| Nucleic Acids Research | Pages |
Cloning, characterization and expression of the gene coding for a cytosine-5-DNA methyltransferase recognizing GpC
Introduction
Materials And Methods
Strains and plasmids
PCR amplification of a fragment with high homology to conserved sequences of 5-meC MTases
Construction and screening of a Chlorella virus NYs-1 genomic library
Cell culture
Determination of methylation specificity of M.CviPI in E.coli and yeast
Enzyme purification and MTase assay
Results And Discussion
Isolation and identification of the M.CviPI gene from Chlorella virus NYs-1
Sequence comparison of M.CviPI and other 5-meC MTases
Purification of M.CviPI protein
Expression of M.CviPI in yeast shows its potential for chromatin mapping
Summary
Acknowledgements
References
Cloning, characterization and expression of the gene coding for a cytosine-5-DNA methyltransferase recognizing GpC
ABSTRACT
INTRODUCTION
In vivo methylation of DNA has been used successfully in Saccharomyces cerevisiae to study protein-DNA interactions in the chromatin of living cells [1-3; for more information about the expression of foreign methyltransferases (MTases) in yeast see 4]. A high frequency of MTase targets is critical for high resolution mapping of chromatin structure. Among currently available MTase probes, the only de novo dinucleotide MTase is M.SssI, which recognizes a CpG site (5). Due to under-representation of the CpG dinucleotide in the genome, the resolution of chromatin structure maps using this enzyme is ~35 bp on average in S.cerevisiae (6). With this moderate level of resolution, M.SssI can possibly serve to detect the presence of a positioned nucleosome, 146 bp in yeast, without the need for introduction of additional CpG sites into native DNA sequences. However, this resolution is insufficient for mapping the interactions of non-histone, regulatory proteins, as the typical length of the target DNA sequence of yeast regulatory proteins is ~20-30 bp or less. For example, yeast TBP recognizes and binds to an 8 bp sequence (7), while the well-characterized transcriptional activator Gal4p binds to a 17 bp consensus sequence (8). Further, methylation of CpG islands has been implicated as an important controlling element for gene regulation in mammalian systems, which may limit the application of M.SssI in higher organisms (9). To address both the limitation of resolution and the possible inability to utilize M.SssI in higher organisms, cloning and expression of cytosine-5-DNA MTases (5-meC MTase) with different specificities but similarly small recognition sites is essential.
A family of double-stranded DNA viruses that infect certain unicellular, eukaryotic, Chlorella-like green algae are reported to be a rich source of restriction/modification systems (10,11). Among the 37 viruses infecting Chlorella NC64A and the five viruses infecting Chlorella Pbi which have been partially characterized, 39 viral DNAs contain 5-methylcytosine, ranging in concentration from 0.1 to 47% of total cytosine (10,11). One cytosine MTase, M.CviJI, has been cloned from Chlorella virus IL-3A and shown to recognize RGC(T/C/G) (12). As determined by the resistance/sensitivity of the viral DNAs to over 70 methylation-sensitive restriction endonucleases, at least five independent 5-meC modification systems are predicted to be encoded by some of the more highly modified viruses, including MTases thought to recognize CpC and RpCpY (10,11). Based on the composition of the yeast genome, on average, one CpC site per 13.9 bp and one RpCpY site per 10.7 bp can be expected in the genome. Achieving this level of resolution would allow mapping the interactions of most non-histone, regulatory proteins. The cloning of MTases from Chlorella viruses could greatly extend the resolution of chromatin mapping as well as allow extension of in vivo chromatin mapping to higher organisms.
Amino acid sequence comparison of most cloned cytosine MTases indicates the presence of conserved motifs in similar order (13). The crystal structures of HhaI and HaeIII cytosine MTases provide physical support for this conservation by showing that the conserved motifs form the core structure of the protein and surround the active site cleft (14,15). The two most conserved motifs are motif I, which serves as the S-adenosyl methionine (Ado-Met) binding pocket, and motif IV, which contains the catalytic site (14). The presence of these highly conserved motifs provides a potential target for cloning new cytosine MTases. Indeed, a putative cytosine MTase has been isolated from Arabidopsis by sequence homology (16).
Here we report the cloning of a new dinucleotide 5-meC MTase gene, called M.CviPI, from Chlorella virus NYs-1. We used the high conservation at motifs I and IV to design primers for PCR amplification of a fragment spanning these two regions. This PCR fragment was then used to probe an NYs-1 genomic library to finally clone the M.CviPI gene. A bacterially expressed protein was purified and demonstrated to methylate GpC in vitro. The M.CviPI gene was also successfully expressed in yeast, indicating that the enzyme can potentially be used to increase the resolution of in vivo chromatin mapping.
MATERIALS AND METHODS
Strains and plasmids
Escherichia coli strain DH5[alpha]mut10 (17), lacking the mcrABC and mrr genes, and the vector pTZ18U were used for all the cloning and sequencing. For expression of M.CviPI in E.coli, the DH5[alpha]mut10 lysogen of bacteriophage DE3 was constructed carrying the T7 RNA polymerase gene under control of the LacUV5 promoter (18). The plasmid pET20b+ is a T7 expression vector which adds a hexahistidine tag at the C-terminus of a cloned protein, while pLysS is a compatible plasmid providing a small amount of lysozyme (Novagen). The entire coding sequence of M.CviPI, with the SV40 nuclear localization signal (ATG CCA AAG AAG AAG AGA AAG GTT) appended at its N-terminus, was cloned into the yeast expression vector pMPK1 via the SfiI and SphI sites (2). The gene was subsequently integrated into the genomes of both YPH500[Delta]L (MAT[alpha] ade2-101° his3-[Delta]200 leu2-[Delta]1 lys2-[Delta]1 trp1-[Delta]63 ura3-52) and YPH499[Delta]L (identical genotype but MATa) cells to create yeast strains MXY107 and MXY108, respectively (2). These cell lines express the M.CviPI gene under the control of a GAL1 promoter.
PCR amplification of a fragment with high homology to conserved sequences of 5-meC MTases
Degenerate primers for PCR were MEC1 (5[prime]-ccggatcCTNTTYGCNGGNAT-3[prime]), located in motif I, and MEC2 (5[prime]-acctgcagRAANCCYTGRCANGGRAANCC-3[prime]), corresponding to motif IV of the conserved amino acid sequence of 5-meC MTase. The sequence was chosen based on either the consensus sequence or, where there was no consensus, the sequence of M.CviJI (12,13). Within the primer sequence, N represents a mixture of all four bases and lower case letters indicate sequence not existing in the MTase gene but introduced for the convenience of cloning (BamHI and PstI sites). Viral genomic DNA was amplified in 50 µl reactions which contained 40 pmol each primer, 10 pmol each dATP, dCTP, dGTP and dTTP, 1 µg DNA, 2.5 U Taq DNA polymerase (Fisher) in a buffer of 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 3 mM MgCl2, 0.05% (v/v) NP-40 and 0.05% (v/v) Tween 20. PCR cycling parameters were as follows: preheating at 94°C for 3 min; five cycles of 94°C for 30 s, 42°C for 30 s and 72°C for 1 min; 20 cycles of 94°C for 30 s, 60°C for 30 s and 72°C for 1 min. Reaction products were separated on a 1.5% agarose gel and fragments ~200-300 bp in size were purified from the gel, digested with BamHI and PstI, and subsequently cloned into pTZ18U. About 20 clones were sequenced and the DNA sequences of insertion fragments were translated into peptide sequences to compare with the amino acid sequence of M.CviJI.
Construction and screening of a Chlorella virus NYs-1 genomic library
A genomic library of Chlorella virus NYs-1 was constructed by partial digestion of viral DNA with Sau3AI, gel electrophoretic separation to enrich for fragments in the 2-4 kb size range and then ligation of the size-selected DNA fragments into BamHI-digested pTZ18U. The resultant plasmids were subsequently transformed into DH5[alpha]mut10 and clones with genomic DNA insertions were screened on plates containing X-Gal. The cloned, PCR-amplified fragment with homology to 5-meC MTase was excised from the vector and random primer labeled in the presence of [[alpha]-32P]dATP. About 1000-2000 colonies of the library were screened with this probe by in situ hybridization to obtain positive clones (19). Plasmid DNAs were purified from positive clones and the inserts sequenced to identify potential 5-meC MTase genes.
Cell culture
Genes with high homology to known 5-meC MTases were cloned in-frame in pET20b+ at the NdeI and EcoRI sites for expression in E.coli. The plasmid was then co-transformed into DH5[alpha]mut10(DE3) with pLysS. Expression was induced as follows. Cells were grown to an OD600 of ~0.4-0.6 and centrifuged prior to resuspension in fresh medium. IPTG was added to 0.4 mM for 2.5-4 h induction. Plasmids were then purified from induced cells and digested with a set of restriction enzymes (HindIII, HaeII, HaeIII, HhaI and AvaII) to detect the presence of MTase activity. For the expression of 5-meC MTases in yeast, a starter culture was grown overnight at 30°C in YPG medium (10 g yeast extract, 20 g peptone, 20 g galactose/l) to an OD600 of ~1. Cells were centrifuged and resuspended in fresh YPG medium for growth at 30°C for an additional 16 h. DNA from ~3 ml cells was rapidly isolated by the glass bead method for deamination as described below (20).
Determination of methylation specificity of M.CviPI in E.coli and yeast
The DNA sequence recognized by M.CviPI was investigated by bisulfite genomic sequencing. Either linearized plasmid (pET20b+ or pET-NYs-1-5) DNA purified from E.coli cells or genomic DNA purified from yeast cells was treated by the method initially developed by Frommer et al. (21) and further modified by Kladde et al. (2). Briefly, DNA is subjected to quantitative deamination with sodium metabisulfite, converting all deoxycytidine residues to deoxyuridine, while 5-meC residues, created by the MTase in vivo, resist deamination. Subsequent PCR amplification of selected DNA regions yields a product that is directly sequenced to provide a positive display of 5-meC. The primers used for the E.coli pET20b+ plasmid were as follows: pair I (Fig.
| Figure 1. M.CviPI methylates GpC sites in E.coli. Plasmid pETNYs-1-5, containing the coding sequence of M.CviPI, or pET20b+, the parent vector, was purified from IPTG-induced E.coli cells. 5-meC residues on the lower strand were identified in isolated DNA by deamination, subsequent PCR amplification and direct cycle sequencing of the purified PCR products. (a and b) Two different sequence regions that were investigated, which include GpC sites with all 16 possible combinations of flanking bases. In (a), lanes 1-4 are the sequencing lanes of the region investigated using the PCR product used in lane 5 as template. Lane 5 is a negative control where the same sequence region was examined in the pET20b+ vector only sample. The last lane is the methylation pattern of pETNYs-1-5 isolated from induced cells expressing M.CviPI. In (b), the first four lanes (lanes 1-4) are sequencing lanes while the last two lanes (lanes 5 and 6) represent the methylation pattern of pETNYs-1-5 isolated from two different induction experiments. In both parts, the arrow represents artefactual primer extension pauses which occurred in samples from both pET20b+ and pETNYs-1-5. The asterisks indicate every GpC site in the sequence that is resolvable on the gel. The flanking bases are listed 5[prime]->3[prime] beside each methylation band. Due to the fact that methylation of the lower strand was analyzed, the sequence context can be ascertained by reading the gel from the top to the bottom. | a
 |
b
 |
Enzyme purification and MTase assay
A 1 l E.coli culture was induced as described above for enzyme purification. Cells were sonicated in 10 ml binding buffer (5 mM imidazole, 0.5 M NaCl and 20 mM Tris-HCl, pH 7.9). Cell debris was removed by centrifugation (20 min at 10 000 g) and the supernatant was applied to a Ni2+-agarose column (1 × 1 cm). Following sequential washes with 10 ml binding buffer,10 ml wash buffer I (60 mM imidazole, 0.5 M NaCl and 20 mM Tris-HCl, pH 7.9) and 1.5 ml wash buffer II (100 mM imidazole, 0.5 M NaCl and 20 mM Tris-HCl, pH 7.9), the bound enzyme was eluted with 1.5 ml elution buffer (500 mM imidazole, 0.5 M NaCl and 20 mM Tris-HCl, pH 7.9). The eluate was applied to a phosphocellulose column (Whatman P11, 1 × 1 cm) equilibrated with 20 mM Tris (pH 8.0), 50 mM NaCl and 1 mM EDTA. Protein was eluted stepwise with 2 ml portions of the same buffer containing NaCl from 0.1 to 1.0 M in 0.1 M increments. Fractions of 0.5 ml were collected and the fractions (0.3-0.4 M NaCl) containing the predicted 41 kDa protein band were pooled and concentrated using a Centricon concentrator (Amicon). The final enzyme solution was kept at -80°C in a buffer of 20 mM Tris (pH 8.0), 1 mM EDTA, 2 mM DTT and 10% (v/v) glycerol.
Cytosine MTase activity was assayed in a 20 µl reaction containing 20 mM Tris (pH 8.0), 1 mM EDTA, 2 mM DTT, 0.32 nM S-adenosyl methionine (New England Biolabs), 1 µg pTZ18U plasmid DNA and 1 µl enzyme fraction. After incubation for 1 h at 37°C, the reaction was stopped by extraction with StrataClean resin (Stratagene). Following ethanol precipitation, the DNA was digested with HaeIII and analyzed on 1% agarose gels. HaeIII cleaves GGCC but not GGmeCC sequences (22,23). Therefore, lack of HaeIII cleavage implies a MTase which methylates the internal cytosine in GGCC sites.
RESULTS AND DISCUSSION
Isolation and identification of the M.CviPI gene from Chlorella virus NYs-1
The Chlorella virus NYs-1 genome contains a very high level of 5-meC (47.5%) (24). The resistance/sensitivity of its DNA to >70 methylation-sensitive restriction endonucleases indicated that the virus probably encodes several 5-meC MTases (10,11). Sequence alignment of M.CviJI, the only cytosine MTase cloned from a Chlorella virus, with the conserved motifs of other 5-meC MTases indicated that the two most conserved motifs of M.CviJI are motif I and motif IV. These motifs, which are usually ~200 bp apart (13), correspond to the Ado-Met binding site and the catalytic site in the crystal structure of the HhaI and HaeIII MTases (14,15). A pair of degenerate primers based on the consensus amino acid sequence within the two motifs was used for PCR with NYs-1 DNA. These primers generated several bands ranging from 150 bp to 2 kb, with a prominent band around 200 bp (data not shown). The ~200 bp fragment was gel purified and cloned into pTZ18U following digestion with PstI and BamHI. Sequence analysis of ~20 different transformants identified six different fragments, each of which encoded part of an ORF with high homology to M.CviJI as well as other 5-meC MTases. Southern blotting with NYs-1 genomic DNA using each of the six fragments as the probe confirmed that these sequences do occur in the viral genome.
To clone the full-length gene of the putative 5-meC MTases, a genomic library of NYs-1 was constructed in pTZ18U, a non-expressing vector, to avoid possible toxicity resulting from a high level of methylation. One of the six cloned PCR fragments was used as a hybridization probe to isolate six positive clones which contained a viral genomic fragment with homology to the probe. Sequence analysis of each of the positive clones revealed that two of them contained a fragment encoding the same, full-length ORF with high homology to 5-meC MTases.
To further characterize this potential MTase gene, the coding sequences were fused to a His6 tag within the expression vector pTZ20b+, cloned into strain DH5[alpha]mut10(DE3) and tested for expression (Materials and Methods). One of the clones, designated pETNYs-1-5, expressed a MTase activity which resulted in plasmid DNA that was resistant to digestion by HhaI, HaeII and HaeIII (data not shown). NYs-1 viral DNA is also resistant to these same enzymes (data not shown).
Figure 2. Protein sequence alignment of three cytosine MTases, M.CviPI, M.CviJI and M.HaeIII, by CLUSTLW (31). The bars above the sequence represent the regions of conserved motifs, as indicated. Residues that are identical in all three proteins are indicated in blue and similar residues are shown in red. All the other residues remain as black letters. The regions utilized to create the two degenerate PCR primers are indicated by arrows (only the portion of the primer that is complementary to the MTase gene is indicated). The methylation specificity of clone pETNYs-1-5 was examined by bisulfite genomic sequencing. Briefly, pETNYs-1-5 purified from induced cells was deaminated and, following PCR amplification, its methylation pattern was determined by thermal cycle sequencing (Fig. 
Sequence comparison of M.CviPI and other 5-meC MTases
The amino acid sequence comparison of M.CviPI with the sequences of other 5-meC MTases supports its identification as a 5-meC MTase. Sequence alignment was performed for M.CviPI (GenBank accession no. AF062394), M.CviJI (GenBank accession no. M27265) and M.HaeIII (GenBank accession no. AF051375), a bacterial MTase recognizing GpGpCpC sequences. The three enzymes have the dinucleotide GpC as the whole or a part of their recognition sequence. As shown in Figure
The high amino acid conservation between M.CviPI and M.CviJI, cytosine MTases isolated from two different viruses, NYs-1 and IL-3A, and their common GpC methylation sites is quite interesting in the context of mechanisms leading to the profusion of restriction/modification systems in Chlorella viruses. It will be interesting if additional enzymes can be isolated from some of the Chlorella viruses which recognize versions of GpC or RGC(T/C/G) sites. The high level of sequence conservation reinforces the possibility of cloning such additional 5-meC MTases from Chlorella viruses by the sequence homology strategy.
Purification of M.CviPI protein
The pET20b+ plasmid, containing the M.CviPI gene, was introduced into E.coli strain DH5[alpha]mut10(DE3) for expression. Lack of the mcrABC and mrr genes in this bacterial strain should decrease any possible toxicity of high levels of cytosine methylation (26). To further reduce the chance of possible deleterious effects of M.CviPI expression, another compatible plasmid, pLysS, was transformed into the same host. pLysS contains a gene encoding lysozyme, which is an inhibitor of T7 RNA polymerase (18). The low level of lysozyme produced from the plasmid inhibits any T7 RNA polymerase activity resulting from leaky repression of the LacUV5 promoter and thereby allows more stringent control of MTase production.
Induction of cells at an OD600 of ~0.4-0.6 for 2.5-4 h with 0.4 mM IPTG led to an increase in a protein of 41 kDa, the predicted size of M.CviPI protein (Fig.
Figure 3. Purification of M.CviPI and assay for activity in vitro. (a) SDS-polyacrylamide gel showing the proteins after each step of purification. From 1 l of non-induced (lane 1) or IPTG-induced (lane 2) cells, total protein from 40 and 20 µl cells, respectively, was extracted and analyzed on the gel. Lane 3 contains 10 µl of the 200 µl total eluate from the Ni2+-agarose column. The final, purified enzyme eluted with 0.3-0.4 M NaCl from the phosphocellulose column was analyzed in lane 4 (10 µl of 200 µl total eluate). The arrow indicates the position of the enzyme band. The molecular weight marker used was the broad range standard from Bio-Rad; the 200, 116, 97.4, 66, 45 and 31 kDa species are visible. (b) In vitro MTase assay of purified M.CviPI after phosphocellulose column chromatography. Purified enzyme [1 µl of the 200 µl phosphocellulose eluate analyzed in lane 4 of (a)] was assayed for MTase activity as described in Materials and Methods. pTZ18U DNA (1 µg) was treated with the enzyme in the presence of Ado-Met and subsequently digested with HaeIII (lane 1). As a control, the same reaction was performed in the absence of Ado-Met (lane 2). The molecular marker is a [lambda]/HindIII plus [phis]X174RF/HaeIII digestion mixture. Note the resistance to digestion by the methylation-sensitive restriction endonuclease in the sample treated with M.CviPI and Ado-Met. Figure 4. Methylation activity of M.CviPI expressed in yeast. Genomic DNA was purified from yeast cells grown in medium containing galactose and subsequently deaminated to determine the methylation pattern. Lanes 1-4 contain the sequencing lanes for 45871-45331 m.u. of chromosome XI, a region near the 3[prime]-end of the coding sequence of STE6. The methylation pattern of DNAs purified from MXY108 yeast cells is shown in lane 5. Every GpC site present in the region (indicated by the asterisks followed by the GpC site and its flanking bases) is modified by the methyltransferase. As a control, DNA isolated from the parental a cell line, lacking the methyltransferase gene, is devoid of methylation (lane 6), demonstrating that the MTase activity is encoded by the M.CviPI gene. The M.CviPI gene was cloned and expressed in S.cerevisae, a eukaryotic organism lacking endogenous methylation of DNA, enabling unambiguous detection of de novo modification. Expression of the protein in yeast was under control of the GAL1 promoter, a strong yeast promoter tightly regulated by carbon source, that is repressed in glucose and induced in galactose (27). Transfection of the cloned gene together with its controllable promoter into the genomic LYS2 locus created a single copy, stable integrant similar to that used in our previous studies of DNA methylation and chromatin structure in S.cerevisiae (2). A region located near the 3[prime]-end of the coding sequence of STE6 was chosen for bisulfite genomic sequencing, as it is known to be devoid of positioned nucleosomes which may obscure determination of the methyltransferase specificity. After induction by growing yeast cells in medium containing galactose, genomic DNA was purified and deaminated to determine the methylation pattern, as previously described (2). As shown in Figure We document the occurrence of a novel cytosine-5-DNA methyltransferase, named M.CviPI, from Chlorella virus NYs-1 that recognizes the sequence GpC. The GpC specificity of M.CviPI is only the second dinucleotide recognition sequence reported for a DNA MTase. Although the genes for ~15 6-meA MTases have been cloned from Chlorella viruses, the M.CviPI gene is only the second 5-meC DNA MTase gene isolated from these viruses. As such, the M.CviPI gene and its product are of interest in terms of enzymatic mechanisms of small site DNA recognition, cytosine methyltransferases and the evolution of modification systems in algal viruses. From a practical point of view, the identification of methyltransferases such as MCviPI, having small recognition sites that occur with a high frequency in eukaryotic genomes, and expressing these enzymes in eukaryotic cells is highly important in on-going efforts to develop high resolution methods for analysis of chromatin structure and protein-DNA interactions in living cells.
Expression of M.CviPI in yeast shows its potential for chromatin mapping
Summary
ACKNOWLEDGEMENTS
We thank members of the Simpson laboratory for helpful discussions, Dwight Burbank for supplying Chlorella viral genomic DNA and Mike Nelson for critical reading of the manuscript. This work was supported by NIH grants GM52908 (R.T.S.) and GM32441 (J.V.E.).
REFERENCES
This article has been cited by other articles:
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 14 Aug 1998
Copyright©Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
S.-h. Chan, Z. Zhu, J. L. Van Etten, and S.-y. Xu
Cloning of CviPII nicking and modification system from chlorella virus NYs-1 and application of Nt.CviPII in random DNA amplification
Nucleic Acids Res.,
November 29, 2004;
32(21):
6187 - 6199.
[Abstract]
[Full Text]
[PDF]
C. D. Carvin, R. D. Parr, and M. P. Kladde
Site-selective in vivo targeting of cytosine-5 DNA methylation by zinc-finger proteins
Nucleic Acids Res.,
November 15, 2003;
31(22):
6493 - 6501.
[Abstract]
[Full Text]
[PDF]
C. D. Carvin, A. Dhasarathy, L. B. Friesenhahn, W. J. Jessen, and M. P. Kladde
Targeted cytosine methylation for in vivo detection of protein-DNA interactions
PNAS,
June 24, 2003;
100(13):
7743 - 7748.
[Abstract]
[Full Text]
[PDF]
X. Wang and R. T. Simpson
Chromatin structure mapping in Saccharomyces cerevisiae in vivo with DNase I
Nucleic Acids Res.,
May 1, 2001;
29(9):
1943 - 1950.
[Abstract]
[Full Text]
[PDF]
This Article 
Abstract
Print PDF (329K)
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (18)
Request Permissions
Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Articles by Xu, M.
Articles by Simpson, R. T.
Search for Related Content
PubMed
PubMed Citation
Articles by Xu, M.
Articles by Simpson, R. T.
Social Bookmarking
What's this?