulation by methylation (from 
-500 to +1 with respect to the transcription start site) in DNA samples from 10 pancreatic cancer cell lines, as well as 5 normal pancreatic tissues. An example of bisulfite sequencing of mesothelin is shown in Fig. 2A
. In all DNA samples from normal pancreata, the CpG sites within the 5' region of mesothelin were largely (60–90%) methylated. In contrast, a subset of cell lines, including AsPC1, were completely unmethylated at all of the CpG sites. Identical results were obtained for claudin4, PSCA, and 14-3-3, i.e., all normal pancreatic tissues were largely or completely methylated, but a significant proportion of pancreatic cancer cell lines were completely unmethylated.
|
. Of note, normal pancreata that were microdissected manually showed amplification of both methylated and unmethylated alleles for several genes, including mesothelin, TFF2, and claudin4, whereas most of the 7 samples of duct epithelia microdissected by LCM were almost completely methylated (Fig. 2B)
. Despite the predominant methylation in these LCM-microdissected ductal cells, some of the genes (claudin4, LCN2, 14-3-3, and TFF2) were unmethylated in an immortal human pancreatic duct epithelial cell line (HPDE). Consistent with this, claudin4, 14-3-3, and TFF2 mRNA were expressed at variable levels in HPDE by RT-PCR (data not shown). A possible explanation for this discrepancy is that the methylation status of these genes may have been altered during immortalization with the E6 and E7 genes of papillomavirus.
Methylation Analysis of Multiple Genes in Pancreatic Cancer Cell Lines.
The 19 genes that harbored methylation of 5' CpGs in normal pancreas were then tested for hypomethylation in a panel of 20 pancreatic cancer cell lines. On the basis of the results from normal pancreatic duct epithelium, we defined a gene as hypomethylated if only the unmethylated templates were amplified. MSP analyses revealed that 8 of 19 genes showed hypomethylation in pancreatic cancer cell lines (Fig. 2C)
. These included 14–3-3 (hypomethylated in 85% of 20 cell lines), claudin4 (85%), LCN2 (85%), TFF2 (65%), S100A4 (50%), mesothelin (40%), PSCA (30%), and CDH11 (20%). Thus, among the 19 genes normally methylated in the pancreas, all 7 of the genes that we were able to confirm as overexpressed in pancreatic cancer showed hypomethylation in pancreatic cancer cell lines. By contrast, only 1 (CDH11) of 12 genes not overexpressed in pancreatic cancer showed hypomethylation and then only in 4 of 20 pancreatic cancer cell lines (P = 0.0002, Fisher’s exact test).
To determine the pattern of this epigenetic alteration for each cell line, we investigated the hypomethylation profile of these 8 genes in 20 pancreatic cancer cell lines (Fig. 3)
. Overall, 19 of 20 (95%) pancreatic cancer cell lines showed hypomethylation in at least one gene, and 18 of 20 (90%) had hypomethylation in 2 or more of the target genes. These results suggest that DNA hypomethylation of overexpressed genes is a frequent epigenetic event in pancreatic cancer cell lines. One pancreatic cancer cell line (Hs766T) that lacked hypomethylated loci is notable for a gene expression profile that closely resembles that of normal ductal epithelium by hierarchical cluster analysis (7)
.
|
(P = 0.3) or mesothelin (P = 0.1; Fig. 4
). For both 14-3-3 and mesothelin, we did observe a trend toward an inverse correlation between methylation and expression, but several cell lines that were unmethylated at these genes lacked expression, suggesting that other mechanisms besides hypomethylation were responsible for this lack of expression (49
, 50)
. Because mRNA expression of CDH11 was detected in only 1 of 18 cell lines tested, we were unable to confirm any correlation between hypomethylation of CDH11 and its expression.
|
, with the DNA methyltransferase inhibitor (5Aza-dC) and/or the histone deacetylase inhibitor (TSA). Treatment of MiaPaCa2 cells, which harbored methylated 14-3-3, with 5Aza-dC (1 µM for 4 days) led to partial loss of methylation at 5' CpG sites of this gene and combined treatment with 5Aza-dC and TSA resulted in almost complete loss of methylation, as evidenced by MSP (Fig. 5A)
. Notably, TSA treatment alone also induced a slight loss of methylation at this gene locus. We also observed that treatment of MiaPaCa2 and Hs766T with 5Aza-dC but not with TSA led to loss of methylation at various degrees at other 6 gene loci (data not shown). We then determined the change in expression of 14-3-3 in MiaPaCa2 cells after treatment with 5Aza-dC and/or TSA using quantitative oligonucleotide microarray analysis (Affymetrix U133A chips). The 14-3-3 transcript was not detectable (called absent) in mock PBS-treated MiaPaCa2 cells, which is consistent with our RT-PCR result on this cell line (data not shown). Treatment with 5Aza-dC led to a substantial increase (35-fold) in the 14-3-3 expression, whereas TSA treatment slightly increased the expression value (Fig. 5B)
. Furthermore, combined treatment with 5Aza-dC and TSA resulted in synergistic induction (80-fold) of 14-3-3 expression, suggesting that DNA methylation and histone deacetylation may be cooperatively associated with the transcriptional silencing of this gene.
|
. Similar to the finding for 14-3-3, many of the other genes (including claudin4, LCN2, and TFF2) were reactivated synergistically after combined treatment with 5Aza-dC and TSA. These results suggest that the combination of DNA methylation and histone deacetylation may be an important mechanism controlling the transcriptional activity of these hypomethylated genes.
Hypomethylation of Multiple Genes in Primary Pancreatic Carcinomas and in Pancreatic Cancer Xenografts.
To test whether the aberrant hypomethylation detected in pancreatic cancer cell lines is also present in primary tumors, normal and tumor tissues were separately microdissected from 8 cases with invasive pancreatic adenocarcinoma and analyzed for hypomethylation of 7 genes (claudin4, LCN2, 14-3-3, TFF2, S100A4, mesothelin, and PSCA) by MSP. In all of the 7 genes, hypomethylation was frequently (75–100%) detected in these primary pancreatic cancers compared with their normal counterparts (Fig. 6A)
. When the relative methylation ratio (the relative proportion of methylated alleles against total intensity of unmethylated and methylated alleles) was measured by densitometry, the mean methylation ratio of tumor DNA was lower than that of the corresponding normal tissue for claudin4 (tumor versus normal, 19 versus 69%, P = 0.002; Student’s t test), LCN2 (15 versus 60%, P = 0.004), 14-3-3 (19 versus 59%, P < 0.0001), TFF2 (16 versus 61%, P = 0.0002), S100A4 (14 versus 55%, P = 0.008), mesothelin (31 versus 74%, P = 0.007), and PSCA (51 versus 64%, P = 0.2). These findings indicate that hypomethylation of these genes occurs in primary tumors as well as in cancer cell lines.
|
, 92% for LCN2, 92% for mesothelin, 89% for claudin 4, 84% for TFF2, 76% for S100A4, 54% for PSCA, and 0% for CDH11 (Fig. 6B)
. A gene hypomethylation profile of pancreatic cancer xenografts is summarized in Fig. 7
. Aberrant hypomethylation of at least 1 of 7 loci was detected in virtually all (36 of 37, 97%) pancreatic cancer xenografts. Notably, one cancer xenograft (PX195) that harbored no DNA hypomethylation at the 8 genes tested was derived from spindle cell carcinoma of the pancreas, suggesting that the extent of DNA hypomethylation in a cancer may vary between certain histological subtypes. Alternatively, this cancer may have a distinct pattern of gene expression that gave rise to hypomethylation of a different set of affected genes.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In examining the hypomethylation patterns of multiple genes in pancreatic cancer, one question we addressed was whether such gene hypomethylation in cancer is specific to certain genes or merely the result of genome-wide hypomethylation. Previous investigators found a correlation between MAGE-1 gene activation and overall 5'-methylcytosine content in a series of tumor cell lines, suggesting that genome-wide hypomethylation might be responsible for the aberrant hypomethylation of MAGE-1 in cancer (22) . However, we found hypomethylation in pancreatic cancers was not random among genes methylated in normal pancreas but was largely restricted to those genes overexpressed in pancreatic cancer and not expressed in normal pancreatic epithelium. These findings provide evidence that aberrant gene hypomethylation in cancer is not simply a random by product of genome-wide hypomethylation.
Our data also showed that the genes that are overexpressed and hypomethylated in pancreatic cancer appeared to have less CpG dinucleotides in their 5' regions (Table 1)
. In fact, none of the 7 overexpressed and hypomethylated genes contains CpG islands in their upstream sequences. These findings raise the possibility that low-density CpG dinucleotides in the 5' region of the gene are more susceptible to hypomethylation than CpG islands. Confirmation of this finding will require analysis of a large number of genes.
Two pancreatic cancer cell lines harbored methylation at the 7 genes that were hypomethylated in most other pancreatic cancer cell lines, and treatment of these two cell lines with a combination of 5Aza-dC and TSA synergistically restored gene expression, suggesting that these genes are regulated by both methylation and histone deacetylation. That TSA alone induced the re-expression of methylated genes supports previous studies suggesting that the state of histone acetylation can influence gene expression. Indeed, in some experimental systems, histone acetylation may also influence the methylation of a gene (51, 52, 53, 54)
, and in our present study, we also observed that in some cases, TSA treatment led to a partial loss of methylation (Fig. 5)
. In contrast, it has been shown that TSA alone is unable to reactivate several tumor suppressor genes with densely methylated CpG islands (55)
. A possible mechanism responsible for the different responses to TSA treatment may relate to the CpG density in the 5' regions of methylated genes; histone deacetylase inhibition induced the transcription of genes having a few methylated CpGs but may be insufficient to reactivate the transcription of genes with densely methylated CpG islands.
By profiling hypomethylation patterns in >50 pancreatic cancers, we found that aberrant DNA hypomethylation in pancreatic cancer usually occurred at 5' CpGs of several overexpressed genes. On the other hand, we also found a few pancreatic cancer cell lines with little or no DNA hypomethylation at these genes, suggesting either that these cancers tend not to develop DNA hypomethylation or that different loci are targeted for hypomethylation in these cancers. In fact, one pancreatic cancer cell line (Hs766T), which lacked DNA hypomethylation at any of the genes we tested, has been characterized as having an atypical gene expression profile (a normoid gene expression profile) by hierarchical cluster analysis (7) . The genetic basis for the normoid gene expression pattern is not known because it has a similar gene mutation pattern to other pancreatic cancers; although Hs766T lacks mutation of p53, it harbors mutated K-ras at codon 61, as well as homozygous deletion of p16 and DPC4 (56) .
In considering how DNA hypomethylation may occur during tumor development, two mechanisms warrant consideration. One possibility is that DNA hypomethylation of a gene in a cancer is a remnant of a hypomethylated gene in a pancreatic epithelial cell that evolved into a cancer cell. Although normal pancreatic epithelium was predominantly methylated at each of the 8 genes found to be hypomethylated in pancreatic cancer, we found some variability in the level of methylation in the samples of LCM nonneoplastic pancreatic epithelium from seven individuals. In addition, 100% methylation was not always observed in these normal pancreatic epithelia. Therefore, it is possible that some pancreatic epithelial cells harboring hypomethylation at a gene locus before the onset of neoplasia could retain that hypomethylation throughout neoplastic evolution. However, we do not believe that this is the likely explanation for the gene hypomethylation we observe in pancreatic cancers. First, for several of the loci, the percentage of pancreatic cancers with hypomethylation was 90%, whereas normal pancreas epithelium was predominantly methylated. Second, because we found that methylation status is associated with the expression of each of the genes tested, it is likely that if hypomethylation originated in a nonneoplastic pancreatic epithelial cell that underwent selection because of other genetic events, one would expect to find that the earliest neoplasms (PanINs) express these genes (57)
. In contrast, we find that the expression of genes such as S100A4, PSCA, and mesothelin only becomes manifest in late-stage PanINs or in the carcinomas (23
, 28
, 29)
. The late expression of genes such as mesothelin and others in PanINs is indirect evidence that hypomethylation is typically not an early feature of pancreatic neoplastic development. Definitive answers to this question will require characterization of hypomethylation patterns of PanINs.
An alternative mechanism for the DNA hypomethylation observed in pancreatic cancer is the occurrence of loss of DNA methylation during tumorigenesis. Little is known about the factors that lead to loss of CpG methylation. As is the case for aberrant DNA hypermethylation in cancer, it is not certain at this time if gene-related methylated changes are a cause or consequence of altered transcriptional activity in cancer cells (58) . Recently, Di Croce et al. (53) showed that methylation-associated silencing of retinoic acid receptor ß2 occurred as a result of binding of oncogenic promyelocytic leukemia-retinoic acid receptor fusion protein to its promoter, whereas the addition of retinoic acid led to partial loss of methylation. Similarly, Bachman et al. (59) demonstrated, in a model system where DNA methyltransferase genes are disrupted in a colorectal cancer cell line, that methylation of histone H3 lysine-9 at the p16 gene locus occurred in conjunction with re-silencing of the gene in the absence of DNA methylation, suggesting that gene silencing because of histone methylation might lead to subsequent DNA methylation. In addition, Hoffman et al. (54) demonstrated that treatment of mouse fibroblasts with a HDAC inhibitor led to partial loss of methylation at the imprinted insulin-like growth factor-2 receptor gene. Cervoni and Szyf (51) have demonstrated that addition of methyl-CpG-binding domain protein-2 and acetylation of histones can induce the demethylation of ectopically methylated DNA. These results and our findings of selective gene hypomethylation raise the possibility that hypomethylation of transcriptionally activated genes in cancer may occur as a consequence of genetic or other events that alter the transcriptional activities of affected promoters.
Aberrant hypomethylation and overexpression of genes such as PSCA and S100A4 may be functionally important in the progression of pancreatic cancer. Overexpression of these genes has been implicated in tumorigenesis and progression of other cancers (60 , 61) . Importantly, we found that these genes when silenced by methylation became expressed after treatment with 5Aza-dC or TSA. Therefore, treatment of cancers methylated at these loci with DNA methyltransferase inhibitors or HDAC inhibitors could result in accelerated tumor progression rather than in growth suppression caused by re-expression of silenced tumor suppressor genes (62 , 63) . Indeed, we have recently found that treatment of pancreatic cancer cell lines with 5Aza-dC resulted in an increase in their invasive properties in association with an enhanced expression of several matrix metalloproteinases (64) . Methylation or gene expression profiling may help to predict the cancers likely to respond to HDAC inhibitors and demethylating agents.
We analyzed hypomethylation patterns in cancer cell lines because it permitted facile comparison of methylation and gene expression patterns in a panel of cancers. In contrast, when primary cancers are studied, it is necessary to completely microdissect the cancer from surrounding stroma, and determining RNA expression on the same primary tissues is difficult. However, aberrant hypermethylation of CpG islands detected in cancer cell lines cannot always be found in primary cancers (65) , thus, it is important to confirm such methylation patterns in primary cancers. Failure to find aberrant methylation in primary cancers may arise for several reasons, including failure to use microdissected cancers, tumor heterogeneity, or changes in methylation arising in culture. We have previously reported that the vast majority of aberrantly hypermethylated genes detected in cancer cell lines could also be found in the primary cancers from which they were derived (66) . Similarly, Suzuki et al. (67) were able to confirm many of the aberrant hypermethylation patterns first identified in a colon cancer cell line in a panel of primary colorectal cancers. In this study, we were able to confirm that the hypomethylation patterns identified in pancreatic cancer cell lines were also present in microdissected primary pancreatic cancers, as well as in xenografts of primary pancreatic cancers for all 7 genes tested.
In summary, we demonstrate that gene hypomethylation is a common event in pancreatic adenocarcinoma and suggest that hypomethylation associated with overexpression of multiple genes contributes to neoplastic progression.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This work is supported by the SPORE in Gastrointestinal Malignancies Grant CA62924, the Grant CA90709, and the Michael Rolfe Foundation. 
2 To whom requests for reprints should be addressed, at Departments of Pathology, Medicine, and Oncology, The Johns Hopkins Medical Institutions, 632 Ross Building, 720 Rutland Avenue, Baltimore, MD. Phone: (410) 955-3511; Fax: (410) 614-0671; E-mail: mgoggins{at}jhmi.edu
3 The abbreviations used are: SAGE, serial analysis of gene expression; LCM, laser capture microdissection; MSP, methylation-specific PCR; RT-PCR, reverse transcription-PCR; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; 5Aza-dC, 5-aza-2'-deoxycytidine; TSA, trichostatin A; PSCA, prostate stem cell antigen; TFF2, trefoil factor 2; CDH11, cadherin11; LCN2, lipocalin2; HDAC, histone deacetylase.
Received 8/19/02. Accepted 5/12/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
gene CpG island promotes its aberrant expression in breast carcinoma and ovarian carcinoma. Cancer Res., 63: 664-673, 2003. is a p53-regulated inhibitor of G2-M progression. Mol. Cell, 1: 3-11, 1997.[Medline]
This article has been cited by other articles:
|
|
 |
|
  T. Arumugam, V. Ramachandran, K. F. Fournier, H. Wang, L. Marquis, J. L. Abbruzzese, G. E. Gallick, C. D. Logsdon, D. J. McConkey, and W. Choi Epithelial to Mesenchymal Transition Contributes to Drug Resistance in Pancreatic Cancer Cancer Res., July 15, 2009; 69(14): 5820 - 5828. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Neupane and M. Korc 14-3-3{sigma} Modulates Pancreatic Cancer Cell Survival and Invasiveness Clin. Cancer Res., December 1, 2008; 14(23): 7614 - 7623. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Cui, L.-Y. Xu, Z.-Y. Shen, Q. Tao, S.-Y. Gao, Z. Lv, Z.-P. Du, W.-K. Fang, and E.-M. Li NGALR Is Overexpressed and Regulated by Hypomethylation in Esophageal Squamous Cell Carcinoma Clin. Cancer Res., December 1, 2008; 14(23): 7674 - 7681. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Yegnasubramanian, M. C. Haffner, Y. Zhang, B. Gurel, T. C. Cornish, Z. Wu, R. A. Irizarry, J. Morgan, J. Hicks, T. L. DeWeese, et al. DNA Hypomethylation Arises Later in Prostate Cancer Progression than CpG Island Hypermethylation and Contributes to Metastatic Tumor Heterogeneity Cancer Res., November 1, 2008; 68(21): 8954 - 8967. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. I. Joensuu, W. M. Abdel-Rahman, M. Ollikainen, S. Ruosaari, S. Knuutila, and P. Peltomaki Epigenetic Signatures of Familial Cancer Are Characteristic of Tumor Type and Family Category Cancer Res., June 15, 2008; 68(12): 4597 - 4605. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. P. Feinberg Epigenetics at the Epicenter of Modern Medicine JAMA, March 19, 2008; 299(11): 1345 - 1350. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Feldmann and A. Maitra Molecular Genetics of Pancreatic Ductal Adenocarcinomas and Recent Implications for Translational Efforts J. Mol. Diagn., March 1, 2008; 10(2): 111 - 122. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Li, U. Bharadwaj, R. Zhang, S. Zhang, H. Mu, W. E. Fisher, F. C. Brunicardi, C. Chen, and Q. Yao Mesothelin is a malignant factor and therapeutic vaccine target for pancreatic cancer Mol. Cancer Ther., February 1, 2008; 7(2): 286 - 296. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Abe, N. Fukushima, K. Brune, C. Boehm, N. Sato, H. Matsubayashi, M. Canto, G. M. Petersen, R. H. Hruban, and M. Goggins Genome-Wide Allelotypes of Familial Pancreatic Adenocarcinomas and Familial and Sporadic Intraductal Papillary Mucinous Neoplasms Clin. Cancer Res., October 15, 2007; 13(20): 6019 - 6025. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. M. Brena and J. F. Costello Genome-epigenome interactions in cancer Hum. Mol. Genet., April 15, 2007; 16(R1): R96 - R105. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Herceg Epigenetics and cancer: towards an evaluation of the impact of environmental and dietary factors Mutagenesis, March 1, 2007; 22(2): 91 - 103. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Jimeno, B. Rubio-Viqueira, M. L. Amador, V. Grunwald, A. Maitra, C. Iacobuzio-Donahue, and M. Hidalgo Dual mitogen-activated protein kinase and epidermal growth factor receptor inhibition in biliary and pancreatic cancer Mol. Cancer Ther., March 1, 2007; 6(3): 1079 - 1088. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Khalkhali-Ellis Maspin: The New Frontier Clin. Cancer Res., December 15, 2006; 12(24): 7279 - 7283. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D.-F. Peng, Y. Kanai, M. Sawada, S. Ushijima, N. Hiraoka, S. Kitazawa, and S. Hirohashi DNA methylation of multiple tumor-related genes in association with overexpression of DNA methyltransferase 1 (DNMT1) during multistage carcinogenesis of the pancreas Carcinogenesis, June 1, 2006; 27(6): 1160 - 1168. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Larsson, N. Hakansson, E. Giovannucci, and A. Wolk Folate intake and pancreatic cancer incidence: a prospective study of Swedish women and men. J Natl Cancer Inst, March 15, 2006; 98(6): 407 - 413. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Rodenhiser and M. Mann Epigenetics and human disease: translating basic biology into clinical applications Can. Med. Assoc. J., January 31, 2006; 174(3): 341 - 348. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Goggins Molecular Markers of Early Pancreatic Cancer J. Clin. Oncol., July 10, 2005; 23(20): 4524 - 4531. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Li, M. Ahmed, Y. Li, L. Jiao, T.-H. Chou, R. A. Wolff, R. Lenzi, D. B. Evans, M. L. Bondy, P. W. Pisters, et al. 5,10-Methylenetetrahydrofolate Reductase Polymorphisms and the Risk of Pancreatic Cancer Cancer Epidemiol. Biomarkers Prev., June 1, 2005; 14(6): 1470 - 1476. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Nishigaki, K. Aoyagi, I. Danjoh, M. Fukaya, K. Yanagihara, H. Sakamoto, T. Yoshida, and H. Sasaki Discovery of Aberrant Expression of R-RAS by Cancer-Linked DNA Hypomethylation in Gastric Cancer Using Microarrays Cancer Res., March 15, 2005; 65(6): 2115 - 2124. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Ogishima, H. Shiina, J. E. Breault, L. Tabatabai, W. W. Bassett, H. Enokida, L.-C. Li, T. Kawakami, S. Urakami, L. A. Ribeiro-Filho, et al. Increased Heparanase Expression Is Caused by Promoter Hypomethylation and Up-Regulation of Transcriptional Factor Early Growth Response-1 in Human Prostate Cancer Clin. Cancer Res., February 1, 2005; 11(3): 1028 - 1036. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Guweidhi, J. Kleeff, N. Giese, J. E. Fitori, K. Ketterer, T. Giese, M. W. Buchler, M. Korc, and H. Friess Enhanced expression of 14-3-3sigma in pancreatic cancer and its role in cell cycle regulation and apoptosis Carcinogenesis, September 1, 2004; 25(9): 1575 - 1585. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Widschwendter, G. Jiang, C. Woods, H. M. Muller, H. Fiegl, G. Goebel, C. Marth, E. Muller-Holzner, A. G. Zeimet, P. W. Laird, et al. DNA Hypomethylation and Ovarian Cancer Biology Cancer Res., July 1, 2004; 64(13): 4472 - 4480. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. A. Puolakkainen, R. A. Brekken, S. Muneer, and E. H. Sage Enhanced Growth of Pancreatic Tumors in SPARC-Null Mice Is Associated With Decreased Deposition of Extracellular Matrix and Reduced Tumor Cell Apoptosis Mol. Cancer Res., April 1, 2004; 2(4): 215 - 224. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Rump, Y. Morikawa, M. Tanaka, S. Minami, N. Umesaki, M. Takeuchi, and A. Miyajima Binding of Ovarian Cancer Antigen CA125/MUC16 to Mesothelin Mediates Cell Adhesion J. Biol. Chem., March 5, 2004; 279(10): 9190 - 9198. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
