Stem Cell Research 15 (2015) 435–443

Contents lists available at ScienceDirect

Stem Cell Research

j ourna l homepage: www.e lsev ie r .com/ locate /scr

Differential regulation of genomic imprinting by TET proteins inembryonic stem cells

Lizhi Liu a,b,1, Shi-Qing Mao c,1, Chelsea Ray a,b,1, Yu Zhang a,b,1, Fong T. Bell a,b, Sheau-Fang Ng a,b,Guo-Liang Xu c, Xiajun Li a,b,⁎a Black Family Stem Cell Institute, Department of Developmental and Regenerative Biology, Graduate School of Biological Sciences, Icahn School of Medicine at Mount Sinai,One Gustave L. Levy Place, New York, NY 10029, USAb Black Family Stem Cell Institute, Department of Oncological Sciences, Graduate School of Biological Sciences, Icahn School of Medicine at Mount Sinai, One Gustave L. Levy Place,New York, NY 10029, USAc State Key Laboratory of Molecular Biology, Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences, Shanghai 200031, China

⁎ Corresponding author at: Black Family Stem CDevelopmental and Regenerative Biology, Graduate SchoSchool of Medicine at Mount Sinai, One Gustave L. Levy P

E-mail address: xiajun.li@mssm.edu (X. Li).1 Equal contribution.

http://dx.doi.org/10.1016/j.scr.2015.08.0101873-5061/© 2015 The Authors. Published by Elsevier B.V

a b s t r a c t

a r t i c l e i n f o

Article history:Received 10 June 2015Received in revised form 25 August 2015Accepted 27 August 2015Available online 29 August 2015

Keywords:Genomic imprintingTETDNA methylationES cellsEBCOBRA analysisBisulfite sequencing

TET proteins have been found to play an important role in active demethylation at CpG sites in mammals. Thereare some reports implicating their functions in removal of DNA methylation imprint at the imprinted regions inthe germline. However, it is not well established whether TET proteins can also be involved in demethylation ofDNA methylation imprint in embryonic stem (ES) cells. Here we report that loss of TET proteins caused a signif-icant increase in DNA methylation at the Igf2–H19 imprinted region in ES cells. We also observed a variableincrease in DNA methylation at the Peg1 imprinted region in the ES clones devoid of TET proteins, in particularin the differentiated ES cells. By contrast, we did not observe a significant increase of DNA methylation imprintat the Peg3, Snrpn and Dlk1–Dio3 imprinted regions in ES cells lacking TET proteins. Interestingly, loss of TETproteins did not result in a significant increase of DNA methylation imprint at the Igf2–H19 and Peg1 imprintedregions in the embryoid bodies (EB). Therefore, TET proteins seem to be differentially involved in maintainingDNA methylation imprint at a subset of imprinted regions in ES cells and EBs.

© 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

Genomic imprinting is an epigenetic phenomenon characterized byparental origin-dependent expression of the imprinted genes (Barlowand Bartolomei, 2014; Bartolomei and Ferguson-Smith, 2011; Lawsonet al., 2013; Li, 2013; Peters, 2014; Tomizawa and Sasaki, 2012). Roughly150 imprinted genes have been identified in mouse and many of themare conserved in humans (Barlow and Bartolomei, 2014; Bartolomeiand Ferguson-Smith, 2011; Kelsey and Bartolomei, 2012). A large num-ber of the imprinted genes are clustered and co-regulated by a cis-actingimprinting control region called ICR (Ben-Porath and Cedar, 2000;Lewis and Reik, 2006; Robertson, 2005). ICRs are marked by germline-derived differential DNA methylation that is present at the CpG siteswithin the ICR of either paternal chromosome ormaternal chromosome(Barlow and Bartolomei, 2014; Bartolomei and Ferguson-Smith, 2011;Ciccone et al., 2009). Allelic differential DNA methylation at the ICRs is

ell Institute, Department ofol of Biological Sciences, Icahnlace, New York, NY 10029, USA.

. This is an open access article under

maintained by DNA methyltransferase complexes to prevent its lossduring cell divisions (Li et al., 1993; Li and Zhang, 2014; Moore et al.,2013; Reik et al., 2001; Tomizawa and Sasaki, 2012). Without DNAmethyltransferase complexes, the newly synthesized DNA will lackDNA methylation at the CpG sites including those at the ICRs (Kanedaet al., 2004; Klose and Bird, 2006; Law and Jacobsen, 2010; Li andZhang, 2014; Okano et al., 1999). ZFP57 and PGC7/Stella are twomaternal-effect genes necessary for the maintenance of DNA methyla-tion imprint at most imprinted regions examined (Li et al., 2008;Mackay et al., 2008; Nakamura et al., 2007; Payer et al., 2003;Quenneville et al., 2011; Strogantsev et al., 2015). Human and mouseZFP57 proteins appear to play similar roles in genomic imprinting(Mackay et al., 2008; Takikawa et al., 2013b). Our previous studieshave demonstrated that ZFP57 interacts with DNA methyltranferasesvia its cofactor KAP1/TRIM28 (Li et al., 2008; Zuo et al., 2012). Indeed,KAP1/TRIM28 also appears to be required for the maintenance of DNAmethylation imprint (Messerschmidt et al., 2012; Zuo et al., 2012).Therefore, the DNA methyltransferase complexes containing ZFP57and KAP1/TRIM28 play a major role in maintaining DNA methylationimprint (Li, 2010, 2013).

DNAmethylation can be passively lost during DNA replication if it isnot maintained by DNA methyltransferases (Chen and Riggs, 2011; Li

the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

436 L. Liu et al. / Stem Cell Research 15 (2015) 435–443

and Zhang, 2014; Moore et al., 2013; Ooi et al., 2009). Besides passiveloss due to DNA replication during cell divisions, DNA methylation canalso be subject to active demethylation via DNA repair pathway(Walsh andXu, 2006;Wuand Zhang, 2010). One recent exciting discov-ery in the epigenetics field is the presence of 5-hydroxymethylcytosinein mammalian DNA (Kriaucionis and Heintz, 2009; Tahiliani et al.,2009). It is well-established that 5-methylcytosine can be catalyticallyconverted to 5-hydroxymethylation by three mammalian TET proteins(Gu et al., 2011; Guo et al., 2014; Guo et al., 2011; He et al., 2011; Itoet al., 2011; Ko et al., 2015; Lee et al., 2014; Pastor et al., 2013;Tahiliani et al., 2009; Williams et al., 2012; Xu and Walsh, 2014). Inter-estingly, it is reported in some studies that DNA methylation imprintmay be partially erased by TET proteins in the germline during theresetting of genomic imprinting (Dawlaty et al., 2013; Hackett et al.,2013; Ko et al., 2015; Nakamura et al., 2012; Piccolo et al., 2013;Yamaguchi et al., 2013). However, passive loss of DNA methylationthrough DNA replication may be more important in erasure of the orig-inal DNA methylation imprint in the germline (Kagiwada et al., 2013).Upon fertilization, both maternal and paternal pronuclear genomesundergo locus-specific passive and active demethylation in the zygote(Guo et al., 2014; Wang et al., 2014; Xu andWalsh, 2014). The patternsof differential DNA methylation at various ICRs are reformed in thezygote, andDNAmethylation imprint is thought to be stablymaintainedin somatic cells after it is established in the germline (Barlow andBartolomei, 2014; Bartolomei and Ferguson-Smith, 2011; Tilghman,1999). Indeed, a recent whole-genome bisulfite sequencing analysishas confirmed this hypothesis although several imprinted regions maybe subject to DNA demethylation during early embryogenesis(Wang et al., 2014). PGC7/Stella has been found to protect DNAmethylation imprint from TET3-catalyzed 5mC oxidation in earlymouse embryos (Nakamura et al., 2012). Despite these advances, itis not clear whether TET proteins could also play a role in maintain-ing genomic imprinting in ES cells. Here we provide evidencesuggesting that loss of TET proteins in embryonic stem (ES) cellsmay affect the steady-state level of DNA methylation imprint at asubset of imprinted regions.

Fig. 1. The undifferentiated ES cell colonies and differentiated mature EBs of TET mutant ES clotiated ES cells and EBs derived from a TET DKO ES clone and a TET TKO ES clone were shown hcolonies grown on top of the irradiated SNL feeder cells. Arrowheads, differentiated mature EB

2. Materials and methods

2.1. Undifferentiated ES cell culture for uES samples

For undifferentiated ES cells (Fig. 1), ES cloneswere culturedwith anES cell growth medium and plated on top of the irradiated SNL feedercells that constitutively express leukemia inhibitory factor (LIF). TheES cell growth medium was made of a DMEM medium plus 15% (v/v)fetal bovine serum (FBS). ES clones were passaged onto 6-well or24-well plates once every 3–5 days. The uES genomic DNA sampleswere harvested from the undifferentiated ES cells grown on 6-wellplates when they became confluent so that only a small portion ofgenomic DNA was derived from feeder cells. We estimated that lessthan 5% of the entire population of the cells might be SNL feedercells and a vast majority of the cells used for DNA preparation shouldbe ES cells.

2.2. Differentiated ES cell culture for dES samples

For dES samples, undifferentiated ES cells grown on top of the irradi-ated SNL feeder cells were plated onto gelatin-coated 6-well or 24-wellplates without SNL feeder cells. The ES clones grown on gelatin-coatedplates for one generation were passaged onto gelatin-coated 6-well or24-well plates without SNL feeder cells for one more generation. TheES clones were mostly differentiated and free of almost all SNL feedercells after two generations on gelatin-coated plates (Fig. S1). Thesedifferentiated ES cells (dES) were harvested for preparation of dES ge-nomic DNA samples when they became confluent on gelatin-coatedplates.

2.3. Embryoid body (EB) culture for EB samples

About 1–2 million of the ES cells for these ES clones grown on top ofthe irradiated SNL feeder cells were plated onto the non-adherent 10-cm tissue culture dish plates coated with poly-hema (Sigma). An EScell growth medium without LIF was used for culturing EBs. Half of

nes displayed similar morphology to those of wild-type parental ES clone. The undifferen-ere, along with those derived from the wild-type ES clone. Arrows, undifferentiated ES cells cultured on non-adherent plates for 9 days.

437L. Liu et al. / Stem Cell Research 15 (2015) 435–443

the medium for the EB culture was changed every 2–3 days byaspiration without removing the EBs in the dish plate. The mature EBsfor each ES clone were harvested for genomic DNA preparation of EBsamples after they had been cultured on non-adherent plates for9 days (Fig. 1).

2.4. Bisulfite mutagenesis

The uES, dES and EB genomic DNA samples for each ES clone weresubjected to bisulfite treatment with the EZ DNA Methylation-Gold™Kit (Zymo Research). The bisulfite-treated DNA was used for COBRAanalysis of the imprinted regions and the non-imprinted IAP repeatregions (Fig. S2).

Fig. 2. COBRA analysis of paternally inherited DNA methylation imprint at two imprinted regiocovering the paternally inherited imprinting control region of two imprinted regions (Fig. S2)sites. uES, undifferentiated ES cell samples derived from the ES cells grown on feeder cells. dESfor two generations. EB, embryoid bodies formed after the ES cells grown on non-adherent 10-cm3, TET DKO#2 ES clone. Lane 4, TET TKO#1 ES clone. Lane 5, TET TKO#2 ES clone. u, the restricmethylated DNA. A, COBRA analysis of the H19 DMR of the Igf2–H19 imprinted region, with aused for digestion. The sizes of the restriction enzyme digestion product for the methylatedBstUI; 349 bp and 112 bp for TaqαI. B, COBRA analysis of the IG-DMR of the Dlk1–Dio3 imprinteenzymes used for digestion. The sizes of the restriction enzyme digestion product for the meth

2.5. Combined bisulfite restriction analysis (COBRA)

COBRAwas used for most analyses of DNAmethylation levels at theimprinted regions and IAP repeats in this study (Eads and Laird, 2002;Xiong and Laird, 1997). After bisulfite mutagenesis, the purifiedmutagenized genomic DNA was subjected to PCR amplification withthe primers covering a portion of the imprinting control region (ICR)for the imprinted regions or a portion of the non-imprinted IAP repeatregions (Takikawa et al., 2013a; Zuo et al., 2012). The resultant PCRproduct was used for restriction digestion for 2–3 h with the restrictionenzymes targeting the CpG sites within the amplified ICR or IAP regions(Fig. S2). Then the digested PCR product was loaded to a gel for electro-phoresis so that the undigested product indicative of unmethylatedtemplate DNA and digested product indicative of methylated template

ns. PCR amplification was performed on bisulfite-treated DNA samples with the primers. Then the PCR product was subjected to restriction enzyme digestion targeting the CpG, differentiated ES cell samples derived from the ES cells passaged on gelatin-coated plates

dish plates for 9 days. Lane 1,wild-type (WT) ES clone. Lane 2, TET DKO#1 ES clone. Lanetion enzyme product of the unmethylated DNA. m, the restriction enzyme product of thePCR product of 461 bp (Fig. S2A). ClaI, RsaI, BstUI and TaqαI are four restriction enzymesDNA are: 350 bp and 111 bp for ClaI; 340 bp and 121 bp for RsaI; 379 bp and 82 bp ford region, with a PCR product of 384 bp (Fig. S2B). HpyCH4IV and TaqαI are two restrictionylated DNA are: 280 bp and 104 bp for ClaI; 267 bp and 117 bp for TaqαI.

Fig. 3. COBRA analysis of maternally inherited DNAmethylation imprint at three imprinted regions. PCR amplification was performed on bisulfite-treated DNA samples with the primerscovering the maternally inherited imprinting control region of three imprinted regions (Fig. S2). Then the PCR product was subjected to restriction enzyme digestion targeting the CpGsites. uES, undifferentiated ES cell samples derived from the ES cells grown on feeder cells. dES, differentiated ES cell samples derived from the ES cells passaged on gelatin-coated platesfor two generations. EB, embryoid bodies formed after the ES cells grown on non-adherent 10-cmdish plates for 9 days. Lane 1,wild-type (WT) ES clone. Lane 2, TET DKO#1 ES clone. Lane3, TET DKO#2 ES clone. Lane 4, TET TKO#1 ES clone. Lane 5, TET TKO#2 ES clone. u, the restriction enzyme product of the unmethylated DNA. m, the restriction enzyme product of themethylated DNA. A, COBRA analysis of the Peg1 imprinted region, with a PCR product of 564 bp (Fig. S2C). ClaI, RsaI and HpyCH4IV are three restriction enzymes used for digestion.The sizes of the restriction enzyme digestion product for the methylated DNA are: 341 bp and 223 bp for ClaI; 348 bp and 216 bp for RsaI; 367 bp and 197 bp for HpyCH4IV. B, COBRAanalysis of the Peg3 imprinted region, with a PCR product of 375 bp (Fig. S2D). BstUI and TaqαI are two restriction enzymes used for digestion. The sizes of the restriction enzyme digestionproduct for the methylated DNA are: 223 bp and 152 bp for BstUI; 248 bp and 127 bp for TaqαI. C, COBRA analysis of the Snrpn imprinted region, with a PCR product of 375 bp (Fig. S2E).BstUI and HhaI are two restriction enzymes used for digestion. The sizes of the restriction enzyme digestion product for themethylated DNA are: 264 bp and 135 bp for BstUI; 265 bp and136 bp for HhaI.

438 L. Liu et al. / Stem Cell Research 15 (2015) 435–443

DNA were separated on the gel if the restriction enzyme sites for theunmethylated template DNA were lost after bisulfite mutagenesis(Figs. 2–3). For each imprinted region, we performed triplicateCOBRA analyses starting from the bisulfite-treated DNA samples forone restriction enzyme. The results for the triplicate COBRA areshown in Supplemental Figs. S3–S7 with statistical analysis dataincluded.

2.6. Bacterial colony bisulfite sequencing

Upon ligation, the purified bisulfite PCR product of the bisulfite-treated DNA samples was cloned into the pGEM-T vector system(Promega). After bacterial transformation, the bacterial colonies onthe dish plateswere sent for direct sequencing (Zuo et al., 2012). The se-quence results for the imprinted regions were analyzed with the web-

439L. Liu et al. / Stem Cell Research 15 (2015) 435–443

based bisulfite DNA sequence analysis program called QUMA (see thewebsite: http://quma.cdb.riken.jp/).

3. Results

TET mutant ES clones were generated in the previous study (Huet al., 2014). One wild-type ES clone, two TET DKO (Tet1−/−Tet2−/−)and two TET TKO (Tet1−/−Tet2−/−Tet3−/−) ES clones were culturedfor the samples of the undifferentiated ES cells (uES) grown on SNLfeeder cells (Fig. 1), differentiated ES cells (dES) grown on gelatin-coated plates for two generations without LIF (Fig. S1), and embryoidbodies (EB) after extended culture for 9 days on non-adherent tissueculture plates coated with poly-hema (Fig. 1). These clones werenamed WT#1, DKO#1, DKO#2, TKO#1 and TKO#2, respectively. Geno-mic DNA samples were isolated from uES cells, dES cells and EBs foreach ES clone. Then we subjected these 15 genomic DNA samples toCOBRA analysis (Takikawa et al., 2013a; Zuo et al., 2012).

3.1. DNA methylation imprint at two paternally imprinted regions

COBRA analysis was performed at the region of H19 DMR and IG-DMR of Dlk1–Dio3 imprinted region, with a PCR product of 461 bp and384 bp, respectively (Figs. S2A and S2B). Compared with that ofthe wild-type parental ES cells (WT#1), H19 DMR appeared to behypermethylated in the uES samples of two TET DKO and two TETTKO ES cell clones based on COBRA analyses of the genomic DNA sam-ples with four different restriction enzymes recognizing the distinctiveCpG sites within the H19 DMR (Fig. 2A). When quantified by ImageJ,hypermethylation was observed at the H19 DMR in the uES samples ofthese TET DKO and TET TKO ES clones compared with that of WT#1(Fig. S3). Based on the triplicate COBRA analyses of the H19 DMR withClaI, H19 DMR was significantly hypermethylated in the undifferentiat-ed ES cells of four TETmutant ES clones in comparison to those ofWT#1(Fig. S3A).

Compared with that of WT#1, the methylation level was not muchdifferent at the H19 DMR in the dES sample of TET DKO#2 (Figs. 2 andS3). By contrast, a significant increase of methylation was observed atthe H19 DMR in the dES samples of TET DKO#1 and two TET TKO ESclones in comparison with WT#1 (Figs. 2 and S3). For EBs, roughlysimilar levels of methylation were observed at the H19 DMR in WT#1,two TET DKO and two TET TKO ES clones (Figs. 2 and S3). These resultswere confirmed by statistical analyses for the COBRA of H19 DMR withClaI (Fig. S3A). Therefore, loss of TET proteins did not cause any signifi-cant increase of methylation at the H19 DMR in the TET mutant ESclones when they differentiated as EBs but it resulted in hypermethyla-tion at the CpG sites of the H19 DMR when TET mutant ES clones wereinduced to differentiate by culturing on gelatin-coated plates for twogenerations.

By large, DNA methylation imprint at the IG-DMR of Dlk1–Dio3imprinted region did not appear to be hypermethylated in the uES,dES and EB samples of two TET DKO and two TET TKO ES clones, com-pared with those of WT#1 (Figs. 2B, S4). This was confirmed by thestatistical analysis of the triplicate COBRA of the IG-DMR by TaqαI(Fig. S4A). Thus, DNA methylation imprint was relatively stable at theIG-DMR in the uES, dES and EB samples of the TET DKO and TKO ESclones, and loss of TET proteins did not apparently lead to hypermethy-lation at the IG-DMR of Dlk1–Dio3 imprinted region in either undiffer-entiated ES cells or their differentiated progeny.

3.2. DNA methylation imprint at three maternally imprinted regions

We also examined DNA methylation imprint by COBRA at threematernally imprinted regions (Peg1, Peg3 and Snrpn), with a PCR prod-uct of 564 bp, 375 bp and 401 bp, respectively (Figs. S2 and 3). Peg1DMR appeared to be mildly hypermethylated, with variable degrees,in some samples derived from TET DKO and TET TKO ES clones in

comparison with WT#1 (Figs. 3A, S5). Based on the results of triplicateCOBRA of Peg1 DMR by HpyCH4IV, a significant increase of methylationwas observed at the Peg1DMR in the uES samples of DKO#1 and TKO#2compared with that of WT#1 (Fig. S5C). Methylation appeared to be in-creased, although not statistically significant, at the Peg1DMR in the uESsamples of DKO#2 and TKO#1 in comparison with WT#1 (Fig. S5C).This increase of methylation was more apparent in the dES samples.Peg1 DMR was significantly hypermethylated in the dES samples ofDKO#1, TKO#1 and TKO#2 compared with that of WT#1 (Fig. S5C).Methylation at the Peg1DMR in the dES sample of DKO#2 also appearedto be increased in comparison with WT#1 although it was not statisti-cally significant (Fig. S5C). Methylation levels were more variable atthe Peg1 DMR in the EB samples of TET DKO and TKO ES clones. Basedon the triplicate COBRA of the Peg1 DMR by HpyCH4IV, methylationappeared to be increased in 3 out of 4 EB samples of TET DKO and TKOES clones although this increase was not statistically significant(Fig. S5C). Surprisingly, we did not observe a similar increase ofmethyl-ation at the Peg1 DMR in COBRA analyses by ClaI and RsaI (Figs. 3A, S5).Peg1 DMR appeared to be similarly methylated at the CpG site recog-nized by ClaI in almost all samples, whereas the CpG site of Peg1 DMRrecognized by RsaI exhibited higher methylation in some TET TKOsamples, in particular the dES samples of TET TKO ES clones (Figs. 3A,S5). It appears that loss of TET proteins caused a variable increase ofmethylation at the Peg1 DMR with some CpG sites being more affectedthan others. Hypermethylation of Peg1 DMR was more apparent in thedifferentiated ES samples without all three TET proteins.

We also examined DNA methylation imprint at the Peg3 imprintedregion (Fig. S2D). Although it seems that BstUI digestion for Peg3 DMRwas not as complete as TaqαI digestion, the results for these two restric-tion digestions of Peg3 DMR are largely consistent with each other(Figs. 3B and S6). Methylation appeared to be increased in the uESand dES samples of TET DKO and TKO ES clones in comparison withWT#1. However, it was not statistically significant based on triplicateCOBRA analyses of Peg3 DMR by TaqαI (Fig. S6B). Similar levels of DNAmethylation were observed at the Peg3 DMR in the EB samples derivedfrom WT#1, DKO and TKO ES clones (Figs. 3B and S6). We also per-formed bacterial colony bisulfite sequencing of the amplified bisulfitePCR product from the Peg3 DMR for these samples of WT#1, DKO#2and TKO#2 ES clones (Fig. 4). The levels of methylation at the Peg3DMR inferred from bacterial colony bisulfite sequencing are generallyconsistent with those obtained in TaqαI digestion of these samples(compare Fig. S6B with Fig. 4). Relatively higher levels of methylationwere obtained from the uES and dES samples but not from the EB sam-ples of DKO#2 and TKO#2 ES clones in comparison with those ofWT#1according to the bisulfite sequencing results of Peg3 DMR (Fig. 4). Wealso determined the sequencing results for only the unique clonesbased on the presence of uniquemethylated CpG sites or unique incom-pletely converted unmethylated cytosine (C) residues in the sequencedDNA molecules (Fig. S8). Similarly, relatively higher levels of methyla-tion were obtained from the uES and dES samples but not from the EBsamples of DKO#2 and TKO#2 ES clones compared with WT#1(Fig. S8). Taken together, loss of TET proteins seemed to cause increasedmethylation at the Peg3 DMR in the uES and dES samples but not in theEB samples of TETmutant ES clones by COBRA and bisulfite sequencing,although it was not statistically significant based on triplicate COBRAanalyses of Peg3 DMR by TaqαI.

DNA methylation imprint at the Snrpn imprinted region was morevariable in these samples based on COBRA (Fig. 3C). Methylationappeared to be increased at Snrpn DMR in some samples of TET DKOand TKO ES clones, in particular the dES samples of TET TKO ES clones(Fig. 3C and Fig. S7). However, there was no significant increase inDNA methylation at Snrpn DMR in TET DKO and TKO ES clones basedon triplicate COBRA analyses by HhaI (Fig. S7B). We also performedbisulfite sequencing analysis of SnrpnDMR for the uES and dES samplesof WT#1, DKO#2 and TKO#2 ES clones (Fig. 5). The methylation levelsat the Snrpn DMR inferred from bacterial colony bisulfite sequencing

Fig. 4. Bisulfite sequencing results of the Peg3 DMR. The bisulfite PCR product was amplified from the Peg3 DMR of the uES, dES and EB samples of the wild-type (WT), one TET DKO(DKO#2) and one TET TKO (TKO#2) ES clones after bisulfite mutagenesis, and then subjected to bacterial colony bisulfite sequencing. Filled circle, methylated CpG. Unfilled circle,unmethylated CpG. Cross (×), CpG sitewithout a clearmethylation status. Each row stands for a sequencedDNA templatemolecule froma single bacterial colony. The number underneatheach sequencing diagram indicates the percentage of all methylated CpG sites over the total number of CpG sites of the sequenced bacterial colonies for each sample.

440 L. Liu et al. / Stem Cell Research 15 (2015) 435–443

are in general agreement with those obtained in COBRA analysis(compare Fig. S7 with Fig. 5). Indeed, methylation at the Snrpn DMRwas roughly similar in the uES and dES samples of DKO#2 and WT#1,whereas it was relatively increased in those of TKO#2 ES clone(Fig. 5). We also determined the sequencing results for the Snrpn DMRof these samples based on only the unique clones containing eitherunique methylated CpG sites or unique incompletely convertedunmethylated cytosine (C) residues (Fig. S9). Compared with WT#1,Snrpn DMR was roughly similarly methylated in the uES of DKO#2but relatively hypermethylated in the dES of DKO#2, whereasboth the uES and dES samples of TKO#2 appeared to be relativelyhypermethylated at Snrpn DMR (Fig. S9). Taken together, Snrpn DMRappeared to be relatively hypermethylated in some uES and dES sam-ples of TET mutant ES clones, in particular the dES samples of TET TKOES clones, by COBRA and bisulfite sequencing. However, this increasewas variable and not statistically significant based on triplicate COBRAanalyses of Snrpn DMR by HhaI.

3.3. DNA methylation at the non-imprinted IAP repeats

In our previous study we found that IAP repeat regions were almostfully methylated in undifferentiated ES cells (Zuo et al., 2012). We per-formed COBRA analysis for IAP repeats in these samples, with a PCRproduct of 257 bp (Fig. S2F). Similarly, we found that IAP repeats werefully methylated in the uES samples derived from WT#1, TET DKO and

TET TKO ES clones (Fig. 6). It remained highly methylated in the dESand EB samples derived from these ES clones,with a slight loss of hyper-methylation observed in WT#1 and TET TKO EB samples (Fig. 6). Itseems that loss of TET proteins did not have a significant effect on thehypermethylation status at the IAP repeat regions in undifferentiatedES cells and their differentiated progeny.

4. Discussion

DNA methylation imprint is erased in the germline and re-established during gametogenesis (Barlow and Bartolomei, 2014;Bartolomei and Ferguson-Smith, 2011; Li, 2013). Upon fertilization,the patterns of differential DNA methylation at the ICRs are reformedin the zygote after germline-derived DNAmethylation imprint is passedthrough the gametes. DNA methylation imprint is stably maintainedduring embryogenesis. It has been documented in a few recent studiesthat TET proteins may play a role in partial erasure of DNAmethylationimprint at some imprinted regions in primordial germ cells (Dawlatyet al., 2013; Hackett et al., 2013; Ko et al., 2015; Nakamura et al.,2012; Yamaguchi et al., 2013). It was also reported in a recent studythat passive demethylation caused by DNA replication during cellcycle in proliferating cells may be more important than active demeth-ylation in erasure of the original DNA methylation imprint in thegermline (Kagiwada et al., 2013). Our current study demonstratedthat loss of TET proteins had a significant effect on DNA methylation

Fig. 5.Bisulfite sequencing results of the SnrpnDMR. Thebisulfite PCRproductwas amplified from the SnrpnDMRof theuES and dES samples of thewild-type (WT), one TETDKO (DKO#2)and one TET TKO (TKO#2) ES clones after bisulfite mutagenesis, and then subjected to bacterial colony bisulfite sequencing. Filled circle, methylated CpG. Unfilled circle, unmethylatedCpG. Cross (×), CpG sitewithout a clearmethylation status. Each row stands for a sequencedDNA templatemolecule from a single bacterial colony. The number underneath each sequenc-ing diagram indicates the percentage of all methylated CpG sites over the total number of CpG sites of the sequenced bacterial colonies for each sample.

441L. Liu et al. / Stem Cell Research 15 (2015) 435–443

at theH19DMR in undifferentiated ES cells and lesser so in differentiat-ed ES cells. DNA methylation imprint was variably hypermethylated atthe Peg1DMR in undifferentiated ES cells without TET proteins, rangingfrom no effect to significant hypermethylation in TET DKO and TKO ESclones. Hypermethylation became more consistently apparent at thePeg1 DMR in the differentiated ES cells derived from TET mutant ESclones, in particular TET TKO ES clones. Hypermethylation was notobserved atH19DMRwhen the TETmutant ES cloneswere differentiat-ed as EBs, and the increase ofmethylation at Peg1DMRbecame insignif-icant in the EBs of TET mutant ES clones. Loss of TET proteins did notsignificantly affect DNA methylation at Peg3 DMR and Snrpn DMRalthough methylation appeared to be increased at these two imprintedregions in the undifferentiated and differentiated ES cells of TETmutantES clones. By contrast, DNA methylation remained relatively stable atthe IG-DMR of the Dlk1–Dio3 imprinted region in either undifferentiat-ed ES cells or differentiated cells lacking TET proteins. Therefore, loss of

Fig. 6. COBRA analysis of non-imprinted IAP repeat regions. PCR amplification was performed oregions (Fig. S2F). Then the PCR product was subjected to restriction enzyme digestion targetifeeder cells. dES, differentiated ES cell samples derived from theES cells passagedon gelatin-coaadherent 10-cm dish plates for 9 days. Lane 1,wild-type (WT) ES clone. Lane 2, TET DKO#1 ES cu, the restriction enzyme product of the unmethylated DNA. m, the restriction enzyme produc

TET proteins has a variable effect on DNAmethylation imprint at differ-ent imprinted regions with some being more affected than others.

A plausible interesting observation coming out from this study isthat DNA methylation imprint at H19 DMR may be more sensitive toloss of TET proteins in undifferentiated ES cells than they are in differen-tiated cells, whereas DNA methylation imprint at Peg1 DMR may bemore prone to hypermethylation in differentiated ES cells rather thanin EBs. These effectsmay be due to the inherent differences in the stabil-ities of epigenetic modifications at these DMRs in undifferentiated EScells, differentiated ES cells or EBs. DNA methylation imprint at H19DMR may be more stably maintained in differentiated cells, whereasmethylation at Peg1 DMR may be more stable in undifferentiated EScells relative to the differentiated ES cells.

In general, the increase of methylation at the imprinted regionsmaybe less apparent in EBs than in ES cells. This could be the result ofgenome-wide DNA methylation during ES cell differentiation, similar

n bisulfite-treated DNA samples with the primers covering the non-imprinted IAP repeatng the CpG sites. uES, undifferentiated ES cell samples derived from the ES cells grown ontedplates for two generations. EB, embryoid bodies formedafter the ES cells grownonnon-lone. Lane 3, TET DKO#2 ES clone. Lane 4, TET TKO#1 ES clone. Lane 5, TET TKO#2 ES clone.t of the methylated DNA.

442 L. Liu et al. / Stem Cell Research 15 (2015) 435–443

to the gain of de novo genome-wide DNA methylation after implanta-tion in mouse embryos (Law and Jacobsen, 2010; Li and Zhang, 2014;Smith et al., 2012). This increase in the levels of de novo DNA methyla-tion during differentiation may stabilize the DNA methylation imprintat the imprinted regions (Gopalakrishnan et al., 2008; Latos et al.,2009), rendering them less susceptible to loss of TET proteins in thedifferentiated cells. These hypotheses could be tested in a future study.

Hypermethylation was more consistently observed at the H19 andPeg1 DMRs in two TET TKO ES clones than in two TET DKO ES clones.This implies that three TET proteins may play a partially redundantrole in DNA methylation imprint in ES cells although TET1 and TET2may be responsible for most TET-mediated demethylation activities inES cells. This observation is consistent with the expression patterns ofthese three TET proteins. Indeed, it is reported that notable expressionof TET1 and TET2 is very much restricted to ES cells whereas TET3 ishighly expressed in the adult tissues but only barely expressed in EScells (Koh et al., 2011; Tsagaratou and Rao, 2013) Dawlaty et al., 2014.

Hypermethylation at theH19DMR in the TETmutant ES clones (TETDKO or TKO ES clones) in our study is consistent with what was report-ed in Tet1−/−Tet2−/− DKO mutant embryos previously published byanother group (Dawlaty et al., 2013). Peg1 DMR was found to be vari-ably hypermethylated in the TET mutant ES clones. Similarly, methyla-tion at the Peg1 (also called Mest) DMR was also reported to bevariably mildly increased in a subset of Tet1−/−Tet2−/− DKO mutantembryos (Dawlaty et al., 2013). These results suggest that loss of TETproteins has similar effects on DNA methylation imprint in ES cellsand mouse embryos, and TET mutant ES clones can be used as amodel system to investigate the functions of TET proteins on genomicimprinting. This is consistent with some other previously publishedstudies using ES cells as a model system for studying genomic imprint-ing (Kohama et al., 2012; Latos et al., 2009; Mann, 2001; Quennevilleet al., 2011; Stelzer et al., 2015; Stelzer et al., 2014; Strogantsev et al.,2015; Takikawa et al., 2013b; Zuo et al., 2012). We will continue toemploy these TETmutant ES clones to dissect the dynamicmaintenancemechanisms of DNA methylation imprint in our future research.

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.scr.2015.08.010.

Acknowledgments

The work in the author's laboratory is currently supported by thegrant from NIH (R01GM093335). XL and GX conceived the study. LL,SM, CR, YZ, FB, SN and XL performed the experiments. XL wrote themanuscript.

References

Barlow, D.P., Bartolomei, M.S., 2014. Genomic imprinting in mammals. Cold Spring Harb.Perspect. Biol. 6.

Bartolomei, M.S., Ferguson-Smith, A.C., 2011. Mammalian genomic imprinting. ColdSpring Harb. Perspect. Biol. 3.

Ben-Porath, I., Cedar, H., 2000. Imprinting: focusing on the center. Curr. Opin. Genet. Dev.10, 550–554.

Chen, Z.X., Riggs, A.D., 2011. DNA methylation and demethylation in mammals. J. Biol.Chem. 286, 18347–18353.

Ciccone, D.N., Su, H., Hevi, S., Gay, F., Lei, H., Bajko, J., Xu, G., Li, E., Chen, T., 2009. KDM1B isa histone H3K4 demethylase required to establish maternal genomic imprints. Na-ture 461, 415–418.

Dawlaty, M.M., Breiling, A., Le, T., Raddatz, G., Barrasa, M.I., Cheng, A.W., Gao, Q., Powell,B.E., Li, Z., Xu, M., et al., 2013. Combined deficiency of Tet1 and Tet2 causes epigeneticabnormalities but is compatible with postnatal development. Dev. Cell 24, 310–323.

Dawlaty, M.M., Breiling, A., Le, T., Barrasa, M.I., Raddatz, G., Gao, Q., Powell, B.E., Cheng,A.W., Faull, K.F., Lyko, F., et al., 2014. Loss of Tet enzymes compromises proper differ-entiation of embryonic stem cells. Dev. Cell 29, 102–111.

Eads, C.A., Laird, P.W., 2002. Combined bisulfite restriction analysis (COBRA). MethodsMol. Biol. 200, 71–85.

Gopalakrishnan, S., Van Emburgh, B.O., Robertson, K.D., 2008. DNA methylation in devel-opment and human disease. Mutat. Res. 647, 30–38.

Gu, T.P., Guo, F., Yang, H., Wu, H.P., Xu, G.F., Liu, W., Xie, Z.G., Shi, L., He, X., Jin, S.G., et al.,2011. The role of Tet3 DNA dioxygenase in epigenetic reprogramming by oocytes.Nature 477, 606–610.

Guo, J.U., Su, Y., Zhong, C., Ming, G.L., Song, H., 2011. Emerging roles of TET proteins and 5-hydroxymethylcytosines in active DNA demethylation and beyond. Cell Cycle 10,2662–2668.

Guo, F., Li, X., Liang, D., Li, T., Zhu, P., Guo, H., Wu, X., Wen, L., Gu, T.P., Hu, B., et al., 2014.Active and passive demethylation of male and female pronuclear DNA in the mam-malian zygote. Cell Stem Cell 15, 447–458.

Hackett, J.A., Sengupta, R., Zylicz, J.J., Murakami, K., Lee, C., Down, T.A., Surani, M.A., 2013.Germline DNA demethylation dynamics and imprint erasure through 5-hydroxymethylcytosine. Science 339, 448–452.

He, Y.F., Li, B.Z., Li, Z., Liu, P., Wang, Y., Tang, Q., Ding, J., Jia, Y., Chen, Z., Li, L., et al., 2011.Tet-mediated formation of 5-carboxylcytosine and its excision by TDG inmammalianDNA. Science 333, 1303–1307.

Hu, X., Zhang, L., Mao, S.Q., Li, Z., Chen, J., Zhang, R.R., Wu, H.P., Gao, J., Guo, F., Liu, W., etal., 2014. Tet and TDG mediate DNA demethylation essential for mesenchymal-to-epithelial transition in somatic cell reprogramming. Cell Stem Cell 14, 512–522.

Ito, S., Shen, L., Dai, Q., Wu, S.C., Collins, L.B., Swenberg, J.A., He, C., Zhang, Y., 2011. Tetproteins can convert 5-methylcytosine to 5-formylcytosine and 5-carboxylcytosine.Science 333, 1300–1303.

Kagiwada, S., Kurimoto, K., Hirota, T., Yamaji, M., Saitou, M., 2013. Replication-coupledpassive DNA demethylation for the erasure of genome imprints in mice. EMBO J.32, 340–353.

Kaneda, M., Okano, M., Hata, K., Sado, T., Tsujimoto, N., Li, E., Sasaki, H., 2004. Essential rolefor de novo DNAmethyltransferase Dnmt3a in paternal andmaternal imprinting. Na-ture 429, 900–903.

Kelsey, G., Bartolomei, M.S., 2012. Imprinted genes … and the number is? PLoS Genet. 8,e1002601.

Klose, R.J., Bird, A.P., 2006. Genomic DNAmethylation: the mark and its mediators. TrendsBiochem. Sci. 31, 89–97.

Ko, M., An, J., Pastor, W.A., Koralov, S.B., Rajewsky, K., Rao, A., 2015. TET proteins and 5-methylcytosine oxidation in hematological cancers. Immunol. Rev. 263, 6–21.

Koh, K.P., Yabuuchi, A., Rao, S., Huang, Y., Cunniff, K., Nardone, J., Laiho, A., Tahiliani, M.,Sommer, C.A., Mostoslavsky, G., et al., 2011. Tet1 and Tet2 regulate 5-hydroxymethylcytosine production and cell lineage specification in mouse embryon-ic stem cells. Cell Stem Cell 8, 200–213.

Kohama, C., Kato, H., Numata, K., Hirose, M., Takemasa, T., Ogura, A., Kiyosawa, H., 2012.ES cell differentiation system recapitulates the establishment of imprinted gene ex-pression in a cell-type-specific manner. Hum. Mol. Genet. 21, 1391–1401.

Kriaucionis, S., Heintz, N., 2009. The nuclear DNA base 5-hydroxymethylcytosine is pres-ent in Purkinje neurons and the brain. Science 324, 929–930.

Latos, P.A., Stricker, S.H., Steenpass, L., Pauler, F.M., Huang, R., Senergin, B.H., Regha, K.,Koerner, M.V., Warczok, K.E., Unger, C., et al., 2009. An in vitro ES cell imprintingmodel shows that imprinted expression of the Igf2r gene arises from an allele-specific expression bias. Development 136, 437–448.

Law, J.A., Jacobsen, S.E., 2010. Establishing, maintaining and modifying DNA methylationpatterns in plants and animals. Nat. Rev. Genet. 11, 204–220.

Lawson, H.A., Cheverud, J.M., Wolf, J.B., 2013. Genomic imprinting and parent-of-origineffects on complex traits. Nat. Rev. Genet. 14, 609–617.

Lee, H.J., Hore, T.A., Reik, W., 2014. Reprogramming the methylome: erasing memory andcreating diversity. Cell Stem Cell 14, 710–719.

Lewis, A., Reik, W., 2006. How imprinting centres work. Cytogenet. Genome Res. 113,81–89.

Li, X., 2010. Extending the maternal–zygotic effect with genomic imprinting. Mol. Hum.Reprod. 16, 695–703.

Li, X., 2013. Genomic imprinting is a parental effect established in mammalian germ cells.Curr. Top. Dev. Biol. 102, 35–59.

Li, E., Zhang, Y., 2014. DNA methylation in mammals. Cold Spring Harb. Perspect. Biol. 6,a019133.

Li, E., Beard, C., Jaenisch, R., 1993. Role for DNA methylation in genomic imprinting.Nature 366, 362–365.

Li, X., Ito, M., Zhou, F., Youngson, N., Zuo, X., Leder, P., Ferguson-Smith, A.C., 2008. Amaternal–zygotic effect gene, Zfp57, maintains both maternal and paternal imprints.Dev. Cell 15, 547–557.

Mackay, D.J., Callaway, J.L., Marks, S.M.,White, H.E., Acerini, C.L., Boonen, S.E., Dayanikli, P.,Firth, H.V., Goodship, J.A., Haemers, A.P., et al., 2008. Hypomethylation of multipleimprinted loci in individuals with transient neonatal diabetes is associated with mu-tations in ZFP57. Nat. Genet. 40, 949–951.

Mann, J.R., 2001. Deriving and propagating mouse embryonic stem cell lines for studyinggenomic imprinting. Methods Mol. Biol. 181, 21–39.

Messerschmidt, D.M., de Vries, W., Ito, M., Solter, D., Ferguson-Smith, A., Knowles, B.B.,2012. Trim28 is required for epigenetic stability during mouse oocyte to embryotransition. Science 335, 1499–1502.

Moore, L.D., Le, T., Fan, G., 2013. DNA methylation and its basic function.Neuropsychopharmacology 38, 23–38.

Nakamura, T., Arai, Y., Umehara, H., Masuhara, M., Kimura, T., Taniguchi, H., Sekimoto, T.,Ikawa, M., Yoneda, Y., Okabe, M., et al., 2007. PGC7/Stella protects against DNA de-methylation in early embryogenesis. Nat. Cell Biol. 9, 64–71.

Nakamura, T., Liu, Y.J., Nakashima, H., Umehara, H., Inoue, K., Matoba, S., Tachibana, M.,Ogura, A., Shinkai, Y., Nakano, T., 2012. PGC7 binds histone H3K9me2 to protectagainst conversion of 5mC to 5hmC in early embryos. Nature 486, 415–419.

Okano, M., Bell, D.W., Haber, D.A., Li, E., 1999. DNA methyltransferases Dnmt3a andDnmt3b are essential for de novo methylation and mammalian development. Cell99, 247–257.

Ooi, S.K., O'Donnell, A.H., Bestor, T.H., 2009. Mammalian cytosine methylation at a glance.J. Cell Sci. 122, 2787–2791.

Pastor, W.A., Aravind, L., Rao, A., 2013. TETonic shift: biological roles of TET proteins inDNA demethylation and transcription. Nat. Rev. Mol. Cell Biol. 14, 341–356.

443L. Liu et al. / Stem Cell Research 15 (2015) 435–443

Payer, B., Saitou, M., Barton, S.C., Thresher, R., Dixon, J.P., Zahn, D., Colledge, W.H., Carlton,M.B., Nakano, T., Surani, M.A., 2003. Stella is a maternal effect gene required for nor-mal early development in mice. Curr. Biol. 13, 2110–2117.

Peters, J., 2014. The role of genomic imprinting in biology and disease: an expandingview. Nat. Rev. Genet. 15, 517–530.

Piccolo, F.M., Bagci, H., Brown, K.E., Landeira, D., Soza-Ried, J., Feytout, A., Mooijman, D.,Hajkova, P., Leitch, H.G., Tada, T., et al., 2013. Different roles for Tet1 and Tet2 proteinsin reprogramming-mediated erasure of imprints induced by EGC fusion. Mol. Cell 49,1023–1033.

Quenneville, S., Verde, G., Corsinotti, A., Kapopoulou, A., Jakobsson, J., Offner, S., Baglivo, I.,Pedone, P.V., Grimaldi, G., Riccio, A., et al., 2011. In embryonic stem cells, ZFP57/KAP1recognize a methylated hexanucleotide to affect chromatin and DNA methylation ofimprinting control regions. Mol. Cell 44, 361–372.

Reik, W., Dean, W., Walter, J., 2001. Epigenetic reprogramming in mammalian develop-ment. Science 293, 1089–1093.

Robertson, K.D., 2005. DNA methylation and human disease. Nat. Rev. Genet. 6, 597–610.Smith, Z.D., Chan, M.M., Mikkelsen, T.S., Gu, H., Gnirke, A., Regev, A., Meissner, A., 2012. A

unique regulatory phase of DNA methylation in the early mammalian embryo. Na-ture 484, 339–344.

Stelzer, Y., Sagi, I., Yanuka, O., Eiges, R., Benvenisty, N., 2014. The noncoding RNA IPW reg-ulates the imprinted DLK1–DIO3 locus in an induced pluripotent stem cell model ofPrader–Willi syndrome. Nat. Genet. 46, 551–557.

Stelzer, Y., Bar, S., Bartok, O., Afik, S., Ronen, D., Kadener, S., Benvenisty, N., 2015. Differen-tiation of human parthenogenetic pluripotent stem cells reveals multiple tissue- andisoform-specific imprinted transcripts. Cell Rep. 11, 308–320.

Strogantsev, R., Krueger, F., Yamazawa, K., Shi, H., Gould, P., Goldman-Roberts, M.,McEwen, K., Sun, B., Pedersen, R., Ferguson-Smith, A.C., 2015. Allele-specific bindingof ZFP57 in the epigenetic regulation of imprinted and non-imprinted monoallelicexpression. Genome Biol. 16, 112.

Tahiliani, M., Koh, K.P., Shen, Y., Pastor, W.A., Bandukwala, H., Brudno, Y., Agarwal, S., Iyer,L.M., Liu, D.R., Aravind, L., et al., 2009. Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1. Science 324,930–935.

Takikawa, S., Ray, C., Wang, X., Shamis, Y., Wu, T.Y., Li, X., 2013a. Genomic imprinting isvariably lost during reprogramming of mouse iPS cells. Stem Cell Res. 11, 861–873.

Takikawa, S., Wang, X., Ray, C., Vakulenko, M., Bell, F.T., Li, X., 2013b. Human and mouseZFP57 proteins are functionally interchangeable in maintaining genomic imprintingat multiple imprinted regions in mouse ES cells. Epigenetics 8, 1268–1279.

Tilghman, S.M., 1999. The sins of the fathers and mothers: genomic imprinting in mam-malian development. Cell 96, 185–193.

Tomizawa, S., Sasaki, H., 2012. Genomic imprinting and its relevance to congenitaldisease, infertility, molar pregnancy and induced pluripotent stem cell. J. Hum.Genet. 57, 84–91.

Tsagaratou, A., Rao, A., 2013. TET proteins and 5-methylcytosine oxidation in the immunesystem. Cold Spring Harb. Symp. Quant. Biol. 78, 1–10.

Walsh, C.P., Xu, G.L., 2006. Cytosine methylation and DNA repair. Curr. Top. Microbiol.Immunol. 301, 283–315.

Wang, L., Zhang, J., Duan, J., Gao, X., Zhu,W., Lu, X., Yang, L., Li, G., Ci, W., Li, W., et al., 2014.Programming and inheritance of parental DNA methylomes in mammals. Cell 157,979–991.

Williams, K., Christensen, J., Helin, K., 2012. DNA methylation: TET proteins-guardians ofCpG islands? EMBO Rep. 13, 28–35.

Wu, S.C., Zhang, Y., 2010. Active DNA demethylation: many roads lead to Rome. Nat. Rev.Mol. Cell Biol. 11, 607–620.

Xiong, Z., Laird, P.W., 1997. COBRA: a sensitive and quantitative DNA methylation assay.Nucleic Acids Res. 25, 2532–2534.

Xu, G.L., Walsh, C.P., 2014. Enzymatic DNA oxidation: mechanisms and biological signifi-cance. BMB Rep. 47, 609–618.

Yamaguchi, S., Shen, L., Liu, Y., Sendler, D., Zhang, Y., 2013. Role of Tet1 in erasure ofgenomic imprinting. Nature 504, 460–464.

Zuo, X., Sheng, J., Lau, H.T., McDonald, C.M., Andrade, M., Cullen, D.E., Bell, F.T., Iacovino,M., Kyba, M., Xu, G., et al., 2012. Zinc finger protein ZFP57 requires its co-factor to re-cruit DNA methyltransferases and maintains DNA methylation imprint in embryonicstem cells via its transcriptional repression domain. J. Biol. Chem. 287, 2107–2118.