DNA methyltransferases from Neisseria meningitidis and Neisseria gonorrhoeae FA1090 associated with mismatch nicking endonucleases

Agnieszka Kwiatek, Monika Kobes, Kamil Olejnik and Andrzej Piekarowicz

Institute of Microbiology, Warsaw University, Miecznikowa 1, 02-096 Warsaw, Poland

Correspondence
Andrzej Piekarowicz
anpiek{at}biol.uw.edu.pl


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The genes encoding the DNA methyltransferases M.NmeDI and M.NmeAI from Neisseria meningitidis associated with the genes encoding putative Vsr endonucleases were overexpressed in Escherichia coli. The enzymes were purified to apparent homogeneity on Ni-NTA agarose columns, yielding proteins of 49±1 kDa and 39·6±1 kDa, respectively, under denaturing conditions. M.NmeDI recognizes the degenerate sequence 5'-RCCGGB-3'. It methylates the first 5' cytosine residue on both strands within the core sequence CCGG. The enzyme shows higher affinity with the hemimethylated degenerate sequence than with the unmethylated degenerate sequence. Comparison of the amino acid sequence of the target-recognizing domain of M.NmeDI with the closest neighbours recognizing the sequence 5'-RCCGGY-3' showed the presence of the homologous domain and an additional domain that may be responsible for recognizing the degenerate sequence. M.NmeAI recognizes the sequence 5'-CCGG-3' and methylates the second 5' cytosine residue on both DNA strands. In Neisseria gonorrhoeae strain FA1090 the homologues of these ORFs are truncated due to a variety of mutations.


Abbreviations: AdoMet, S-adenosyl-L-methionine; MTase, methyltransferase; m5C-MTase, DNA methyltransferase methylating the endocyclic C-5 of cytosine; ODN, oligonucleotide; R-M, restriction and modification; TRD, target recognition domain; Vsr, very short patch repair


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Enzymic transfer of the methyl group from S-adenosyl-L-methionine (AdoMet) to certain nucleotides in DNA is the most common form of biological DNA modification. This reaction is catalysed by DNA methyltransferases (MTases). The most common DNA MTases are associated with a cognate restriction endonuclease (Wilson & Murray, 1991). However, the solitary MTases modify DNA sequences for purposes other than restriction. DNA MTases can be divided into those that methylate the exocyclic amino group of adenines and cytosines (amino-MTases) and those that methylate the endocyclic C-5 of cytosine (m5C-MTases). The m5C-MTases show the presence of 10 sequence motifs (I through X) that are located in the amino- and carboxyl-terminal protein regions and are responsible for AdoMet binding and transfer of the methyl group to a cytosine within the DNA target sequence. Motifs VIII and IX are separated by a variable target recognition domain (TRD) (Posfai et al., 1989; Lauster, 1989). TRDs determine the sequence specificity in both monospecific (Klimasauskas et al., 1991) and multispecific m5C-MTases (Balganesh et al., 1987; Wilke et al., 1988; Walter et al., 1992; Trautner et al., 1996).

The presence of m5C-MTases is the source of the T-G mismatch through spontaneous deamination of 5-methylcytosine residues (Bhagwat & Lieb, 2002). These T-G mismatches are repaired by a very short patch (VSP) repair system. The presence of an active VSP system was described in Escherichia coli K-12 (Lieb & Bhagwat, 1996) and in Bacillus stearothermophilus (Laging et al., 2003). Nucleotide sequence homologous to genes ecoKDcmV and ecoKDcmM of E. coli K-12 has been reported in several enteric pathogens, e.g. Shigella sonnei, Salmonella typhimurium, Salmonella enteritidis, Enterobacter cloacae (Lieb & Bhagwat, 1996) and Haemophilus parainfluenzae (http://rebase.neb.com/rebase). In two studied cases, the specificity of the mismatch nicking endonuclease is the same as found for the associated DNA m5C-MTase (Hennecke et al., 1991; Laging et al., 2003).

The genus Neisseria belongs to a group of bacteria encoding as many as 10 or more different MTases, including at least several m5C-MTases (Roberts & Macelis, 2002). Claus et al. (2000), using representational difference analysis (RDA), isolated from Neisseria meningitidis the DNA fragments encoding three m5C-MTases. These DNA fragments were isolated from strain MC58 (MTase M.NmeBI), strain Z2491 (MTase M.NmeAI) and strain 2120 (MTase M.NmeDI). The MTase M.NmeBI was cloned and expressed by these authors. The sequence of the DNA fragments isolated from strains Z2491 and 2120 was used for comparison with other sequences, which allowed identification of these ORFs as m5C-MTases. However, neither their biological activity nor their recognition specificity was determined, although the authors showed that the M.NmeAI does not recognize the sequence 5'-GTCAGC-3' recognized by M.HgiDII despite the strong homology between these two MTases.

In the vicinity of the genes encoding these two m5C-MTases, we have found the presence of ORFs encoding putative Vsr-like enzymes (A. Piekarowicz, unpublished data). Since the specificity of these Vsr-like enzymes should be the same as the associated DNA MTases we have cloned the genes encoding the M.NmeDI and M.NmeAI MTases into expression vectors and purified both enzymes. In this paper we present data indicating that these genes encode biologically active enzymes that recognize the specific DNA sequences 5'-RCCGGB-3' (M.NmeDI) and 5'-CCGG-3' (M.NmeAI).


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, plasmids and growth conditions.
Escherichia coli K-12 strains XL-1 Blue MRF' [{Delta}(mcrA)183 {Delta}(mcrCB–hsdSMR-mrr)173 endA1 supE44 thi-1 recA1 gyrA96 relA1 lac [F' proAB lacIqZ{Delta}M15 Tn10 (Tetr)]], GM2163 [dam-13 : : Tn9(CmR) dcm-6 hsdR2 () leuB6 hisG4 thi-1 araC lacY1 galK2 galT22 xylA5 mtl-1 rpsL] and ER2566 [F fhuA2 [lon] ompT lacZ : : T7 gene 1 gal sulA11 (mcrC mrr)114 : : IS10R(mcr-73 : : miniTn10-tetS)2R(zgb-210 : : Tn10(TetS) endA1 dcm] were used. These strains and their derivatives were grown at 37 °C in Luria–Bertani (LB) medium (Sambrook et al., 1989). Antibiotics included in media were used at the following final concentrations (µg ml–1): ampicillin 100, chloramphenicol 10, tetracycline 10. Chromosomal DNAs of Neisseria gonorrhoeae FA1090 and Neisseria meningitidis strain Z2491 (serogroup A, subgroup IV-1, The Gambia, 1983) and 2120 (serogroup C, ST-11 complex, Germany, 1997) were obtained from Dr D. C. Stein (University of Maryland, USA). Plasmid pQE-30 was purchased from Qiagen, and pET15b from Novagen.

Cloning of N. gonorrhoeae and N. meningitidis DNA fragments carrying the mismatch nicking endonuclease and accompanying R-M system genes.
DNA fragments carrying the mismatch nicking endonuclease and accompanying R-M system genes were amplified by PCR. The fragment of the chromosomal DNA of N. meningitidis strain 2120 that encodes the M.NmeDI MTase was amplified using primer 3 (5'-GCAGGGATCCTCGTTAAAATACAACC-3') and primer 4 (5'-GCAGAAGCTTTTAGCAGCCGTCAG-3'). This amplicon of 1257 bp was cloned into the R.BamHI and R.HindIII sites of pQE-30, resulting in the formation of plasmid pAK3. This plasmid contains the sequence encoding the M.NmeDI MTase fused into the His-Tag sequence encoded by pQE-30 DNA. The 2784 bp fragment of the chromosomal DNA of N. meningitidis strain 2120 encoding ORFs of the NmeDI R-M system together with the endonuclease V.NmeDI was amplified using primer 5 (5'-GACTCTAGACTTCCTGCGCGATTG-3') and primer 6 (5'-CGCGAAGCTTGGTAACCGAGTGTA-3') and cloned into the R.XbaI and R.HindIII sites of pUC19 DNA, creating plasmid pAK4. The fragment of the chromosomal DNA of N. meningitidis strain Z2491 encoding the M.NmeAI MTase was amplified using primer 1 (5'-CGGCCCTCGAGATGAAAAACAGTAAG-3') and primer 2 (5'-GCGCGGATCCCTAACATTCGATATTATC-3'). This amplicon of 1056 bp was cloned into the R.XhoI and R.BamHI sites of pET15b, resulting in the formation of plasmid pAK1. This plasmid contains the sequence encoding the M.NmeAI MTase fused to the His-Tag sequence encoded by pET15b. All the oligonucleotides used to amplify the above fragments are located in the same regions of the chromosomes of N. meningitidis strains used in studies by Claus et al. (2000). The oligonucleotides used to amplify the fragment of the chromosomal DNA of N. gonorrhoeae strain FA1090 encoding M.NgoAORFC713P (nomenclature according to REBASE) together with the ORFs encoding the putative Vsr-like enzyme (V.NgoAORFC713P) and NgoAORFC713P were primer 7 (5'-ATAGCATGGATCCTCAGACGGCATCTTTTATTTCCTC-3') and primer 8 (5'-CACGCCTGCAGATAGAAATGAAAAACAGTAAGTTAAAG-3'). The amplicon of 3148 bp was cloned into the R.BamHI and R.PstI sites of pUC19 DNA, creating plasmid pAK5.

Oligonucleotides used to amplify the fragment of the chromosomal DNA of N. gonorrhoeae strain FA1090 encoding M.NgoAORFC703P, together with the ORFs encoding the putative Vsr-like enzyme (V.NgoAORFC703P) and restriction endonuclease (NgoAORFC703P), were primer 9 (5'-CGTCTCTAGACTTGAAGCAGTTTG-3') and primer 10 (5'-GCGTGTCGACGATATTTACTCAAC-3'). The amplicon of 1767 bp was cloned into the R.XbaI and R.SalI restriction sites of pLMPMT4{Omega} DNA (Mayer, 1995) creating plasmid pAK6.

The primers for PCR amplification were obtained from IBB Poland. All the PCR reactions were carried out using Pfu DNA polymerase (MBI Fermentas) and used according to the manufacturer's recommendations. All routine cloning procedures were carried out in accordance with protocols described by Sambrook et al. (1989). The expression vectors (pQE-30 or pET15b) carrying the cloned genes were always isolated from cultures induced by IPTG. The published sequences of pQE-30 (www.qiagen.com), pET15b (www.novagen.com), pUC19 (www.neb.com) and the cloned fragments of the chromosomal DNA of N. meningitidis and N. gonorrhoeae used for construction of the plasmids pAK1, pAK3, pAK4, pAK5 and pAK6 served to determine their genetic and restriction maps using the computer program CLONE (Scientific & Educational Software, Durham, NC 27722-2045, USA).

Purification of M.NmeAI and M.NmeDI m5C-MTases.
To purify the M.NmeAI or M.NmeDI MTases single colonies of fresh transformants of E. coli ER2566(pAK1) or of E. coli XL-1 Blue MRF'(pAK3) were used to inoculate 100 ml LB broth (Sambrook et al., 1989) and incubated at 37 °C. When the OD600 of the culture reached 0·6, IPTG (MBI Fermentas) was added to a final concentration of 1 mM and incubation was continued for an additional 4 h at 30 °C. The culture was then used to isolate the plasmid DNA and for purification of the MTase. The culture was centrifuged and bacteria suspended in 1 ml buffer containing 50 mM NaH2PO4, 300 mM NaCl and 10 mM imidazole. After sonication the cellular debris was removed by centrifugation at 40 000 g for 1 h and then the supernatant was applied to a 5 ml Ni-NTA agarose column previously equilibrated with 100 ml of the above buffer. Proteins were eluted according to the manufacturer (Qiagen). The M.NmeAI MTase was eluted at 0·2–0·25 M imidazole and the M.NmeDI MTase at 0·25–0·3 M. The homogeneity of the enzymes was determined by 10 % SDS-PAGE. Proteins used as standards were {beta}-galactosidase (116 kDa), bovine serum albumin (66·2 kDa), ovalbumin (45 kDa) and lactate dehydrogenase (35 kDa).

MTase assays.
During the purification, MTase activity was detected by measurement of transfer of -CH3 groups from [methyl-3H]AdoMet to {lambda} DNA. One microlitre of the prepared enzyme was used in a total volume of 20 µl to methylate 1 µg {lambda} DNA in the presence of 50 mM Tris/HCl (pH 7·5), 10 mM EDTA, 10 mM 2-mercaptoethanol and 7·4x104 Bq [methyl-3H]AdoMet. After 2 h incubation at 37 °C, reaction mixtures were spotted onto DE81 paper (Whatman), washed as described by Sambrook et al. (1989), then dried and their radioactivity measured in a Wallac 1400 liquid scintillation counter. To methylate oligonucleotides (ODNs), the reaction mixtures contained a total volume of 20 µl: 1 µg ODN substrate, 50 mM Tris/HCl (pH 7·5), 10 mM EDTA, 10 mM 2-mercaptoethanol, 7·4x104 Bq [methyl-3H]AdoMet and 1 µg of the purified enzyme. The extent of methylation was assayed as described by Renbaum & Razin (1995).

Determination of the methylation of the separate DNA strands.
Samples (5 µg) of the synthetic double-stranded ODN were methylated as described for MTase assays and separated on a denaturing polyacrylamide gel (Landry et al., 1992). The bands were excised, and the gel slices were crushed in microfuge tubes and incubated overnight at room temperature in 300 µl 0·5 M ammonium acetate, 1 mM EDTA. The slices were microfuged and the supernatant collected. The ammonium acetate supernatants were mixed with water washes and the 3H radioactivity was measured using 10 ml scintillation cocktail (Rotiszint).

Enzymes and chemicals.
Restriction enzymes were purchased from MBI Fermentas and New England Biolabs. T4 DNA ligase, Pfu DNA polymerase and DNA and protein size markers were purchased from MBI Fermentas. Kits for the DNA clean-up and plasmid DNA isolation were purchased from A&A Biotechnology, Gdansk, Poland. Ni-NTA agarose was purchased from Qiagen and [methyl-3H]AdoMet from PerkinElmer Life Sciences. All the chemicals used were reagent grade or better and they were obtained from Sigma and ICN, unless otherwise noted.

Computer analysis.
DNA and protein sequences were compared with the GenBank and SWISS-PROT databases on the BLAST server hosted by the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/blast). The N. gonorrhoeae strain FA1090 genomic sequence was obtained from the University of Oklahoma's Advanced Center for Genome Technology (http://www.genome.ou.edu/gono.html) and the N. meningitidis strain Z2491 (serogroup A) genomic sequence from the Sanger Institute (http://www.sanger.ac.uk/Projects/N_meningitidis). We used the genomic sequence of the N. meningitidis strain FAM 18 (http://www.sanger.ac.uk/Projects/N_meningitidis), which like strain 2120 belongs to serogroup C (complex ST-11), and the DNA sequence (accession no. AJ238948) published by Claus et al. (2000) to characterize the organization of the R-M NmeDI region.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Localization of the genes encoding the MTases associated with potential Vsr-like endonucleases on the chromosome of N. meningitidis and N. gonorrhoeae
In the chromosome of N. meningitidis strain 2120 (serogroup C, ST-11 complex) (Fig. 1), two ORFs encoding the m5C-MTases associated with putative Vsr-like enzymes are present. The nmeDIMP gene (position 699 439 bp to 700 701 bp) encodes a 420 aa protein. M.NmeDI MTase shares a very high homology to various m5C-MTases recognizing the sequence 5'-RCCGGY-3' (http://rebase.neb.com/rebase), like M.NpuORFC230P (54 % identity, 68 % similarity; accession no. AAK68641), M.AvaIX (55 % identity, 79 % similarity; accession no. AAF75232), M.Cfr10I (54 % identity, 70 % similarity; accession no. AAL03947) and M.BsrFI (36 % identity, 54 % similarity; accession no. AAG01143) and much lower homology to other MTases recognizing other sequences, like M.Phi3TII (recognition sequence 5'-TCGA-3'; 27 % identity, 44 % similarity; accession no. CAB56493) and M.HpaII (recognition sequence 5'-CCGG-3'; 25 % identity, 41 % similarity; accession no. AAA20481). It also shares a very high homology to a putative m5C-MTase from Vibrio cholerae (56 % identity, 72 % similarity; accession no. AAF96111) whose specificity is not known.



View larger version (20K):
[in this window]
[in a new window]
 
Fig. 1. Schematic representation of the chromosome regions of N. meningitidis and N. gonorrhoeae FA1090 containing the ORFs encoding the DNA MTases associated with the gene encoding the putative Vsr-like endonucleases. The arrows represent the genes of MTases and other enzymes.

 
The second ORF encoding a potential MTase (position 2 037 650 bp to 2 038 705 bp) associated with the ORF encoding a putative vsr gene (position 2 035 564 bp to 2 035 986 bp) is located between ORF 1 (position 2 038 970 bp to 2 039 113 bp) and ORF 2 (Fig. 1). This ORF M.NmeAI shares 100 % identity with the putative DNA MTase M.NmeAI (ORF NMA0427) from N. meningitidis Z2491.

The DNA sequence analysis of the chromosome of N. meningitidis strain Z2491 (serogroup A) showed the presence of the ORF NMA0427 encoding MTase (M.NmeAI) (position 398 752 bp to 399 807 bp) associated with the ORF NMA0429 encoding the potential Vsr-like protein (V.NmeAI) and the ORF NMA0428 (Fig. 1). This region (position 398 752 bp to 401 893 bp) is flanked by ORF NMA0426 (position 398 238 bp to 398 486 bp) on one side and by ORF NMA0430 (position 401 900 bp to 405 041 bp), encoding the putative DNA helicase, on the other. ORF NMA0427 encodes a 351 aa protein that shares similarity with various m5C-MTases. The M.NmeAI MTase shares homology with various m5C-MTases recognizing the sequence 5'-GTCGAC-3', such as M.HgiDII (38 % identity, 53 % similarity; accession no. CAA38941), M.LmoAP (38 % identity, 53 % similarity; accession no. CAC22275) and M.TerORFS122P (34 % identity, 48 % similarity; accession no. ZP00072165). It also shares a high homology with m5C-MTases recognizing different sequences, such as M.Kpn2I (recognition sequence 5'-TCCGGA-3'; 32 % identity, 48 % similarity; accession no. CAC41108), M.NspI (recognition sequence 5'-RCATGY-3'; 30 % identity, 45 % similarity; accession no. AAC97190) and M.HphI (recognition sequence 5'-TCACC-3'; 30 % identity, 45 % similarity; accession no. CAA59690).

Computer analysis suggests that ORF NMA0428 shares homology with MutL proteins: e.g. 50 % identity and 60 % similarity to mismatch repair protein MutL of Heliobacillus mobilis (accession no. AAN87382), 42 % identity and 64 % similarity to MutL of Xanthomonas axonopodis (accession no. AAM37257), 41 % identity and 62 % similarity to MutL of Caulobacter crescentus (accession no. AAK22680), and 40 % identity and 66 % similarity to MutL of Ralstonia solanacearum (accession no. CAD16270).

In the chromosome of N. gonorrhoeae strain FA1090 two homologous ORFs encoding potential m5C-MTases associated with mismatch nicking endonuclease were detected. However, both ORFs M.NgoAORFC703P (position 301 718 bp to 302 119 bp) and M.NgoAORFC713P (position 1 112 994 bp to 1 113 970 bp) encode truncated forms of proteins due to the presence of mutations. The truncated forms of putative MTases encoded by M.NgoAORFC713P and M.NgoAORFC703P share a very high level of homology to M.NmeAI and M.NmeDI respectively (around 98 % identity).

Purification and characterization of the M.NmeDI and M.NmeAI MTases
The M.NmeDI and M.NmeAI MTases were purified to near-homogeneity from E. coli XL-1 Blue MRF'(pAK3) or E. coli ER2566(pAK1) using single-step purification on Ni-NTA agarose columns. As shown in Fig. 2 the molecular masses of these proteins (39·6±1 kDa for M.NmeAI and 49±1 kDa for M.NmeDI) are close to those predicted on the basis of the nucleotide sequence.



View larger version (69K):
[in this window]
[in a new window]
 
Fig. 2. Coomassie Brilliant Blue R250-stained 10 % SDS-PAGE gel of purified M.NmeAI and M.NmeDI. M, molecular mass markers.

 
Specificity and properties of the M.NmeDI MTase
The high homology of the M.NmeDI to m5C-MTases recognizing the sequence 5'-RCCGGY-3' and the presence of all 10 characteristic motifs for m5C-MTases (data not shown) allow us to assume that this enzyme will recognize the same sequence. The substrate specificity of the M.NmeDI was then analysed by testing the susceptibility of pAK3 DNA to the enzyme R.BsrFI, which recognizes the sequence 5'-RCCGGY-3' and will not cleave the substrate DNA when the second 5' cytosine residues are methylated (http://rebase.neb.com/rebase). The plasmid pAK3 has one sequence 5'-RCCGGY-3' at position 3771 nt (5'-ACCGGC-3') that is not cleaved by this restriction endonuclease (Fig. 3). On the other hand the pAK3 DNA is cleaved by R.Kpn2I (recognition sequence 5'-TCCGGA-3'; positions 588 nt and 1776 nt) and R.BsaWI (recognition sequence 5'-WCCGGW-3') at all sites present in this DNA (588, 1776, 3004, 3151 and 3982 nt) containing the sequence (ACCGGA, TCCGGA or TCCGGT) (data not shown). We have also shown that pAK4 DNA is not cleaved by R.NaeI (recognition sequence 5'-GCCGGC-3') (position 2467 nt), which will not cleave when any of the cytosine residues are methylated (http://rebase.neb.com/rebase), and is cleaved by R.SmaI within the sequence 5'-TACCCGGGGGC/GCCCCCGGGTA-3' (position 5435 nt) (data not shown).



View larger version (62K):
[in this window]
[in a new window]
 
Fig. 3. Susceptibility of the pAK3 plasmid encoding M.NmeDI m5C-MTase isolated from E. coli XL-1 Blue MRF', to cleavage by restriction enzymes. M, DNA size standards (sizes in bp indicated on the left). The grey and black arrows show the positions of the missing and fused fragments respectively, generated after digestion with R.HpaII and R.MspI. The asterisk (*) indicates pAK3 isolated from E. coli GM2163 (Dam Dcm).

 
pAK3 is also not cleaved to completion by the restriction enzymes that recognize the sequences not directly overlapping the sequence 5'-RCCGGY-3'. After cleavage of pAK3 by R.ScrFI (recognition sequence 5'-CCNGG-3') isolated from E. coli XL-1 Blue MRF' (Dcm+), only the linear form of the plasmid DNA is produced instead of the generation of the six fragments of 1958, 944, 699, 696, 351 and 35 bp (Fig. 3). This linear form is the result of the cleavage at one particular position located at 2444 nt of the pAK3 DNA as indicated by the double cleavage with R.ScrFI and R.HindIII. Since R.ScrFI does not cleave at the sites modified by the M.EcoKDcm MTase (recognition site 5'-CCWGG-3') (http://rebase.neb.com/rebase) we also tested digestion by this enzyme of pAK3 DNA isolated from E. coli GM2163 (Dam Dcm) cells. R.ScrFI generates fragments of 3679, 565, 248, 121, 56 and 13 bp (on Fig. 3 only the three largest fragments are visible) instead of the fragments of 1958, 696, 457, 351, 347, 248, 218, 183, 121, 56, 35 and 13 bp that would be obtained after complete digestion. If M.NmeDI also methylated at the sequence 5'-CCWGG-3', then R.ScrFI would cleave pAK3 isolated from E. coli GM2163 at only one position and generate the linear form.

R.HpaII and R.MspI generate the same patterns of fragments after digestion of pAK3 DNA (Fig. 3). The complete digestion of the pAK3 DNA should generate 16 fragments, of 1047, 910, 539, 527, 404, 242, 236, 190, 147, 128, 110, 67, 42, 34, 34 and 26 bp. Instead, fused fragments of 1290, 1146, 561 and 434 bp (black arrows in Fig. 3) and the absence of fragments of 1047, 910, 527 or 404 bp (grey arrows in Fig. 3) are observed. The analysis of the fragment profile of pAK3 DNA obtained after digestion with R.ScrFI, R.HpaII and R.MspI indicates the common sequence of 5'-RCCGGB-3' that is methylated by M.NmeDI (Table 1).


View this table:
[in this window]
[in a new window]
 
Table 1. Analysis of the restriction pattern of pAK3 DNA obtained after digestion with R.ScrFI, R.BsrFI, R.HpaII and R.MspI

 
To get more information on the specificity of M.NmeDI, a series of double-stranded ODNs were tested for methylation in vitro by the purified enzyme in the presence of [methyl-3H]AdoMet (Table 2). A comparison of the sequences of the ODNs M.NmeDI/M.NmeDIkr, Cfr10L/Cfr10P (the best substrates), the ODNs that are not methylated (KS1/KS2, VSR30/VSR31, AP108/AP109, AK3/AK4) and the ODNs AK1/AK2, KANSDIL/KANSDIP and AK5/AK6 methylated about twenty times less efficiently (Table 2) than the ODNs having sequence 5'-RCCGGY-3' indicates that the enzyme recognizes the sequence 5'-RCCGGB-3' but preferentially methylates the sequence 5'-RCCGGY-3' over the fully palindromic sequence 5'-RCCGGB-3'. In the recognized sequences the enzyme methylates both strands with the same efficiency (Table 3). Comparison of the methylation efficiency of MK7/MK5 and AK1/AK2 ODNs indicates that M.NmeDI methylates the sequence 5'-RCCGGG/CCCGGY-3' more efficiently when it is in the hemimethylated as opposed to the unmethylated state although not as efficiently as the sequence 5'-RCCGGY-3'. Under the conditions where the enzyme showed the highest specificity for the 5'-RCCGGY-3' sequence, the rate of methylation of hemimethylated degenerate sequence was about five times lower than for unmethylated sequence.


View this table:
[in this window]
[in a new window]
 
Table 2. Efficiency of the methylation of double-stranded ODNs by M.NmeDI and M.NmeAI MTases

See Methods for details of the methylation assay.

 

View this table:
[in this window]
[in a new window]
 
Table 3. Methylation of the separate strands of duplex ODNs by M.NmeDI MTase

The substrate sequence for M.NmeDI MTase is shown in bold. M denotes the C-5 methylated cytosine residues. The reaction was carried out as described in Methods.

 
The resistance of pAK3 to cleavage by R.BsrFI, R.HpaII and R.MspI at position 3772 nt containing the sequence 5'-RCCGGY-3' indicates that the M.NmeDI methylates the first 5' cytosine residue within the sequence. According to Roberts & Macelis (2002), only the modification of this cytosine protects against cleavage by both R.MspI and R.HpaII. The same cytosine residue is methylated in the upper strand of the degenerate 5'-RCCGGG/CCCGGY-3' sequence (shown in bold). This conclusion is based on the facts that: (1) R.MspI and R.HpaII (sensitive to methylation of the first 5' cytosine residue in the sequence 5'-CCGG-3') generate the same pattern of fragments after digestion of pAK3 DNA; (2) pAK3 is completely cleaved by R.NciI (sensitive to methylation of the cytosine residue in the CpG context within the recognized sequence 5'-CCSGG-3'; Roberts & Macelis, 2002), generating all expected fragments (1958, 944, 699, 696, 351 and 35 bp) (Fig. 3); and (3) methylation by M.NmeDI blocks digestion by R.ScrFI (sensitive to methylation of the cytosine residue in the CNG context within the sequence 5'-CCNGG-3'; Roberts & Macelis, 2002). To find which of the cytosine residues in the lower strand of the recognition sequence is methylated, the hemimethylated ODN duplexes (AK9/MK7 and MK2/MK7) containing the substrate sequence were methylated in vitro in the presence of [methyl-H3]AdoMet. Measurement of the extent of the methylation of each DNA strand showed that in the lower strand the second cytosine has to be methylated by M.NmeDI (Table 3). Thus M.NmeDI methylates the first 5' cytosine residues in both strands in the sequence 5'-RCCGGY-3' and the first 5' cytosine residue in the upper strand and the second 5' cytosine residue in the lower strand within the degenerate sequence 5'-RCCGGB/YGGCCC-3'.

Specificity and properties of the M.NmeAI MTase
The high homology of the M.NmeAI to m5C-MTases recognizing the sequence 5'-GTCGAC-3' suggested that this enzyme recognizes the same sequence. Since the plasmid pAK1 encoding the M.NmeAI does not contain the presumptive specific recognition sequence 5'-GTCGAC-3', the ability to recognize it was tested by the methylation of the specific ODNs. The results presented in Table 3 indicate that an ODN containing this sequence is not methylated by the purified MTase. However, as in the case of M.NmeDI, we have noticed that pAK1 is not cleaved by several restriction enzymes whose activity is inhibited by the presence of C-5 methylated cytosine residues in the DNA (Roberts & Macelis, 2002). Among the restriction enzymes tested [R.HpaII (recognition sequence 5'-CCGG-3'), R.MspI (recognition sequence 5'-CCGG-3'), R.NciI (recognition sequence 5'-CCSGG-3')], R.HpaII was totally unable to cleave pAK1 DNA, R.NciI showed partial cleavage, while R.MspI cleaved completely (Fig. 4). These results indicate that M.NmeAI recognizes the sequence 5'-CCGG-3', which was confirmed by the methylation of the ODNs containing this sequence (Table 2).



View larger version (52K):
[in this window]
[in a new window]
 
Fig. 4. Susceptibility of pAK1 DNA, encoding M.NmeAI m5C-MTase isolated from E. coli ER2566, to cleavage by restriction enzymes. M, DNA size standards (sizes in bp indicated on the right).

 
R.HpaII and R.MspI recognize the same 5'-CCGG-3' sequence but R.MspI will not cleave when only the first 5' cytosine is methylated at the C-5 position while R.HpaII will not cleave when either of the cytosine residues is methylated (Roberts & Macelis, 2002). pAK1 is partially protected against cleavage by R.NciI. The activity of this enzyme is blocked by the modification of the cytosine at the C-5 position located in the CpG context (Roberts & Macelis, 2002). The purified enzyme methylates both DNA strands and, like M.NmeDI, shows higher efficiency of methylation for hemimethylated DNA (data not shown). These results indicate that M.NmeAI m5C-MTase recognizes the sequence 5'-CCGG-3' and methylates the second 5' cytosine residue in both DNA strands.

Cloning of the N. gonorrhoeae putative homologues of M.NmeAI and M.NmeDI MTases
The organization and the types of the ORFs in N. gonorrhoeae strain FA1090 present between the pheS and pheT genes in the chromosome sequence are the same as in N. meningitidis strain 2120. The analysis of the amino acid sequence of the ORF encoding a potential MTase homologous to M.NmeAI indicated the presence of an additional stop codon and a frameshift mutation (position 1 113 500 bp, insertion of G: GACCTT[G]GGTCAG) that divided the amino acid sequence into three separate putative proteins. After cloning the chromosome fragments containing these pseudogenes, we were unable to detect any specific methylase activity using different assay methods. Similarly, we have shown the presence of the frameshift mutation in the amino acid sequence of the ORF encoding a potential MTase homologous to M.NmeDI, and cells harbouring pMPMT4{Omega}, carrying the cloned chromosome fragment encoding this ORF, did not show any specific methylase activity.

Sequence analysis of the variable region (TRD) of M.NmeDI
M.NmeDI shows the presence of characteristics of the 10 conserved motifs (I through X) of m5C-MTases (Posfai et al., 1989; Cheng et al., 1993; Kumar et al., 1994; Lauster, 1989) with the same linear order and the TRD region located between the conserved motifs VIII and IX (data not presented). The TRD region of M.NmeDI MTase contains about 110 aa. In this region residues present in the C-terminal part (TRD-C) (about 75 aa) share almost 90 % identity, while the N-terminal part has only about 28 % identity with its closest neighbours which recognize the sequence 5'-RCCGGY-3' (Fig. 5), suggesting the possibility that they represent two different TRD sequences. To confirm that the TRD-C is responsible for the recognition of the sequence 5'-RCCGGY-3', while the N-terminal part is responsible for the recognition of the degenerate sequence, we generated mutations in the C-terminal part of TRD (amino acids 323 to 335). However, the two deletion constructs studied had lost the capacity to methylate both the 5'-RCCGGY-3' and the degenerate sequence, arguing strongly against the presence of the two different TRD regions (data not shown).



View larger version (23K):
[in this window]
[in a new window]
 
Fig. 5. Sequence alignment of the TRD regions of M.NmeDI (accession no. CAB59897), M.NpuORFC230P (accession no. AAK68641) and M.AvaIX (accession no. AAF75232) MTases. Protein sequences were compared with the GenBank and SWISS-PROT databases on the BLAST server hosted by the National Center for Biotechnology Information.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
We have determined the recognition sequence of the two m5C-MTases from N. meningitidis that are associated with the genes encoding the putative Vsr endonucleases. Our results indicate that M.NmeDI recognizes the degenerate sequence 5'-RCCGGB-3'. This conclusion is based on the analysis of the protection against cleavage of the plasmid DNA encoding the M.NmeDI by the restriction endonucleases and the methylation efficiency of the specific ODNs by the purified M.NmeDI MTase. Of the 16 possible flanking sites of the sequence CCGG we have tested 14, which excludes any bias against any particular sequence. Among them the sequences with the flanking bases GT, GC, AC were the best substrates in the in vitro methylation reactions as well as in a full protection against appropriate restriction endonucleases. The sequence CCGG with the flanking sites AG or GG gave a good protection against cleavage but was less efficiently methylated in vitro. The enzyme does not recognize efficiently more degenerate sequences like 5'-CCWGG-3', 5'-CCSGG-3' or just 5'-CCGG-3' if they do not overlap with the specific sequence, since the ODNs AP108/AP109, AK3/AK4, VSR30/VSR31 (Table 2) are not methylated. The sequence 5'-CCGG-3' is not recognized since not all sites present in pAK3 DNA are protected against cleavage by R.HpaII and R.MspI. However, the presence of some additional fragments of about 1900, 1490 or 600 bp (Fig. 3) produced with a different molar ratio indicates that the M.NmeDI MTase can probably methylate more degenerate sequences with lower efficiency.

In the most efficiently methylated sequence 5'-RCCGGY-3' and its degenerate form 5'-RCCGGB-3' the M.NmeDI methylates the same cytosine residues. The TRD region of M.NmeDI MTase shows the presence of two distinct regions, of which the carboxyl-terminal part shares very high identity with the MTases recognizing the sequence 5'-RCCGGY-3', while the amino-terminal part shares less then 30 %. We do not know whether the amino-terminal part of TRD regions of such MTases like M.NpuI or M.Cfr10I reflects their ability to recognize different degenerate forms of the same basic 5'-RCCGGY-3' sequence as recognized by M.NmeDI. If this were true then the lack of homology in the N-terminal parts of their TRD regions would be connected to differences in the degeneracy of the recognition sequence. However, it was shown that M.AvaIX, which shares high homology with its C-terminal part of the TRD with M.NmeDI, methylates only within the sequence 5'-RCCGGY-3' (Matveyev et al., 2001). Although DNA MTases are viewed as highly sequence specific (Dryden, 1999), recent observations suggest that methylation of non-canonical sites may be a common feature of these enzymes (Bandaru et al., 1996; Beck et al., 2001; Cohen et al., 2002; Friedrich et al., 2000). What could be the biological role of the ability to recognize the non-canonical form of the recognition sequence? If the R.NmeDI REase associated with M.NmeDI makes mistakes and is able to cleave the DNA in the degenerate sequences, then such activity will be suicidal for the cell. The ability of the MTase to recognize the same degenerate form of the sequence will then protect the DNA against the cleavage and suicidal death. Further studies of the activity of the R.NmeDI are needed to determine if this assumption is correct.

M.NmeAI is the only MTase characterized in N. meningitidis and N. gonorrhoeae that recognizes the sequence 5'-CCGG-3' (Roberts & Macelis, 2002). The regions of N. meningitidis Z2491 and N. gonorrhoeae FA1090 encoding the M.NmeAI and its homologue had no genes nearby that could conceivably encode a restriction endonuclease. We can argue that M.NmeAI is a solitary MTase whose function in these bacteria is not known. It could be that this MTase was a part of an R-M system that was transferred horizontally (the region is flanked by the uptake sequences) and later the gene encoding the restriction enzyme was lost. However, the fact that this MTase does not methylate the chromosomal DNA of N. meningiditis uniformly (data not presented) may indicate some important role in the life of these bacteria.

The 100 % identity of the M.NmeAI present in N. meningitidis serogroup C strain 2120 to M.NmeAI in N. meningitidis serogroup A and the same localization in the chromosome strongly suggest that these two MTases recognize the same 5'-CCGG-3' sequence.


   ACKNOWLEDGEMENTS
 
This work was supported by KBN grant no. 6 P04A 037 18 and KBN grant no. 3 P04A 040 23.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Balganesh, T. S., Reiners, L., Lauster, R., Noyer-Weidner, M., Wilke, K. & Trautner, T. A. (1987). Construction and use of chimeric SPR/phi 3T DNA methyltransferases in the definition of sequence recognizing enzyme regions. EMBO J 11, 3543–3549.

Bandaru, B., Gopal, J. & Bhagwat, A. S. (1996). Overproduction of DNA cytosine methyltransferases causes methylation and C>T mutations at non-canonical sites. J Biol Chem 271, 7851–7859.[Abstract/Free Full Text]

Beck, C., Cranz, S., Solmaz, M., Roth, M. & Jeltsch, A. (2001). How does a DNA interacting enzyme change its specificity during molecular evolution? A site-directed mutagenesis study at the DNA binding site of the DNA-(adenine-N6) methyltransferase EcoRV. Biochemistry 40, 10956–10965.[CrossRef][Medline]

Bhagwat, A. S. & Lieb, M. (2002). Cooperation and competition in mismatch repair: very short-patch repair and methyl-directed mismatch repair in Escherichia coli. Mol Microbiol 44, 1421–1426.[CrossRef][Medline]

Cheng, X., Kumar, S., Posfai, J., Pflugrath, J. W. & Roberts, R. J. (1993). Crystal structure of the HhaI DNA methyltransferase complexed with S-adenosyl-methionine. Cell 74, 299–307.[Medline]

Claus, H., Friedrich, A., Frosch, M. & Vogel, U. (2000). Differential distribution of novel restriction-modification systems in clonal lineages of Neisseria meningitidis. J Bacteriol 182, 1296–1303.[Abstract/Free Full Text]

Cohen, H. M., Tawfik, D. S. & Griffiths, A. D. (2002). Promiscuous methylation of non-canonical DNA sites by HaeIII methyltransferase. Nucleic Acids Res 30, 3850–3855.

Dryden, D. T. F. (1999). Bacterial DNA methyltransferases. In S-Adenosylmethionine-dependent Methyltransferases, pp. 283–340. Edited by X. Cheng & R. M. Blumenthal. Singapore: World Scientific.

Friedrich, T., Fatemi, M., Gowhar, H., Leismann, O. & Jeltsch, A. (2000). Specificity of DNA binding and methylation by the M.FokI DNA methyltransferase. Biochim Biophys Acta 1480, 145–159.[Medline]

Hennecke, F., Kolmar, H., Bründl, K. & Fritz, H. J. (1991). The vsr gene product of Escherichia coli K-12 strain is a strand- and sequence-specific DNA mismatch endonuclease. Nature 353, 776–778.[CrossRef][Medline]

Klimasauskas, S., Nelson, J. L. & Roberts, R. J. (1991). The sequence specificity domain of cytosine-C5 methylases. Nucleic Acids Res 19, 6183–6190.[Abstract]

Kumar, S., Cheng, X., Klimasauskas, S., Mi, S., Posfai, J., Roberts, R. J. & Wilson, G. G. (1994). The DNA (cytosine-5) methyltransferases. Nucleic Acids Res 22, 1–10.[Abstract]

Laging, M., Lindner, E., Fritz, H. J. & Kramer, W. (2003). Repair of hydrolytic DNA deamination damage in thermophilic bacteria: cloning and characterization of a Vsr endonuclease homologue from Bacillus stearothermophilus. Nucleic Acids Res 31, 1913–1920.[Abstract/Free Full Text]

Landry, D., Barsomian, J. M., Feehery, G. R. & Wilson, G. G. (1992). Characterization of type II DNA-methyltransferase. Methods Enzymol 216, 244–254.[Medline]

Lauster, R. (1989). Evolution of type II DNA methyltransferases. A gene duplication model. J Mol Biol 206, 313–321.[Medline]

Lieb, M. & Bhagwat, S. A. (1996). Very short patch repair: reducing the cost of cytosine methylation. Mol Microbiol 20, 467–473.[CrossRef][Medline]

Matveyev, A. V., Young, K. T., Meng, A. & Elhai, J. (2001). DNA methyltransferses of the cyanobacterium Anabaena PCC 7120. Nucleic Acids Res 29, 1491–1506.[Abstract/Free Full Text]

Mayer, M. P. (1995). A new set of useful cloning and expression vectors derived from pBlueScript. Gene 163, 41–46.[CrossRef][Medline]

Posfai, J., Bhagwat, A. S., Posfai, G. & Roberts, R. J. (1989). Predictive motifs derived from cytosine methyltransferases. Nucleic Acids Res 17, 2421–2435.[Abstract]

Renbaum, P. & Razin, A. (1995). Interaction of M.SssI and M.HhaI with single-base mismatched oligodeoxynucleotide duplexes. Gene 157, 177–179.[CrossRef][Medline]

Roberts, R. J. & Macelis, D. (2002). REBASE: restriction enzymes and methyltransferases. Nucleic Acids Res 31, 418–420 (see also http://rebase.neb.com/rebase).[CrossRef]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Trautner, T. A., Pawlek, B., Behrens, B. & Willert, J. (1996). Exact size and organization of DNA target-recognizing domains of multispecific DNA-(cytosine-C5) methyltransferases. EMBO J 15, 1434–1442.[Abstract]

Walter, J., Trautner, T. A. & Noyer-Weidner, M. (1992). High plasticity of multispecific DNA methyltransferases in the region carrying DNA target recognizing enzyme modules. EMBO J 12, 4445–4450.

Wilke, K., Rauhut, E., Noyer-Weidner, M., Lauster, R., Pawlek, B., Behrens, B. & Trautner, T. A. (1988). Sequential order of target-recognizng domains in multispecific DNA-methyltransferases. EMBO J 7, 2601–2609.[Abstract]

Wilson, G. G. & Murray, N. E. (1991). Restriction and modification systems. Annu Rev Genet 25, 585–627.[CrossRef][Medline]

Received 29 December 2003; revised 24 February 2004; accepted 5 March 2004.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Kwiatek, A.
Articles by Piekarowicz, A.
Articles citing this Article
PubMed
PubMed Citation
Articles by Kwiatek, A.
Articles by Piekarowicz, A.
Agricola
Articles by Kwiatek, A.
Articles by Piekarowicz, A.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2004 Society for General Microbiology.