Analysis of DNA methylation landscape reveals the roles of DNA methylation in the regulation of drug metabolizing enzymes
© Habano et al. 2015
Received: 28 July 2015
Accepted: 15 September 2015
Published: 28 September 2015
Drug metabolizing enzymes (DMEs) exhibit dramatic inter- and intra-individual variability in expression and activity. However, the mechanisms determining this variability have not been fully elucidated. The aim of this study was to evaluate the biological significance of DNA methylation in the regulation of DME genes by genome-wide integrative analysis.
DNA methylation and mRNA expression profiles of human tissues and hepatoma cells were examined by microarrays. The data were combined with GEO datasets of liver tissues, and integrative analysis was performed on selected DME genes. Detailed DNA methylation statuses at individual CpG sites were evaluated by DNA methylation mapping. From analysis of 20 liver tissues, highly variable DNA methylation was observed in 37 DME genes, 7 of which showed significant inverse correlations between DNA methylation and mRNA expression. In hepatoma cells, treatment with a demethylating agent resulted in upregulation of 5 DME genes, which could be explained by DNA methylation status. Interestingly, some DMEs were suggested to act as tumor-suppressor or housekeeper based on their unique DNA methylation features. Moreover, tissue-specific and age-dependent expression of UDP-glucuronosyltransferase 1A splicing variants was associated with DNA methylation status of individual first exons.
Some DME genes were regulated by DNA methylation, potentially resulting in inter- and intra-individual differences in drug metabolism. Analysis of DNA methylation landscape facilitated elucidation of the role of DNA methylation in the regulation of DME genes, such as mediator of inter-individual variability, guide for correct alternative splicing, and potential tumor-suppressor or housekeeper.
KeywordsDNA methylation mapping Differences in drug metabolism Integrative analysis Alternative splicing
Inter-individual differences in responses to drug therapy vary widely. Such differences are due in part to variable pharmacokinetics, which can be explained by diversity in genes affecting drug absorption, distribution, metabolism, and excretion (ADME) . For example, genetic polymorphisms in cytochrome P450 (CYP) genes, such as CYP2C9, CYP2C19, and CYP2D6, affect drug metabolizing activity and lead to different drug responses [2–4]. Inter-individual differences are also observed in CYP3A4 expression and activity [5, 6] but are rarely associated with any detectable polymorphisms in the CYP3A4. Additionally, UDP-glucuronosyltransferase 1A1 (UGT1A1) mutant genotypes *28 and *6 are related to poor glucuronidation activity and have been shown to act as useful indictors of adverse reactions to cancer chemotherapy with irinotecan (CPT-11) [7, 8]. However, some individuals who do not have these mutant genotypes also suffer from adverse reactions to irinotecan .
Intra-individual differences are also involved in the regulation of ADME-related genes. For example, the expression and metabolizing activity of UGT1A isoforms vary among tissues and during normal development [10, 11]. The UGT1A locus encodes nine functional isoforms through an exon sharing mechanism in which the transcripts of individual first exon cassettes are spliced to exons 2–5, leading to the expression of individual UGT1A isoforms . Although the tissue-specific and age-dependent expressions of the UGT1A isoforms are important determinants of drug efficacy and adverse reactions, the regulatory mechanisms involved in UGT1A expression cannot be explained by genetic polymorphisms. Thus, the mechanisms underlying inter- and intra-individual differences in responses to drug therapy have not been fully elucidated.
In order to examine these mechanisms, we investigated the involvement of epigenetics, the mechanism of heritable changes in gene regulation without DNA sequence alteration. We previously found that epigenetic mechanisms, such as DNA methylation, are involved in the regulation of drug metabolizing enzymes (DMEs) in colon cancer cells [13, 14]. The CYP1B1 and CYP3A4 genes were shown to be transcriptionally upregulated by treatment with a demethylating agent, and such upregulation could be explained by hypermethylation at CpG sites located in the 5′-promoters of the CYP1B1 and pregnane X receptor (PXR) genes. To date, more than 50 ADME-related genes have been reported as targets for epigenetic regulation [15, 16]. Although these ADME-related genes were found to be aberrantly regulated by epigenetic mechanisms in tumor cells, most ADME-related genes have been discovered unexpectedly by multiple independent studies during the search for tumor suppressor genes. Therefore, we still have a limited understanding of the biological significance of epigenetic mechanisms in the regulation of ADME-related genes.
In the present study, we examined global DNA methylation and mRNA expression profiles of human tissues and hepatoma cell lines using microarray platforms. These two omic datasets were combined with similar corresponding datasets derived from the Gene Expression Omnibus (GEO) database at the National Center for Biotechnology Information (NCBI). In order to evaluate the significance of DNA methylation in the regulation of DME genes, we examined which DME genes were regulated by DNA methylation in normal liver tissues and hepatoma cells, whether the tissue-specific and age-dependent expression of UGT1A isoforms could be regulated by DNA methylation, and whether DNA methylation profiles could be used to elucidate the specific roles of DME genes. To this end, we performed DNA methylation mapping to determine the methylation levels and variations at individual CpG sites in relation to the structure of individual DME genes. Recent studies demonstrated that global and detailed analysis of dynamic DNA methylation profiles facilitated the finding of the informative fraction of CpG sites [17, 18]. Therefore, we would be able to identify the role of DNA methylation of DME genes by analysis of DNA methylation landscapes.
DNA methylation profiles of DME genes in human tissues
DME genes regulated by DNA methylation in adult livers
DME genes regulated by DNA methylation in three hepatoma cell lines
DNA methylation and alternative splicing of UGT1A isoforms
Classification of DME genes based on their DNA methylation landscape
We used GEO datasets previously registered elsewhere  and combine these datasets with our original data. We demonstrated that genome-wide integrative analysis using GEO datasets was a powerful tool for identification of novel roles of DNA methylation. We cannot rule out the possibility of batch effects, which are the technical artifacts such as laboratory conditions, experiment time, reagent lots, and/or laboratory personnel differences . However, the datasets obtained from the same laboratory were used for the study of methylation mapping and correlation analysis. Therefore, we think that such effect was reduced to the lowest level in our study.
The DNA methylation statuses of DME genes varied among normal liver tissues. We found seven DME genes (i.e., CYP1A2, CYP2C19, CYP2D6, GSTA4, GSTM5, GSTT1, and SULT1A1) that showed significant inverse correlations between DNA methylation and mRNA expression levels. Some other DME genes also tended to show such correlations, but statistical significance was not reached. These results suggested that a small but significant fraction of DME genes were transcriptionally regulated by DNA methylation, resulting in different mRNA expression levels among individuals. Further investigations with larger sample sizes may allow for identification of additional targets for DNA methylation. Recent studies demonstrated that considerable numbers of hepatic genes are also regulated by 5-hydroxymethylcytosine (5hmC) . Therefore, the levels of 5hmC should be evaluated in our next studies.
Interestingly, CYP1A2, CYP2C19, CYP2D6, and SULT1A1 tend to show inter-individual variations in activity, and this has become a major problem in clinical practice. Our data supported that these variations may result from the variable DNA methylation statuses of these DME genes. For example, DNA methylation status could explain the discrepancy in the relatively variable expression of CYP2C19 among Caucasian populations in which there is a low frequency of mutant genotypes . Although considerable inter-individual differences were observed in CYP3A4 expression (CV 22.9 %; Fig. 2b), we did not detect a significant relationship between DNA methylation and mRNA expression. As reported by Kacevska et al. , DNA methylation at more 5′ distal regions (position −1547 or −10,762) may be associated with CYP3A4 regulation. Alternatively, PXR gene methylation, but not CYP3A4 gene methylation, may contribute to CYP3A4 downregulation in colon cancer cells, as previously described by our laboratory .
We also found five DME genes (CYP1B1, CYP8B1, GSTM2, GSTP1, and UGT3A2) that were likely to be downregulated by DNA methylation. Further analyses, such as chromatin immunoprecipitation, will be required to establish functional evidence for this relationship. Interestingly, the methylation statuses of these five DME genes were highly stable in normal livers, suggesting that DNA methylation may play different roles in the regulation of DME genes between normal and cancerous livers. As expected, DNA methylation in the 5′-regulatory region of housekeeping and tumor suppressor genes was stable and restricted to low levels. In contrast, the methylation levels of these genes were relatively high within the exons. These results were supported by a previous study showing that DNA methylation plays critical roles as markers of exon-intron boundaries and transcriptional silencing [25, 26].
In this study, we demonstrated that DNA methylation could function as a guide for correct alternative splicing of the UGT1A gene in a tissue-specific manner. Despite the limited cases in our study, our findings were supported by recent genome-wide analyses showing that many genes are regulated by alternative splicing via DNA methylation [18, 25, 26]. Oda et al. also reported that tissue-specific UGT1A10 expression is regulated by DNA methylation . We found that the relatively low expression levels of DME genes in the fetal liver could be associated with the higher methylation levels observed in these genes. Thus, DNA methylation mapping may provide novel insights into the tissue-specific and age-dependent splicing switch of DME genes regulated by DNA methylation.
From these results, we propose that the roles of DME genes in the liver depend on the DNA methylation landscape (Fig. 7). Variable methylation of the DME gene observed in normal livers may be one of the mechanisms mediating inter-individual variation in drug metabolizing activity and drug efficacy if the methylation status is associated with mRNA expression. Therefore, the DNA methylation statuses of the seven HVM-type genes may be indicators for prediction of drug efficacy and safety. Although there were considerable number of DME genes that showed highly variable DNA methylation (maximum β R > 0.296) without inverse correlation to gene expression, such DNA methylation was probably unrelated to gene expression and other mechanisms might be involved in the regulation of these DME genes. The cutoff value as maximum β R > 0.296 may not be the best for categorizing HVM type, because maximum β R can depend on the number of CpG sites examined for each gene. However, some DME genes such as CYP1A2 are regulated by DNA methylation at exclusively confined CpG sites . In order to detect any informative signals without averaging effect, therefore, we evaluated β R value with single (maximum) CpG site rather than with the fraction of CpG sites. We also found five TSG-type genes that were thought to act as tumor suppressor genes because they exhibited DNA methylation features similar to those observed in the BMP4 and IGFBP3 genes. Therefore, these TSG-type genes may have novel functions related to gatekeeping or genome stability in tumor cells. On the other hand, some HKG-type genes were suggested to function as housekeeping genes. Based on this extremely stable DNA methylation status, we predicted that the expression of these HKG-type genes should be tightly regulated, which may explain why DME genes with stable DNA methylation had conserved CpG islands. Interestingly, DMEs with a small range of variation in DNA methylation were known to metabolize endogenous substrates rather than xenobiotics. In contrast, DMEs with highly variable DNA methylation (including CYP1A2, CYP2C19, and CYP2D6) catalyze the metabolism of many xenobiotics, including drugs used in the clinical setting. These types of DMEs may flexibly modify their DNA methylation and expression profiles to metabolize and excrete various xenobiotics. We hypothesize that DNA methylation may play different roles in the regulation of DME genes depending on context. For example, HKG-type genes are regulated by “rigid” DNA modifications to strictly retain the epigenome. On the other hand, HVM-type genes are regulated by “plastic” DNA modifications to flexibly rewrite the epigenome for adjustment to new environments. Thus, HVM-type genes tend to show higher inter-individual variations in DNA methylation status.
We described the global DNA methylation landscape of DME genes in human tissues and demonstrated that a small but specific fraction of DME genes was regulated by DNA methylation. Variations in DNA methylation may result in inter-individual differences in the efficacies and toxicities of many drugs. DNA methylation of DME genes may represent landmarks of tissue-specific and age-dependent splicing switches and may be useful indicators for predicting the unknown functions, such as tumor-suppressor or housekeeper.
Cell culture and treatment
Human hepatoma HepG2, HuH7, and JHH1 cells were obtained from American Type Culture Collection (Manassas, VA, USA) and cultured in Dulbecco’s minimal essential medium (Invitrogen, Carlsbad, CA, USA) containing 10 % fetal bovine serum at 37 °C in a humidified atmosphere containing 5 % CO2. In order to reverse DNA methylation, the cells were treated with 0.5 or 5 μM 5-aza-2′-deoxycytidine (DAC; Sigma-Aldrich, St. Louis, MO, USA) for 72 h. After DAC treatment, the cells were treated with trichostatin A (TSA; Sigma-Aldrich) at 200 nM (HepG2 and HuH7 cells) or 20 nM (JHH1 cells) for 24 h.
DNA and RNA samples
Genomic DNA was extracted from hepatoma cells using a standard proteinase K/sodium dodecyl sulfate and phenol/chloroform method. Total cellular RNA was isolated using an RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Genomic DNA and total RNA of human adult liver tissues (NLA and NL2), fetal liver tissue (NLF), adult small intestinal tissue (NSI), and adult colonic tissue (NC) were obtained from commercially available products (Capital Biosciences, Gaithersburg, MD, USA and BioChain Institute, Newark, CA, USA). Because we had no information other than age, gender, and race (Chinese) for these samples, we did not submit our research proposal to an ethical committee.
DNA methylation analysis
Genomic DNA from tissues and hepatoma cells was subjected to sodium bisulfite modification using an EpiTect Plus DNA Bisulfite Kit (Qiagen) and examined for DNA methylation using an Infinium HumanMethylation450 Bead Chip (Illumina, San Diego, CA, USA), which interrogates over 480,000 CpG sites in the genome . The level of DNA methylation for each CpG site was reported as the β value, which ranged from 0 (fully unmethylated) to 1 (fully methylated). Data have been deposited in the GEO repository with accession numbers GSE67477 and GSE67484 (Additional file 10: Table S2). DNA methylation data of 55 CYP genes and 62 phase II DME genes, including 17 glutathione S-transferase (GST), 10 N-acetyltransferase (NAT), 13 sulfotransferase (SULT), and 22 UGT genes, were selected and used for subsequent analyses. For the control, four housekeeping genes (ACTB, B2M, GAPDH, TBP), two tumor suppressor genes (BMP4 and IGFBP3), and two DNA repair genes (MLH1 and MGMT) were also examined. The results were visualized in heat maps using Cluster 3.0 software (http://bonsai.hgc.jp/~mdehoon/software/cluster/software.htm#ctv). The DNA methylation status of candidate DME genes was validated by COBRA method . Briefly, after polymerase chain reaction (PCR) using a TaKaRa EpiTaq HS kit (Takara Bio Inc., Shiga, Japan), the resulting products were digested with an appropriate restriction enzyme, such as HpyCH4IV, TaqI, or BstUI (New England Biolabs, Ipswich, MA, USA). The digested products were electrophoresed on 2 % agarose gels followed and visualized by ethidium bromide staining. Primers used are listed in Additional file 11: Table S3.
mRNA expression analysis
First-strand cDNA was synthesized from total RNA samples using a Transcriptor First Strand cDNA Synthesis Kit (Roche Diagnostics, Mannheim, Germany) according to the manufacturer’s protocol. The mRNA expression profiles were examined using a SurePrint G3 Human Gene Expression 8 × 60K v2 microarray (Agilent Technologies, Palo Alto, CA, USA). The signal intensity for each probe was normalized by the 75th percentile. The data have been deposited in the GEO repository with the accession number GSE67318 (Additional file 10: Table S2). The mRNA expression data for the CYP gene, phase II DME genes, and control genes were selected and used for subsequent analyses. The level of each transcript was validated by quantitative real-time PCR analysis using a FirstStart Universal SYBR Green Master (ROX) kit (Roche Diagnostics). Transcript levels of individual UGT1A isoforms and ACTB were evaluated by TaqMan Gene Expression Assays (Life Technologies, Gaithersburg, MD, USA). All primer sets used were described in Additional file 11: Table S3. Each real-time PCR analysis was performed in triplicate.
GEO datasets used for integrative analysis
DNA methylation datasets of liver tissues derived from 18 healthy German individuals were found in GEO records (GSE48325) . These data were obtained in another study using the same HumanMethylation450 platform (GPL13534) and could be combined with our data for two adult liver tissues (NLA and NL2). Moreover, transcript datasets were also found for 10 of the 18 liver tissues in GEO records (GSE48452) and used for correlation analysis between DNA methylation and mRNA expression.
DNA methylation mapping
DNA methylation mapping was carried out for two purposes. First, in order to estimate the variable DNA methylation statuses of the 20 adult liver tissues, the β value of each CpG site was expressed as a box-and-whisker plot. Second, we aimed to identify informative CpG sites in which DNA methylation levels were different between tissues or cell lines. We constructed line graphs connecting each CpG site arranged in the 5′ to 3′ direction according to the relative location and distance of each CpG site.
Correlations between DNA methylation levels of each CpG site were examined for all CpG sites in the 5′ regulatory region by Spearman’s rank correlation coefficient in 10 liver tissues. In the DNA methylation mapping of the UGT1A isoforms, Wilcoxon signed rank test was used to compare median of 18 individual β R values of normal adult liver and one normal small intestine (NSI) or one normal fetal liver (NLF). These analyses were performed using GraphPad Prism 5 software (GraphPad Software, La Jolla, CA, USA), and differences or correlations with p values of less than 0.05 were considered significant.
Availability of supporting data
GSE67477 (DNA methylation)
GSE67484 (DNA methylation)
GSE67318 (mRNA expression)
This series is linked to GSE67485.
bone morphogenetic protein 4
combined bisulfite restriction analysis
drug metabolizing enzyme
Gene Expression Omnibus
insulin-like growth factor binding protein 3
MutL homologue 1
TATA box-binding protein
This work was supported by a Grant-in-Aid for Scientific Research (C) (24590451) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (2012–2014) and by Medical Innovation by Advanced Science and Technology (MIAST): a Grant-in-Aid for Strategic Medical Science Research Center for the Ministry of Education, Culture, Sports, Science, and Technology of Japan (2010–2014).
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. 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.
- Evans WE, Relling MV. Pharmacogenomics: translating functional genomics into rational therapeutics. Science. 1999;286:487–91.View ArticlePubMedGoogle Scholar
- Van Booven D, Marsh S, McLeod H, Carrillo MW, Sangkuhl K, Klein TE, et al. Cytochrome P450 2C9-CYP2C9. Pharmacogenet Genomics. 2010;20:277–81.PubMed CentralPubMedGoogle Scholar
- Scott SA, Sangkuhl K, Shuldiner AR, Hulotn J-S, Thorn CF, Altman RB, et al. PharmGKB summary: very important pharmacogene information for cytochrome P450, family 2, subfamily C, polypeptide 19. Pharmacogenet Genomics. 2012;22:159–65.PubMed CentralView ArticlePubMedGoogle Scholar
- Owen RP, Sangkuhl K, Klein TE, Altman RB. Cytochrome P450 2D6. Pharmacogenet Genomics. 2009;19:559–62.PubMed CentralView ArticlePubMedGoogle Scholar
- Shimada T, Yamazaki H, Mimura M, Inui Y, Guengerich P. Inter-individual variations in human liver cytochrome P-450 enzymes involved in the oxidation of drugs, carcinogens and toxic chemicals: studies with liver microsomes of 30 Japanese and 30 Caucasians. J Pharmacol Exp Ther. 1994;270:414–23.PubMedGoogle Scholar
- Lamba JK, Lin YS, Schuetz EG, Thummel KE. Genetic contribution to variable human CYP3A-mediated metabolism. Adv Drug Deliv Rev. 2002;54:1271–94.View ArticlePubMedGoogle Scholar
- Minami H, Sai K, Saeki M, Saito Y, Ozawa S, Suzuki K, et al. Irinotecan pharmacokinetics/pharmacodynamics and UGT1A genetic polymorphisms in Japanese: roles of UGT1A1*6 and *28. Pharmacogenet Genomics. 2007;17:497–504.View ArticlePubMedGoogle Scholar
- Sai K, Saeki M, Saito Y, Ozawa S, Katori N, Jinno H, et al. UGT1A1 haplotypes associated with reduced glucuronidation and increased serum bilirubin in irinotecan-administered Japanese patients with cancer. Clin Pharmacol Ther. 2004;75:501–15.View ArticlePubMedGoogle Scholar
- Li M, Wang Z, Guo J, Liu J, Li C, Liu L, et al. Clinical significance of UGT1A1 gene polymorphisms on irinotecan-based regimens as the treatment in metastatic colorectal cancer. Onco Targets Ther. 2014;7:1653–61.PubMed CentralPubMedGoogle Scholar
- Nakamura A, Nakajima M, Yamanaka H, Fujiwara R, Yokoi T. Expression of UGT1A and UGT2B mRNA in human normal tissues and various cell lines. Drug Metab Dispos. 2008;36:1461–4.View ArticlePubMedGoogle Scholar
- Strassburg CP, Strassburg A, Kneip S, Barut A, Tukey RH, Rodeck B, et al. Developmental aspects of human hepatic drug glucuronidation in young children and adults. Gut. 2002;50:259–65.PubMed CentralView ArticlePubMedGoogle Scholar
- Tukey RH, Strassburg CP. Human UDP-glucuronosyltransferases: metabolism, expression, and disease. Annu Rev Pharmacol Toxicol. 2000;40:581–616.View ArticlePubMedGoogle Scholar
- Habano W, Gamo T, Sugai T, Otsuka K, Wakabayashi G, Ozawa S. CYP1B1, but not CYP1A1, is downregulated by promoter methylation in colorectal cancers. Int J Oncol. 2009;34:1085–91.View ArticlePubMedGoogle Scholar
- Habano W, Gamo T, Terashima J, Sugai T, Otsuka K, Wakabayashi G, et al. Involvement of promoter methylation in the regulation of pregnane X receptor in colon cancer cells. BMC Cancer. 2011;11:81.PubMed CentralView ArticlePubMedGoogle Scholar
- Kim IW, Han N, Burckart GJ, Oh JM. Epigenetic changes in gene expression for drug-metabolizing enzymes and transporters. Pharmacotherapy. 2014;34:140–50.View ArticlePubMedGoogle Scholar
- Ivanov M, Kacevska M, Ingelman-Sundberg M. Epigenomics and inter-individual differences in drug response. Clin Pharmacol Ther. 2012;92:727–36.View ArticlePubMedGoogle Scholar
- Ziller MJ, Gu H, Müller 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
- Anastasiadou C, Malousi A, Maglaveras N, Kouidou S. Human epigenome data reveal increased CpG methylation in alternatively spliced sites and putative exonic splicing enhancers. DNA Cell Biol. 2011;30:267–75.View ArticlePubMedGoogle Scholar
- Edgar R, Tan PP, Portales-Casamar E, Pavlidis P. Meta-analysis of human methylomes reveals stably methylated sequences surrounding CpG islands associated with high gene expression. Epingenetics Chromatin. 2014;7:28.View ArticleGoogle Scholar
- Bièche I, Narjoz C, Asselah T, Vacher S, Marcellin P, Lidereau R, et al. Reverse transcriptase-PCR quantification of mRNA levels from cytochrome (CYP)1, CYP2 and CYP3 families in 22 different human tissues. Pharmacogenet Genomics. 2007;17:731–42.View ArticlePubMedGoogle Scholar
- Ahrens M, Ammerpohl O, von Schönfels W, Kolarova J, Bens S, Itzel T, et al. DNA methylation analysis in nonalcoholic fatty liver disease suggests distinct disease-specific and remodeling signatures after bariatric surgery. Cell Metab. 2013;18:296–302.View ArticlePubMedGoogle Scholar
- Sun Z, Chai HS, Wu Y, White WM, Donkena KV, Klein CJ, et al. Batch effect correction for genome-wide methylation data with Illumina Infinium platform. BMC Med Genomics. 2011;4:84.PubMed CentralView ArticlePubMedGoogle Scholar
- Ivanov M, Kals M, Kacevska M, Barragan I, Kasuga K, Rane A, et al. Ontogeny, distribution and potential roles of 5-hydroxymethylcytosine in human liver function. Genome Biol. 2013;14:R83.PubMed CentralView ArticlePubMedGoogle Scholar
- Kacevska M, Ivanov M, Wyss A, Kasela S, Milani L, Rane A, et al. DNA methylation dynamics in the hepatic CYP3A4 gene promoter. Biochimie. 2012;94:2338–44.View ArticlePubMedGoogle Scholar
- Choi JK. Contrasting chromatin organization of CpG islands and exons in the human genome. Genome Biol. 2010;11:R70.PubMed CentralView ArticlePubMedGoogle Scholar
- Gelfman S, Cohen N, Yearim A, Ast G. DNA-methylation effect on cotranscriptional splicing is dependent on GC architecture of the exon-intron structure. Genome Res. 2013;23:789–99.PubMed CentralView ArticlePubMedGoogle Scholar
- Oda S, Fukami T, Yokoi T, Nakajima M. Epigenetic regulation of the tissue-specific expression of human UDP-glucuronosyltransferase (UGT) 1A10. Biochem Pharmacol. 2014;87:660–7.View ArticlePubMedGoogle Scholar
- Miyajima A, Furihata T, Chiba K. Functional analysis of GC Box and its CpG methylation in the regulation of CYP1A2 gene expression. Drug Metab Pharmacokinet. 2009;24:269–76.View ArticlePubMedGoogle Scholar
- Bibikova M, Barnes B, Tsan C, Ho V, Klotzle B, Le JM, et al. High density DNA methylation array with single CpG site resolution. Genomics. 2011;98:288–95.View ArticlePubMedGoogle Scholar
- Xiong Z, Laird PW. COBRA: a sensitive and quantitative DNA methylation assay. Nucleic Acids Res. 1997;25:2532–4.PubMed CentralView ArticlePubMedGoogle Scholar