Abstract
Abnormal patterns of DNA methylation are observed in several types of human cancer. While localized DNA methylation of CpG islands has been associated with gene silencing, the effect that genome-wide loss of methylation has on tumorigenesis is not completely known. To examine its effect on tumorigenesis, we induced DNA demethylation in a rat model of human chondrosarcoma using 5-aza-2-deoxycytidine. Rat specific pyrosequencing assays were utilized to assess the methylation levels in both LINEs and satellite DNA sequences following 5-aza-2-deoxycytidine treatment. Loss of DNA methylation was accompanied by an increase in invasiveness of the rat chondrosarcoma cells, in vitro, as well as by an increase in tumor growth in vivo. Subsequent microarray analysis provided insight into the gene expression changes that result from 5-aza-2-deoxycytidine induced DNA demethylation. In particular, two genes that may function in tumorigenesis, sox-2 and midkine, were expressed at low levels in control cells but upon 5-aza-2-deoxycytidine treatment these genes became overexpressed. Promoter region DNA analysis revealed that these genes were methylated in control cells but became demethylated following 5-aza-2-deoxycytidine treatment. Following withdrawal of 5-aza-2-deoxycytidine, the rat chondrosarcoma cells reestablished global DNA methylation levels that were comparable to that of control cells. Concurrently, invasiveness of the rat chondrosarcoma cells, in vitro, decreased to a level indistinguishable to that of control cells. Taken together these experiments demonstrate that global DNA hypomethylation induced by 5-aza-2-deoxycytidine may promote specific aspects of tumorigenesis in rat chondrosarcoma cells.
Introduction
Aberrant DNA methylation is thought to play an integral role in the complex process of tumorigenesis [1]. Abnormal hypermethylation may result in the silencing of genes that are members of pathways ranging from cell division to tumor suppression. Reintroducing the expression of abnormally silenced genes may restore control of these various signaling and regulatory pathways.
Current epigenetic therapies are aimed at bringing on hypomethylation with the goal of reverting hypermethylation-induced gene silencing [2]. One such therapeutic agent is 5-Aza-2-deoxycytidine, which is a deoxycytidine analog that becomes incorporated into DNA and inhibits the activity of DNA methyltransferases [3], [4]. The incorporation of 5-Aza-2-deoxycytidine and subsequent inhibition of DNA methyltransferases results in reduced levels of DNA methylation [5], [6]. 5-Aza-2-deoxycytidine has been shown to have clinical benefits in the treatment of myelodysplastic syndrome and it has also been shown to be effective in the treatment of other myeloid malignancies [7].
Despite the potential benefits of 5-aza-2-deoxycytidine, the complete ramifications of treating tumor cells with a global DNA hypomethylating agent are unknown. The phenomenon of global DNA hypomethylation has been observed in several types of cancer [8], and DNA hypomethylation has also been associated with tumor aggressiveness [9]. Evidence also suggests that DNA hypomethylation may play a causal role in tumorigenesis [10], [11].
To study the impact of global DNA hypomethylation on the behavior of tumor cells we treated swarm rat chondrosarcoma (SRC) cells with 5-aza-2-deoxycytidine and monitored its effect both in vitro and in vivo. We selected the SRC tumor model based on its extensive characterization and the ability of the SRC cells to be grown and studied both in vitro and in vivo [12], [13], [14], [15], [16].
SRC cells were treated in vitro with a low dose of 5-Aza-2-deoxycytidine for 30 days to induce genome-wide hypomethylation and their level of methylation was assessed using rat specific pyrosequencing assays. Treatment with 5-Aza-2-deoxycytidine led to demethylation of both LINE and microsatellite regions throughout the genome. The effects of long-term exposure to epigenetic agents are not completely known, and a potential concern is that such treatment may lead to the expression of genes that are normally epigenetically silenced. In addition, it may cause illegitimate transcription events [17], [18]. Indeed, invasion assays performed with treated and untreated SRC cells indicated that loss of methylation is accompanied by an increase in invasiveness. Furthermore, microarray analysis revealed that 5-Aza-2-deoxycytidine treatment leads to alterations in expression of several developmentally regulated genes.
Based on their differential expression, two of these genes, midkine and sox-2, were selected for additional expression and epigenetic analyses. Midkine, a growth factor [19], and sox-2, a pluripotent transcription factor [20], are expressed at higher levels following 5-Aza-2-deoxycytidine treatment. Treatment with 5-Aza-2-deoxycytidine leads to loss of methylation in the promoter regions of both midkine and sox-2 genes, thus suggesting that methylation may play a role in the transcriptional regulation of these genes.
Since 5-Aza-2-deoxycytidine-induced hypomethylation resulted in several phenotypic changes in the SRC cells in vitro, we wanted to determine if the treatment would affect cell growth in vivo. Following subcutaneous transplantation, the 5-Aza-2-deoxycytidine treated SRC cells formed larger tumors than the corresponding untreated SRC cells. Methylation and expression analyses of the in vivo SRC cells revealed that the effect of 5-Aza-2-deoxycytidine could be observed for at least 60 days following treatment discontinuation.
Altogether the in vivo and in vitro results suggest that induction of genome-wide hypomethylation by 5-Aza-2-deoxycytidine results in an increase in the tumorigenicity of the SRC cells. The SRC experiments also highlight the importance that epigenetic modifications may have in cancer and suggest that DNA hypomethylation may have a functional role in tumor progression.
Materials and Methods
Ethics Statement
All animals were handled in strict accordance with good animal practice as defined by the relevant national and/or local animal welfare bodies, and all animal work was approved by the Institutional Animal Care and Use Committee (Children's Memorial Research Center; protocol IACUC #2006-30).
Establishment of a Bioluminescent Rat Chondrosarcoma Cell Line
A Murine Stem Cell Virus-Luciferase-Internal ribosomal entry site-Hygromycin (MSCV-Luc-I-Hygro) retroviral vector was prepared by transfecting 293T cells with three plasmids; pMSCV-Luc-I-Hyrgo (which encodes the Luciferase and hygromycin phosphotransferase), pEQ-Pam3(-E) (which encodes retroviral gag and pol) and pSRα-G (which encodes glycoprotein G from Vesicular Stomatitis Virus) [21]. Forty-eight hours post-transfection media containing retroviral vector was collected, aliquoted, frozen, and stored at −80°C. This vector was then used to transduce the Swarm rat chondrosarcoma cell line (SRC-LTC (Long Term Culture) [13], [obtained from Jeff W. Stevens, University of Iowa]), in the presence of 5 µg/ml polybrene on three successive days allowing the cells to recover in the media generally used overnight. Transduced cells were selected by incubation with hygromycin B (Sigma-Aldrich, St. Louis, MO) at a concentration of 500 µg/ml for 14 days. Once the hygromycin resistant population was established, the cells were maintained in media containing 500 µg/ml of hygromycin B. The newly established rat chondrosarcoma cell line was named SRC-MSCV3-LTC.
Cell Culture Conditions and 5-Aza-2-Deoxycytidine Treatment
SRC-MSCV3-LTC cells were cultured in DMEM high glucose (4.5 g glucose/ml) supplemented with 10% FBS and Penicillin/Streptomycin. Cells were plated at 2.5.×104 cells with 6 ml of media in a 25 cm2 T flask. Cells were grown until they became 80–90% confluent (6 days), and at this time the cells were trypsinized/split and plated as described. For the 5-Aza-2-deoxycytidine treatment, the media was supplemented with 0.1 uM 5-Aza-2-deoxycytidine on the day that the cells were passaged. Cell viability was assessed 72 hours and 144 hours following 5-Aza-2-deoxycytidine treatment (0.1, 0.3, 1.0, 3.0, 10, 30, 100, and 300 µM). For the 72-hour treatment, the cells were treated with 5-Aza-2-deoxycytidine at day 0 and 72 hours later the viability was examined (Figure S1A). For the 144-hour treatment, cells were treated with 5-Aza-2-deoxycytidine every 24 hours without changing the media as previously described [22], and 144 hours after the initial treatment the viability was examined (Figure S1B). Viability was determined using the Guava EasyCyte Mini Flow Cytometry System, and the Guava ViaCount Reagent (Millipore; cat no. 4000-0040).
The Swarm rat chondrosarcoma line SRC-MSCV3-LTC, was treated with 0.1 uM 5-Aza-2-deoxycytidine for 5 passages (30 days). Control cells were grown for 5 passages without 5-Aza-2-deoxycytidine. After 5 passages, cells were either frozen for subsequent DNA and RNA analysis, or they were passaged for five additional passages (30 days) without any drug treatment after which they were frozen for future analysis.
For in vivo experiments, cells were grown in vitro for 5 passages with or without 5-Aza-2-deoxycytidine. For the treated cells, the 5-Aza-2-deoxycytidine treatment was removed on the day of the injection and the cells did not receive further 5-Aza-2-deoxycytidine treatment.
Tumor Inductions
Following growth for 5 passages, cells were injected subcutaneously into the lower lumbar region of 4 week old nude mice(Males; Charles River, Strain code: 088). The SRC cells were grown until they were 80% confluent, the cells were then washed with PBS, and then cells were removed from the plate using TrypLE Express (GIBCO cat#: 12605-010) according to manufactures instructions. Following removal of SRC cells from plates, the cells were washed with PBS, centrifuged, and resuspended in PBS. Either 1×106, 5×106, or 10×106 cells were injected subcutaneously. For each experiment one animal was injected with untreated control SRC cells, and one animal was injected with 5-Aza-2-deoxycytidine treated cells. The animals did not receive any dose of 5-Aza-2-deoxycytidine.
Following the injection, the animals were monitored twice weekly for 60 days. After 60 days the animals were euthanized by CO2 gas inhalation followed by cervical dislocation. Immediately following euthanization, tumors and other tissues were frozen in liquid nitrogen or placed in paraformaldehyde for histology.
In Vivo Imaging
All in vivo imaging was performed with the Xenogen IVIS 200 imaging system. Ten minutes prior to imaging, D-luciferin (150 mg/kg of body weight) was injected into the intraperitoneal cavity of the mice. During the image acquisition animals were anesthetized with isoflurane inhalation at 1 to 2%.
Primer Design and Pyrosequencing
Rat genome sequence (rn4/version 3.4, Nov. 2004) and the annotation for repetitive elements were obtained from the UCSC Genome Database. Based on genomic co-ordinates of LINE elements provided by the UCSC database, 899,092 LINE sequences were extracted and subjected to in silico bisulfite treatment. 8,460 L1 elements with length over 6000 bp were identified and used for alignment to generate LINE nucleotide base matrix. A region within L1 elements with dense CpG dinucleotides was selected for PCR primer design. An electronic PCR was performed with the novel primers designed for rat LINEs. 827 LINE elements in the rat genome would be targeted in PCR reactions with the primer set designed. With two sequencing primers, a total number of 7 CpG dinucleotides were sequenced for each LINE element targeted. The global methylation data generated was derived from a minimum number of 5,700 CpG dinucleotides in LINE elements. A similar approach was taken to design novel primers for rat satellite repeats. Primers targeting a minimum number of 137 distinct Satellite I elements and five distinct Satellite II elements were designed. For each Satellite element targeted, the methylation profiles were determined for three CpG dinucleotides. Primer sequence and reaction conditions are available in Table S1.
Microarray
Microarray analysis was used to examine the gene expression profiles of the SRC cells (+) or (−)5-Aza-2-deoxycytidine treatment. Microarray was carried out using the NimbleGen microarray service. The Rattus norvegicus 1-plex array (14 probes/target; 26739 genes; cat#: A6184-00-01) was used for each hybridization. Two hybridizations were performed on 5-Aza-2-deoxycytidine treated SRC cells and three hybridizations were performed on untreated control SRC cells. Data were processed and displayed using Genespring software (Agilent Technologies). Genespring was used to identify differentially expressed genes that had a 5-fold difference between the 5-Aza-2-deoxycytidine treated samples and the untreated control samples. Additionally, Genespring was used to create a gene tree (Pearson coefficient) to graphically represent the data.
The list of differentially expressed genes was analyzed using a pathway-mapping program (Ingenuity Pathway Analysis version 7.0). Ingenuity was used to sort the list of differentially expressed genes (977genes) based on their role in cellular function and disease. Ingenuity identified 135 cancer related genes. For the heat map, the cancer gene list was further filtered by requiring a minimum expression level of at least 1,000 relative fluorescence units in at least 2 different hybridizations.
Genespring was used for hierarchical clustering, to create a gene tree (Pearson coefficient), and to generate the heat map used to graphically represent the data.
The list of genes with a 5-fold difference (977genes) was also analyzed using GeneGo to identify pathways that were altered following 5-Aza-2-deoxycytidine treatment.
All presented microarray data is MIAME compliant. The raw microarray data has been deposited in a MIAME compliant database. The microarray data has been deposited at GEO (GEO accession number: GSE17598).
Real-Time Quantitative PCR
Total RNA was isolated using Trizol; RNA was treated with TURBO DNA-free (Ambion Cat# AM1907). Total RNA (1 µg) was used to make cDNA with the iScript cDNA Synthesis kit (BioRad). Rat Midkine real time PCR was performed with the iQ SYBR Green Supermix (BioRad), and midkine rat specific primers (Forward: CCCAAGATGTAACCCACCAG; Reverse: GCTCACTTCCCAGAATCCC). For SYBR green PCR's, 18S-RNA was used as a reference gene [23] (Forward: GGGAGGTAGTGACGAAAAATAACAAT; Reverse: TTGCCCTCCAATGGATCCT).
Rat sox-2 real time PCR was performed with iQ Supermix (biorad) using Roche universal probe #119 (cat. no. 04693531001) and rat specific primers(forward: ATTACCCGCAGCAAAATGAC and Reverse: TTTTTGCGTTAATTTGGATGG). For PCR's with the Roche probes 18S-RNA was used as a reference gene (Probe #22 (cat. no. 04686969001 with primers: Forward: GGTGCATGGCCGTTCTTA; Reverse: TCGTTCGTTATCGGAATTAACC).
The Pfaffl method was used to calculate the normalized gene expression [24]. For each real time PCR analysis the individual sample being examined was used as the test sample in the Pfaffl method. The calibrator sample, for the Pfaffl method, was an equal mixture of cDNA from SRC control cells and 5-Aza-2-deoxycytidine SRC cells. All real time qPCR results are displayed as a ratio of the target gene relative to the reference gene, in a specific test sample, compared to the expression of the target gene relative to the reference gene in the calibrator sample.
CpG Island Identification
CpG islands in were located by searching the midkine and sox-2 genes in BLAT [25]. Each CpG island was more closely examined using “CpG Island Searcher” [26] and each island was classified as either a high-CpG promoter, an intermediate CpG promoter, or as a low-CpG promoter as previously described [27].
Analysis of DNA Methylation by Sequencing of Sodium Bisulfite-Treated DNA
Genomic DNA was obtained by digestion with proteinase K (Quiagen) followed by phenol/chloroform extraction, and was subjected to sodium bisulfite treatment to modify unmethylated cytosine to uracil using the ‘CpGenome™ DNA Modification Kit’ (Chemicon International, CA). Bissulfite-treated DNA was amplified by a nested-PCR protocol using the primers described in Table S2. PCR was performed in a volume of 25 µl containing PCR Buffer (Qiagen); 1.5 mM of MgCl2 (Qiagen); 200 µM of dNTPs (Invitrogen); 0.32 µM of each primer and 1 U of Hot Start Taq Plus DNA Polymerase (Qiagen). The PCR conditions were: 94°C for 10 min, 94°C for 3 min, 48°C for 3 min, 72°C for 2 min, one cycle; 94°C for 3 min, 50°C for 3 min, 72°C for 2 min, five cycles; and 94°C for 1 min, 52°C for 1 min, 72°C for 1 min, 35 cycles for the first reaction and the same annealing temperatures (48°, 50° and 52°C) for the nested reaction. Amplified products were purified using the Gel Purification Kit (Qiagen) and were ligated to a vector using the TOPO TA Cloning Kit (Invitrogen). Twenty-four positive clones were sequenced for each sample using the vector's forward and reverse primers. DNA sequencing reactions were performed using the ‘DNA dRhodamine Terminator Cycle Sequencing Ready reaction’ kit (Applied Biosystems) and an ABI3730xl sequencer (Applied Biosystems) according to the manufacturer's instructions.
Invasion Assay
A Membrane Invasion Culture System (MICS) was used to measure the in vitro invasiveness of all SRC cell lines as previously described [28]. Briefly, a polycarbonate membrane with 10-um pores was uniformly coated with a defined matrix. Both upper and lower wells of the chamber were filled with RPMI. SRC cells were seeded into upper wells at a concentration of 5×105 cells per well. After a 24-hour incubation in a humidified incubator at 37°C with 5% CO2, cells that had invaded through the basement membrane were collected, stained, and counted by light microscopy [29].
Statistical Analysis
Analysis of Variance (ANOVA) or two-sample t-test was used to analyze changes of DNA methylation level among different treatment conditions for Line1-S1, Line1-S2, Satellites 1 and 2, respectively. Tukey's method or Dunnett method was used to adjust p-values due to multiple comparisons in ANOVA analysis.
A linear regression method was applied to analyze tumor weight between two tumor groups (SRC Control and SRC 5AZA) after adjusting for the number of cells injected. Tumor weight and number of cells were transformed using the logarithm so that data distribution was appropriate for the analysis methods used.
We used 0.05 as the significance level for comparisons. SAS 9.1 and R software was used for data analysis and graphing.
Results
5-Aza-2-Deoxycytidine Induces Hypomethylation of LINE1 and Satellites 1 and 2
Methylation levels of cytosines in CpG dinucleotides of repetitive elements has been used as a surrogate marker for genome-wide methylation [30],in this study, LINE (Long Interspersed Element) 1 and Satellites 1 and 2 were selected as surrogate methylation markers. Rat specific pyrosequencing assays were designed to examine methylation levels of these repetitive sequences throughout the genome. The pyrosequencing assays were used to determine the methylation levels in untreated cells, in cells treated with a low dose of 5-Aza-2-deoxycytidine, and in cells that were treated with 5-Aza-2-deoxycytidine followed by an additional recovery period without the drug (30 days).
The level of DNA methylation in LINE, Satellite 1 and Satellite 2 regions of SRC cells decreases following 5-Aza-2-deoxycytidine treatment (Figure 1). The SRC cells were grown for 30 additional days after removal of 5-Aza-2-deoxycytidine. Following withdrawal of the drug, methylation was restored to levels that were indistinguishable from those of control cells based on the LINE1 and Satellite 2 assays (Figure 1). Methylation was partially restored for Satellite 1, but it did not completely regain the level that was observed in control cells. This result suggests that the demethylating effects of 5-Aza-2-deoxycytidine treatment may persist after five additional passages (30 days) without the drug.
Invasion Assay
The invasiveness of the SRC cells increased 40% following 5-Aza-2-deoxycytidine treatment (Figure 2). Thirty days post-removal of treatment, the invasive activity dropped to a level that was indistinguishable from that of control cells. The invasion assays demonstrated that 5-Aza-2-deoxycytidine-induced DNA hypomethylation leads to an increase in the in vitro invasiveness of SRC cells, and that following withdrawal of the drug, the invasive activity of the SRC cells returns to the levels observed for control cells.
Microarray Analysis
Based on the invasion assays it was hypothesized that the 5-Aza-2-deoxycytidine may alter gene expression of the SRC cells. Microarray analysis was carried out to identify changes in gene expression in untreated and in treated cells. The expression level of several genes increased after treatment with 5-Aza-2-deoxycytidine (Figure 3). Data analysis revealed that 977 genes (603 genes upregulated and 374 downregulated) exhibited a 5-fold expression difference in the untreated and treated SRC cells (see Table S3 for gene list). The pathway-mapping program, Ingenuity, was used to analyze the group of genes with a 5-fold difference in expression (See Table S4 for separate pathway analysis of up- and down regulated genes). Ingenuity revealed that the differentially expressed genes might play a role in several cancer-relevant pathways (Table 1). The top pathway, Cancer (135 genes), was selected for further analysis. A subset of the cancer related genes with a 5-fold difference are shown in a heat map (Figure 3). As illustrated by the heat map, 5-Aza-2-deoxycytidine treatment can lead to the alterations in the expression of genes that may play a role in different aspects of cancer, ranging from cell growth and proliferation, to cell cycle control, and to cell death.
Table 1. Top pathways altered following 5-Aza-2-deoxycytidine treatment.
Rank | Function and Diseases | p-value | #Molecules |
1 | Cancer | 9.84E-06 | 135 |
2 | Cellular Growth and Proliferation | 9.84E-06 | 68 |
3 | Gastrointestinal Disease | 9.84E-06 | 14 |
4 | Nervous System Development and Function | 2.23E-05 | 28 |
5 | Ophthalmic Disease | 1.21E-04 | 11 |
6 | Cellular Function and Maintenance | 3.74E-04 | 17 |
7 | Reproductive System Development and Function | 7.29E-04 | 11 |
8 | Reproductive System Disease | 7.40E-04 | 55 |
9 | Cell Cycle | 9.11E-04 | 20 |
10 | Cell Death | 1.16E-03 | 34 |
The list of genes with a 5-fold difference (977 genes) in gene expression between untreated SRC control cells and 5-Aza-2-deoxycytidine treated cells was analyzed using Ingenuity Pathway Analysis software. The top 10 functions and diseases are shown in the table. The function, its associated p-value, and the number of molecules in the specific pathway are shown.
One potential explanation of the microarray results is that 5-Aza-2-deoxycytidine treatment results in the derepression of genes that were epigenetically silenced. As a result, we may observe an increase or a decrease in expression (e.g. derepression of a negative regulator). To examine this possibility, expression and methylation analyses were performed on two of the cancer related genes (Figures 4 and 5). The genes, midkine and sox-2, were selected on the basis of their differential expression compared to control cells, and because they have CpG islands in their promoter regions (we have previously demonstrated that both sox-2 and midkine are not expressed in the control tissue, normal rat articular cartilage, data not shown; GEO: GSM25926). Midkine and sox-2 were also selected because they are developmentally regulated genes, and studies have indicated that these genes may play functional roles in stem cells [31], [32]. Therefore, we wanted to determine if 5-Aza-2-deoxycytidine-induced DNA hypomethylation could lead to the expression of stem cell related genes in the SRC cells.
Midkine and Sox-2
The increase in expression of midkine (Figure 4A) and sox-2 (Figure 5A), following exposure to 5-Aza-2-deoxycytidine, was confirmed by real-time quantitative RT-PCR. The expression of each midkine and sox-2 decreased to a level that is slightly higher than that of control cells thirty days following discontinuation of 5-Aza-2-deoxycytidine treatment. These data suggest that midkine and sox-2 expression increases as a result of exposure to 5-Aza-2-deoxycytidine, and that 30 days after discontinuation of 5-Aza-2-deoxycytidine in vitro, the expression of these genes begins to decrease. Although their expression level decreases, both midkine and sox-2 are expressed at levels that are higher than those observed in untreated control cells (Figures 4A and 5A). This suggests that SRC cells may continue to express midkine and sox-2 at high levels for at least 30 days following removal of 5-Aza-2-deoxycytidine.
Both midkine and sox-2 contain CpG islands at their transcription start sites. The CpG islands in the promoters of midkine and sox-2 can be classified as intermediate CpG islands, which is relevant because the activity of promoters containing intermediate CpG islands correlates negatively with their methylation status [27]. Additionally, intermediate CpG islands may be preferential targets for de novo methylation in somatic cells during development [27].
The methylation statuses of both midkine and sox-2 CpG islands were examined using bisulfite sequencing (Figures 4B and 5B). Two CpG islands were identified in the rat midkine gene. One CpG island encompasses the midkine transcriptional start site and the other is slightly downstream from it (Figure 4A). Both midkine CpG islands were heavily methylated in untreated SRC cells, and they became hypomethylated in 5-Aza-2-deoxycytidine treated cells (Figure 4B).
Two CpG islands were also examined for sox-2 (Figure 5A). The CpG island encompassing the sox-2 transcriptional start site (CpG 154) was not methylated in either control or treated cells (Figure S4). However, a CpG island (CpG 47) located less than 1 kb upstream of the sox-2 transcriptional start site was methylated in untreated SRC cells. Following 5-Aza-2-deoxycytidine, this CpG island became hypomethylated (Figure 5B).
These results suggest that 5-Aza-2-deoxycytidine treatment can lead to the demethylation of CpG islands at or near the transcriptional start site of midkine (Figures 4B) and sox-2 genes (Figure 5B). The decrease in CpG island methylation was accompanied by an increase in the expression of sox-2 and midkine (Figures 4A and 5A), consistent with the hypothesis that methylation may play a role in the regulation of these genes.
In Vivo Tumor Formation
SRC cells were transplanted into nude mice to test tumorigenicity following 5-Aza-2-deoxycytidine-induced DNA hypomethylation. The SRC cell line used for all aforementioned experiments stably expresses luciferase, which enables tumor growth to be examined in vivo. SRC cells treated with 5-Aza-2-deoxycytidine produced larger tumors than those induced with control SRC cells (Figure 6A and 6B).
Subcutaneous tumors were induced with 1×106, 5×106, 7×106, or 10×106 cells. In preparation for injections, SRC cells were treated in vitro for 30 days with 5-Aza-2-deoxycytidine, after which the treatment was stopped and the cells were transplanted into nude mice. Control tumors were induced with untreated SRC cells. Tumors were resected sixty days after transplantation. As it has been documented with human chondrosarcoma [33], [34], the SRC cells produced tumors with varying degrees of heterogeneity (Figure 7A and 7B). Albeit relevant, histological grading is not predictive of outcome [35], and no markers of prognostic value have been identified to date for human chondrosarcoma [36]. Hence, we examined another characteristic of the tumor, tumor weight, which demonstrated that 5-Aza-2-deoxycytidine treated cells produced larger tumors than those derived from untreated cells (Figure 6B and Table S5).
Histological analysis detected the presence of SRC cells in the lungs of mice injected with control cells and in the lungs of mice injected with 5-Aza-2-deoxycytidine treated cells (Table S5). However, it should be noted that SRC cells were detected macroscopically in the lungs of mice injected with 5-Aza-2-deoxycytidine-treated cells (3 out of 9 mice; Figure S2). SRC cells could not be detected macroscopically in the lungs of mice injected with untreated control cells. These results indicated that 5-Aza-2-deoxycytidine-treated SRC cells may grow more aggressively than untreated SRC cells in the lung of subcutaneously injected mice.
Taken together, the in vivo analyses demonstrated that DNA hypomethylation, induced by 5-Aza-2-deoxycytidine, led to the formation of more aggressive tumors than the tumors formed from untreated control SRC cells both locally, at the site of injection, and distantly, in the lungs.
Methylation of SRC Cells In Vivo
The tumors derived from untreated SRC cells were more methylated than the tumors derived from 5-Aza-2-deoxycytidine treated cells (Figures 8). In vivo the 5-Aza-2-deoxycytidine treated cells were injected and allowed to grow in vivo without treatment for 60 days. After 60 days of growth in vivo, tumors derived from 5-Aza-2-deoxycytidine-treated SRC cells exhibited a significantly lower level of methylation (Figure 8; LINE-1, LINE1-S2, Satellite 1, and Satellite 2) than that of the tumors derived from untreated control cells. This result is of note since in vitro the SRC cells were treated for 30 days and subsequently grown in vitro for 30 days without treatment, but these cells did reestablish methylation levels that were similar to control cells (Figure 1; LINE-1, LINE1-S2, and Satellite2). These results suggest that 5-Aza-2-deoxycytidine treated SRC cells more efficiently reestablish hypermethylation in vitro than in vivo.
Discussion
Aberrant DNA hypermethylation has been observed in a variety of cancers including chondrosarcoma [37], [38], [39]. 5-Aza-2-deoxycytidine treatment is thought to lead to the reactivation of aberrantly hypermethylated genes [40], and treatment of leukemias with 5-Aza-2-deoxycytidine has been shown to have clinical benefits [7]. However, genome-wide hypometylation has also been observed in several types of cancer[8], and it has been suggested that DNA hypomethylation may play a role in tumorigenesis [10], [11]. Although 5-Aza-2-deoxycytidine does have clinical benefits, one potential concern with a drug that induces DNA hypomethylation is the possibility that, in addition to reintroducing the expression of abnormally silenced genes, it may also lead to the expression of genes that are normally epigenetically silenced or it may lead to an increase of illegitimate transcription events [18]. Genome-wide derepression of transcription of genes that are normally epigenetically silenced is likely to have a dramatic impact in cancer cells that already possess abnormal genetic, epigenetic, or gene expression profiles.
In this study, we examined the effect of 5-Aza-2-deoxycytidine-induced genome-wide hypomethylation on SRC cells in vitro and in vivo, using pyrosequencing assays. As expected, treatment with 5-Aza-2-deoxycytidine led to a decrease in the global methylation levels of SRC cells. This decrease in methylation was accompanied by an increase in the invasiveness of the SRC cells in vitro. Subsequent global gene expression analysis revealed that 5-Aza-2-deoxycytidine treatment leads to (abnormal) expression of several cancer related genes.
More detailed analysis of two of the cancer related genes, sox-2 and midkine, confirmed that their expression levels increased following 5-Aza-2-deoxycytidine treatment. Methylation analysis of CpG islands at their transcription start site revealed that these genes were methylated in control cells and that they lost methylation after treatment with 5-Aza-2-deoxycytidine. This result suggests that loss of methylation may play a role in the activation of both sox-2 and midkine in 5-Aza-2-deoxycytidine-treated SRC cells. It is noteworthy that sox-2 and midkine contain “intermediate-type” CpG islands [27]. Transcriptional activity of genes with such type of CpG islands is known to correlate negatively with their level of methylation[27]. The expression of these genes may, at least in part, explain the increase in invasiveness following 5-Aza-2-deoxycytidine treatment, as both sox-2 and midkine may play roles in tumor progression [41], [42], [43].
Sox-2 and midkine may also have a function in stem cells. Midkine is involved in the growth of neuronal stem cells [32], and the expression of sox-2 has been shown to be an important factor for restoring somatic cells to a pluripotent state [31]. An intriguing possibility is that 5-Aza-2-deoxycytidine may induce the expression of genes or networks that allow the cells to acquire stem cell-like properties. Pathway analysis of the differentially expressed genes led to the identification of a network of stem cell related genes that became upregulated following 5-Aza-2-deoxycytidine treatment (Figure S5). Based on the network, sox-2 plays a role in the regulation of Dppa5 [44], Alpha crystallin B [45] and P-cadherin [46], and this is consistent with the microarray data (see Figure 3 and Table S3). Although products of these genes may have functions in stem cells, their role in the SRC cells is not known. It is important to note that these genes may have additional cellular functions. For example, besides their putative roles in stem cells, midkine and Alpha crystallin B may also play a role in drug resistance [47], [48]. These examples demonstrate the complex nature of gene expression changes that occur following 5-Aza-2-deoxycytidine treatment.
A number of genes are also downregulated following 5-Aza-2-deoxycytidine treatment. Among these are genes with diverse cellular functions (see Figure 3 and Table S3). One potential explanation for this observation is that 5-Aza-2-deoxycytidine treatment may lead to the activation of genes that negatively regulate other genes. These negative regulators may include protein-coding genes, as well as microRNA genes that could negatively regulate gene expression [49]. Another intriguing possibility is that 5-Aza-2-deoxycytidine treatment might lead to the expression of other noncoding antisense RNAs that could in turn negatively regulate gene expression [50], [51], [52], [53].
5-Aza-2-deoxycytidine-induced changes in gene expression had a significant impact on the phenotype of the SRC cells as tumors derived from 5-Aza-2-deoxycytidine treated cells produced larger tumors than tumors derived from untreated cells (Figure 6A and 6B). Tumors derived from 5-Aza-2-deoxycytidine treated cells had a lower level of methylation (60 days after tumor induction) than those derived from control untreated cells (Figure 8 This is a notable observation because in vitro, 30 days following 5-Aza-2-deoxycytidine removal, the SRC cells had reestablished a methylation level that was similar to that of control cells (Figure 1). It is possible that the in vivo microenvironment may provide more favorable growth conditions that would allow selection and/or propagation of hypomethylated cells. In vitro cells do not encounter the same selective pressure as in vivo cells do, and this difference in selective pressure may provide an explanation as to why the tumor cells maintain a lower level of methylation.
While it can be speculated that the microenvironment may exert some selective pressure on the SRC cells in vivo, the possibility cannot be ruled out that the in vitro cells may have a faster doubling time than that of the cells in vivo. The faster doubling time would presumably allow the in vitro cells to more quickly regain DNA methylation, whereas the cells in vivo may have a slower doubling time and therefore would require more time for the cells to reestablish the same methylation level that was observed in vitro.
It is important to note that 5-Aza-2-deoxycytidine is not currently used for the treatment of human chondrosarcoma, and the treatment schedule presented in this paper was designed for the treatment of cells in vitro and therefore it does not match a standard clinical treatment schedule. Consequently, the results obtained with 5-Aza-2-deoxycytidine may be specific for the SRC cell line and the conditions of the experiment.
Despite the nontraditional use of 5-Aza-2-deoxycytidine, our results suggest that genome-wide DNA hypomethylation, induced by 5-Aza-2-deoxycytidine, may actually promote certain aspects of tumorigenesis in SRC cells. This observation may initially seem counterintuitive based on the use of 5-Aza-2-deoxycytidine as a chemotherapeutic agent, but previous studies have demonstrated that 5-Aza-2-deoxycytidine can be mutagenic [54], and that DNA hypomethylation can promote the formation of tumors [10], [11]. Recent studies have also shown that chromatin modifying agents, including 5-Aza-2-deoxycytidine, are capable of inducing pluripotency associated genes [55] and it is possible to speculate that the activation of pluripotency associated genes may have a substantial impact on tumor cells. However, further studies are needed to attain a greater understanding of the effect that these epigenetic modifying drugs have on tumor cells. Finally, additional studies are needed to investigate the specific mechanisms by which genome-wide loss of methylation may promote tumorigenesis.
Supporting Information
Acknowledgments
We thank Dr. Nikia Laurie and Carl Radosevich for their assistance with experimental design and technical support with the cell viability assays.
Footnotes
Competing Interests: The authors have declared that no competing interests exist.
Funding: The Everett/O'Connor Charitable Trust; Dr. Ralph and Marian C. Falk Medical Research Trust; Gus Foundation; The Maeve McNicholas Memorial Foundation; Medical Research Institute Council. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
References
- 1.Brena RM, Costello JF. Genome-epigenome interactions in cancer. Hum Mol Genet 16 Spec No. 2007;1:R96–105. doi: 10.1093/hmg/ddm073. [DOI] [PubMed] [Google Scholar]
- 2.Mund C, Brueckner B, Lyko F. Reactivation of epigenetically silenced genes by DNA methyltransferase inhibitors: basic concepts and clinical applications. Epigenetics. 2006;1:7–13. doi: 10.4161/epi.1.1.2375. [DOI] [PubMed] [Google Scholar]
- 3.Chen L, MacMillan AM, Chang W, Ezaz-Nikpay K, Lane WS, et al. Direct identification of the active-site nucleophile in a DNA (cytosine-5)-methyltransferase. Biochemistry. 1991;30:11018–11025. doi: 10.1021/bi00110a002. [DOI] [PubMed] [Google Scholar]
- 4.Santi DV, Norment A, Garrett CE. Covalent bond formation between a DNA-cytosine methyltransferase and DNA containing 5-azacytosine. Proc Natl Acad Sci U S A. 1984;81:6993–6997. doi: 10.1073/pnas.81.22.6993. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Mund C, Hackanson B, Stresemann C, Lubbert M, Lyko F. Characterization of DNA demethylation effects induced by 5-Aza-2′-deoxycytidine in patients with myelodysplastic syndrome. Cancer Res. 2005;65:7086–7090. doi: 10.1158/0008-5472.CAN-05-0695. [DOI] [PubMed] [Google Scholar]
- 6.Yang AS, Doshi KD, Choi SW, Mason JB, Mannari RK, et al. DNA methylation changes after 5-aza-2′-deoxycytidine therapy in patients with leukemia. Cancer Res. 2006;66:5495–5503. doi: 10.1158/0008-5472.CAN-05-2385. [DOI] [PubMed] [Google Scholar]
- 7.Issa JP, Garcia-Manero G, Giles FJ, Mannari R, Thomas D, et al. Phase 1 study of low-dose prolonged exposure schedules of the hypomethylating agent 5-aza-2′-deoxycytidine (decitabine) in hematopoietic malignancies. Blood. 2004;103:1635–1640. doi: 10.1182/blood-2003-03-0687. [DOI] [PubMed] [Google Scholar]
- 8.Hoffmann MJ, Schulz WA. Causes and consequences of DNA hypomethylation in human cancer. Biochem Cell Biol. 2005;83:296–321. doi: 10.1139/o05-036. [DOI] [PubMed] [Google Scholar]
- 9.Fraga MF, Herranz M, Espada J, Ballestar E, Paz MF, et al. A mouse skin multistage carcinogenesis model reflects the aberrant DNA methylation patterns of human tumors. Cancer Res. 2004;64:5527–5534. doi: 10.1158/0008-5472.CAN-03-4061. [DOI] [PubMed] [Google Scholar]
- 10.Eden A, Gaudet F, Waghmare A, Jaenisch R. Chromosomal instability and tumors promoted by DNA hypomethylation. Science. 2003;300:455. doi: 10.1126/science.1083557. [DOI] [PubMed] [Google Scholar]
- 11.Gaudet F, Hodgson JG, Eden A, Jackson-Grusby L, Dausman J, et al. Induction of tumors in mice by genomic hypomethylation. Science. 2003;300:489–492. doi: 10.1126/science.1083558. [DOI] [PubMed] [Google Scholar]
- 12.Choi HU, Meyer K, Swarm R. Mucopolysaccharide and protein–polysaccharide of a transplantable rat chondrosarcoma. Proc Natl Acad Sci U S A. 1971;68:877–879. doi: 10.1073/pnas.68.5.877. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.King KB, Kimura JH. The establishment and characterization of an immortal cell line with a stable chondrocytic phenotype. J Cell Biochem. 2003;89:992–1004. doi: 10.1002/jcb.10571. [DOI] [PubMed] [Google Scholar]
- 14.Maibenco HC, Krehbiel RH, Nelson D. Transplantable osteogenic tumor in the rat. Cancer Res. 1967;27:362–366. [PubMed] [Google Scholar]
- 15.Morcuende JA, Huang XD, Stevens J, Kucaba TA, Brown B, et al. Identification and initial characterization of 6,000 expressed sequenced tags (ESTs) from rat normal-growing cartilage and swarm rat chondrosarcoma cDNA libraries. Iowa Orthop J. 2002;22:28–34. [PMC free article] [PubMed] [Google Scholar]
- 16.Stevens JW, Patil SR, Jordan DK, Kimura JH, Morcuende JA. Cytogenetics of swarm rat chondrosarcoma. Iowa Orthop J. 2005;25:135–140. [PMC free article] [PubMed] [Google Scholar]
- 17.Costa FF, Paixao VA, Cavalher FP, Ribeiro KB, Cunha IW, et al. SATR-1 hypomethylation is a common and early event in breast cancer. Cancer Genet Cytogenet. 2006;165:135–143. doi: 10.1016/j.cancergencyto.2005.07.023. [DOI] [PubMed] [Google Scholar]
- 18.Stresemann C, Lyko F. Modes of action of the DNA methyltransferase inhibitors azacytidine and decitabine. Int J Cancer. 2008;123:8–13. doi: 10.1002/ijc.23607. [DOI] [PubMed] [Google Scholar]
- 19.Kadomatsu K, Muramatsu T. Midkine and pleiotrophin in neural development and cancer. Cancer Lett. 2004;204:127–143. doi: 10.1016/S0304-3835(03)00450-6. [DOI] [PubMed] [Google Scholar]
- 20.Niwa H. How is pluripotency determined and maintained? Development. 2007;134:635–646. doi: 10.1242/dev.02787. [DOI] [PubMed] [Google Scholar]
- 21.Rose JK, Gallione CJ. Nucleotide sequences of the mRNA's encoding the vesicular stomatitis virus G and M proteins determined from cDNA clones containing the complete coding regions. J Virol. 1981;39:519–528. doi: 10.1128/jvi.39.2.519-528.1981. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Qin T, Youssef EM, Jelinek J, Chen R, Yang AS, et al. Effect of cytarabine and decitabine in combination in human leukemic cell lines. Clin Cancer Res. 2007;13:4225–4232. doi: 10.1158/1078-0432.CCR-06-2762. [DOI] [PubMed] [Google Scholar]
- 23.Zhu LJ, Altmann SW. mRNA and 18S-RNA coapplication-reverse transcription for quantitative gene expression analysis. Anal Biochem. 2005;345:102–109. doi: 10.1016/j.ab.2005.07.028. [DOI] [PubMed] [Google Scholar]
- 24.Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 2001;29:e45. doi: 10.1093/nar/29.9.e45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Kent WJ, Sugnet CW, Furey TS, Roskin KM, Pringle TH, et al. The human genome browser at UCSC. Genome Res. 2002;12:996–1006. doi: 10.1101/gr.229102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Takai D, Jones PA. The CpG island searcher: a new WWW resource. In Silico Biol. 2003;3:235–240. [PubMed] [Google Scholar]
- 27.Weber M, Hellmann I, Stadler MB, Ramos L, Paabo S, et al. Distribution, silencing potential and evolutionary impact of promoter DNA methylation in the human genome. Nat Genet. 2007;39:457–466. doi: 10.1038/ng1990. [DOI] [PubMed] [Google Scholar]
- 28.Hendrix MJ, Seftor EA, Seftor RE, Fidler IJ. A simple quantitative assay for studying the invasive potential of high and low human metastatic variants. Cancer Lett. 1987;38:137–147. doi: 10.1016/0304-3835(87)90209-6. [DOI] [PubMed] [Google Scholar]
- 29.Sood AK, Coffin JE, Schneider GB, Fletcher MS, DeYoung BR, et al. Biological significance of focal adhesion kinase in ovarian cancer: role in migration and invasion. Am J Pathol. 2004;165:1087–1095. doi: 10.1016/S0002-9440(10)63370-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Yang AS, Estecio MR, Doshi K, Kondo Y, Tajara EH, et al. A simple method for estimating global DNA methylation using bisulfite PCR of repetitive DNA elements. Nucleic Acids Res. 2004;32:e38. doi: 10.1093/nar/gnh032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Park IH, Zhao R, West JA, Yabuuchi A, Huo H, et al. Reprogramming of human somatic cells to pluripotency with defined factors. Nature. 2008;451:141–146. doi: 10.1038/nature06534. [DOI] [PubMed] [Google Scholar]
- 32.Zou P, Muramatsu H, Miyata T, Muramatsu T. Midkine, a heparin-binding growth factor, is expressed in neural precursor cells and promotes their growth. J Neurochem. 2006;99:1470–1479. doi: 10.1111/j.1471-4159.2006.04138.x. [DOI] [PubMed] [Google Scholar]
- 33.Soderstrom M, Aro HT, Ahonen M, Johansson N, Aho A, et al. Expression of matrix metalloproteinases and tissue inhibitors of metalloproteinases in human chondrosarcomas. Apmis. 2001;109:305–315. doi: 10.1034/j.1600-0463.2001.d01-125.x. [DOI] [PubMed] [Google Scholar]
- 34.Soderstrom M, Ekfors T, Bohling T, Aho A, Aro HT, et al. Cysteine proteinases in chondrosarcomas. Matrix Biol. 2001;19:717–725. doi: 10.1016/s0945-053x(00)00124-4. [DOI] [PubMed] [Google Scholar]
- 35.Aigner T. Towards a new understanding and classification of chondrogenic neoplasias of the skeleton–biochemistry and cell biology of chondrosarcoma and its variants. Virchows Arch. 2002;441:219–230. doi: 10.1007/s00428-002-0641-x. [DOI] [PubMed] [Google Scholar]
- 36.Lee FY, Mankin HJ, Fondren G, Gebhardt MC, Springfield DS, et al. Chondrosarcoma of bone: an assessment of outcome. J Bone Joint Surg Am. 1999;81:326–338. doi: 10.2106/00004623-199903000-00004. [DOI] [PubMed] [Google Scholar]
- 37.Asp J, Inerot S, Block JA, Lindahl A. Alterations in the regulatory pathway involving p16, pRb and cdk4 in human chondrosarcoma. J Orthop Res. 2001;19:149–154. doi: 10.1016/S0736-0266(00)00022-X. [DOI] [PubMed] [Google Scholar]
- 38.Asp J, Sangiorgi L, Inerot SE, Lindahl A, Molendini L, et al. Changes of the p16 gene but not the p53 gene in human chondrosarcoma tissues. Int J Cancer. 2000;85:782–786. doi: 10.1002/(sici)1097-0215(20000315)85:6<782::aid-ijc7>3.0.co;2-o. [DOI] [PubMed] [Google Scholar]
- 39.Ropke M, Boltze C, Neumann HW, Roessner A, Schneider-Stock R. Genetic and epigenetic alterations in tumor progression in a dedifferentiated chondrosarcoma. Pathol Res Pract. 2003;199:437–444. doi: 10.1078/0344-0338-00443. [DOI] [PubMed] [Google Scholar]
- 40.Karpf AR, Jones DA. Reactivating the expression of methylation silenced genes in human cancer. Oncogene. 2002;21:5496–5503. doi: 10.1038/sj.onc.1205602. [DOI] [PubMed] [Google Scholar]
- 41.Kato M, Maeta H, Kato S, Shinozawa T, Terada T. Immunohistochemical and in situ hybridization analyses of midkine expression in thyroid papillary carcinoma. Mod Pathol. 2000;13:1060–1065. doi: 10.1038/modpathol.3880195. [DOI] [PubMed] [Google Scholar]
- 42.Sanada Y, Yoshida K, Konishi K, Oeda M, Ohara M, et al. Expression of gastric mucin MUC5AC and gastric transcription factor SOX2 in ampulla of vater adenocarcinoma: comparison between expression patterns and histologic subtypes. Oncol Rep. 2006;15:1157–1161. [PubMed] [Google Scholar]
- 43.Tanabe K, Matsumoto M, Ikematsu S, Nagase S, Hatakeyama A, et al. Midkine and its clinical significance in endometrial carcinoma. Cancer Sci. 2008;99:1125–1130. doi: 10.1111/j.1349-7006.2008.00796.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Tanaka TS, Kunath T, Kimber WL, Jaradat SA, Stagg CA, et al. Gene expression profiling of embryo-derived stem cells reveals candidate genes associated with pluripotency and lineage specificity. Genome Res. 2002;12:1921–1928. doi: 10.1101/gr.670002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Ijichi N, Tsujimoto N, Iwaki T, Fukumaki Y, Iwaki A. Distal Sox binding elements of the alphaB-crystallin gene show lens enhancer activity in transgenic mouse embryos. J Biochem. 2004;135:413–420. doi: 10.1093/jb/mvh049. [DOI] [PubMed] [Google Scholar]
- 46.Boyer LA, Lee TI, Cole MF, Johnstone SE, Levine SS, et al. Core transcriptional regulatory circuitry in human embryonic stem cells. Cell. 2005;122:947–956. doi: 10.1016/j.cell.2005.08.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Ivanov O, Chen F, Wiley EL, Keswani A, Diaz LK, et al. alphaB-crystallin is a novel predictor of resistance to neoadjuvant chemotherapy in breast cancer. Breast Cancer Res Treat. 2008;111:411–417. doi: 10.1007/s10549-007-9796-0. [DOI] [PubMed] [Google Scholar]
- 48.Mirkin BL, Clark S, Zheng X, Chu F, White BD, et al. Identification of midkine as a mediator for intercellular transfer of drug resistance. Oncogene. 2005;24:4965–4974. doi: 10.1038/sj.onc.1208671. [DOI] [PubMed] [Google Scholar]
- 49.Lujambio A, Calin GA, Villanueva A, Ropero S, Sanchez-Cespedes M, et al. A microRNA DNA methylation signature for human cancer metastasis. Proc Natl Acad Sci U S A. 2008;105:13556–13561. doi: 10.1073/pnas.0803055105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Cayre A, Rossignol F, Clottes E, Penault-Llorca F. aHIF but not HIF-1alpha transcript is a poor prognostic marker in human breast cancer. Breast Cancer Res. 2003;5:R223–230. doi: 10.1186/bcr652. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Stuart JJ, Egry LA, Wong GH, Kaspar RL. The 3′ UTR of human MnSOD mRNA hybridizes to a small cytoplasmic RNA and inhibits gene expression. Biochem Biophys Res Commun. 2000;274:641–648. doi: 10.1006/bbrc.2000.3189. [DOI] [PubMed] [Google Scholar]
- 52.Tufarelli C, Stanley JA, Garrick D, Sharpe JA, Ayyub H, et al. Transcription of antisense RNA leading to gene silencing and methylation as a novel cause of human genetic disease. Nat Genet. 2003;34:157–165. doi: 10.1038/ng1157. [DOI] [PubMed] [Google Scholar]
- 53.Costa FF. Non-coding RNAs, epigenetics and complexity. Gene. 2008;410:9–17. doi: 10.1016/j.gene.2007.12.008. [DOI] [PubMed] [Google Scholar]
- 54.Jackson-Grusby L, Laird PW, Magge SN, Moeller BJ, Jaenisch R. Mutagenicity of 5-aza-2′-deoxycytidine is mediated by the mammalian DNA methyltransferase. Proc Natl Acad Sci U S A. 1997;94:4681–4685. doi: 10.1073/pnas.94.9.4681. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Ruau D, Ensenat-Waser R, Dinger TC, Vallabhapurapu DS, Rolletschek A, et al. Pluripotency associated genes are reactivated by chromatin-modifying agents in neurosphere cells. Stem Cells. 2008;26:920–926. doi: 10.1634/stemcells.2007-0649. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.