Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow Print PDF (110K) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (20)
Right arrowRequest Permissions
Right arrow Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Right arrow Articles by Kong, H.
Right arrow Articles by Smith, C. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kong, H.
Right arrow Articles by Smith, C. L.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© 1995 Oxford University Press 3687-3692

Substrate DNA and cofactor regulate the activities of a multi-functional restriction-modification enzyme, BcgI

Substrate DNA and cofactor regulate the activities of a multi-functional restriction-modification enzyme, Bcg I Huimin Kong* and Cassandra L. Smith1

New England Biolabs, 32 Tozer Road, Beverly, MA 01915, USA and 1Center for Advanced Biotechnology and Pharmacology, Boston University, Boston, MA 02215, USA

Received May 8, 1997; Revised and Accepted August 1, 1997

ABSTRACT

The BcgI restriction-modification system consists of two subunits, A and B. It is a bifunctional protein complex which can cleave or methylate DNA. The regulation of these competing activities is determined by the DNA substrates and cofactors. BcgI is an active endonuclease and a poor methyltransferase on unmodified DNA substrates. In contrast, BcgI is an active methyltransferase and an inactive endonuclease on hemimethylated DNA substrates. The cleavage and methylation reactions share cofactors. While BcgI requires Mg2+ and S-adenosyl methionine (AdoMet) for DNA cleavage, its methylation reaction requires only AdoMet and yet is significantly stimulated by Mg2+. Site-directed mutagenesis was carried out to investigate the relationship between AdoMet binding and BcgI DNA cleavage/methylation activities. Most substitutions of conserved residues forming the AdoMet binding pocket in the A subunit abolished both methylation and cleavage activities, indicating that AdoMet binding is an early common step required for both cleavage and methylation. However, one mutation (Y439A) abolished only the methylation activity, not the DNA cleavage activity. This mutant protein was purified and its methylation, cleavage and AdoMet binding activities were tested in vitro. BcgI-Y439A had no detectable methylation activity, but it retained 40% of the AdoMet binding and DNA cleavage activities.

INTRODUCTION

Restriction-modification (R-M) systems comprise pairs of competing intracellular enzyme activities: an endonuclease activity which cleaves at its target site and a methyltransferase activity which prevents endogenous DNA from endonuclease digestion. The methyltransferase prevents endonuclease digestion by modifying the same target sequence, that is usually hemimethylated following DNA replication, with a methyl group from S-adenosyl methionine (AdoMet). More than 2700 restriction endonucleases have been isolated from bacterial sources and most of them can be grouped into three classes based on subunit composition, cofactor requirements and type of DNA cleavage (1 ,2 ).

Type II systems are the simplest systems. They have two enzymes: an endonuclease and a corresponding methyltransferase. The endonuclease coordinates Mg2+ with acidic residues to polarize the reactive phosphate group and/or to stabilize the transition state (3 ). In the case of EcoRV, Mg2+ is also required for specific DNA binding (4 ). The methyltransferase transfers the methyl group from AdoMet to a target adenine or cytosine within the DNA recognition site. All DNA methyltransferases use AdoMet as their methyl donor. The AdoMet is bound in a hydrophobic pocket formed by two conserved motifs in the DNA methyltransferase, as is seen in the structure of M.TaqI (5 ).

Type I restriction endonucleases are multifunctional protein complexes of three subunits that have both DNA cleavage and DNA methylation activities. Biochemical studies revealed that the early steps of AdoMet binding and enzyme activation are common to both the restriction and modification activities of type I enzymes, such as EcoKI (6 ,7 ). Once the EcoKI is bound with AdoMet, it becomes activated and it will either cleave or methylate DNA. As a maintenance methyltransferase, EcoKI has very low methylation activity on unmodified DNA compared with hemimethylated DNA (8 ). However, results from comparative gel retardation experiments show that the binding affinities for three substrate DNAs (unmodified, hemimethylated and fully modified) are very similar (9 ). This suggests that for the EcoKI methyltransferase the major discrimination between DNA substrates of different methylation states occurs at the level of catalysis rather than substrate binding (9 ).

Type III R-M systems are also multimeric systems with M and R subunits (10 ). The M subunit alone is an active methyltransferase, and determines the specificity for both methylation and restriction. The R subunit alone is an inactive protein. When it forms an heterodimer with M, it cleaves on one side of the asymmetric recognition ~25 bp away in the presence of ATP, AdoMet and Mg2+.

BcgI (11 ) and the other BcgI-like restriction endonucleases, such as BaeI (12 ), Bsp24I (13 ), CjeI and CjePI (14 ) are distinct from all other types of restriction endonucleases in their unique cleavage pattern and cofactor requirement. They all cleave DNA on both sides of their recognition sequences to excise a short DNA fragment including the recognition site in the middle (~30 bp). They all require AdoMet and Mg2+ for DNA cleavage. The genes encoding for BcgI proteins have been cloned and sequenced (15 ). It was found that the BcgI R-M system consists of two adjacent, similarly oriented genes. The BcgIA gene encodes for a 637 aa protein that contains AdoMet binding motifs like those seen in m6A-methyltransferases such as M.TaqI and the M subunit of EcoKI. The BcgIB gene encodes for a 341 aa protein (molecular mass = 39.2 kDa). The acidic A subunit (I.P = 5.07) and basic B subunit (I.P. = 9.66) form a very tight complex that cleaves or methylates DNA. Although the A subunit contains the two conserved methylase motifs, it lacks any methylation activity by itself. Both the A and B subunits are required to bind, cleave and methylate target DNA (15 ). However, it is not known how BcgI coordinates its competing activities.

In this report, the methylation and cleavage activities of the BcgI protein complex were studied using a set of synthetic oligonucleotides which form unmethylated, hemimethylated and fully methylated DNAs. In addition, the effects of enzymatic cofactors on the cleavage/methylation activities were also studied. To further investigate the relationship between AdoMet binding and the catalytic activities of BcgI, we have substituted conserved residues in the A subunit that forms the AdoMet binding pocket and characterized the phenotypes of the mutant enzymes.

MATERIALS AND METHODS

Enzymes and other reagents

Restriction enzymes and DNA polymerase I Klenow fragment were obtained from New England Biolabs. All single-stranded oligodeoxynucleotides were made by New England Biolabs Organic Synthesis division. Both [32P]dATP (3000 Ci/mmol) and [3H]AdoMet (10 Ci/mmol) were purchased from DuPont New England Nuclear.

Substrate DNAs with various methylation

Double-stranded duplex DNA was generated by heating/annealing complementary single-stranded oligodeoxynucleotides (93oC for 3 min, then slowly cooling to 25oC over 1 h).

The DNA duplex A contains a single BcgI recognition sequence (shown in bold):

5' TTTGAGAATAGTGTATGCGGCGAGGATCCTGCTCTTGCCCGGCGTCAATA 3'3' ACTGTTATCACATACGCCGCTCCTAGGACGAGAACGGGCCGCAGTTATGG 5'

Hemimethylated DNA duplexes were generated by annealing one unmethylated strand with its complementary strand containing an N6-methyl adenine (shown as underlined) within the BcgI recognition sequence.

The duplex B also contains a single BcgI recognition sequence, but its surrounding nucleotides differ from duplex A: 5' AAACGTCATCACCGAATTCCGCGATCCAGCTGCAGTAAAGCTCATCAGC 3'3' CAGTAGTGGCTTAAGGCGCTAGGTCGACGTCATTTCGAGTAGTCGCG 5'

Site-directed mutagenesis

Nine mutant BcgI proteins that differ from the wild-type by only a single amino acid were made using the corresponding oligonucleotides. Site-directed mutagenesis was carried out as described previously (16 ). Positive clones were screened by restriction enzyme analysis and verified by DNA sequencing.

Purification of BcgI-Y439A

This mutant was purified from Escherichia coli ER2504-DE3- plysS containing pETBcgI-Y439A, a recombinant plasmid containing the BcgI R-M genes downstream of the T7 promoter (17 ). All operations were performed at 4oC unless otherwise noted. Frozen cells (20 g), from cultures grown at 37oC, were thawed and suspended in buffer A (20 mM Tris-HCl pH 7.5, 0.1 mM Na2EDTA, 1 mM dithiothreitol) containing 100 mM NaCl. The suspension was sonicated. Following cell rupture, the supernatant was applied to a buffer A-equilibrated heparin-Sepharose column. DNA cleavage activity was eluted with a linear gradient of 0.1-1 M NaCl in buffer A. Fractions with BcgI activity were pooled, dialyzed and applied to a mono-Q column (Pharmacia). Active fractions from mono-Q were applied to a heparin Tsk column (Toso Haas) equilibrated with buffer A. Fractions with BcgI endonuclease activity eluted from Heparin Tsk were dialyzed against storage buffer (10 mM Tris-HCl pH 7.4, 100 mM NaCl, 1 mM dithiothreitol, 0.1 mM Na2EDTA and 50% glycerol), and kept at -20oC.

In vitro DNA cleavage and methylation activities

BcgI DNA cleavage activity was determined by incubating BcgI protein with [lambda] DNA in BcgI digestion buffer (New England Biolabs) at 37oC for 15 min. The digestion products were analyzed by agarose gel electrophoresis. BcgI methylation activity was determined by incubating BcgI with DNA duplexes and [3H]AdoMet in BcgI DNA methylation buffer (10 mM Tris-HCl pH 8.0, 5 mM Na2EDTA and 1 mM dithiothreitol). Aliquots (40 [mu]l) were withdrawn following incubation at 37oC and spotted on to a 3 MM filter paper (Whatman). The filters were washed in 10% trichloroacetic acid to remove free [3H]AdoMet, while the incorporated 3H-methyl group was quantitated by liquid scintillation counting as acid-insoluble radioactivity.

AdoMet binding assay

BcgI-AdoMet binding activity was measured using a filter binding assay which had been modified (J.S. Benner, unpublished). BcgI was incubated with [3H]AdoMet at 37oC for 10 min. The BcgI-[3H]AdoMet was trapped on a HAWP 02500 filter (Millipore). Unbound [3H]AdoMet was washed away with wash buffer (20 mM Tris-HCl pH 8, 50 mM NaCl, 5 mM Na2EDTA and 1 mM dithiothreitol). Subsequently, the filter was dried and the amount of trapped [3H]AdoMet was quantified by liquid scintillation counting.

BcgI is a maintenance methyltransferase

In the presence of Mg2+ and AdoMet, BcgI preferentially cleaves DNA on both sides of the target sequence (CGAN6TGC) to produce a 34 bp fragment with an intact recognition sequence. Following cleavage, BcgI remains bound to the 34 bp fragment and methylates it at an extremely slow rate: the turnover number is 0.03 methyl-transfer by BcgI per hour (11, Kong, unpublished observations).

The amino acid sequence of the BcgI A subunit is most similar to the amino acid sequences of the type I M subunits, some of which, such as EcoKI, are maintenance methyltransferases preferring hemimethylated DNA (8 ,15 ). Therefore, a set of synthetic oligonucleotides including unmethylated, hemimethylated and fully methylated DNAs were used to test whether BcgI can function as a maintenance methyltransferase. BcgI had 5- to 15-fold higher methylation activities on hemimethylated DNA as compared to unmethylated DNA (Table 1 , compare line 1 with lines 2 and 3).

Table 1 . Comparison of BcgI methylation activity on substrate DNAs with different methylation states
Line number Substrate Mtase activitya
1 -/- -------CGAGGATCCTGC------- 474
  -------GCTCCTAGGACG-------  
               m  
2 +/- -------CGAGGATCCTGC------- 2606
  -------GCTCCTAGGACG-------  
3 -/+ -------CGAGGATCCTGC------- 7096
  -------GCTCCTAGGACG-------  
                                     m  
                  m  
4 +/+ -------CGAGGATCCTGC------- u.d.
  -------GCTCCTAGGACG-------
                                     m  
  [up arrow] Duplex A  
-----------------------------------------------------------
  <=> Duplex B  
5 -/- -------CGATCCAGCTGC------- 406
  -------GCTAGGTCGACG-------  
                  m  
6 +/- -------CGATCCAGCTGC------- 1209
  -------GCTAGGTCGACG-------  
7 -/+ -------CGATCCAGCTGC------- 3448
  -------GCTAGGTCGACG-------  
                                     m  
                  m  
8 +/+ -------CGATCCAGCTGC------- u.d.
  -------GCTAGGTCGACG-------  
                                     m  
aThe methylation activities were determined by incubating annealed duplexes (0.15 [mu]M) with BcgI (0.16 U/[mu]l) in the presence of [3H]AdoMet (1 [mu]M), and BcgI methylation buffer (Materials and Methods). The acid-insoluble radioactivity was quantified by liquid scintillation counting as detailed in Materials and Methods. The methyltransferase activities on different substrate DNAs are presented by the number of scintillation counts per minute (c.p.m.); u.d. indicates undetectable activity compared with background.

There was also a significant difference in BcgI methylation activities with the two different hemimethylated substrates (Table 1 , lines 2 and 3). The difference could be due to the nature of the asymmetric recognition sequence (CGAN6TGC versus GCAN6TCG), such that BcgI might methylate one unmethylated sequence (CGAN6TGC) more efficiently than the other (GCAN6TCG). Alternatively, the difference could simply be due to the influence of surrounding nucleotide sequences. To distinguish between these two possibilities, the same methylation assay was carried out using DNA duplexes with different surrounding nucleotides sequences (duplex B, Materials and Methods). Again, BcgI had higher methylation activities on hemimethylated duplex B with unmethylated CGAN6TGC sequence (Table 1 , compare lines 7 and 6), indicating that BcgI is consistently a more active methylase on one asymmetric recognition sequence (CGAN6TGC) than its complementary sequence (GCAN6TCG) in the presence of AdoMet. Thus, the surrounding nucleotide sequence does not appear to have much influence on this asymmetric methylation activity.

Mg2+ stimulates BcgI methylation activity

To study the influence of Mg2+ on BcgI methylation activity, a comparison methylation assay with and without Mg2+ was carried out. Mg2+ stimulates BcgI methylation activity on one hemimethylated DNA (Fig. 1 , +/-) by 4.3-fold and on the other hemimethylated DNA (Fig. 1 , -/+) by 17.2-fold. Moreover, the discrimination of BcgI methylation activities on the two different hemimethylated target sequences disappeared sharply in the presence of Mg2+. Stimulation was also observed with unmodified DNA (Fig. 1 , -/-).


Figure 1. The influence of Mg2+ on BcgI DNA methylation activity. The in vitro methylation assays were carried out as described in Table 1, except in the case of plus Mg2+ (filled bar), where the 5 mM Na2EDTA was replaced by 10 mM MgCl2. The blank bars represent the methylation activity in the absence of Mg2+. Symbols, +/-, -/+, -/- and +/+, represent different methylation states of substrate DNA, as detailed in Table 1.

The dual role of AdoMet in DNA methylation

Previous work, in a different R-M system, revealed cooperativity of AdoMet in the EcoKI DNA methylation reaction (6 ). Similar observations have been reported in DNA-adenine-methyltransferase of E.coli, which underwent DNA-binding affinity and conformational changes upon binding of AdoMet (18 ).

This allosteric activation role of AdoMet was also observed in a kinetic study in BcgI methylation. The apparent Km value of AdoMet for BcgI methylation was measured at ~300 nM (Fig. 2 ). The dependence of BcgI methylation on AdoMet concentration is a typical sigmoidal curve, indicating the presence of cooperativity of AdoMet in the DNA methylation reaction. In contrast, the Km value of AdoMet in the BcgI cleavage reaction was measured at 100 nM, and no cooperativity was observed in the dependence of DNA cleavage on AdoMet concentration (11 ). This difference is probably due to the different role that AdoMet plays in the methylation reaction compared to the cleavage reaction. In addition to serving as an allosteric effector in both methylation and cleavage reactions, AdoMet is also the substrate for DNA methylation, and this dual role of AdoMet in methylation results in the sigmoidal curve in Figure 2 .


Figure 2. Kinetic study in BcgI methylation. BcgI methylation activity was assayed in BcgI methylation buffer (as described in Fig. 1) containing variable amounts of AdoMet (from 800 to 50 nM). DNA duplex concentration was fixed at 150 nM and the BcgI enzyme concentration was also fixed at 5 nM. Samples were withdrawn as a function of time; the incorporated 3H-methyl groups were determined by liquid scintillation counting. Plots of the reaction time course were analyzed to determine the initial rates, and the data was then replotted as shown.

Site-directed mutagenesis in the AdoMet-binding pocket

Site-directed mutagenesis on conserved residues that form the AdoMet binding pocket was carried out to further investigate the dual role of AdoMet and to study the relationship between AdoMet binding and BcgI enzymatic activities. AdoMet is inserted into a hydrophobic pocket formed by two conserved amino acid segments in adenine methyltransferases, such as M.TaqI (5 ). Nine mutations in two sequence motifs of BcgI (motif I, DPACGTG and motif IV, NPPY) were made using nine corresponding oligonucleotides. The DNA cleavage/methylation activities of all mutant BcgI proteins are summarized in Table 2 .

Table 2 . The phenotypes of BcgI mutants in the two conserved methyltransferase motifs
Mutants DNA cleavage (%)a DNA methylationb
WT 100 ++
G355A 90 ++
G355D 0 -
N436A 0 -
N436D 0 -
N436Q 0 -
N436S 20 +
Y439A 20 -*
Y439F 90 ++
Y439W 20 +
aBcgI DNA cleavage activity was determined as follows: cells containing plasmid pET21at-BcgIAB were induced with IPTG and crude cell extracts were prepared by sonication and centrifugation. The DNA cleavage phenotypes of these mutants were determined by measuring the cleavage activity of the mutant proteins in supernatant on [lambda] DNA in BcgI digestion buffer.
bThe methylation phenotypes were determined by measuring the protection of the BcgI recognition sites in the mutant BcgI plasmid constructs. If the mutant BcgI still has methylation activity in vivo, it will modify the BcgI recognition sites in the plasmid so that the plasmid DNA becomes resistant to BcgI digestion. The plasmid DNA, which contains several BcgI sites, was then purified following IPTG induction. The DNA was then cleaved by BcgI endonuclease to determine whether the BcgI sites were methylated or not. ++, plasmid DNA completely resisted BcgI digestion; +, plasmid DNA was partially degraded, indicating incomplete methylation; -, plasmid DNA can be cleaved by BcgI; -*, plasmid DNA was degraded before BcgI digestion, probably due to the R+M- phenotype. After induction, the R+M- BcgI cleaved the plasmid and the BcgI digested DNA fragments were further degraded in vivo by non-specific nucleases.

All DNA methyltransferases contain a conserved Gly in motif I except M.TaqI which has an Ala in this position based on sequence alignment (19 ). Changing this most conserved Gly to an Asp completely abolishes BcgI methylation and cleavage activities. However, the Gly to Ala substitution (G355A) had little effect on either BcgI DNA methylation or DNA cleavage activities.


Figure 3. In vitro methylation activity of BcgI-Y439A. Wild-type BcgI or BcgI-Y439A were incubated with hemimethylated DNA duplex (0.15 [mu]M) and 2.5 [mu]M [3H]AdoMet in BcgI methylation buffer at 37oC. Aliquots (40 [mu]l) were withdrawn over a 2 h time course and spotted on to 3 MM filter paper (Whatman). The filter was washed in 10% trichloroacetic acid. Acid-insoluble radioactivity was quantitated by liquid scintillation counting.

All m6A and m4C methyltransferases contain a conserved (N/D/S)PP(Y/F) segment of motif IV. Substitution of the Asn in motif IV of BcgI (NPPY) to Ala, Asp or Gln abolished both methylation and cleavage activities of BcgI, even though the Asp is found in some m6A methyltransferases such as M.HhaII and M.EcoRV. The Asn to Ser change resulted in partial enzyme activity.

The last residue in motif IV (Tyr) was changed to Phe, Trp and Ala. Change of Tyr to Phe (Y439F) left a similar aromatic ring, but removed a hydroxyl group. This substitution had no effect on BcgI methylation and cleavage activities, indicating that the hydroxyl group is not directly involved in the catalysis. Changing the Tyr to Trp (Y439W), an aromatic amino acid residue with a bulkier side chain, resulted in partial enzyme activity. The Tyr to Ala change (Y439A) abolished methylation activity completely, but still retained some of the cleavage activity (Table 2 ).

Characterization of the BcgI-Y439A protein

The BcgI-Y439A protein also appears to be an active endonuclease and an inactive methyltransferase in vivo. This statement is based on the observation that the plasmid carrying the BcgI-Y439A genes was degraded upon induction of the genes. This is probably due to the fact that the BcgI-Y439A protein could not methylate the BcgI recognition sites in the plasmid and thus the plasmid was cleaved by the active endonuclease (data not shown). The cleavage- plus/methylation-minus (R+M-) phenotype of BcgI-Y439A was further confirmed in vitro after column chromatographic purification of this mutant protein (Materials and Methods). A hemimethylated double-stranded DNA was used for an in vitro methylation assay with purified BcgI-Y439A. No detectable methyl-transfer was observed with the mutant BcgI-Y439A, while 30% of the sites in the DNA substrate were modified with a 3H-methyl group from AdoMet by wild-type BcgI (Fig. 3 ). The DNA cleavage activity of purified mutant BcgI-Y439A was also assessed. It cleaved DNA specifically like wild-type BcgI with a lower specific activity (40% of wild-type).

A filter-binding assay was performed to test the AdoMet binding ability of this methylation-deficient mutant. The quantitative binding of AdoMet was examined by measuring the amount of BcgI bound to [3H]AdoMet using a filter. BcgI-Y439A was still able to bind AdoMet (Fig. 4 ) and its target DNA (unpublished observation). This suggests that BcgI-Y439A appears to be a catalytic mutant, which can bind both substrates but fails to catalyze the reaction. Compared to the wild-type, the Y439A mutant showed 42% binding affinity for AdoMet (Fig. 4 ), Since BcgI requires AdoMet for DNA cleavage, the reduced AdoMet binding affinity might be the cause of the reduced specific activity in the DNA cleavage reaction of the Y439A mutant.


Figure 4. BcgI-AdoMet filter-binding assay. [3H]AdoMet (1.5 [mu]M) was incubated with 0.2 [mu]M BcgI protein in 25 [mu]l BcgI methylation buffer at 37oC for 10 min. The reaction mixture was then diluted to 250 [mu]l with wash buffer (20 mM Tris-HCl pH 8.0, 50 mM NaCl, 5 mM Na2EDTA, 1 mM dithiothreitol) and applied to a Millipore HAWP 02500 filter in a filtration unit. The filter was then washed with 15 ml wash buffer. The radioactivity of captured [3H]AdoMet was determined by liquid scintillation counting.

DISCUSSION

The results from kinetic studies of BcgI methylation suggest that AdoMet acts both as a methyl donor and an allosteric effector. This observation agrees with the data from mutagenesis studies on the AdoMet binding motifs of BcgI. The DNA methylation and cleavage activities of the BcgI enzyme are closely related to each other. This was reflected by the fact that most mutants either lost the ability to methylate and cleave DNA or retained both activities (Table 2 ). Figure 5 is a schematic diagram showing the possible reaction pathways of wild-type BcgI along with mutant proteins. Aggressive substitutions (G355D, N436A, D, Q) in AdoMet binding motifs, that may possibly interfere with AdoMet binding, abolished both downstream methylation and cleavage activities (Fig. 5 C). Some conservative changes, such as G355A retain both cleavage and methylation activities. These results suggest that methylation and cleavage pathways diverge after the first common step of AdoMet binding (Fig. 5 A).


Figure 5. Schematic diagram of BcgI reaction pathways. +/-, hemimethylated DNA; -/-DNA, unmodified DNA. Thick line with arrow indicates reactive pathway. Thin line represents inactive pathway. The pathway blockage is indicated by cross lines.

This hypothesis was strengthened by a mutagenesis study on Tyr439 in motif IV (NPPY). In this case, the change only affected the methylation step, but not the upstream AdoMet binding step or the parallel cleavage step, thereby allowed the mutant BcgI to still cleave DNA (Fig. 5 D).

Mg2+ is the only other cofactor for BcgI in addition to AdoMet. It is very interesting that the two competing activities share the same cofactors: AdoMet, the methyl donor for DNA methyltransferase, is required for DNA cleavage by BcgI; and Mg2+, the divalent cation for endonucleolytic digestion, can greatly stimulate the methylation activity of BcgI. We think while AdoMet serves as a major allosteric activator for BcgI, Mg2+ might serve as a minor activator as well. When binding with both AdoMet and Mg2+, BcgI is probably in its most active form which can cleave or methylate DNA (Fig. 5 A). When binding with only AdoMet, BcgI is probably in a partially active form which cannot cleave DNA but can methylate DNA at a reduced activity level (Fig. 5 B). Because of being in this partially active conformation, the two asymmetric recognition sequences were treated differentially (Table 1 ). The binding of Mg2+ probably induces a conformation change, so that BcgI becomes a more active methyltransferase which no longer discriminates between the two asymmetric sequences (Fig. 1 ).Finally, when binding with Mg2+ alone, BcgI is probably in a inactive form which lacks both activities (equivalent to the BcgI mutants that cannot bind AdoMet, Fig. 5 C). Since the BcgI cleavage and methylation activities are closely associated with each other, we are not surprised to see that cleavage cofactor (Mg2+) influences methylation activity. Indeed a recent study mapping the functional domain of BcgI revealed that both the cleavage domain and the methylation domain reside in the same polypeptide (Kong and Smith, manuscript in preparation).

In addition to the cofactors, substrate DNA behaves as an allosteric effector to coordinate the dual activities of BcgI as well. In the presence of unmethylated DNA substrates, BcgI is an active endonuclease and a poor methyltransferase. As a result, any unmethylated exogenous DNA, such as invading phage DNA, will be preferentially cleaved. When hemimethylated DNA substrates are present, BcgI is an active methyltransferase and an inactive endonuclease so that its host endogenous DNA can be methylated efficiently following replication.

The crystal structure of M.TaqI with bound sinefungin, an analog of AdoMet, shows that two conserved motifs form the AdoMet binding pocket, with the exception of the Tyr in motif IV (NPPY) which sticks out of the pocket by a sharp turn due to the preceding prolines (5 ). The role of this aromatic residue in M.TaqI was further defined by structural comparison of the adenine methyltransferase M.TaqI with the cytosine methylase M.HhaI (20 ,21 ). Although M.HhaI and M.TaqI belong to different families of DNA methyltransferases and are believed to have different catalytic mechanisms, the structural comparison of these AdoMet-dependent methyltransferases reveals that the catalytic domains of M.HhaI and M.TaqI exhibit a very similar 3-D structure (22 ). Indeed, the Y in the NPPY motif of M.TaqI occupies the position analogous to Cys81, the key catalytic residue for M.HhaI (23 ), indicating that the Tyr in motif IV of adenine methyltransferase probably is an important catalytic residue (24 ). In addition, the structural modeling of EcoKI predicted that this aromatic residue in motif IV is poised at the edge of the active site where it can interact with the target base if it is flipped out of the DNA (25 ).

Mutagenesis studies on the two conserved adenine methyltransferase motifs that have been carried out on Dam methyltransferase (26 ,27 ), EcoKI methyltransferase (28 ) and BcgI methyltransferase (this study) generally agree with the structural data. A change of the most conserved Gly in motif I of EcoKI (G177D) completely abolishes AdoMet binding and methylation activity (28 ), while substitution of this equivalent residue in BcgI (G355D) abolishes methylation, cleavage and presumably AdoMet binding activities. These data show that motif I of adenine methylases is very important for AdoMet binding.

In the case of motif IV, previous studies on Dam methyltransferase (motif IV = DPPY) have shown that substitutions of the first residue (D to S, N and G) abolish both AdoMet binding and methylation activities (26 ). In addition, change of the first Pro also resulted in an increased Km for AdoMet, indicating that motif IV is also important for AdoMet binding (27 ). However, mutagenesis on motif IV of EcoKI (NPPF) yielded a different result. Mutant EcoKI-N266D fails to methylate DNA but still binds AdoMet, suggesting that Asn in motif IV is involved in methylation catalysis but is not involved in AdoMet binding. In the case of BcgI, substitutions of the first residue in the NPPY motif to Ala, Asp and Gln completely abolish both downstream activities and an Asn to Ser change results in a partially active BcgI with reduced methylation and cleavage activities. Our data suggests that Asn in motif IV of BcgI is involved in an early step required for both catalytic activities. Presumably, this asparagine is important for AdoMet binding.

The essential catalytic role of the aromatic residue in motif IV has also been explored by mutagenesis studies. Substitutions of F to Y and W in EcoKI result in an active enzyme and changes of F to G and C abolished methylation activity but left cofactor binding activity unaltered (28 ). Similar results were obtained from this study in which substitutions of Y439 with other aromatic residues result in an active BcgI whereas the Y to A change abolishes the methylation but not the AdoMet binding activity. These results suggest that the aromatic residue in motif IV of adenine methyltransferases is a catalytic residue which is important for methyltransfer and is not essential for AdoMet binding.

ACKNOWLEDGEMENTS

We thank Dr Richard Roberts, Dr Charles Cantor, Dr Edward Loechler, Dr Thomas Gilmore and Dr Ira Schildkraut for helpful advice and critical reading of this manuscript; Dr Jack Benner for help with the AdoMet binding assay; Lauren Sears and Leigh Olmsted for help in preparation of this article; and Dr Donald Comb for his support of this project.

REFERENCES

1 Roberts, R.J. and Macelis, D. (1996) Nucleic Acids Res., 23, 223-235.

2 Yuan, R. (1981) Annu. Rev. Biochem., 50, 285-315. MEDLINE Abstract

3 Winkler, F.K., Banner, D.W., Oefner, C., Tsernoglou, D., Brown, R.S. and Heathman, S.P., Bryan, R.K., Martin, P.D., Petratos, K. and Wilson, K.S. (1993) EMBO J., 12, 1781-1795. MEDLINE Abstract

4 Thielking, V., Selent, U., Kohler, E., Landgraf, A., Wolfes, H., Alves, J. and Pingound, A. (1992) Biochemistry, 31, 3727-3732. MEDLINE Abstract

5 Labahn, J., Granzin, J., Schluchebier, G., Robinson, D.P., Jack, W.E. and Schildkraut, I. (1994) Proc. Natl. Acad. Sci. USA, 91, 10957-10961. MEDLINE Abstract

6 Burckhardt, J., Weisemann, J., Hamilton, D.L. and Yuan, R. (1981) J. Mol. Biol., 153, 425-440. MEDLINE Abstract

7 Burckhardt, J., Weisemann, J. and Yuan, R. (1981) J. Biol. Chem., 256, 4024-4032. MEDLINE Abstract

8 Suri, B., Nagaraja, V. and Bickle, T.A. (1984) Curr. Topics Microbiol. Immunol., 108, 1-9.

9 Powell, L.M., Dryden, T.F., Willcock, D.F., Pain, R.H. and Murrray, N.E. (1993) J. Mol. Biol., 234, 60-71. MEDLINE Abstract

10 Roberts, R.J. (1990) Nucleic Acids Res., 18, 2331-2365. MEDLINE Abstract

11 Kong, H., Morgan, R.D., Maunus, R.E. and Schildkraut, I. (1993) Nucleic Acids Res., 21, 987-991. MEDLINE Abstract

12 Sears, L.E., Zhou, B., Aliotta, J.M., Morgan, R.D. and Kong, H. (1996) Nucleic Acids Res., 24, 3590-3592. MEDLINE Abstract

13 Degtyarev, S.K., Rechkunova, N.I., Zernov, U.P., Dedkov, V.S., Chizikov, V.E., Calligan,M.V., Williams, R. and Murray, E. (1993) Gene, 131, 93-95. MEDLINE Abstract

14 Vitor, J.M.B. and Morgan, R.D. (1995) Gene, 157, 109-110.

15 Kong, H., Roemer, S.E., Waite-Rees, P.A., Benner, J.S., Wilson, G.G. and Nwankwo, D.O. (1994) J. Biol. Chem., 269, 683-690. MEDLINE Abstract

16 Kunkel, T.A., Roberts, J.D. and Zakour, R.A. (1987) Methods Enzymol., 154, 367-382. MEDLINE Abstract

17 Studier, F.W., Rosenberg, A.H., Dunn, J.J. and Dubendorff, J.W. (1990) Methods Enzymol., 185, 60-89. MEDLINE Abstract

18 Bergerat, A. and Guschlbauer, W. (1990) Nucleic Acids Res., 18, 4369-4375. MEDLINE Abstract

19 Klimasauskas, S., Timinskas, A., Menkevicius, S., Butkiene, D., Butkus, V. and Janulaitis, A. (1989) Nucleic Acids Res., 17, 9823-9831. MEDLINE Abstract

20 Cheng, X., Kumar, S., Klimasauskas, S. and Roberts, R.J. (1993) Cold Spring Harbor Symp. Quant. Biol., 58, 331-338. MEDLINE Abstract

21 Klimasauskas, S., Kumar, S., Roberts, R.J. and Cheng, X. (1994) Cell, 76, 357-369. MEDLINE Abstract

22 Schluckebier, G., O'Gara, M., Saenger, W. and Cheng, X. (1995) J. Mol. Biol., 247, 16-20. MEDLINE Abstract

23 Wu, J.C. and Santi, D.V. (1987) J. Biol. Chem., 262, 4778-4786. MEDLINE Abstract

24 Cheng, X. (1995) Annu. Rev. Biophys. Biomol. Struct., 24, 293-318. MEDLINE Abstract

25 Dryden, D.T.F., Sturrock, S.S. and Winter, M (1995) Nature Struct. Biol., 2, 632-635.

26 Guyot, J.B., Grassi, J., Hahn, U. and Guschlbauer, W. (1993) Nucleic Acids Res., 21, 3183-3190. MEDLINE Abstract

27 Kossykh, V.G., Schlagman, S.L. and Hattman, S. (1993) Nucleic Acids Res., 21, 4659-4662. MEDLINE Abstract

28 Willcock, D.F., Dryden, D.T.F. and Murray, N.E. (1994) EMBO J., 13, 3902-3908. MEDLINE Abstract

*To whom correspondence should be addressed. Tel: +1 508 927 5054; Fax: +1 508 921 1350; Email: kong@neb.com
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Infect. Immun.Home page
T. E. Erova, A. A. Fadl, J. Sha, B. K. Khajanchi, L. L. Pillai, E. V. Kozlova, and A. K. Chopra
Mutations within the Catalytic Motif of DNA Adenine Methyltransferase (Dam) of Aeromonas hydrophila Cause the Virulence of the Dam-Overproducing Strain To Revert to That of the Wild-Type Phenotype.
Infect. Immun., October 1, 2006; 74(10): 5763 - 5772.
[Abstract] [Full Text] [PDF]

Home page
 MicrobiologyHome page
R. M. Dempsey, D. Carroll, H. Kong, L. Higgins, C. T. Keane, and D. C. Coleman
Sau42I, a BcgI-like restriction-modification system encoded by the Staphylococcus aureus quadruple-converting phage {pi}42
Microbiology, April 1, 2005; 151(4): 1301 - 1311.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
P. Bist and D. N. Rao
Identification and Mutational Analysis of Mg2+ Binding Site in EcoP15I DNA Methyltransferase: INVOLVEMENT IN TARGET BASE EVERSION
J. Biol. Chem., October 24, 2003; 278(43): 41837 - 41848.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
N. K. Raghavendra and D. N. Rao
Functional cooperation between exonucleases and endonucleases--basis for the evolution of restriction enzymes
Nucleic Acids Res., April 1, 2003; 31(7): 1888 - 1896.
[Abstract] [Full Text] [PDF]

Home page
 Nucleic Acids ResHome page
A. Pingoud and A. Jeltsch
Structure and function of type II restriction endonucleases
Nucleic Acids Res., September 15, 2001; 29(18): 3705 - 3727.
[Abstract] [Full Text] [PDF]

Home page
 J. Bacteriol.Home page
P.-c. Hsieh, J.-p. Xiao, D. O'loane, and S.-y. Xu
Cloning, Expression, and Purification of a Thermostable Nonhomodimeric Restriction Enzyme, BslI
J. Bacteriol., February 15, 2000; 182(4): 949 - 955.
[Abstract] [Full Text]

Home page
 J. Biol. Chem.Home page
M. Roth, S. Helm-Kruse, T. Friedrich, and A. Jeltsch
Functional Roles of Conserved Amino Acid Residues in DNA Methyltransferases Investigated by Site-directed Mutagenesis of the EcoRV Adenine-N6-methyltransferase
J. Biol. Chem., July 10, 1998; 273(28): 17333 - 17342.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Print PDF (110K) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (20)
Right arrowRequest Permissions
Right arrow Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Right arrow Articles by Kong, H.
Right arrow Articles by Smith, C. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kong, H.
Right arrow Articles by Smith, C. L.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?