Biochemical Journal

Research article

A nucleotide insertion between two adjacent methyltransferases in Helicobacter pylori results in a bifunctional DNA methyltransferase

Ritesh Kumar , Desirazu N. Rao

Abstract

Helicobacter pylori has a dynamic R-M (restriction–modification) system. It is capable of acquiring new R-M systems from the environment in the form of DNA released from other bacteria or other H. pylori strains. Random mutations in R-M genes can result in non-functional R-M systems or R-M systems with new properties. hpyAVIAM and hpyAVIBM are two solitary DNA MTase (methyltransferase) genes adjacent to each other and lacking a cognate restriction enzyme gene in H. pylori strain 26695. Interestingly, in an Indian strain D27, hpyAVIAMhpyAVIBM encodes a single bifunctional polypeptide due to insertion of a nucleotide just before the stop codon of hpyAVIBM and, when a similar mutation was made in hpyAVIAMhpyAVIBM from strain 26695, a functional MTase with an N-terminal C5-cytosine MTase domain and a C-terminal N6-adenine MTase domain was constructed. Mutations in the AdoMet (S-adenosylmethionine)-binding motif or in the catalytic motif of M.HpyAVIA or M.HpyAVIB selectively abrogated the C5-cytosine or N6-adenine methylation activity of M.HpyAVIA–M.HpyAVIB fusion protein. The present study highlights the ability of H. pylori to evolve genes with unique functions and thus generate variability. For organisms such as H. pylori, which have a small genome, these adaptations could be important for their survival in the hostile host environment.

  • DNA methylation
  • Helicobacter pylori
  • methyltransferase
  • restriction–modification system

INTRODUCTION

The presence of highly polymorphic R-M (restriction–modification) enzymes in Helicobacter pylori suggests that it has evolving R-M systems [1]. It is naturally competent and thus able to take up DNA from the surroundings. It has the ability to accumulate R-M systems, presumably by integrating them into inactive positions of the genome, but it can also, by random mutations, phase-variation and recombination, inactivate R-M systems [2]. It has been shown that R-M systems in H. pylori play a critical role in regulating gene expression. Several reports have indicated that H. pylori R-M systems have a function beyond genome protection [3,4]. A number of H. pylori strains have been sequenced, and each sequenced genome has more than 20 R-M systems. H. pylori R-M systems belong to Types I, II and III, but a large number of the R-M genes code for Type II enzymes [1]. Several subtypes of Type II enzymes (Types IIB, IIE, IIF, IIG, IIS, IIT, IIM and IIP) have been defined [5,6].

Type II MTases (methyltransferases) which recognize symmetrical sequences modify either adenine or cytosine residues in both strands of the recognition sequence. Enzymes that recognize asymmetrical sequences usually require two MTases to modify the two strands of DNA. For instance, in the case of M.FokI, it has been demonstrated that the same enzyme methylates adenine residues in both strands of the target. However, different protein segments of M.FokI are involved in base modification of complementary strands [7]. Another Type IIS R-M system, HgaI, consists of two separate cytosine MTases which are responsible for methylation of different DNA strands [8]. Some Type IIS enzymes recognize sequences where one strand contains no adenine and the other strand contains no cytosine, e.g. MboII- 5′-GAAGA-3′ [9] or Ksp632I-5′-CTCTTC-3′ [10]. One-strand methylation has been postulated for M.NgoBI [11] which recognizes the sequence 5′-GGTGA-3′. This is a cytosine MTase and methylates only one strand because the other strand contains no cytosine. M.Alw26I, M.Eco31I and M.Esp3I methylate both strands of asymmetrical recognition sites yielding C5-methylcytosine in the upper strand and N6-methyladenine in the lower strand. These MTases are represented by proteins of unusual structural and functional organization. Both M.Alw26I and M.Esp3I are represented by a single bifunctional protein which is composed of an N6-adenine MTase domain fused to a C5-cytosine MTase domain [12]. In contrast, two separate genes encode the N6-adenine MTase and C5-cytosine MTase in M.Eco31I.

In H. pylori strain 26695, hpyAVIAM encodes an N6-adenine MTase, and hpyAVIBM encodes a C5-cytosine MTase. It has been shown that M.HpyAVIA and M.HpyAVIB MTases recognize the same sequence 5′-CCTC-3′/3′-GGAG-5′, where M.HpyAVIA MTase methylates the adenine base in the lower strand and M.HpyAVIB methylates the cytosine base in the upper strand [13,14]. This is also true in the case of the HpyAV R-M system, where the MTase(s) modify a cytosine of the upper strand and an adenine of the lower strand in the target sequence. Bioinformatic analysis has shown that the MTase gene of the HpyAV R-M system is a fusion of a C5-cytosine MTase and an N6-adenine MTase, highly homologous with M1.Hin4II and M2.Hin4II respectively [15].

The significance of DNA methylation in gene regulation is well established in a number of bacterial pathogens [16,17]. The M.HpyAVIA and M.HpyAVIB MTases share homology with the MnlI system, a Type IIS R-M system having one restriction enzyme and two cognate MTases [18]. It has been shown previously that both the MTases (M.HpyAVIA and M.HpyAVIB) lack an active cognate restriction enzyme [19]. In H. pylori strain J99, the hpyAVIAM homologue is present, but the hpyAVIBM homologue is absent [1]. Thus it is possible that hpyAVIAM and hpyAVIBM together represent a decaying R-M system. In H. pylori strain 26695, hpyAVIAM and hpyAVIBM are encoded by same mRNA [20] and hpyAVIAM overlaps the hpyAVIBM ORF (open reading frame) by one nucleotide. The distribution of hpyAVIAM [14] and hpyAVIBM (R. Kumar, A. K. Mukhopadhyay and D. N. Rao, unpublished work) was studied in number of clinical isolates and a number of hpyAVIAM and hpyAVIBM homologues were sequenced. Interestingly, in H. pylori strain D27 (isolated from a patient suffering from gastritis), an insertion just before the stop codon of hpyAVIBM causes a frameshift mutation and thus both the MTases are in one frame coding for a single polypeptide. In the present study, we found that the fused hpyAVIAM and hpyAVIBM genes (GenBank® accession number HQ 389436) produce a single polypeptide with a functional C5-cytosine MTase NTD (N-terminal domain) and an N6-adenine MTase CTD (C-terminal domain). Using site-directed mutagenesis a similar mutation was made in hpyAVIAM and hpyAVIBM from strain 26695 and a single polypeptide was obtained with a functional C5-cytosine MTase NTD and an N6-adenine MTase CTD. The fusion of hpyAVIAM and hpyAVIBM highlights the fluidity of the H. pylori genome where random mutations and genomic rearrangements may result in evolution of new genes with unique features.

MATERIALS AND METHODS

Strains and plasmids

H. pylori 26695 strain (cagA+ iceA1 vacAs1m1) genomic DNA was obtained as a gift from New England Biolabs. H. pylori strains PG227 (cagA+vacA), D27(cagA+vacA), J99 (cagA+vacA) and SS1 (cagA+ iceA1 vacAs1m2) were grown on Petri dishes containing BHI (brain/heart infusion) agar (Difco). Chromosomal DNA from bacterial pellets was prepared from confluent growth on BHI agar plate cultures using the cetyltrimethylammonium bromide extraction method [21]. Escherichia coli strain DH5α {F′/endA1 hsdR17 (rk mk) glnV44 thi1 recA1 gyrA (NalR) relA1 Δ(lacIZYA-argF) U169 deoR [Φ80dlac Δ(lacZ)M15]} was used as a host for preparation of plasmid DNA. E. coli BL21(DE3) F ompT hsdSB (rB mB) gal (dcm)(lon) (DE3) cells were used for expression of wild-type and mutant M.HpyAVIA–M.HpyAVIB fusion proteins.

RNA isolation and RT (reverse transcription)–PCR

Bacterial RNA was stabilized in vivo by using RNA Protect Bacteria Reagent (Qiagen). Total RNA was isolated by using RNeasy Kits for RNA purification (Qiagen) following the manufacturer's protocol. RT was performed on 2 μg of total RNA by using the RevertAid™ H Minus First Strand cDNA synthesis kit (Fermentas) following the manufacturer's protocol. Of the cDNA, 2 μl was used in separate PCR volumes of 20 μl for each gene. To exclude the presence of DNA, for each sample, the complete RT–PCR procedure was also carried out without adding reverse transcriptase. Primers used in the RT–PCR are described in Supplementary Table S1 at http://www.BiochemJ.org/bj/433/bj4330487add.htm.

PCR amplification and cloning of hpyAVIAM and hpyAVIBM ORFs from H. pylori strains

The 1763 bp DNA fragment corresponding to the hpyAVIAM and hpyAVIBM gene was amplified from genomic DNA of different strains of H. pylori by the PCR with Pfu polymerase using primers 1 and 2 (see Supplementary Table S2 at http://www.BiochemJ.org/bj/433/bj4330487add.htm). The primers were designed with the help of the annotated complete genome sequence of H. pylori 26695 considering the putative gene sequence of hpyAVIAM and hpyAVIBM obtained from TIGR (http://www.tigr.org/tdb/tgi/). The amplified PCR fragment corresponding to hpyAVIAM and hpyAVIBM was ligated into the SmaI site of pUC19 and then inserted into the bacterial expression vector pET28a/pGEX4T2 at the BamHI and XhoI sites.

Purification of M.HpyAVIA–M.HpyAVIB fusion MTase of H. pylori

E. coli strain BL21(DE3) cells were separately transformed with the pET28a-hpyAVIAMhpyAVIBM and the pGEX4T2-hpyAVIAM–hpyAVIBM construct plasmids. E. coli BL21(DE3) cells harbouring pET28a-hpyAVIAMhpyAVIBM were grown at 30 °C in 600 ml of LB (Luria–Bertani) broth containing 50 μg/ml kanamycin to a D600 of 0.6, and protein expression was induced by the addition of isopropyl β-D-thiogalactopyranoside to a final concentration of 0.5 mM. After 1 h of induction at 30 °C, the culture was cooled on ice, and the cells were harvested by centrifugation at 8000 rev./min for 30 min at 4 °C using a Kubota 6500 rotor. All purification steps were carried out at 4 °C. The cell pellet was resuspended in extraction buffer (10 mM Tris/HCl, pH 8.0, 0.05% Triton X-100, 100 mM L-arginine, 100 mM NaCl and 50 mM imidazole) and lysed by sonication. The cell lysate was centrifuged at 8000 rev./min for 1 h at 4 °C using a Kubota 6500 rotor. The supernatant was collected and was loaded on to an Ni-NTA (Ni2+-nitrilotriacetate) column which was previously equilibrated with the above-mentioned buffer. The M.HpyAVIA–M.HpyAVIB enzyme was eluted by using 10 ml of 10 mM Tris/HCl (pH 8.0) containing 100 mM NaCl and 200 mM imidazole and dialysed overnight at 4 °C against 10 mM Tris/HCl (pH 8.0) containing 100 mM NaCl, 10 mM 2-mercaptoethanol and 30% (v/v) glycerol. The purity of the protein preparations were determined by SDS/PAGE (10% gels) with Coomassie Brilliant Blue staining [22]. Protein was estimated using the method of Bradford using BSA as a standard [23]. The N-terminal His6 tag was removed by using a thrombin CleanCleave™ kit (Sigma). GST (glutathione transferase)-tagged M.HpyAVIA–M.HpyAVIB fusion protein was overexpressed and purified using the method described previously [24]. The pGEX4T2-hpyAVIAMhpyAVIBM plasmid (200 ng) was used as a template in the 25 μl reaction volume and the fusion protein was synthesized by using a PURExpress™ in vitro protein synthesis kit (NEB). The reaction mixture was incubated for 2 h at 37 °C. Translated protein was purified following the manufacturer's instructions.

Methylation activity

All methylation assays monitored incorporation of tritiated methyl groups in to DNA by using a modified ion-exchange filter-binding assay [25]. Methylation assays were carried out in a reaction mixture (25 μl) containing supercoiled pUC19 plasmid DNA or duplex DNA (duplexes 1–4 as a substrate for M.HpyAVIA–M.HpyAVIB fusion MTase) which harbour a single recognition sequence or modified recognition sequence (Table 1), [3H]AdoMet (S-adenosylmethionine) (specific activity of 66 Ci/mmol) and purified protein in the reaction buffer (10 mM Tris/HCl, pH 8.0, and 5 mM 2-mercaptoethanol). After incubation at 37 °C for 30 min, reactions were stopped by snap-freezing in liquid nitrogen. Background counts were subtracted and the data were analysed. All methylation experiments were carried out at least in triplicate and the results were averaged. The S.D. values of the average methylation rates were below 10%.

View this table:
Table 1 Duplex DNA used in the present study

mA, N6-methyladenine; mC, C5-methylcytosine. MTase recognition site is indicated in bold.

Determination of kinetic parameters

Kinetic studies were carried out using a 26-mer duplex. Methylation assays were carried out as described previously [14] in a series of similar reactions containing DNA MTase (50 nM), [3H]AdoMet (5.0 μM) and 26-mer duplexes 1–3 (0.5–20 μM) (Table 1). The velocities were fitted with the Michaelis–Menten equation: Embedded Image

All points were analysed by non-linear regression and the Km (DNA) and Vmax were determined. kcat was calculated as the ratio of Vmax/[E]. Similarly, initial velocity experiments were carried out by varying the concentration of [3H]AdoMet in the range 0.3–15 μM, while keeping the DNA concentration fixed at 5 μM and keeping other reaction conditions identical. Data were plotted by non-linear regression analysis using GraphPad Prism 5. All methylation experiments were carried out in triplicate and the results were averaged. S.D. values of the average methylation rates were less than 10%.

Site-directed mutagenesis

A DNA fragment corresponding to the hpyAVIAM and hpyAVIBM genes was amplified from the genomic DNA of H. pylori strain 26695 and ligated into either pET28a or pGEX4T2 as described above. A fusion was created by site-directed mutagenesis [26]. For the construction of the hpyAVIAM and hpyAVIBM fusion, primers 3 and 4 (see Supplementary Table S2) were used for the synthesis of a megaprimer (primer 3 was mutagenic). The mutagenic primer was designed in such a way that a nucleotide insertion will disrupt a HindIII site allowing the resultant plasmids to be easily screened. The resultant plasmids were used for expression and purification of M.HpyAVIA–M.HpyAVIB fusion protein. Mutations were made in the AdoMet-binding motif and catalytic motif of N-terminal C5-cytosine MTase. Site-directed mutagenesis was performed using a PCR-based technique to replace the required amino acids [26]. Mutations were introduced in hpyAVIBM by using the two-stage megaprimer PCR method. PCRs were carried out with Phusion DNA polymerase (Finnzymes). For each substitution, a mutagenic primer and appropriate second primer was used.

Phenylalanine (shown in bold) in the AdoMet-binding motif sequence, FXGXG, was replaced by using primers 5 and 7 (see Supplementary Table S2, primer 5 was mutagenic). By substituting asparagine for phenylalanine, it was possible to introduce a convenient restriction site (DraI), thus allowing screening for F9N mutants. Similarly, site-directed mutagenesis was performed to replace the cysteine residue (shown in bold) in the catalytic site sequence, PCQ, by using primers 6 and 7 (see Supplementary Table S2, primer 6 was mutagenic). By substituting tryptophan for cysteine, an NcoI site was created and this property was used for the screening for C82W mutants. Mutations were made in the AdoMet-binding motif and catalytic motif of the N6-adenine MTase CTD as described previously [14]. The full-length PCR product was obtained in the second round of the PCR by extension of the megaprimer by using pET28a-hpyAVIAMhpyAVIBM or pGEX4T2-hpyAVIAMhpyAVIBM (strain 26695) as a template. The PCR product thus obtained was purified and digested with the DpnI restriction enzyme to cleave the methylated template DNA, E. coli DH5α was transformed with this plasmid and plated on to LB agar containing kanamycin (50 μg/ml) or ampicillin (100 μg/ml), and the resultant plasmids were used for expression and purification of M.HpyAVIA–M.HpyAVIB fusion protein.

Dot blot assay for methylation activity

Methylation activity was measured in a dot blot assay using rabbit primary antibodies raised against DNA with N6-methyladenine (NEB) and monoclonal antibodies raised against C5-methylcytosine. To investigate methylation by M.HpyAVIA–M.HpyAVIB fusion wild-type or mutant MTases (F9N or C82W or Y387L or F550S), 5 μM duplex 1 (Table 1) containing one GAGG site was incubated with 5 μM AdoMet and 50 nM purified protein (wild-type or mutant) in reaction buffer (10 mM Tris/HCl, pH 8.0, and 5 mM 2-mercaptoethanol) for 5 h at 37 °C followed by protein inactivation at 95 °C for 20 min. DNA was purified and spotted on to a PVDF membrane (Immobilon-N; Millipore) and fixed by UV cross-linking (1.2 mJ/cm2 for 30 s). The dot blot assay was performed as described previously [2].

RESULTS AND DISCUSSION

PCR amplification and cloning of hpyAVIAM and hpyAVIBM from H. pylori strain D27

hpyAVIAM and hpyAVIBM encode two MTases from H. pylori strain 26695 lacking a cognate restriction enzyme [19]. The homologues of hpyAVIAM and hpyAVIBM in H. pylori strain HPAGI are associated with chronic atrophic gastritis [27]. RT–PCR was performed by using primers specific to hpyAVIBM (primers 1 and 2, Supplementary Table S1 and Figure 1A) and hpyAVIAM (primer 3, Supplementary Table S1 and Figure 1A). An amplicon of 940 bp or 330 bp was obtained by using primers 1 and 3 or 2 and 3 respectively, suggesting that, in all H. pylori strains tested, a single mRNA codes for M.HpyAVIA and M.HpyAVIB MTases (Figure 1B). The DNA fragments corresponding to the homologues of hpyAVIAM and hpyAVIBM were amplified by PCR (Figure 2A) and sequenced from a number of clinical isolates such as SSI, 26695, D27 and PG227. H. pylori strain J99 was used as a negative control as it lacks the hpyAVIBM homologue. It was observed that, in strain D27, hpyAVIBM has an insertion mutation just before its stop codon (Figure 2B). The insertion removes the stop codon for hpyAVIBM and brings both the ORFs (hpyAVIAM and hpyAVIBM) in-frame without causing any changes in the amino acid sequences of either protein (Figures 3A–3B). This mutation can therefore generate a fusion protein containing an N-terminal M.HpyAVIB C5-cytosine MTase NTD and an M.HpyAVIA N6-adenine MTase CTD (Figure 3C). In order to investigate whether the fusion protein is functional or not, hpyAVIAMhpyAVIBM was PCR-amplified from the genomic DNA of H. pylori strain D27 and inserted into pET28a and pGEX4T2 expression vectors. The D27-M.HpyAVIA–M.HpyAVIB MTase was overexpressed and purified as described in the Materials and methods section (see Supplementary Figure S1 at http://www.BiochemJ.org/bj/433/bj4330487add.htm).

Figure 1 A single mRNA codes for M.HpyAVIA and M.HpyAVIB

(A) Positions of the primers used in RT–PCR. (B) RT–PCR analysis by using primers 1 and 3 (see Supplementary Table S1 at http://www.BiochemJ.org/bj/433/bj4330487add.htm) and cDNA from no RT control (26695, cagA+ vacA+s1m1) (lane 1) and strains 26695 (lane 2), D27 (lane 3), PG227 (lane 4) and SSI (lane 5). RT–PCR analysis by using primers 2 and 3 (see Supplementary Table S1) and cDNA from strains 26695 (lane 6), D27 (lane 7), PG227 (lane 8) and SS1 (lane 9).

Figure 2 Amplification of hpyAVIAMhpyAVIBM from H. pylori strains

(A) PCR amplification of hpyAVIAMhpyAVIBM from H. pylori strains. Lane 1, J99 (negative control); lane 2: markers; lane 3, 26695; lane 4, D27; lane 5, SSI; lane 6, PG227. (B) Sequence of hpyAVIAMhpyAVIBM from H. pylori strain D27 showing the insertion mutation.

Figure 3 hpyAVIAMhpyAVIBM from H. pylori strains 26695 and D27

(A) DNA sequence showing the arrangement of hpyAVIAMhpyAVIBM in H. pylori strain 26695. The start codon of hpyAVIAM is shown in bold and the stop codon of hpyAVIBM is shown in bold and is underlined. (B) Sequence of hpyAVIAMhpyAVIBM in H. pylori strain D27, showing nucleotide insertion before stop codon of hpyAVIBM. The insertion mutation is shown in bold. (C) Protein sequence of M.HpyAVIA–M.HpyAVIB fusion protein. The protein sequence of M.HpyAVIA is shown in bold. The AdoMet-binding motif and catalytic motif of M.HpyAVIA and M.HpyAVIB are boxed and underlined respectively.

Construction of M.HpyAVIA–M.HpyAVIB fusion protein from H. pylori strain 26695

hpyAVIAM and hpyAVIBM ORFs were amplified by PCR from the genomic DNA of H. pylori strain 26695 by using the primers 1 and 2 (see Supplementary Table S2) and ligated into pET28a and pGEX4T2 as described in the Materials and methods section. An insertion mutation was created by site-directed mutagenesis of hpyAVIAMhpyAVIBM from H. pylori strain 26695 to give a fusion protein and mutation was confirmed by HindIII digestion (see Supplementary Figure S2 at http://www.BiochemJ.org/bj/433/bj4330487add.htm). The mutant fusion MTase was overexpressed and purified as a GST-tagged and a His6-tagged protein as described in the Materials and methods section. The M.HpyAVIA–M.HpyAVIB fusion protein was also synthesized by in vitro transcription and translation. The pGEX4T2-hpyAVIAMhpyAVIBM plasmid was used as the template for in vitro protein synthesis (see Supplementary Figure S3 at http://www.BiochemJ.org/bj/433/bj4330487add.htm).

Methylation activity of M.HpyAVIA–M.HpyAVIB fusion proteins

It was shown previously that the M.HpyAVIA MTase from H. pylori strain 26695 recognizes and methylates GAGG, GGAG and GAAG, with GAGG being the preferred substrate [14]. M.HpyAVIB MTase, on the other hand, recognizes CCTC and methylates the first cytosine [13]. The methylation activity of the M.HpyAVIA–M.HpyAVIB fusion proteins from H. pylori strains 26695 and D27 was determined using four different 26-mer duplex substrates: CCTC/GAGG (duplex 1), mCCTC/GAGG (duplex 2), CCTC/GmAGG (duplex 3) or mCCTC/GmAGG (duplex 4) (Table 1). M.HpyAVIA–M.HpyAVIB fusion protein (50 nM) was added to the reaction mixture containing duplex DNA (5.0 μM) and [3H]AdoMet (5.0 μM). The reaction mixture was incubated at 37 °C and the transfer of methyl groups on to DNA was monitored. It was found that the M.HpyAVIA–M.HpyAVIB fusion MTase from H. pylori strains 26695 and D27 not only recognized CCTC and methylated the first cytosine, but also recognized GAGG and methylated the adenine residue (Table 2 and see Supplementary Figure S4 at http://www.BiochemJ.org/bj/433/bj4330487add.htm). When the methylation reactions were carried out with the in vitro translated fusion MTase, similar results were obtained (see Supplementary Figure S4). On the other hand, M.HpyAVIB MTase alone methylated only the cytosine and M.HpyAVIA MTase alone only methylated the adenine in the cognate sequence (Table 2 and see Supplementary Figure S4). These results can be compared with those from other MTases. M.Alw26I and M.Esp3I are single bifunctional proteins, composed of an N6-adenine MTase domain fused to a C5-cytosine MTase domain [12]. It has been shown that M.FokI has an NTD and a CTD with independent methylation activity. The NTD methylates adenine in strands carrying the sequence 5′-GGATG-3′ and the CTD methylates adenine in strands carrying the sequence 3′-CCTAC-5′ [7].

View this table:
Table 2 Methylation activity of wild-type and mutant M.HpyAVIA–M.HpyAVIB fusion MTase

mA, N6-methyladenine; mC, C5-methylcytosine; + indicates methylation.

Kinetics of the methylation reaction

To establish the relationship between the initial velocity of the reaction and enzyme concentration, the rate of DNA methylation was determined. Different concentrations of M.HpyAVIA–M.HpyAVIB fusion MTase (20–100 nM) were added to the reaction mixture containing pUC19 DNA (100 nM) and AdoMet (5.0 μM) and incubated at 37 °C. When initial velocities were plotted against increasing enzyme concentrations, a linear relationship was obtained (Figure 4A). This indicated that the initial velocity of the reaction was directly proportional to the enzyme concentration, suggesting that the M.HpyAVIA–M.HpyAVIB fusion MTase-catalysed reaction is of the first order with respect to the enzyme concentration. In the presence of 500 nM sinefungin, a known competitive inhibitor of DNA MTases, inhibition was observed (Figure 4A). Next initial velocities were determined at various concentrations of the substrates, [3H]AdoMet and duplexes 1–3 (Table 1). For the determination of Km (DNA), a series of similar reactions containing M.HpyAVIA–M.HpyAVIB fusion MTase (50 nM), [3H]AdoMet (5.0 μM) and increasing concentrations of duplex DNA (duplexes 1–3, Table 1) (0.5–20 μM) were performed and a hyperbolic curve was created. A non-linear regression analysis allowed the determination of Km (DNA) values of 10±0.2, 12±0.2 and 16±0.5 μM for duplexes 1, 2 and 3 respectively (Figure 4B). Comparison between the kinetic constants of M.HpyAVIA, M.HpyAVIB and M.HpyAVIA–M.HpyAVIB fusion MTase indicates that the fusion MTase has 4- and 14-fold lower specificity constants compared with M.HpyAVIA and M.HpyAVIB MTases (Table 3).

Figure 4 Kinetics of methylation

(A) Initial velocity against M.HpyAVIA–M.HpyAVIB fusion MTase concentration. (B) Determination of Km (DNA). (C) Initial velocity against AdoMet concentration.

View this table:
Table 3 Comparision between kinetic parameters for M.HpyAVIA, M.HpyAVIB and M.HpyAVIA–M.HpyAVIB fusion MTase

ND, not determined.

It is a well established fact that DNA MTases play important roles in regulating gene expression, possibly by promoter methylation affecting protein–DNA interactions [17]. A lower specificity constant of the fusion MTase could have a significant effect on the methylation pattern of the H. pylori genome, which in turn would affect the expression profile of certain genes. It has been shown that deletion of the gene coding for the M.HpyAVIA MTase in H. pylori strain PG227 has a significant effect on the growth of the strain [14]. In addition, deletion of the gene for the M.HpyAVIB MTase in strains 26695 and SSI resulted in alterations of gene expression profiles (R. Kumar, A. K. Mukhopadhyay and D. N. Rao, unpublished work).

Next, for the determination of Km (AdoMet), a series of reactions containing M.HpyAVIA–M.HpyAVIB fusion MTase (50 nM), duplex 1 (5 μM) and increasing concentrations of [3H]AdoMet (0.3–15 μM) were performed. Increasing the concentration of AdoMet led to a progressive stimulation of the reaction rate. Whereas the initial portion of the concentrationdependence curve corresponded approximately to a conventional hyperbolic dependence, saturation was never achieved (Figure 4C). Similar observations have been reported for T4 Dam and EcoDam [28,29]. It was proposed that the lack of saturation could be because of an efficient mechanism which includes the release of the end-product AdoHcy and co-ordination with the binding of AdoMet in a single concerted event [28]. It has been shown that the M.HpyAVIA MTase did not show saturation with increasing concentrations of [3H]AdoMet [14]. It is therefore possible that a progressive stimulation of the reaction rate with increasing AdoMet is because of M.HpyAVIA MTase activity in the M.HpyAVIA–M.HpyAVIB fusion MTase.

Methylation activity of mutant M.HpyAVIA–M.HpyAVIB fusion proteins

All DNA MTases have conserved characteristic motifs, namely the AdoMet-binding motif and a catalytic motif. The AdoMet-binding motif is represented by an FXGXG sequence [30]. In N6-adenine MTases, the catalytic motif is DPPY and PCQ is the catalytic motif in C5-cytosine MTases [31]. In order to investigate the role of motifs in each domain in the methylation reaction, four mutants were made in the hpyAVIAMhpyAVIBM fusion from H. pylori strain 26695. M.HpyAVIA–M.HpyAVIB fusion MTase has two domains containing independent AdoMet-binding and catalytic motifs. To understand the role of each motif, AdoMet-binding and catalytic motif mutants were made in both the NTD and the CTD. The F9N and C82W mutations were constructed in the C5-cytosine MTase NTD and the F550S and Y387L mutations were constructed in the N6-adenine MTase CTD as described previously [14].

It was observed that mutations in the AdoMet-binding motif (FXGXG) and the catalytic motif (PCQ) in the C5-cytosine MTase NTD completely abrogated the cytosine methylation activity, but did not affect adenine methylation activity. The F9N and C82W mutants were able to methylate duplexes 1 and 2, but were unable to methylate duplex 3 (Table 2, and see Supplementary Figure S4). The mutations in the AdoMet-binding motif and catalytic motif (DPPY) in the N6-adenine MTase CTD completely abolished the adenine methylation activity of the fusion protein without affecting its cytosine methylation activity. The Y387L and F550S mutants were able to methylate duplexes 1 and 3, but were unable to methylate duplex 2 (Table 2 and see Supplementary Figure S4). The observation was confirmed further by a dot blot assay using antibodies against N6-methyladenine (Figure 5) or C5-methylcytosine (Figure 6). Duplex 1 was used as a substrate and 50 nM fusion protein (wild-type or mutant) was incubated with duplex 1 (5 μM) and AdoMet (5 μM) for 3 h at 37 °C. It was observed that the Y387L and F550S mutant proteins were not able to methylate adenine (lanes A and C, Figure 5), but retained the cytosine methylation activity (lanes D and G, Figure 6), whereas the F9N and C82W mutant proteins retained adenine methylation activity (lanes E and F, Figure 5), but were not able to methylate cytosine (lanes C and E, Figure 6). M.HpyAVIA alone was able to methylate adenine, but not cytosine (lane H, Figure 5; lane F, Figure 6) and M.HpyAVIB alone was able to methylate cytosine, but not adenine (lanes B and H, Figure 6; lane B, Figure 5). In contrast, M.HpyAVIA–M.HpyAVIB fusion MTase was able to methylate both adenine and cytosine in the cognate sequence as detected using antibodies (lanes D and G, Figure 5; lane A, Figure 6). The data clearly demonstrate the fact that both ‘domains’ of the fusion MTase are independent of each other and require their own catalytic motif and importantly their own AdoMet-binding motif.

Figure 5 Effect of point mutations on adenine methylation activity of M.HpyAVIA–M.HpyAVIB

I: schematic presentation of M.HpyAVIA–M.HpyAVIB with domains and active/mutated sites in different motifs. II: dot blot assay to measure the methylation activity using rabbit primary antibodies against N6-methyladenine. Duplex 1 was used as a substrate in the methylation assay with M.HpyAVIA–M.HpyAVIB fusion wild-type or mutant MTase from H. pylori strain 26695. Lane A, M.HpyAVIA–M.HpyAVIB (Y387L); lane B, M.HpyAVIB; lane C, M.HpyAVIA–M.HpyAVIB (F550S); lane D, M.HpyAVIA–M.HpyAVIB; lane E, M.HpyAVIA–M.HpyAVIB (F9N); lane F, M.HpyAVIA–M.HpyAVIB (C82W); lane G, M.HpyAVIA–M.HpyAVIB; lane H, M.HpyAVIA.

Figure 6 Effect of point mutations on cytosine methylation activity of M.HpyAVIA–M.HpyAVIB

I: schematic presentation of M.HpyAVIA–M.HpyAVIB with domains and active/mutated sites in different motifs. II: dot blot assay to measure the methylation activity using monoclonal antibodies against C5-methylcytosine. Duplex 1 was used as a substrate in the methylation assay with M.HpyAVIA–M.HpyAVIB fusion wild-type or mutant MTase from H. pylori strain 26695. Lane A, M.HpyAVIA–M.HpyAVIB; lane B, M.HpyAVIB; lane C, M.HpyAVIA–M.HpyAVIB (F9N); lane D, M.HpyAVIA–M.HpyAVIB (F550S); lane E, M.HpyAVIA–M.HpyAVIB (C82W); lane F, M.HpyAVIA; lane G, M.HpyAVIA–M.HpyAVIB (Y387L); lane H, M.HpyAVIB.

The hpyAVIAMhpyAVIBM ORFs represent an interesting case study which highlights the dynamic nature of R-M systems in H. pylori. It has been postulated previously that the plastic nature of the H. pylori genome could result in the evolution of new R-M systems with new properties. H. pylori can acquire new R-M systems by horizontal gene transfer, but can also lose an R-M system by random mutation and genetic recombination [32,33]. Strain-specific variations in R-M systems could be responsible for strain-to-strain variations in H. pylori [34]. It could be possible that H. pylori had acquired the hpyAVIAMhpyAVIBM ORFs system via horizontal gene transfer from some other bacterial species. Comparative study of the loci hpyAVIAM and hpyAVIBM in strains 26695, J99 and D27 has shown that, whereas both the genes are present in strain 26695 adjacent to each other, hpyAVIBM is absent from strain J99 and both are fused in strain D27 (Figure 7). Interestingly, hp0053hp0054 (hpyAVRhpyAVM) represents another R-M system where hp0053 encodes a restriction enzyme and hp0054 encodes a fused MTase with a C5-cytosine MTase and N6-adenine MTase domain [15]. It has been suggested that these pathogenic bacteria acquired a prototypical R-M system through lateral transfer [1], which provided a survival advantage to the receiving bacteria. Strain-specific methylation patterns could be involved in determining the strain-specific characteristics such as virulence, host adaptability and survival in a changing environment. Our data strongly suggest the notion that the hpyAVIAMhpyAVIBM homologue in strain D27 is the result of fusion of two genes evolved from independent ancestors. The methylation activity of the M.HpyAVIA–M.HpyAVIB fusion protein clearly demonstrates that the DNA interaction of both domains is mutually exclusive and both domains do not functionally influence each other. Bifunctional MTases such as M.HpyAVIA–M.HpyAVIB represent an exceptionally attractive group of MTases to study structure–function relationships in DNA recognition and proteins that display an unusual structural and functional organization. This arrangement is more likely to be an ancestral trait of the common progenitor rather than the consequence of independent events. The present study highlights an interesting case of two MTases that clearly demonstrates the dynamics of R-M system in H. pylori.

Figure 7 Dynamics of the R-M system in H. pylori: hpyAVIAMhpyAVIBM

AUTHOR CONTRIBUTION

Ritesh Kumar designed and performed the experiments. Ritesh Kumar and Desirazu Rao wrote the paper.

FUNDING

R.K. thanks the Council of Scientific and Industrial Research for a Senior Research Fellowship. D.N.R. acknowledges the Department of Science and Technology for a J.C. Bose Fellowship.

Acknowledgments

Genomic DNA of H. pylori strains J99 and 26695, and antibodies against N6-methyladenine were kindly provided by New England Biolabs. We thank Dr Deepti Deobagkar for the antibodies against C5-methylcytosine. We thank Dr Asish K. Mukhopadhyay for providing genomic DNA of H. pylori strains D27, SSI and PG227. We thank Dr Anand Swaroop for the methylated oligonucleotides. Members of the D.N.R. laboratory are acknowledged for critical reading of the paper and useful discussions.

Abbreviations: AdoMet, S-adenosylmethionine; BHI, brain/heart infusion; CTD, C-terminal domain; GST, glutathione transferase; LB, Luria–Bertani; MTase, methyltransferase; NTD, N-terminal domain; ORF, open reading frame; R-M, restriction–modification; RT, reverse transcription

References

View Abstract