Dnmt2 is not required for de novo and maintenance methylation of viral DNA in embryonic stem cells (original) (raw)

Journal Article

,

Cardiovascular Research Center, Massachusetts General Hospital, Department of Medicine,Harvard Medical School

,

149, 13th Street, Charlestown, MA 02129, USA

Search for other works by this author on:

,

Cardiovascular Research Center, Massachusetts General Hospital, Department of Medicine,Harvard Medical School

,

149, 13th Street, Charlestown, MA 02129, USA

Search for other works by this author on:

Cardiovascular Research Center, Massachusetts General Hospital, Department of Medicine,Harvard Medical School

,

149, 13th Street, Charlestown, MA 02129, USA

*To whom correspondence should be addressed. Tel: +1 617 726 4345; Fax: +1 617 726 5806; Email: en@cvrc.mgh.harvard.edu

Search for other works by this author on:

• PDF
• Split View
• • Article contents
  • Figures & tables
  • Video
  • Audio
  • Supplementary Data
• Cite

Cite

Masaki Okano, Shaoping Xie, En Li, Dnmt2 is not required for de novo and maintenance methylation of viral DNA in embryonic stem cells, Nucleic Acids Research, Volume 26, Issue 11, 1 June 1998, Pages 2536–2540, https://doi.org/10.1093/nar/26.11.2536
Close

Navbar Search Filter Mobile Enter search term Search

Abstract

We have shown previously that de novo methylation activities persist in mouse embryonic stem (ES) cells homozygous for a null mutation of Dnmt1 that encodes the major DNA cytosine methyltransferase. In this study, we have cloned a putative mammalian DNA methyltransferase gene, termed Dnmt2, that is homologous to pmt1 of fission yeast. Different from pmt1 in which the catalytic Pro-Pro-Cys (PPC) motif is ‘mutated’ to Pro-Ser-Cys, Dnmt2 contains all the conserved methyltransferase motifs, thus likely encoding a functional cytosine methyltransferase. However, baculovirus-expressed Dnmt2 protein failed to methylate DNA in vitro. To investigate whether Dnmt2 functions as a DNA methyltransferase in vivo, we inactivated the Dnmt2 gene by targeted deletion of the putative catalytic PPC motif in ES cells. We showed that endogenous virus was fully methylated in _Dnmt2_-deficient mutant ES cells. Furthermore, newly integrated retrovirus DNA was methylated de novo in infected mutant ES cells as efficiently as in wild-type cells. These results indicate that Dnmt2 is not essential for global de novo or maintenance methylation of DNA in ES cells.

Introduction

DNA methylation at the C-5 position of cytosine in CpG dinucleotides is the major form of DNA modification in vertebrate animals. DNA methylation has been shown to be essential for mammalian development as inactivation of Dnmt1, a major maintenance DNA cytosine methyltransferase, results in genome-wide demethylation and embryonic lethality (1,2). The function of DNA methylation has been implicated in a diverse range of biological processes. Molecular and genetic studies have demonstrated that DNA methylation plays critical roles in regulation of parent-origin-specific expression of imprinted genes (3–5) and X chromosome inactivation (6,7). Recently, growing evidence also suggests that DNA methylation is involved in carcinogenesis (8–10).

While the function of DNA methylation has been studied extensively, mechanisms by which DNA methylation is regulated and tissue specific DNA methylation patterns are established during development are poorly understood. The major obstacle has been the lack of information about the enzymes that catalyze de novo methylation and demethylation (11). The enzyme encoded by Dnmt1 functions primarily as a maintenance methyltransferase which transfers methyl groups to cytosine in hemi-methylated CpG sites after DNA replication (12). Although this enzyme can also methylate unmethylated DNA in vitro, no evidence has been established so far for its role as a de novo methyltransferase in vivo. Recently, we showed that ES cells homozygous for a null mutation of Dnmt1 contained residual levels of methyl cytosine and retained the ability to methylate provirus DNA de novo (2). This result provides the first genetic evidence for the existence of an independently encoded de novo DNA methyltransferase in mammalian cells.

In this study, we report the cloning of a mammalian gene Dnmt2 that shares homology with the pmt1 gene of fission yeast (13), and encodes a protein which contains all the conserved methyltransferase motifs. We provide genetic evidence that Dnmt2 is not essential for maintenance methylation nor for de novo methylation of viral DNA in ES cells.

Materials and Methods

Cloning of the mammalian Dnmt2 gene

A search of the dbEST database was performed with the TBLASTN program (14) using bacterial cytosine methyltransferases as queries. Two human EST clones (GenBank accession nos N31314 and R95731) were found to match the M.HgiGI sequences. The clones were obtained from American Type Culture Collection (ATCC, MD) and sequenced by the MGH sequencing core facility. The deduced amino acid sequences of the clones share a significant homology with the yeast pmt1 (13). The insert DNA of these clones were cut out by _Eco_RI/_Not_I digestion and used as probes for screening cDNA libraries.

Nine positive clones were obtained by screening a mouse ES cell cDNA library (Clontech, CA) and sequenced. Two of them contained uninterrupted ORFs corresponding to the entire ORF of pmt1, but lacked a stop codon upstream of the first ATG. The human cDNA clones were obtained by screening a human heart cDNA library (Clontech, CA). Of 14 positive clones, one showed a continuous ORF with a stop codon upstream of the putative initiation codon.

RNA preparation and northern analysis

Total RNA was prepared from ES cells, ovary and testis using the GTC-CsCl2 centrifugation method (15), fractionated on a formaldehyde denatured 1% agarose gel by electrophoresis, and transferred to nylon membranes. A poly A+ RNA blot of mouse tissues was obtained from Clontech, CA. All blots were hybridized with a random-primed 540 bp _Eco_RI-_Pst_I cDNA fragment of the mouse Dnmt2 in a standard hybridization solution containing 50% formamide at 42°C, washed with 0.2× SSC, 0.1% SDS at 65°C, and exposed to X-ray film.

Construction of the gene targeting vector

A 14 kb _Sma_I-_Xho_I genomic DNA fragment of the Dnmt2 gene was isolated from a 129/Sv genomic DNA library and subcloned into the pBluescript vector. A 1 kb _Stu_I-_Sna_BI genomic DNA fragment containing exons encoding the putative catalytic domain (the PPC motif) was removed and replaced by the IRES-βgeo cassette with a splicing accept site (16). The resulting gene targeting vector contains a 9.2 kb fragment upstream and a 3.7 kb fragment downstream of the IRES-βgeo cassette (Fig. 3).

Generation of Dnmt2-deficient mutant ES cell lines

Transfection of J1 ES cells with linearized targeting vector DNA and subsequent G418 selection and cloning of drug-resistant colonies were carried out as described previously (1). Genomic DNA from G418-resistant clones was digested with _Bam_HI and analyzed by Southern blot hybridization using the probe pXhR5 (Fig. 3). To generate ES cell lines homozygous for the mutation, cells of a heterozygous clone were subject to selection in medium containing a high concentration of G418 (0.5 mg/ml of pure form) (17).

Analysis of de novo and maintenance methylation of provirus DNA

Infection of wild-type and _Dnmt2_-deficient ES cells with the MoMuLVsup-1 virus, DNA preparation from infected ES cells, analysis of de novo and maintenance methylation of endogenous or newly integrated viral DNA by Southern analysis of _Hpa_II/_Msp_I digested DNA were carried out as described previously (2).

Results and Discussion

Cloning of the mammalian homologs of the yeast pmt1 gene

One of the approaches that we took in search of a de novo DNA methyltransferase in mammalian cells was to screen the dbEST database using the amino acid sequences of different prokaryotic methyltransferases as query sequences (14). When sequences of the bacterial restriction methyltransferase M.HgiGI were used as query sequences, two EST clones of the same gene (GenBank accession nos N31314 and R95731) were found to give significant matches. Sequencing analysis of the EST clones revealed that they contained three of the highly conserved methyltransferase motifs. Multiple cDNA clones of the gene were isolated subsequently by screening human and mouse cDNA libraries using the EST clones as probes, and a full length cDNA was constructed with overlapping cDNA fragments after DNA sequencing. The deduced amino acid sequences of the human and mouse cDNA showed 81% identity and revealed that both genes contained all the conserved cytosine methyltransferase motifs (Fig. 1) (18). The same gene was independently cloned by Yoder and Bestor, and was named Dnmt2 (19). A BLAST search of GenBank with human and mouse cDNA sequences identified pmt1 of fission yeast Schizosaccharomyces pombe as the most closely related sequences, sharing 42% identity at the amino acid level. The yeast pmt1 contains all other conserved methyltransferase motifs except that the catalytic Pro-Pro-Cys motif was ‘mutated’ to Pro-Ser-Cys (13).

Figure 1

The comparison of the deduced amino acid sequences between the mammalian Dnmt2 and the yeast pmt1. The identical amino acids are shadowed. The conserved DNA methyltransferase motifs (I–X) are marked with roman numerals. The stop codons are indicated with asterisks (*) and gaps with dots (…).

To determine whether Dnmt2 has methyltransferase activity, the mouse cDNA was expressed in Escherichia coli or in insect cells using the baculovirus expression system. Methyltransferase activity assay was carried out using either poly(dI-dC) or λ phage DNA as substrates under a standard assay condition which could detect residual levels of enzyme activity in protein extracts prepared from the Dnmt1 null mutant ES cells (2). Despite the presence of large amounts of Dnmt2 protein in both bacterial and insect cell extracts, no methyltransferase activities were detected so far (data not shown). At the moment, it is not clear why the recombinant proteins have no detectable activities. The following two possibilities are considered: (i) the recombinant Dnmt2 protein may lack the natural conformation or modification, or is very unstable in vitro; (ii) Dnmt2 may require cofactors to catalyze methylation reaction. Further studies are necessary to investigate these possibilities.

Figure 2

Dnmt2 expression in organs and ES cells. A blot with 2 µg of poly A+ RNA from mouse tissues (Clontech, CA) is shown on the left, and a blot with 20 µg of total RNA from ES cells, ovary and testis is shown on the right. He, heart; Br, brain; Sp, spleen; Lu, lung; Li, liver; Mu, skeletal muscle; Ki, kidney; Te, testis; ES, ES cells; and Ov, ovary. Note that three Dnmt2 transcripts of sizes1.6, 2.6 and 4.0 kb were detected in mouse tissues, and the 1.6 kb transcript was the most abundant one in the organs examined.

Dnmt2 expression in mouse organs and ES cells

Dnmt2 expression in mouse ES cell lines and various organs were analyzed by northern hybridization using a full length cDNA fragment as probes. We showed that three Dnmt2 transcripts of 1.6, 2.6 and 4.0 kb were detected in mouse tissues, and the 1.6 kb transcript was the most abundant one in most tissues examined (Fig. 2). Dnmt2 appeared to express ubiquitously but at very low levels in mouse tissues, with relatively high levels in the heart, lung, kidney and testis (Fig. 2). Dnmt2 expression was also detected in mouse ES cells (Fig. 2), suggesting that Dnmt2 might be responsible for the residual methyltransferase activity detected in Dnmt1 null ES cells.

_Dnmt2_-deficient ES cells are viable

To investigate the role of Dnmt2 in development, we generated a putative null allele of Dnmt2, termed Dnmt2 m1, by deletion of the exons encoding the putative catalytic PPC motif through homologous recombination in ES cells (Fig. 3A). Of 85 G418-resistant colonies analyzed by Southern blot hybridization, six were positive for homologous recombination (Fig. 3B). To generate ES cell lines homozygous for the mutation, cells of a heterozygous ES cell line were cultured in medium containing a high concentration of G418 (0.5 mg/ml) for 14 days. Of 29 colonies analyzed, two were homozygous for the mutant allele (Fig. 3C). The Dnmt2 homozygous ES cells appeared to be normal in growth and morphology after consecutive passaging for more than 20 generations (data not shown), suggesting that Dnmt2 function is not essential.

De novo and maintenance methylation of provirus DNA in Dnmt2 m1/Dnmt2 m1 ES cells

Since Dnmt2 transcripts were detected in ES cells, we speculated that Dnmt2 might be required for de novo methylation. We showed previously that ES cells homozygous for a Dnmt1 null mutation were able to methylate provirus DNA de novo. We carried out a similar analysis of de novo methylation of integrated provirus DNA in infected Dnmt2 mutant ES cells.

First, we examined methylation status of endogenous virus in Dnmt2 m1/Dnmt2m1 ES. DNA isolated from wild-type and Dnmt2 m1/Dnmt2m1 ES cells was digested with the methylation-sensitive restriction enzyme _Hpa_II or its isoschizomer _Msp_I that cuts CCGG sequences regardless of whether CpG sites are methylated or not, and was then subject to Southern blot hybridization with a MoMuLV cDNA probe that hybridizes with endogenous provirus DNA (1). We showed that endogenous virus DNA in Dnmt2 m1/Dnmt2m1 ES cells was methylated to the same levels as in wild-type cells (Fig. 4). This result indicates that Dnmt2 is not required for the maintenance methylation of genomic DNA.

Figure 3

Targeted disruption of the Dnmt2 gene. (A) The wild-type Dnmt2 genomic locus (top), the targeting vector (middle), and the targeted allele (bottom). The location of the exons (solid bars), PC motif, ENV motif and the IRES-βgeo cassette are shown. The 1.3 kb _Xho_I-_Eco_RV genomic fragment was used as a probe for Southern analysis, and the 10.6 and 9.0 kb _Bam_HI fragments from wild-type and targeted alleles, respectively, are indicated as dashed lines. Sm, _Sma_I; B, _Bam_HI; Xh, _Xho_I; St, _Stu_I; Sn, _Sna_BI; Rv, _Eco_RV; SA, splicing acceptor; and pA, poly (A) signal. (B) Southern blot hybridization of genomic DNA from wild-type and targeted ES cell clones. DNA was digested with _Bam_HI, blotted and hybridized to the probe shown in (A). (C) Southern analysis of genomic DNA from a heterozygous and two homozygous mutant ES cell clones.

Figure 4

Methylation of endogenous provirus DNA in the Dnmt2 null mutant ES cells. Genomic DNA was isolated from ES cells, digested with _Hpa_II (H) or _Msp_I (M), blotted and hybridized to the MoMuLV cDNA probe (1). +/− and −/− are Dnmt2 m1/+and Dnmt2 m1/Dnmt2 m1 cells while n/n and c/c are Dnmt1 n/Dnmt1 n and Dnmt1 c/Dnmt1 c ES cells, respectively (2).

To examine whether Dnmt2 m1/Dnmt2m1 ES cells were able to methylate foreign DNA such as newly integrated provirus DNA, we infected Dnmt2 mutant ES cells with the MoMuLVsup-1 retrovirus and analyzed the methylation status of newly integrated provirus DNA 2–4 days after infection. DNA was digested with _Kpn_I and _Hpa_II, or with _Kpn_I and _Msp_I as controls, and analyzed by Southern blot hybridization using the πAN7 probe that would recognize a 1.45 kb _Kpn_I fragment of infected viral DNA but not the endogenous proviruses (Fig. 5A). We found that the newly integrated virus DNA was methylated in Dnmt2 m1/Dnmt2m1 ES cells as efficiently as in wild-type cells as shown by the presence of an _Hpa_II-resistant 1.45 kb fragment (Fig. 5B), indicating that Dnmt2 is not an essential component of the de novo methyltransferases.

The lack of detectable methyltransferase activities in vitro and in vivo raises interesting possibilities that Dnmt2 might encode a sequence-specific DNA methyltransferase which methylates only a small number of target sequences in the genome, or it may methylate cytosine in non-CpG sequences such as CpNpG. It is also possible that Dnmt2 is simply not a functional cytosine DNA methyltransferase, despite having all the conserved DNA methyltransferase motifs. Dnmt2 may be involved in cellular processes other than DNA methylation, such as DNA repair by binding to mismatched nucleotides as the bacterial cytosine methyltransferases (21–23), DNA recombination and carcinogenesis.

Since Dnmt2 is not essential for de novo methylation in ES cells, additional DNA methyltransferases that catalyze de novo methylation are predicted to be present in mammalian cells. It is formally possible that both Dnmt1 and Dnmt2 are de novo methyltransferases and can functionally compensate each other. Recently, a gene known as masc1 was cloned through homology-based screening using a PCR amplification method, and genetic analysis has revealed that masc1 is involved in de novo methylation in Ascobolus (24). The protein encoded by masc1 contains all the conserved methyltransferase motifs except that motif VI has an EET sequence rather than the ENV sequence that is conserved in almost all the known DNA cytosine methyltransferases. The methyltransferase activity of masc1 encoded proteins has not been reported. Sequence analysis indicates that masc1 is distantly related to Dnmt1 and Dnmt2 (data not shown). It remains to be seen whether a mammalian homologue of masc1 exists, and whether it functions as a de novo DNA methyltransferase.

Figure 5

De novo methylation of provirus DNA in Dnmt2 mutant ES cells. (A) Schematic diagrams of the MoMuLVsup-1 provirus genome (top), the 3′ LTR region (middle), the size marker, the location of the πAN7 probe and the five _Hpa_II/_Msp_I sites (bottom) (2). (B) Genomic DNA was isolated from infected 3T3 cells (lanes 1–3), infected wild-type (lanes 4–6 and 9–11), uninfected wild-type (lanes 7 and 8), infected heterozygous mutant (lanes 12 and 13) and infected homozygous mutant (lanes 14 and 15) at day 0 (lane 9), day 2 (lanes 10, 12 and 14 ) and day 4 (lanes 4–6, 11, 13 and 15) post-infection. DNA was digested with _Msp_I/_Kpn_I (lanes 1, 4 and 7), _Hpa_II/_Kpn_I (lanes 2, 5 and 8–15), or _Kpn_I alone (lanes 3 and 6), blotted and hybridized to the πAN7 probe. Mov, MoMuLVsup-1 virus infected; M, _Msp_I; H, _Hpa_II; K, _Kpn_I.

Acknowledgements

We thank Dr Austin Smith for the plasmid GT1.8Iresβgeo(Sal), Lian Yu for excellent technical assistance, and members of our laboratory for discussion. This work was supported by grants from Bristol-Myers/Squibb and NIH (GM52106 to E.L.). M.O. was a fellow of the Japanese Society for the Promotion of Science.

References

1

,

Cell

,

1992

, vol.

69

(pg.

915

-

926

)

2

,

Development

,

1996

, vol.

122

(pg.

3195

-

3205

)

3

,

Nature

,

1993

, vol.

366

(pg.

362

-

365

)

4

,

Genomic Imprinting: Frontiers in Molecular Biology

,

1997

Oxford

IRL Press

(pg.

53

-

69

)

5

,

Curr. Opin. Genet. Dev.

,

1996

, vol.

6

(pg.

159

-

163

)

6

,

Dev. Genet.

,

1995

, vol.

17

(pg.

188

-

197

)

7

,

Genomic Imprinting: Frontiers in Molecular Biology

,

1997

Oxford

IRL Press

(pg.

191

-

210

)

8

,

Annu. Rev. Genet.

,

1996

, vol.

30

(pg.

441

-

464

)

9

,

Adv. Cancer Res.

,

1998

, vol.

72

(pg.

141

-

196

)

10

,

Proc. Natl. Acad. Sci. USA

,

1997

, vol.

94

(pg.

2103

-

2105

)

11

,

Genomic Imprinting: Frontiers in Molecular Biology

,

1997

Oxford

IRL Press

(pg.

1

-

20

)

12

,

EMBO J.

,

1992

, vol.

11

(pg.

2611

-

2617

)

13

,

Nucleic Acids Res.

,

1995

, vol.

23

(pg.

203

-

210

)

14

,

J. Mol. Biol.

,

1990

, vol.

215

(pg.

403

-

410

)

15

,

Anal. Biochem.

,

1991

, vol.

162

(pg.

156

-

159

)

16

,

Proc. Natl. Acad. Sci. USA

,

1994

, vol.

91

(pg.

4303

-

4307

)

17

Geisterfer-Lowrance

A. A.

,

Mol. Cell. Biol.

,

1992

, vol.

12

(pg.

2391

-

2395

)

18

,

Nucleic Acids Res.

,

1994

, vol.

22

(pg.

1

-

10

)

19

,

Hum. Mol. Genet.

,

1998

, vol.

7

(pg.

279

-

284

)

20

,

J. Mol. Biol.

,

1996

, vol.

257

(pg.

804

-

813

)

21

,

Nucleic Acids Res.

,

1995

, vol.

23

(pg.

1388

-

1395

)

22

,

Nucleic Acids Res.

,

1995

, vol.

23

(pg.

1380

-

1387

)

23

,

Gene

,

1995

, vol.

157

(pg.

177

-

179

)

24

,

Cell

,

1997

, vol.

91

(pg.

281

-

290

)

© 1998 Oxford University Press

I agree to the terms and conditions. You must accept the terms and conditions.

Submit a comment

Name

Affiliations

Comment title

Comment

You have entered an invalid code

Thank you for submitting a comment on this article. Your comment will be reviewed and published at the journal's discretion. Please check for further notifications by email.

Citations

Views

Altmetric

Metrics

Total Views 1,327

897 Pageviews

430 PDF Downloads

Since 12/1/2016

Citations

310 Web of Science

×

Email alerts

Citing articles via

More from Oxford Academic