Development of a new in vitro assay system for evaluating the effects of chemicals on DNA methylation

Abstract

Epigenetic toxicity, a phenomenon in which chemicals exert epigenetic effects and produce toxicity, has been attracting attention in recent years due to advances in toxicology accompanying the development of life sciences. However, it has been difficult to identify epigenetic toxicants due to the lack of a simple experimental system to evaluate epigenetic toxicity. In this study, we developed a prototype of an in vitro reporter assay system for assessing the effects of chemicals on DNA methylation using two promoters showing different degrees of DNA methylation, Agouti IAP and Daz1 promoters, and a luciferase reporter. The system successfully detected DNA demethylating activity using 5-azacytidine, a chemical having DNA demethylation activity, as a positive control chemical, and demethylation of cytosine of CpG in the promoter was confirmed by pyrosequencing analysis. Next, in order to improve the detection sensitivity of the DNA demethylating activity of this system, we tried to increase the basal level of methylation of the Daz1 promoter by pre-methylase treatment of the reporter vectors. As a result, the detection sensitivity of the system was successfully improved in cells where the basal level of methylation was indeed increased by methylase treatment. Thus, the developed assay system here is effective for the simple evaluation of chemicals that affect DNA methylation.

INTRODUCTION

Currently, there are more than 100 million chemicals registered with the Chemical Abstracts Service (CAS), and the number is increasing year by year (Ideta-Otsuka et al., 2017). The safety of such a wide variety of chemicals is basically well controlled by regulations based on the judgement of toxicity profile consisting of a couple of genotoxicity and short-term general toxicity studies. However, there are a number of chemicals that have been reported to show potentials of life-long disability or fetal effect without relation to their genotoxic potentials.

Advances in life science research have brought attention to epigenetics as a mechanism for semi-permanent control of gene expression without changing the DNA sequence (Cavalli and Heard, 2019). There are various reports of cases in which the disruption of epigenetic regulation is closely related to diseases, and there is growing concern about the disruption of epigenetic regulation by chemicals and the resulting toxic effects. It has been pointed out that it is necessary to consider a new concept of toxicity called “epigenetic toxicity”, in which chemicals disrupt epigenetic regulation and exert toxic effects (Marczylo et al., 2016; Ideta-Otsuka et al., 2017). Therefore, it is of value to establish an effective means to screen “epigenetic toxicants”. It has been known that DNA methylation and histone modification are key processes for epigenetic regulation. However, the analysis of such epigenetic state of cells requires the use of various biochemical methods. For example, bisulfite treatment is necessary for detection of DNA methylation (Li and Tollefsbol, 2011), and chromatin immunoprecipitation is necessary for detection of histone modification (Heintzman et al., 2009). These analyses are incompatible with conventional toxicity tests, and not many chemicals have been reported to have epigenetic toxicity.

In the present study, we report a prototype system for evaluating the effects of chemicals on DNA methylation using a reporter system that employs quantitatively measurable luciferase as a reporter and two types of promoters showing different degrees of DNA methylation as target promoters to detect both increase and decrease in DNA methylation. In order to investigate the validity of this system, i.e., whether it can detect the DNA methylation effect of chemicals, we examined detectability of the DNA demethylation activity of chemicals using 5-azacytidine (5-azaC), which is known to have DNA demethylation activity (Okochi-Takada et al., 2018; Stresemann and Lyko, 2008). As a next step, in order to further increase the sensitivity of detection of 5-azaC activity, we attempted to increase the basal level of methylation by treating the reporter vectors with methylase before introducing them into cells. In this paper, we provide the design and validation data of our new assay system and discuss improvement points for future use.

MATERIALS AND METHODS

Cell culture and reagent preparation

Neuro-2a cells were cultured in minimum essential medium [D-MEM (High Glucose) with L-Glutamine, Phenol Red and Sodium Pyruvate; FUJIFILM Wako Pure Chemical, Osaka, Japan] supplemented with 1% Penicillin-Streptomycin (Thermo Fisher Scientific, Waltham, MA, USA) and 10% fetal bovine serum Qualified (Thermo Fisher Scientific), hereafter referred to as the cell maintenance medium. Cells were passaged every 1-2 days, and the cells with passage numbers of 3-5 were used for all the experiments. For each experiment, Neuro-2a cells cultured with the experimental medium were simultaneously exposed to 5-azaC (Sigma-Aldrich Co. LLC., St. Louis, MO, USA) at a concentration of 0, 0.08, 0.16, 0.31, 0.63, 1.25, 2.5, 5, 10, 20 μM for three consecutive days. 5-AzaC was dissolved in the cell maintenance medium at a concentration of 2 mM. 5-AzaC was then diluted with the experimental medium to yield the above-mentioned doses.

Acquisition of cell viability data and measurement of luciferase activity

The effects of 5-azaC on the viability of Neuro-2a cells were assessed using a CellTiter-Glo 2.0 Assay kit (Promega, Madison, WI, USA). In this assay, the number of viable cells in culture was determined by quantifying ATP, which indicates the presence of metabolically active cells. Luminescence activity was measured because luminescence readout is directly proportional to the number of viable cells in culture. Neuro-2a cells were seeded on 96-well plates (1562 cells/well), which were white microplate (Greiner Bio-One GmbH, Frickenhausen, Germany), and cultured in 60 μL of cell maintenance media. The plate was then kept in an incubator overnight at 37°C and in an atmosphere of 5% CO₂. Ten μL of culture medium containing 5-azaC was added on the first day, and 5 μL was added on the second and third days. The assay was performed the next day with a total of 80 μL volume of culture medium. Glo Max (Promega) was used to measure the luminescence of ATP, an indicator of cell viability. The data was obtained as the ratio of viability between the untreated control cells and cells treated with 5-azaC.

Reporter design

We used two types of promoters: one is Agouti IAP, which has a low to moderate level of basal methylation in mice (Dolinoy et al., 2007), and the other is Daz1 promoter, which has a high level of methylation in non-reproductive tissues (Liu et al., 2016). The use of these promoters may allow us to detect chemicals that increase DNA methylation with Agouti and chemicals that decrease DNA methylation with Daz1. To construct the reporter vectors, promoter genes for Agouti and Daz1, respectively, were linked to vectors expressing luciferase [NanoLuc (Promega)] for quantitative measurements and fluorescent protein tdTomato for simultaneous imaging detection. The cell viability data obtained with the CellTiter Glo kit was used to standardize the luciferase activity for creating an assay system to measure the effect of chemicals on DNA methylation.

Transfection of reporter vectors into cells

The reporter vectors were transfected into Neuro-2a cells with a combination of (luc-Agouti, td-Agouti) for the Neuro-2a-Agouti cell and (luc-Daz1, td-Daz1) for the Neuro-2a-Daz1 cell. The PiggyBac transposon system was used to introduce the vectors into the cells to expect that the complete set of reporter vectors would be incorporated into the genome.

Luciferase Assay

Neuro-2a-Agouti-d and Neuro-2a-Daz1-d cells were seeded in 96-well plates (1562 cells/well), which were made of white microplates with transparent bottoms (Greiner Bio-one), and cultured in 60 μL of cell maintenance medium. The plate was then kept in an incubator overnight at 37°C and in an atmosphere of 5% CO₂. 5-AzaC was added with the volume of 10 µL for 3 consecutive days. The assay was performed the next day with a total volume of 80 μL. The activities of two luciferases, Firefly and NanoLuc, in a single sample were quantified with the Nano-Glo Dual Luciferase Reporter Assay System (Promega) and GloMax Discover System (Promega).

Pyrosequencing analysis

The pyrosequencing technique was used for quantitative analysis of DNA methylation. Cell samples for the pyrosequencing were prepared as follows: Neuro-2a-Agouti and Neuro-2a-Daz1 cells were seeded onto 10 cm plates (0.3 x 10⁶ cells/cm) and incubated with 6 mL of cell maintenance medium at 37°C and 5% CO₂ overnight in an incubator. Next day, 1 mL of 5-azaC was added to each culture and the cells were further cultured for 3 days. After incubation, DNA and RNA were purified (All Prep DNA/RNA mini KIT; QIAGEN, Venlo, Netherlands), and 1 μg of the purified DNA was used for bisulfite treatment (Methyleasy Xceed kit, Human Genetic Signatures, Newtown, NSW, Australia). The DNA sequence to be analyzed was amplified from the bisulfite product using biotin-labeled one-sided PCR primers and made into a template. The template was used for pyrosequencing using PyroMark Q24 (QIAGEN). PyroMarkGoldQ24 (QIAGEN) was used as a reagent, and PyroMarkQ24 software (QIAGEN) was used to analyze the methylation data.

Methylase treatment of reporter vectors

Six μg of Agouti IAP and Daz1 promoter vectors were treated with 4 U of CpG methyltransferase (M.SssI) (New England Biolabs Japan Inc., Tokyo, Japan) for 30 min, and the enzyme was inactivated by heating at 65°C for 20 min. The vector was then purified by column (QIAGEN) and transfected into Neuro-2a cells.

RESULTS AND DISCUSSION

Design and construction of reporter vectors

In order to construct a cell system for easy and reliable quantitative measurement of the effects of chemicals on DNA methylation, we used Agouti IAP and Daz1 promoter as target promoters. The reason for the use of these promoters is that the basal level of methylation of Agouti IAP is expected to be low to moderate, while that of Daz1 promoter is expected to be high. Therefore, we thought we could detect the DNA hypermethylating effect of the chemical, i.e., the effect of increasing low-level methylation, using Agouti IAP, and the DNA hypomethylating effect of the chemical, i.e., the effect of decreasing high-level methylation, using Daz1 promoter. We also used NanoLuc, a luciferase with high luminescence intensity, as the reporter to enable quantitative analysis and to make the measurement as sensitive as possible. In addition, Snrpn was used as a basic minimal promoter that reflects the DNA methylation status of upstream promoters (Stelzer et al., 2015; Liu et al., 2016), and was added to the downstream of each promoter. In order to integrate the reporter DNA into the genome of the cell, the PiggyBac transposon system (Wen et al., 2014) was used, and a Core Insulator and 5’, or 3’-ITR (Inverted Terminal Repeat Sequences) were added to both ends. The reporter system constructed as described above is shown in Fig. 1A and B. The constructed reporter vectors are luc-Agouti with Agouti IAP as reporter and luc-Daz1 with Daz1 promoter as reporter. In addition, td-Agouti and td-Daz1 were also constructed using td-Tomato as a reporter that can detect DNA methylation effects of chemicals by fluorescence instead of luciferase.

Fig. 1

Schematic diagram of the DNA methylation reporter units in this study. (A) Agouti IAP promoter unit and (B) Daz1p promoter unit. (C) Illustration of how PiggyBac transposon system integrates the reporter into the TTAA sequence of the genome.

The integration of the PiggyBac transposon system into the genome is outlined in Fig. 1C. In this system, a transposase is expressed from a simultaneously transfected transposase expression vector, binds to the ITR of the reporter vector, excises the reporter cassette, and randomly incorporates it into the TTAA sequence in the genome. In this way, production of cells in which the reporter construct is integrated into the genome can be efficiently performed. We used Neuro-2a, a cell line derived from mouse neuroblastoma (Olmsted et al., 1970), as reporter cells. Neuro-2a cells containing two types of Agouti IAP vectors were named Neuro-2a-Agouti cells, and Neuro-2a cells containing two types of Daz1 promoter vectors were named Neuro-2a-Daz1 cells.

Validation of the constructed reporter system

We examined whether the constructed reporter system functions as designed by using cells in which various reporter vectors were introduced and 5-azaC as a positive chemical to have DNA demethylation effect (Tsai et al., 2012; Stresemann and Lyko, 2008). In addition, 2-HG, a substance that inhibits the activity of TET, an enzyme that hydroxylates methylated cytosine (Yang et al., 2017), was selected in anticipation of its DNA hypomethylating effect. First, 5-azaC was diluted (2-fold) from 20 μM to 0.08 μM, and the cells were treated according to the schedule shown in Fig. 2A. One plate of 5-azaC-treated cells was used to obtain cell viability data, and the other plate was used to measure luciferase activity. Luciferase activity was standardized by cell viability data, and a graph of luciferase activity (ratio standardized by cell viability data) against 5-azaC concentration was created (Fig. 2B). The results showed that the luciferase activity increased from 10 μM of 5-azaC in both Neuro-2a-Agouti cell and Neuro-2a-Daz1 cell, suggesting that the DNA demethylation activity of 5-azaC was successfully detected. Next, we performed the same analysis for 2-HG, an inhibitor of TET activity, in the hope of detecting the DNA hypermethylation effect of the chemical, but we could not detect any change in luciferase activity even in the concentration range where 2-HG completely inhibits TET activity. The reason for this is that TET is an enzyme used to change DNA methylation to demethylation in a specific genomic region (Yang et al., 2017), and Agouti IAP and Daz1 promoter are not the genomic regions on which TET acts. However, we have no candidate chemicals that have the opposite effect of 5-azaC, i.e., increase DNA methylation for various DNA regions. From the above, we concluded that although we could not detect the DNA hypermethylating effect of chemicals, we were able to show the DNA demethylating effect of chemicals.

Fig. 2

Detection of DNA demethylation activity using reporter cells. (A) Flow chart of the assay for DNA methylation effects of chemicals using reporter cells. (B) Detection of 5-azacytidine (5-azaC) activity using Neuro-2a reporter cells. NanoLuc activities driven by Agouti and Daz1 promoters were normalized by CellTiter Glo data under identical conditions. Individual data represent the mean ± standard deviation (SD) of six data sets. **P < 0.01, compared with control (0 nM).

Next, we examined whether the increase in luciferase activity detected for 5-azaC was indeed accompanied by DNA demethylation of the Agouti IAP or Daz1 promoter using pyrosequencing. The Neuro-2a-Agouti cell and Neuro-2a-Daz1 cell were treated with 10 μM 5-azaC for 3 days as shown in Fig. 2A. The genomic DNA was purified, bisulfite treated and subjected to pyrosequencing. The CpG sites measured by pyrosequencing are shown in Fig. 3A for Agouti IAP and Fig. 4A for the Daz1 promoter (cytosines in bold and underlined). As shown in Fig. 3B for Agouti IAPs, the DNA methylation of CpG1 to CpG5 tended to decrease after 5-azaC treatment, while CpG6 to CpG8 tended to increase slightly, but the reason for this is unknown. As shown in Fig. 4B, DNA methylation was significantly decreased in all CpGs measured for the Daz1 promoter. The Daz1 promoter was difficult to design primers for pyrosequencing, so only some CpG sites could be measured. In summary, the cell system constructed using the reporter vectors in this study successfully captured the demethylation activity of 5-azaC as a change in luciferase activity.

Fig. 3

Methylation changes of Agouti IAP by 5-azaC treatment. Agouti reporter-transfected Neuro-2a cells were treated with 5-azaC 10 μM for three days and the degree of DNA methylation of Agouti IAPs was quantitatively measured by pyrosequencing.

Fig. 4

Methylation changes of Daz1 promoter by 5-azaC treatment. Daz1 promoter reporter-transfected Neuro2a cells were treated with 5-azaC 10 μM for three days and the degree of DNA methylation of Daz1 promoter was quantitatively measured by pyrosequencing. Individual data represent the mean ± standard deviation (SD) of three data sets. *P < 0.05, compared with control (0 nM).

Attempt to improve the sensitivity of DNA demethylation activity detection

As shown in Fig. 2B, the detection sensitivity of 5-azaC activity is 10 μM. We judged that this detection sensitivity was still not sufficient. From Fig. 3B and Fig. 4B, the basal level of methylation of Agouti IAP and Daz1 promoter (methylation level in Control) was around 10%, which is not very high. This means that even if chemicals having DNA demethylation activity act on the promoters, DNA methylation will not be reduced to the extent that gene expression is altered, which may not lead to changes in reporter activity. Therefore, in order to increase the basal level of methylation of the Daz1 promoter, the Daz1 promoter vector was pretreated with various amounts of methylase and introduced into Neuro-2a cells by the PiggyBac transposase system. The methylation level of the Daz1 promoter was determined by pyrosequencing. As a result, the concentration-dependent increase of DNA methylation level was observed for both CpG1 and CpG2 of the Daz1 promoter as shown in Fig. 5A. 5-AzaC was used to treat Neuro-2a-Daz1 cell with various amounts of methylase pretreatment to increase the basal level of methylation, and the change in detection sensitivity was examined according to the schedule shown in Fig. 2A. In Fig. 5B, only the cases with 0U and 4U of methylase are shown. Four U methylase pretreatment indeed increased the detection sensitivity of DNA demethylation activity of 5-azaC to 5 μM. These results indicate that pre-methylase treatment of reporter vectors is effective as a strategy to improve the sensitivity of detection of DNA demethylation activity of chemicals.

Fig. 5

Improvement of detection sensitivity of 5-azaC activity by prior methylase treatment. (A) Methylase concentration-dependent increase in the degree of methylation of Daz1 promoter CpG sites. (B) Improvement of detection sensitivity of 5-azaC activity by 4U methylase pretreatment. Neuro-2a cells prepared by pre-treatment of Daz1 vector with 4U methylase were treated with 5-azaC for three days and NanoLuc activity was measured. NanoLuc activity was normalized by CellTiter Glo data. Individual data represent the mean ± standard deviation (SD) of six data sets. *P < 0.05, **P < 0.01, compared with control (0 nM).

Improving points for future use

In this study, we attempted to construct a cell system that can easily detect the DNA methylation effects of chemicals by using a reporter. We constructed a reporter vector using two types of target promoters, Agouti IAP and Daz1 promoter, and luciferase and tdTomato as reporters, and generated reporter cells by integrating the reporter cassettes into the cell genome using the PiggyBac transposase system. We found that the reporter cells using Neuro-2a can detect the DNA demethylation activity of 5-azaC as a change in luciferase activity, demonstrating that this system actually functions in cultured cells. The results suggest that it is possible to introduce this reporter system into individual mice and produce reporter mice to study the effects of chemicals on DNA methylation in vivo.

At the beginning of this study, we used HEK293T (DuBridge et al., 1987) as a cell model, but we could not detect the DNA demethylation activity of 5-azaC. We hypothesized that the reporter was preferentially introduced into the open chromatin regions in HEK293T, resulting in a low basal level of methylation, and thus the DNA demethylation activity of the chemical could not be detected. In fact, quantitative measurements of methylation levels of Agouti IAP and Daz1 promoter in reporter-transfected HEK293T by pyrosequencing showed lower levels than those in Neuro-2a (data not shown), suggesting that the open chromatin hypothesis is not entirely wrong.

After the reporter vector is integrated into the genome, the basal level of methylation of the Daz1 promoter in particular seems to increase during culture. We took this change in the basal level of methylation into account and used the relatively small number of passages of cells after integration into the genome to examine the activity of chemicals. If the methylation level is stable at a high level after a certain number of passages, it is necessary to examine the detection sensitivity of 5-azaC. In the present study, we found that a sufficiently high level of methylation at the basal level is critical for improving the detection sensitivity of DNA demethylation activity of chemicals, and therefore it may be possible to improve the detection sensitivity by increasing the number of passages.

In the present study, we used the PiggyBac transposase system to integrate the reporter vector into the cell genome; however, this system has a tendency to preferentially incorporate the reporter cassette into the TTAA sequence present in the open chromatin regions of the genome. Therefore, this system may not be suitable for improving the detection sensitivity of DNA demethylation activity of chemicals. In future, based on the results of the aforementioned study on the number of passages, we believe that investigating the transformation method of the reporter vector may also be useful for improving the sensitivity of detection of epigenetic effects of chemicals.

Initially, we used Firefly luciferase (Luc2) connected to the promoter downstream of the housekeeping gene PGK to standardize NanoLuc activity. However, since TSA (Trichostatin A) activated the PGK promoter and showed a concentration-dependent increase in Luc2 activity, we decided to use the cell viability data for standardization. In fact, it has been reported that TSA activates the promoters of housekeeping genes such as CMV (Hayashi et al., 2011), so we thought that data standardization using the promoters of housekeeping genes was not ideal. Therefore, in order to obtain cell viability data, we had to prepare separate cell plates treated under the same conditions, which resulted in increased experimental effort. In future, we would like to introduce a method for obtaining cell viability data from the same plate to improve the efficiency of experimental operations.

In the present study, we introduced Agouti IAP and Daz1 promoter into different cells to examine the activity of the chemicals. In this case, we need to prepare two types of culture plates for one chemical, one for Agouti IAP cells and one for Daz1 promoter cells. In future, we would like to investigate the possibility of using Luc2 and Renilla luciferase as quantitative luciferases other than NanoLuc to construct cells that can detect the activity of two different promoters in the same cell. If all of the above can be accomplished, it will be possible to obtain cell viability data and two types of promoter activity simultaneously with one type of reporter cell, which will greatly improve the efficiency of experimental operations.

In the present study, in addition to luciferase, a fluorescent protein, tdTomato, was used as a reporter. The reason for this is that we thought the effect of chemicals on DNA methylation can be quantified with fluorescent proteins, but measuring the activity of chemicals has proven to be difficult because fluorescent proteins are difficult to measure quantitatively with general-purpose fluorescence plate reader instruments. On the other hand, in the case of reporter mice for studying the effects of chemicals in vivo, it is expected that fluorescent proteins can be used to identify the reporter expressing cells, and we believe that tdTomato will be useful for imaging analysis of the effects of chemicals in various organs.

In conclusion, we here reported the construction of a prototype of reporter cell system that can easily detect the effects of chemicals on DNA methylation. For practical use in future, it is necessary to improve the detection sensitivity and further simplify the experimental operation by optimizing the various endpoints mentioned above. It is also necessary to construct and examine reporter mice to study the in vivo effects of chemicals. We hope that this assay system will provide useful tools for future epigenetic toxicity studies.

ACKNOWLEDGMENTS

The authors would like to thank Dr. Yoshinori Kato for his helpful suggestions on this study. This work was supported by Grant-in-Aid for Scientific Research (B) from the Japan Society for the Promotion of Science (JSPS; grant No. 17H01882) and Health Sciences Research Grants from the Ministry of Health, Labour and Welfare, Japan.

Conflict of interest

The authors declare that there is no conflict of interest.

REFERENCES

• Cavalli, G. and Heard, E. (2019): Advances in epigenetics link genetics to the environment and disease. Nature, 571, 489-499.
• Dolinoy, D.C., Huang, D. and Jirtle, R.L. (2007): Maternal nutrient supplementation counteracts bisphenol A-induced DNA hypomethylation in early development. Proc. Natl. Acad. Sci. USA, 104, 13056-13061.
• DuBridge, R.B., Tang, P., Hsia, H.C., Leong, P.M., Miller, J.H. and Calos, M.P. (1987): Analysis of mutation in human cells by using an Epstein-Barr virus shuttle system. Mol. Cell. Biol., 7, 379-387.
• Hayashi, H., Ma, Y., Kohno, T., Igarashi, M., Yasui, K., Jiew, K., Kubo, Y., Ishibashi, M., Urae, R., Irie, S. and Matsuyam, T. (2011): Effective Transgene Constructs to Enhance Gene Therapy with Trichostatin A., Targets in Gene Therapy, Yongping You, IntechOpen, DOI: 10.5772/18523. Available from: https://www.intechopen.com/books/targets-in-gene-therapy/effective-transgene-constructs-to-enhance-gene-therapy-with-trichostatin-a
• Heintzman, N.D., Hon, G.C., Hawkins, R.D., Kheradpour, P., Stark, A., Harp, L.F., Ye, Z., Lee, L.K., Stuart, R.K., Ching, C.W., Ching, K.A., Antosiewicz-Bourget, J.E., Liu, H., Zhang, X., Green, R.D., Lobanenkov, V.V., Stewart, R., Thomson, J.A., Crawford, G.E., Kellis, M. and Ren, B. (2009): Histone modifications at human enhancers reflect global cell-type-specific gene expression. Nature, 459, 108-112.
• Ideta-Otsuka, M., Igarashi, K., Narita, M. and Hirabayashi, Y. (2017): Epigenetic toxicity of environmental chemicals upon exposure during development - Bisphenol A and valproic acid may have epigenetic effects. Food Chem. Toxicol., 109, 812-816.
• Li, Y. and Tollefsbol, T.O. (2011): DNA methylation detection: bisulfite genomic sequencing analysis. Methods Mol. Biol., 791, 11-21.
• Liu, X.S., Wu, H., Ji, X., Stelzer, Y., Wu, X., Czauderna, S., Shu, J., Dadon, D., Young, R.A. and Jaenisch, R. (2016): Editing DNA Methylation in the Mammalian Genome. Cell, 167, 233-247.e17.
• Marczylo, E.L., Jacobs, M.N. and Gant, T.W. (2016): Environmentally induced epigenetic toxicity: potential public health concerns. Crit. Rev. Toxicol., 46, 676-700.
• Okochi-Takada, E., Hattori, N., Ito, A., Niwa, T., Wakabayashi, M., Kimura, K., Yoshida, M. and Ushijima, T. (2018): Establishment of a high-throughput detection system for DNA demethylating agents. Epigenetics, 13, 147-155.
• Olmsted, J.B., Carlson, K., Klebe, R., Ruddle, F. and Rosenbaum, J. (1970): Isolation of microtubule protein from cultured mouse neuroblastoma cells. Proc. Natl. Acad. Sci. USA, 65, 129-136.
• Stelzer, Y., Shivalila, C.S., Soldner, F., Markoulaki, S. and Jaenisch, R. (2015): Tracing dynamic changes of DNA methylation at single-cell resolution. Cell, 163, 218-229.
• Stresemann, C. and Lyko, F. (2008): Modes of action of the DNA methyltransferase inhibitors azacytidine and decitabine. Int. J. Cancer, 123, 8-13.
• Tsai, H.C., Li, H., Van Neste, L., Cai, Y., Robert, C., Rassool, F.V., Shin, J.J., Harbom, K.M., Beaty, R., Pappou, E., Harris, J., Yen, R.W., Ahuja, N., Brock, M.V., Stearns, V., Feller-Kopman, D., Yarmus, L.B., Lin, Y.C., Welm, A.L., Issa, J.P., Minn, I., Matsui, W., Jang, Y.Y., Sharkis, S.J., Baylin, S.B. and Zahnow, C.A. (2012): Transient low doses of DNA-demethylating agents exert durable antitumor effects on hematological and epithelial tumor cells. Cancer Cell, 21, 430-446.
• Wen, S., Zhang, H., Li, Y., Wang, N., Zhang, W., Yang, K., Wu, N., Chen, X., Deng, F., Liao, Z., Zhang, J., Zhang, Q., Yan, Z., Liu, W., Zhang, Z., Ye, J., Deng, Y., Zhou, G., Luu, H.H., Haydon, R.C., Shi, L.L., He, T.C. and Wei, G. (2014): Characterization of constitutive promoters for piggyBac transposon-mediated stable transgene expression in mesenchymal stem cells (MSCs). PLoS One, 9, e94397.
• Yang, Z., Jiang, B., Wang, Y., Ni, H., Zhang, J., Xia, J., Shi, M., Hung, L.M., Ruan, J., Mak, T.W., Li, Q. and Han, J. (2017): 2-HG Inhibits Necroptosis by Stimulating DNMT1-Dependent Hypermethylation of the RIP3 Promoter. Cell Rep., 19, 1846-1857.