Genome-wide DNA methylation analysis reveals estrogen-mediated epigenetic repression of metallothionein-1 gene cluster in breast cancer
© Jadhav et al.; licensee BioMed Central. 2015
Received: 24 August 2014
Accepted: 13 January 2015
Published: 24 February 2015
Recent genome-wide analysis has shown that DNA methylation spans long stretches of chromosome regions consisting of clusters of contiguous CpG islands or gene families. Hypermethylation of various gene clusters has been reported in many types of cancer. In this study, we conducted methyl-binding domain capture (MBDCap) sequencing (MBD-seq) analysis on a breast cancer cohort consisting of 77 patients and 10 normal controls, as well as a panel of 38 breast cancer cell lines.
Bioinformatics analysis determined seven gene clusters with a significant difference in overall survival (OS) and further revealed a distinct feature that the conservation of a large gene cluster (approximately 70 kb) metallothionein-1 (MT1) among 45 species is much lower than the average of all RefSeq genes. Furthermore, we found that DNA methylation is an important epigenetic regulator contributing to gene repression of MT1 gene cluster in both ERα positive (ERα+) and ERα negative (ERα−) breast tumors. In silico analysis revealed much lower gene expression of this cluster in The Cancer Genome Atlas (TCGA) cohort for ERα + tumors. To further investigate the role of estrogen, we conducted 17β-estradiol (E2) and demethylating agent 5-aza-2′-deoxycytidine (DAC) treatment in various breast cancer cell types. Cell proliferation and invasion assays suggested MT1F and MT1M may play an anti-oncogenic role in breast cancer.
Our data suggests that DNA methylation in large contiguous gene clusters can be potential prognostic markers of breast cancer. Further investigation of these clusters revealed that estrogen mediates epigenetic repression of MT1 cluster in ERα + breast cancer cell lines. In all, our studies identify thousands of breast tumor hypermethylated regions for the first time, in particular, discovering seven large contiguous hypermethylated gene clusters.
Aberrant epigenetic changes, including DNA methylation and histone modifications, have been known to be the hallmark of cancer . These changes usually disrupt the regulation of many oncogenes or tumor suppressor genes in tumors, resulting in their abnormal expression. DNA methylation occurs mainly at CpG-rich ‘CpG islands’ and surrounding ‘CpG-shore’ regions, where more than 60% of them are located in 5′ promoters . De novo hypermethylation of these regions, which is often associated with silencing of many tumor suppressor genes, has been shown to play a crucial role in the development of many types of human cancers [3-6]. Many studies, including ours [7-10], used a quantitative approach based on statistical methods or machine learning algorithms to quantify methylation differences and identify differentially methylated regions (DMRs) from genome-wide methylation profiles in many different tissue or cancer patient cohorts. Such quantitative approaches are thus able to provide more insights into the role of DNA methylation in the development of various diseases such as cancer.
Recent genome-wide analysis of DNA methylation has revealed that this epigenetic process is not only a site specific event but also spans long stretches of chromosome regions consisting of clusters of contiguous CpG islands [11,12] or a gene family [13-15]. Extensive hypermethylation of various gene clusters has previously been reported. For example, hypermethylation of HOXA gene clusters was found in breast and lung cancers [16,17], protocadherin (PCDH) in Wilms’ tumor, the region across chromosome 2q14.2 in colorectal cancer and many others [18-20]. The findings in all these studies warrant a novel gene cluster centric approach towards the investigation of DNA methylation. In an effort to further investigate the mechanism responsible in this long-range epigenetic silencing (LRES), our laboratory previously elucidated the role of estrogen in coordinate repression of these gene clusters in breast cancer . The study revealed that persistent estrogen-mediated LRES leads to recruitment of H3K27me3 repressive chromatin marks, which are accompanied by accumulation of DNA methylation in a gene cluster located at 16p11.2.
In this study, we conducted MBDCap sequencing (MBD-seq) analysis on a breast cancer cohort consisting of 77 patients and 10 normal controls, as well as a panel of 38 breast cancer cell lines. Survival analysis conducted on 60 unique gene clusters determined seven clusters with a significant difference in overall survival (OS) by using methylation levels of genes in the cluster for all patients. Bioinformatics analysis further revealed a distinct feature that the conservation of a large gene cluster (approximately 70 kb) metallothionein-1 (MT1) among 45 species is much lower than the average of all RefSeq genes. We also found that DNA methylation is an important factor contributing to gene repression of MT1 gene cluster regardless of the ERα status. In silico analysis using the public domain The Cancer Genome Atlas (TCGA) data revealed that ERα positive (ERα+) breast cancer patients show lower levels of expression for MT1 genes. To investigate if estrogen regulates repression of MT1 cluster in ERα + breast cancer cell types, we conducted 17β-estradiol (E2) and demethylating agent 5-Aza-2′-deoxycytidine (DAC) treatment in various breast cancer cell lines. Our data suggested that both estrogen and DNA methylation mediate repression of the MT1 gene cluster in ERα + breast cancer cell lines. Cell proliferation and invasion assays suggested MT1F and MT1M may have anti-oncogenic roles in breast cancer.
MBD-seq identifies differential methylated patterns in breast primary tumors
Survival analysis determines significant hypermethylated gene clusters in breast cancer
Bioinformatics analysis reveals a distinct feature of the MT1 gene cluster
We further performed the bioinformatics analysis on these seven gene clusters in order to gain an insight into their underlying characterization, such as guanine-cytosine (GC) contents and phylogenetic conservations. As expected, the highest mean GC contents are around 5′TSS regions for all gene clusters, except PCDH gene family, with a peak around downstream 2 kb of 5′TSS while two gene clusters, HIST1 and PCDH, showed lower mean GC contents than the average of all RefSeq genes (Additional file 1: Figure S9A). Plots of phylogenetic conservation among 45 species (Additional file 1: Figure S9B) showed that phastCons scores of the MT1 gene cluster are below 0.1 at 5′TSS which is lower than the average of all RefSeq genes. Two relatively high conserved regions (two peaks downstream 5′TSS) might be the first exons. Our analysis is in line with a finding that the evolution of the lineage that led to human MT1 has undergone further duplication events that have resulted in 13 younger duplicate isoforms  and the divergence of the MT family in mammals. We also observed the ZNF gene cluster that has relatively lower scores of 0.2 but at the same level as the average.
Hypermethylation of the MT1 gene cluster is validated in different breast cancer cell lines
As the hypermethylated status for the MT1 gene cluster was observed in primary tumor samples, we further hypothesized that they would also be hypermethylated in breast cancer cell lines since these cell lines were isolated, cultured, and homogenized from the primary tumor. In order to confirm this hypothesis, we conducted the MBD-seq analysis on a panel of 38 breast cancer cell lines (Additional file 1: Figure S1). Overall, the mean methylation level for six of the gene clusters in the cell lines was higher than that in normal tissue samples but lower than that in tumor samples (Additional file 1: Figure S11). For the MT1 gene cluster, we found that the cell line data had the highest methylation levels followed by ERα + tumor, ERα − tumor, and normal samples (Figure 3C). A detailed visualized analysis along the MT1 gene cluster further revealed a hypermethylation pattern in CGIs that was associated with most MT1 TSS sites in the breast cancer samples relative to normal breast tissue (Figure 4C). Specifically, hypermethylation was observed in promoter CGIs of MT1L, E, M, A, G, and H as well as non-promoter CGIs M-A and F-G (Figure 4C; hypermethylated regions are denoted by dashed squares). The visualization of other gene clusters is shown in Additional file 1: Figures S3-8. Taken together, our result validated the hypermethylated status of the MT1 gene cluster in many different breast cancer cell lines.
Estrogen mediates epigenetic repression of the MT1 gene cluster in ERα + breast cancer cell lines
MT1F and MT1M expression exerts anti-oncogenic effects in breast cancer
Our current studies quantitatively analyzed differential methylation patterns at a genome-wide scale on a breast patient cohort and identified many large contiguous hypermethylated regions mainly consisting of gene clusters. Our results re-assert a newly emerging perspective that DNA methylation goes beyond a discrete gene event and often spans long stretches of chromosome regions. Although this phenomenon has lately been observed by several other studies in different types of cancer [12-14,16,17], this study reported for the first time that the tumor hypermethylation levels of a gene cluster (as many as seven gene clusters) (Figure 2) are significantly associated with overall survival in breast cancer patients. More strikingly, the hypermethylation status for seven clusters identified in the patient cohort was recapitulated in a panel of 38 breast cancer cell lines using the same MBD-seq protocol. Since the selection of cell lines, which includes several sub-types of breast cancer, is purely based on the availability at the time conducting the experiments, our cell line data not only validate our patient data but also further suggest that this large contiguous hypermethylation across seven gene clusters may be a distinguishing characteristic of breast cancer and commonly exist across many different sub-types.
Our analysis of intrinsic genomic features on these gene clusters revealed a distinct feature of the MT1 gene cluster. Compared to other gene clusters, such as HOXA, HOXC, HOXD, HISTI1, and ZNF, which are highly conserved among 45 species, the conservation for MT1 cluster is lower and even much below the average of all RefSeq genes. Although an evolutionary study indicates that the MT1 family is mammal-specific with 13 new isoforms in humans, an earlier study investigating promoter DNA methylation has reported one of the MT1 genes (MT1G) to be associated with breast cancer progression . Another study indicated that the expression of the genes in this cluster led to poor overall survival in a subset of invasive breast cancer patients . How these genomic properties affect the biological functions and biological processes of the genes in this cluster, in relation with specific organs, as well as its role in cancer development and progression, is worth further exploration. Interestingly, our correlation analysis of the methylation levels with the gene expression levels using a larger TCGA breast cancer cohort data found that the MT1 gene cluster exerts clear differential expression patterns among ERα + tumor samples and ERα − tumor samples (Figure 3B). Although the HOXD gene cluster showed a similar pattern, this cluster has previously been reported to be hypermethylated in a large cohort of melanomas  as well as in astrocytomas , a lethal human brain tumor, implying that it is a rather common phenomenon for many types of cancer. So far, only this study has reported the negative correlation of the MT1 cluster in breast cancer, that is, its gene expression levels are decreased upon its hypermethylation, and thus, we speculate that the hypermethylation of this cluster may be breast cancer specific. We also observed some gene clusters having a positive correlation, that is, their gene expression levels were increased upon their hypermethylation in breast tumors.
Despite our data supporting a notion that estrogen mediates epigenetic repression of the MT1 gene cluster in MCF7 cells (Figure 6), the underlying mechanism of what triggers this long-range coordinated repression process remains obscured. Our previous study  has shown that persistent estrogen-mediated long-range repression leads to recruitment of H3K27me3 repressive chromatin marks, which are accompanied by the accumulation of DNA methylation in a gene cluster located at 16p11.2. Recent studies have also shown that estrogen and ERα positively regulate the expression of various methyltransferases (DNMTs), thereby contributing to the malignant transformation of cells in various estrogen responsive breast and endometrial cancers [39,40]. We thus speculate that estrogen mediates the recruiting of some chromatin modifying enzymes, such as polycomb complex, then estrogen further recruits DNA methylation machinery, thus triggering a DNA methylation process at a single embedded gene, which then spreads to other neighboring genes due to their closeness and eventually methylation of a whole cluster occurs.
One important finding in our studies is that we demonstrated the invasiveness of MT1F and MT1M in MCF7 cells (Figure 7). Despite other studies finding that loss of MT1 or one of MT1 member genes was significantly correlated with invasiveness in other tumor types , our study depicts the anti-oncogenic role in breast cancer cell line. Our results offer mechanistic insights into breast tumorigenesis, suggesting that methylation of MT1 gene cluster is involved in oncogenic events. We have used MCF7 as a model cell line for tumors showing ERα + phenotype, and our findings provide a compelling evidence for a comprehensive mechanistic study of estrogen-mediated epigenetic repression of MT1 cluster in other breast cancer cell lines representing various breast cancer subtypes.
Our studies used a sequence-based methylation protocol (MBD-seq) to identify thousands of breast tumor hypermethylated regions, in particular, discovering seven large contiguous hypermethylated gene clusters from a breast cancer patient cohort. Importantly, we were able to use the methylation levels of the cluster to stratify the patients for overall survival, pointing to the potential prognostic and therapeutic significance for this epigenetic modification. As these gene clusters were also selected based on their association with cancer development, their prognostic potential in determining overall survival provides a compelling case for future mechanistic studies.
DNA samples of 77 breast tumors (n = 77) and 10 normal breast tissues from healthy individuals (n = 10) (Additional file 1: Table S1) and 38 ICBP breast cancer cell lines (n = 38) were isolated for subsequent DNA methylation analysis. The breast tumor samples were collected from patients at Chile, IRB # 11-11-3239 approved by UTHSCSA. These tissues were in various stages of tumor advancement and from patients representing different ER and PR statuses. The normal samples were obtained from normal individuals undergoing reduction mammoplasty (Additional file 3). All tissues were obtained following approval of the Institutional Review Board committee.
Most of ICBP panel breast cell lines were obtained from NCI Cancer Biology Program at NCI in November 2008, except HMEC, MCF7, MDAMB134, BT474, BT20, and MDA-MB231 were obtained from American Type Culture Collection (ATCC) (ATCC Breast Cancer Cell Panel, Manassas, VA, USA). All cell lines have been tested and authenticated by ATCC and maintained in our laboratory for less than 6 months during which all experiments were conducted. All cell lines were cultured in ATCC recommended media and conditions. Cell lines MCF7, MDAMB134, BT474, MDA-MB231, and BT20 were cultured in DMEM culture medium (Life Technologies, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS). For all the treatments with E2 and DAC, RPMI culture medium without phenol red was used and supplemented with 10% charcoal stripped-heat inactivated (CSHI) FBS.
MBDCap sequencing analysis
where N Read,t is the normalized read number of the ith bin and U Read,t is the uniquely mapped read number of the ith bin, is the total uniquely mapped reads number. ‘INT’ function rounds the element (in the parenthesis) to the nearest integers towards minus infinity, and ‘^’ means the power operator.
Identification of differentially methylated regions
To further characterize methylation patterns within each DMR, methylation values from 8-kb window were divided into three segments equally representing CGI regions and flanking shore regions. For each segment, it was considered an individual DMR if the P value is smaller than the thresholds. Within 8-kb windows, the center is the TSS for each transcript or is in the middle of its CGI for each intragenic or intergenic region.
The lifetime of OS was defined as the time between the first operation or first-line chemotherapy and end of follow-up or death due to breast cancer. The third quartile (equal 75th percentile) of methylation levels was used to identify high methylation (HM) and low methylation (LM) of each promoter region. HMs and LMs were converted to 1 and 0 binary codes to do Kaplan-Meier survival and Cox proportional hazard model analysis. We select genes as clusters, who are part of HUGO gene families and which consist of genes located on the same chromosome and consecutive geographic region. The selection criteria for combination of family genes were that they were grouped by hazard ratio higher or lower than one with log-rank test P < 0.3 calculated for individual genes in the family (here, we assume that including borderline significant genes will collaboratively contribute towards significant association with disease outcome). In a combination of the same family genes, the receiver operating characteristic (ROC) curve was used to select an optimal cutoff value for the numerous HM of promoter regions. Bootstrapping was performed 200 times, and ‘best’ methods were used to classify dead and alive patients by pROC package of R (version 2.13).
E2 and DAC treatment
Cell lines were incubated in charcoal stripped media for 24 h and then treated with DAC (1 μM) for 72 h. The cells were further treated with E2 (70 nM) for 36 h with and without DAC treatment. RNA was isolated from harvested cells, which were then harvested and subjected to real-time RT-qPCR to examine expression of the MT1 genes. PCR primers are outlined in Additional file 1: Table S4, and conditions are described previously . Relative expression was determined by the formula 2−ΔΔCt .
ERE luciferase assay
MCF7 cells were transfected in triplicate with ERE and Renilla vectors, kindly supplied by Dr. Rong Li, Department of Molecular Medicine, University of Texas Health Science Center (50:1 to make total DNA of 110 ng/well) for 24 h. Cells were incubated with charcoal strip media for 24 h and then treated with E2 for 24 h. The ERE and Renilla activity was calculated with Dual-Glo Luciferase Assay Kit (Promega, Madison, WI, USA).
Cell invasion assay and proliferation assays
Specific siRNA was used to knockdown MT1F and MT1M expression (Thermo Scientific). MCF7 (approximately 50,000 cells) was transfected with MT1F and MT1M siRNA seeded onto the top insert (layered with Matrigel) of an invasion chamber (BD Biosciences, Franklin Lakes, NJ, USA). The invasion chambers were then incubated at 37°C in 5% CO2 for 20 h. Cells that did not invade through the Matrigel (on the upper surface of the insert membrane) were mechanically removed with cotton tip applications and several washes with PBS. Invaded cells on the bottom of the coated membranes were visualized using a fluorescence microscope with a × 20 objective after incubation with Hoechst stain (Life Technologies). Images were obtained from four standardized non-overlapping fields. Invaded cells were counted using the Image J software (http://rsbweb.nih.gov/ij/). Invasion assays were done in triplicate; images of four fields per well (covering about 85% of the well) were taken for counting invaded cells. Cellular proliferation was assayed using CellTiter-Glo® Luminescence Kit (Promega, Madison, WI, USA) according to the manufacturer’s directions as described earlier .
Bioinformatics and statistical analysis
We considered statistical significance as P < 0.05 for all analyses unless explicitly stated. Student’s t-test was used to compare RT-qPCR, and invasion assay results in different treatment and control groups. Statistical significance was assigned as * if P < 0.05, ** if P < 0.01, and *** if P < 0.001.
GC content and phastCons score: For each gene, we retrieved a genomic region with a length of 8-kb DNA sequence spanning from 4-kb upstream to 4-kb downstream around 5′TSS. The regions were then divided into 160 bins with a bin size of 50 bp. The GC content is calculated for each bin as the number of G + C in that particular bin divided by the bin size of 50. The phastCons score is calculated for each bin using the values from UCSC genome browser (http://hgdownload.soe.ucsc.edu/goldenPath/hg18/phastCons44way/). The phastCons score represents conservation information among 45 species. For each gene cluster, we simply calculated the mean and standard deviation values from all member genes in the cluster for all 160 bins. We also used the whole RefSeq gene set as a background control.
In silico gene expression correlation: The clinical information and gene expression values for the breast tumors and normal tissues were downloaded from the TCGA website (http://cancergenome.nih.gov/). A total of 106 normal tissue samples and 988 primary tumor samples with mRNA expression information are currently available for the correlation analysis. Of 988 tumor samples, 207 tumor samples are ERα−, 704 tumor samples are ERα+, and 77 tumor samples are classified as unknown due to missing clinical information. We then used the gene expression values for our seven gene clusters and created an expression matrix to examine their expression pattern. We applied a z-score normalization method on each individual gene on their log-transformed RPKM values across the sample set. In addition, we have calculated the log2 fold change values for normal group vs. tumor group and ERα − group vs. ERα−, respectively. A visualization map for the matrix was constructed by Python library package.
In silico correlation between expression and DNA methylation: RNA-seq and DNA methylation values were retrieved from TCGA using CGDSR library in R. The patient samples showing values for both RNA-seq and DNA methylation for each patient were retained. Scatterplots showing correlation between RNA-seq and DNA methylation were created after excluding the outliers falling in top and bottom 5 quartile of the data.
This study was supported by funds from the Department of Molecular Medicine at University of Texas Health Science Center at San Antonio (VXJ) and by National Institutes of Health [grant numbers: U54CA113001 to THH, VXJ] and P30 CA054174 (Cancer Center Support Grant).
- Jones PA, Baylin SB. The epigenomics of cancer. Cell. 2007;128:683–92.View ArticlePubMed CentralPubMedGoogle Scholar
- Antequera F. Structure, function and evolution of CpG island promoters. Cell Mol Life Sci. 2003;60:1647–58.View ArticlePubMedGoogle Scholar
- Hill VK, Ricketts C, Bieche I, Vacher S, Gentle D, Lewis C, et al. Genome-wide DNA methylation profiling of CpG islands in breast cancer identifies novel genes associated with tumorigenicity. Cancer Res. 2011;71:2988–99.View ArticlePubMedGoogle Scholar
- McGarvey KM, Van Neste L, Cope L, Ohm JE, Herman JG, Van Criekinge W, et al. Defining a chromatin pattern that characterizes DNA-hypermethylated genes in colon cancer cells. Cancer Res. 2008;68:5753–9.View ArticlePubMed CentralPubMedGoogle Scholar
- Sandoval J, Mendez-Gonzalez J, Nadal E, Chen G, Carmona FJ, Sayols S, et al. A prognostic DNA methylation signature for stage I non-small-cell lung cancer. J Clin Oncol Off J Am Soc Clin Oncol. 2013;31:4140–7.View ArticleGoogle Scholar
- Huang RL, Gu F, Kirma NB, Ruan J, Chen CL, Wang HC, et al. Comprehensive methylome analysis of ovarian tumors reveals hedgehog signaling pathway regulators as prognostic DNA methylation biomarkers. Epigenetics. 2013;8:624–34.View ArticlePubMed CentralPubMedGoogle Scholar
- Lister R, Pelizzola M, Kida YS, Hawkins RD, Nery JR, Hon G, et al. Hotspots of aberrant epigenomic reprogramming in human induced pluripotent stem cells. Nature. 2011;471:68–73.View ArticlePubMed CentralPubMedGoogle Scholar
- Rakyan VK, Down TA, Thorne NP, Flicek P, Kulesha E, Graf S, et al. An integrated resource for genome-wide identification and analysis of human tissue-specific differentially methylated regions (tDMRs). Genome Res. 2008;18:1518–29.View ArticlePubMed CentralPubMedGoogle Scholar
- Lan X, Adams C, Landers M, Dudas M, Krissinger D, Marnellos G, et al. High resolution detection and analysis of CpG dinucleotides methylation using MBD-Seq technology. PLoS One. 2011;6:e22226.View ArticlePubMed CentralPubMedGoogle Scholar
- Gu F, Doderer MS, Huang YW, Roa JC, Goodfellow PJ, Kizer EL, et al. CMS: a web-based system for visualization and analysis of genome-wide methylation data of human cancers. PLoS One. 2013;8:e60980.View ArticlePubMed CentralPubMedGoogle Scholar
- Clark SJ. Action at a distance: epigenetic silencing of large chromosomal regions in carcinogenesis. Hum Mol Genet. 2007;16 Spec(1):R88–95.View ArticleGoogle Scholar
- Xie W, Schultz MD, Lister R, Hou Z, Rajagopal N, Ray P, et al. Epigenomic analysis of multilineage differentiation of human embryonic stem cells. Cell. 2013;153:1134–48.View ArticlePubMed CentralPubMedGoogle Scholar
- Zouridis H, Deng N, Ivanova T, Zhu Y, Wong B, Huang D, et al. Methylation subtypes and large-scale epigenetic alterations in gastric cancer. Sci Transl Med. 2012;4:156ra–140.View ArticleGoogle Scholar
- Hsu YT, Gu F, Huang YW, Liu J, Ruan J, Huang RL, et al. Promoter hypomethylation of EpCAM-regulated bone morphogenetic protein gene family in recurrent endometrial cancer. Clin Cancer Res. 2013;19:6272–85.View ArticlePubMed CentralPubMedGoogle Scholar
- Hansen KD, Timp W, Bravo HC, Sabunciyan S, Langmead B, McDonald OG, et al. Increased methylation variation in epigenetic domains across cancer types. Nat Genet. 2011;43:768–75.View ArticlePubMed CentralPubMedGoogle Scholar
- Novak P, Jensen T, Oshiro MM, Wozniak RJ, Nouzova M, Watts GS, et al. Epigenetic inactivation of the HOXA gene cluster in breast cancer. Cancer Res. 2006;66:10664–70.View ArticlePubMedGoogle Scholar
- Rauch T, Wang Z, Zhang X, Zhong X, Wu X, Lau SK, et al. Homeobox gene methylation in lung cancer studied by genome-wide analysis with a microarray-based methylated CpG island recovery assay. Proc Natl Acad Sci U S A. 2007;104:5527–32.View ArticlePubMed CentralPubMedGoogle Scholar
- Frigola J, Song J, Stirzaker C, Hinshelwood RA, Peinado MA, Clark SJ. Epigenetic remodeling in colorectal cancer results in coordinate gene suppression across an entire chromosome band. Nat Genet. 2006;38:540–9.View ArticlePubMedGoogle Scholar
- Dallosso AR, Hancock AL, Szemes M, Moorwood K, Chilukamarri L, Tsai HH, et al. Frequent long-range epigenetic silencing of protocadherin gene clusters on chromosome 5q31 in Wilms’ tumor. PLoS Genet. 2009;5:e1000745.View ArticlePubMed CentralPubMedGoogle Scholar
- Shiraishi M, Sekiguchi A, Oates AJ, Terry MJ, Miyamoto Y. HOX gene clusters are hotspots of de novo methylation in CpG islands of human lung adenocarcinomas. Oncogene. 2002;21:3659–62.View ArticlePubMedGoogle Scholar
- Hsu PY, Hsu HK, Singer GA, Yan PS, Rodriguez BA, Liu JC, et al. Estrogen-mediated epigenetic repression of large chromosomal regions through DNA looping. Genome Res. 2010;20:733–44.View ArticlePubMed CentralPubMedGoogle Scholar
- Irizarry RA, Ladd-Acosta C, Wen B, Wu Z, Montano C, Onyango P, et al. The human colon cancer methylome shows similar hypo- and hypermethylation at conserved tissue-specific CpG island shores. Nat Genet. 2009;41:178–86.View ArticlePubMed CentralPubMedGoogle Scholar
- Rossi M, Abbondanza C, D’Arcangelo A, Gazzerro P, Medici N, Moncharmont B, et al. The Zn-finger domain of RIZ protein promotes MCF-7 cell proliferation. Cancer Lett. 2004;215:229–37.View ArticlePubMedGoogle Scholar
- Friedline JA, Garrett SH, Somji S, Todd JH, Sens DA. Differential expression of the MT-1E gene in estrogen-receptor-positive and -negative human breast cancer cell lines. Am J Pathol. 1998;152:23–7.PubMed CentralPubMedGoogle Scholar
- Oyama T, Take H, Hikino T, Iino Y, Nakajima T. Immunohistochemical expression of metallothionein in invasive breast cancer in relation to proliferative activity, histology and prognosis. Oncology. 1996;53:112–7.View ArticlePubMedGoogle Scholar
- Flagiello D, Gibaud A, Dutrillaux B, Poupon MF, Malfoy B. Distinct patterns of all-trans retinoic acid dependent expression of HOXB and HOXC homeogenes in human embryonal and small-cell lung carcinoma cell lines. FEBS Lett. 1997;415:263–7.View ArticlePubMedGoogle Scholar
- Miyamoto K, Fukutomi T, Akashi-Tanaka S, Hasegawa T, Asahara T, Sugimura T, et al. Identification of 20 genes aberrantly methylated in human breast cancers. Int J Cancer. 2005;116:407–14.View ArticlePubMedGoogle Scholar
- Yang Y, Liu J, Li X, Li JC. PCDH17 gene promoter demethylation and cell cycle arrest by genistein in gastric cancer. Histol Histopathol. 2012;27:217–24.PubMedGoogle Scholar
- Moleirinho A, Carneiro J, Matthiesen R, Silva RM, Amorim A, Azevedo L. Gains, losses and changes of function after gene duplication: study of the metallothionein family. PLoS One. 2011;6:e18487.View ArticlePubMed CentralPubMedGoogle Scholar
- Laurent L, Wong E, Li G, Huynh T, Tsirigos A, Ong CT, et al. Dynamic changes in the human methylome during differentiation. Genome Res. 2010;20:320–31.View ArticlePubMed CentralPubMedGoogle Scholar
- Ziller MJ, Gu H, Muller F, Donaghey J, Tsai LT, Kohlbacher O, et al. Charting a dynamic DNA methylation landscape of the human genome. Nature. 2013;500:477–81.View ArticlePubMedGoogle Scholar
- Nordlund J, Backlin CL, Wahlberg P, Busche S, Berglund EC, Eloranta ML, et al. Genome-wide signatures of differential DNA methylation in pediatric acute lymphoblastic leukemia. Genome Biol. 2013;14:r105.View ArticlePubMed CentralPubMedGoogle Scholar
- Chin K, DeVries S, Fridlyand J, Spellman PT, Roydasgupta R, Kuo WL, et al. Genomic and transcriptional aberrations linked to breast cancer pathophysiologies. Cancer Cell. 2006;10:529–41.View ArticlePubMedGoogle Scholar
- Wang LH, Yang XY, Zhang X, An P, Kim HJ, Huang J, et al. Disruption of estrogen receptor DNA-binding domain and related intramolecular communication restores tamoxifen sensitivity in resistant breast cancer. Cancer Cell. 2006;10:487–99.View ArticlePubMedGoogle Scholar
- Mao J, Yu H, Wang C, Sun L, Jiang W, Zhang P, et al. Metallothionein MT1M is a tumor suppressor of human hepatocellular carcinomas. Carcinogenesis. 2012;33:2568–77.View ArticlePubMedGoogle Scholar
- Park SY, Kwon HJ, Lee HE, Ryu HS, Kim SW, Kim JH, et al. Promoter CpG island hypermethylation during breast cancer progression. Virchows Arch. 2011;458:73–84.View ArticlePubMedGoogle Scholar
- Marzese DM, Scolyer RA, Huynh JL, Huang SK, Hirose H, Chong KK, et al. Epigenome-wide DNA methylation landscape of melanoma progression to brain metastasis reveals aberrations on homeobox D cluster associated with prognosis. Hum Mol Genet. 2014;23:226–38.View ArticlePubMed CentralPubMedGoogle Scholar
- Wu X, Rauch TA, Zhong X, Bennett WP, Latif F, Krex D, et al. CpG island hypermethylation in human astrocytomas. Cancer Res. 2010;70:2718–27.View ArticlePubMed CentralPubMedGoogle Scholar
- Shi JF, Li XJ, Si XX, Li AD, Ding HJ, Han X, et al. ERalpha positively regulated DNMT1 expression by binding to the gene promoter region in human breast cancer MCF-7 cells. Biochem Biophys Res Commun. 2012;427:47–53.View ArticlePubMedGoogle Scholar
- Cui M, Wen Z, Yang Z, Chen J, Wang F. Estrogen regulates DNA methyltransferase 3B expression in Ishikawa endometrial adenocarcinoma cells. Mol Biol Rep. 2009;36:2201–7.View ArticlePubMedGoogle Scholar
- Tai SK, Tan OJ, Chow VT, Jin R, Jones JL, Tan PH, et al. Differential expression of metallothionein 1 and 2 isoforms in breast cancer lines with different invasive potential: identification of a novel nonsilent metallothionein-1H mutant variant. Am J Pathol. 2003;163:2009–19.View ArticlePubMed CentralPubMedGoogle Scholar
- Cheng AS, Jin VX, Fan M, Smith LT, Liyanarachchi S, Yan PS, et al. Combinatorial analysis of transcription factor partners reveals recruitment of c-MYC to estrogen receptor-alpha responsive promoters. Mol Cell. 2006;21:393–404.View ArticlePubMedGoogle Scholar
- Hsu PY, Hsu HK, Lan X, Juan L, Yan PS, Labanowska J, et al. Amplification of distant estrogen response elements deregulates target genes associated with tamoxifen resistance in breast cancer. Cancer Cell. 2013;24:197–212.View ArticlePubMed CentralPubMedGoogle Scholar
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.