HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted 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 Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gormley, N. A. Articles by Halford, S. E. Search for Related Content PubMed PubMed Citation Articles by Gormley, N. A. Articles by Halford, S. E. Social Bookmarking                      What's this?

J Biol Chem, Vol. 275, Issue 10, 6928-6936, March 10, 2000

Reactions of BglI and Other Type II Restriction Endonucleases with Discontinuous Recognition Sites^*

Niall A. Gormley, Abigail J. Bath, and Stephen E. Halford

From the Department of Biochemistry, School of Medical Sciences, University of Bristol, University Walk, Bristol, BS8 1TD, United Kingdom

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Type II restriction enzymes generally recognize continuous sequences of 4-8 consecutive base pairs on DNA, but some recognize discontinuous sites where the specified sequence is interrupted by a defined length of nonspecific DNA. To date, a mechanism has been established for only one type II endonuclease with a discontinuous site, SfiI at GGCCNNNNNGGCC (where N is any base). In contrast to orthodox enzymes such as EcoRV, dimeric proteins that act at a single site, SfiI is a tetramer that interacts with two sites before cleaving DNA. BglI has a similar recognition sequence (GCCNNNNNGGC) to SfiI but a crystal structure like EcoRV. BglI and several other endonucleases with discontinuous sites were examined to see if they need two sites for their DNA cleavage reactions. The enzymes included some with sites containing lengthy segments of nonspecific DNA, such as XcmI (CCANNNNNNNNNTGG). In all cases, they acted at individual sites. Elongated recognition sites do not necessitate unusual reaction mechanisms. Other experiments on BglI showed that it bound to and cleaved DNA in the same manner as EcoRV, thus further delineating a distinct group of restriction enzymes with similar structures and a common reaction mechanism.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Type II restriction endonucleases recognize specific sequences in DNA and cut both strands at fixed locations within or adjacent to the recognition sequence, in reactions that need only Mg²⁺ as a cofactor (1). The value of these enzymes as tools for the analysis and manipulation of DNA has prompted extensive searches for new enzymes of this type, and over 3000 have been identified (2). The recognition sites for most restriction enzymes are palindromic sequences of 4, 6, or 8 consecutive bp.¹ Almost all of the relatively small number of type II enzymes that have been analyzed to date, with respect to their reaction mechanisms and/or structures, recognize continuous sequences of this sort: viz. BamHI, EcoRI, EcoRV, MunI, PvuII, and TaqI (3-8). All of these examples are homodimeric proteins that interact symmetrically with their respective recognition sites so that the two active sites are located on the two target phosphodiester bonds, one in each strand. The enzyme can then cut both strands at one site in a single DNA-binding event (9, 10). On a DNA with two or more sites, they usually act in a distributive manner (11), catalyzing separate reactions at each site, but they sometimes act processively, translocating to a second site and cutting it before leaving the DNA (10, 12).

With a few exceptions, the type II restriction enzymes have dissimilar amino acid sequences (13) but their three-dimensional structures can be similar to one another (14). For instance, the tertiary fold of EcoRI is very similar to BamHI (15); likewise EcoRV is similar to PvuII (16), though the overall structures of EcoRI and EcoRV differ considerably. The mechanisms of these enzymes also show similarities in some instances and differences in other instances. For example, EcoRI and BamHI bind to DNA in the absence of Mg²⁺ preferentially at their recognition sites (3, 4, 12). In contrast, under their optimal reaction conditions but for the absence of Mg²⁺, EcoRV, TaqI, and several others bind equally well to all DNA sequences (17-20). In the presence of Mg²⁺, the latter enzymes still cleave DNA specifically at their recognition sites (21, 22), but they also need a divalent metal ion for specific binding; Ca²⁺ can function in this respect (7, 8, 23, 24). The similarities and the differences among the type II restriction enzymes have led to the proposal that many, including BamHI, can be classified as EcoRI-like enzymes, whereas many others, including PvuII and TaqI, can be classified as EcoRV-like enzymes (10, 14, 20, 22). However, an alternative view maintains that there is no clear-cut distinction between the EcoRI- and the EcoRV-like enzymes and that these enzymes are essentially similar to each other in terms of both mechanism and core structure (6, 19, 24-26).

The recognition sites for some type II restriction enzymes are not the continuous sequences of 4-8 consecutive bp noted above, but are instead discontinuous sequences in which the palindromic elements recognized by the enzyme are separated by a fixed length of nonspecific DNA (2). The intervening DNA can be as long as 9 bp, viz. XcmI (Table I), or as short as 1 bp, viz. HhaII (GANTC). Apart from some early studies on HhaII (27), SfiI is the only type II enzyme with a discontinuous recognition site whose reaction mechanism has been analyzed (28-33). SfiI differs from the orthodox enzymes such as EcoRV in several ways. First, its recognition site contains 8 specified bp interrupted by 5 bp of nonspecific DNA and thus covers 13 bp (Table I), which is longer than usual for a restriction site (34). Second, SfiI is a tetramer of identical subunits instead of the normal dimer. Third, no DNA cleavage arises from the binding of the tetramer to one recognition site. Instead, SfiI is only active when bound to two copies of its site. It displays its optimal activity with two sites on the same DNA, where it loops out the DNA between the sites; but it can also, albeit less readily, cleave DNA with one SfiI site by bridging two such molecules. Fourth, once bound to the two sites, SfiI usually cuts both strands at both sites before leaving the DNA. Another subset of the type II enzymes, known as type IIe and typified by EcoRII and NaeI, also needs two sites for the DNA cleavage reactions but these differ from SfiI in that they seem to use one site to activate the reaction at the other site; the activator DNA is not cleaved (35-37). Nonetheless, concerted cleavage of two recognition sites in the manner of SfiI has been noted with both Cfr10I and SgrAI (11, 38). Unlike SfiI, Cfr10I and SgrAI recognize continuous sequences, RCCGGY and GRCCGGYC, respectively. The 8-bp site for SgrAI is identical to the 6-bp site for Cfr10I except for one extra bp at each end.

In this study, the reactions of BglI (39) and several other restriction enzymes with discontinuous recognition sites (Table I) were analyzed on plasmids with either one or two target sites, to determine whether enzymes with elongated sites behave like the orthodox enzymes or whether they interact with two sites, like SfiI. The recognition sequence for BglI (40) is identical to that for SfiI except that it is one bp shorter at each end (Table I), i.e. the same relationship as that between the Cfr10I and SgrAI sites. This raises the possibility that BglI might act concertedly at two sites in the same way as SfiI. However, the structure of BglI bound to its recognition sequence was recently determined by x-ray crystallography, and this shows a dimeric protein bound symmetrically to one DNA duplex (41). Moreover, the DNA recognition and catalytic functions in each subunit of BglI are similar to those in EcoRV, though the subunit interface in BglI differs from EcoRV. The amino acid sequences of BglI and EcoRV are not homologous (13). In EcoRV, which cleaves GATATC as marked, the recognition/catalytic functions of the two subunits are located almost opposite each other across the axis of the DNA (42). In contrast, the recognition/catalytic functions of the two subunits in BglI are displaced relative to each other along the axis of the DNA, thus allowing it to recognize two specific segments of DNA separated by 5 bp of nonspecific DNA (41). The main question posed here is whether BglI acts like SfiI, as might be expected from its recognition sequence, or like EcoRV, as might be expected from its crystal structure.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Proteins-- BglI endonuclease, purified to homogeneity (41) by I. Schildkraut and colleagues (New England Biolabs, Beverly, MA), was a gift from M. Newman (Imperial Cancer Research Fund, London, UK). Its concentration was determined by the method of Bradford (43) and is given in terms of molarity of the dimeric protein of M_r 70,000. All other enzymes were purchased from New England Biolabs; concentrations of the latter are given in terms of units of enzyme activity, as defined by the supplier.

DNA-- Plasmids pUC19 (44), pBR322 (45), pAT153 (46), and pNEB193 (New England Biolabs) were manipulated by standard procedures (47). To make pBGL1 (Fig. 1a), pUC19 was cleaved with EcoRI and KasI, and the large fragment was purified by electrophoresis prior to the removal of its single strand extensions with mung bean nuclease and re-circularization with T4 DNA ligase. To make pML1 (Fig. 1b), pBR322 was cleaved with EcoRI and HindIII and ligated to a 43-bp duplex that had single strand extensions that matched an EcoRI terminus at one end and a HindIII terminus at the other end; the duplex was made from two self-complementary oligodeoxynucleotides, both 47 bases long, with recognition sites for Tth111I, PshAI, and AhdI that each had the same flanking and intervening sequences as the native site in pBR322. To make pAB2 (Fig. 1c), pNEB193 was cleaved with PstI and HindIII and ligated to a 25-bp duplex with 4-base single strand extensions that matched a PstI terminus at one end and a HindIII terminus at the other end; the duplex was made from two self-complementary oligonucleotides, of 25 and 33 bases, with recognition sequences for PflMI, BstXI, and XcmI. To make pAB3 (Fig. 1c), the small DNA fragment obtained by cleaving pAB2 with PvuII was cloned at the SspI site on pAB2; the two copies of the PvuII fragment present in pAB3 are in inverted orientation. The plasmids were used to transform recA strains of Escherichia coli, either XL-blue or HB101 (47). The transformants were cultured in M9 minimal medium with 1 mCi/l [methyl-³H]thymidine and the covalently closed form of the plasmid purified by density gradient centrifugations (48). The preparations were largely supercoiled monomeric plasmid, with <10% as either dimeric plasmid or nicked open circle DNA.

DNA Cleavage-- Restriction enzymes were assayed by adding an aliquot (typically 5 µl) to a ³H-labeled plasmid (5 or 10 nM) in 200 µl of the buffer recommended by the supplier of the enzyme (or modifications thereof). For BglI, the aliquots were samples diluted to the requisite concentration in 10 mM Tris·HCl (pH 7.4), 300 mM NaCl, 0.1 mM EDTA, 1 mM DTT, 500 µg/ml bovine serum albumin, and 50% (v/v) glycerol. For the other enzymes, 1-15 units of the purchased stocks were added directly to the reactions. At various times after adding the enzyme, 15-µl samples were removed from the reaction and mixed immediately with 10 µl of stop mix (21). The samples were analyzed by electrophoresis through agarose to separate the supercoiled substrate and each of the reaction products. The segments of agarose that encompassed the substrate and each product were dissolved in 5 M sodium perchlorate and analyzed individually by scintillation counting to yield the concentration of each form of the DNA at each time point (48, 49). For plasmids with two recognition sites, the two fragments arising from cleavage at both sites were counted together to obtain one value for the concentration of doubly cut DNA. Values of v₁ and v_2A, the rates for the utilization of the one-site and two-site substrates, respectively, were evaluated from the initial linear decline in the concentration of the supercoiled substrate with time, whereas v_2B, the rate for cutting the second site on the two-site substrate, was assessed relative to v_2A by the curve-fitting procedure described previously (11).

DNA Binding-- A 465-bp fragment with one BglI site was obtained by cleaving pUC19 with AatII and EcoRI (Fig. 1a). An isogenic 465-bp fragment that lacked a BglI site was obtained by an AatII/EcoRI digest of a mutated form of pUC19 where BglI site A, GCCATTCAGGC, had been changed to GCCATTCAGAC by using the Quikchange Mutagenesis kit (Stratagene) with primers CGCCATTCAGACTGCGCAACTG and its complement. The fragments were isolated by electrophoresis through agarose, purified with the Qiagen gel-purification kit, and labeled by using Klenow polymerase with [-³²P]dATP and dTTP (47). Varied amounts of BglI (diluted in binding buffer) were added to the fragments to give 10-µl samples containing ~50 pM DNA and 0-12.5 nM enzyme. The binding buffers contained the required concentration of NaCl in either EDTA buffer (50 mM Tris·HCl (pH 7.5), 10 mM -mercaptoethanol, 100 µg/ml bovine serum albumin, 0.1 mM EDTA) or Ca²⁺ buffer (the same supplemented with 5 mM CaCl₂). After 20 min at room temperature, loading buffer (5 µl) was added to each sample, and the samples were applied to 6% polyacrylamide gels (17). The loading buffer was the same as that for the binding reaction, with either EDTA or CaCl₂, augmented in both cases with 40% glycerol and 0.01% (w/v) bromphenol blue. For samples in EDTA, the gel was prepared and run in 0.089 M Tris base, 0.089 M boric acid, and 2 mM EDTA. For samples with Ca²⁺, the gel buffer was as above except for 5 mM CaCl₂ in place of the EDTA (23). After electrophoresis, the radioactivity on the gel was measured with a 400B PhosphorImager and analyzed by ImageQuant software (Molecular Dynamics, Sunnyvale, CA).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

DNA Cleavage by BglI-- A comparison of the reactions of a restriction enzyme on substrates with either one or two recognition sites can distinguish the following schemes (11): independent reactions at each site, in the conventional distributive manner for a type II enzyme; processive reactions, where the enzyme translocates from one site to another without leaving the DNA; activation by a second site, as proposed for the type IIe enzymes; concerted action at two sites, as noted with SfiI. A type II enzyme will cleave a supercoiled (SC) plasmid with one site first in one strand to give the open circle (OC) form of the DNA and then in the other strand, to give the full-length linear (FLL) form (Fig. 2a). The cutting of both strands is often faster than the dissociation of the enzyme from the DNA (9, 10), and steady-state reactions then yield none of the OC form (48). A plasmid with two sites is cleaved by an enzyme that acts independently at each site in sequential steps, leading first to the transient formation of FLL DNA as one site is cut and then, after a lag phase, to the final products cut at both sites, L1 and L2 (Fig. 2b). If the recognition sites on the one-site and two-site substrates are equally susceptible to the enzyme, then independent action at each site results in the same steady-state rates for the utilization of the one- and two-site substrates (v₁ and v_2A, respectively) and for the conversion of the FLL form of the two-site DNA to L1 and L2 (v_2B). If v_2A = v_2B, the maximal amount of FLL DNA formed during the reaction on the two-site substrate will be 40% of the total DNA. On the other hand, both processive and concerted reactions yield less of the FLL DNA, whereas an enzyme employing a second site, either as an activator or for a concerted reaction, would consume the two-site substrate at a faster rate than the one-site substrate (11).

To determine which scheme applies to the BglI enzyme, the kinetics of DNA cleavage were analyzed on plasmids with one or two BglI sites, pBGL1 and pUC19, respectively (Fig. 1a). Under steady-state conditions with lower concentrations of enzyme than DNA (Fig. 2), the rates increased linearly with increasing concentrations of the enzyme (data not shown), and both substrates were converted completely to the final products expected from the cleavage of all BglI sites: FLL DNA from pBGL1 and L1 and L2 from pUC19. Hence, BglI carries out multiple catalytic turnovers, as expected for a type II endonuclease (1). The reaction of BglI on the one-site plasmid yielded none of the OC DNA and instead progressed directly to the FLL form cut in both strands (Fig. 2a). BglI thus cleaves both DNA strands at a single recognition site at rates that are faster than its dissociation from the cleaved DNA. The two-site substrate was cleaved in sequential stages; first at one site to give FLL DNA and then, after a lag phase, at the other site to give L1 and L2 (Fig. 2b). Moreover, the rate of utilization of the two-site substrate (v_2A = 6.3 ± 0.3 min¹) was similar to the one-site substrate (v₁ = 5.4 ± 0.6 min¹). The BglI endonuclease therefore cleaves each recognition site in an independent reaction, in the orthodox manner for a type II enzyme.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Plasmid substrates. a, pUC19 contains two sites for BglI, marked A and B. The removal of site A, by deleting the DNA between the KasI and the EcoRI sites, yielded a plasmid with only BglI site B, pBGL1. b, pBR322 has single sites for Tth111I, PshAI, and AhdI. A plasmid with two sites for each of these enzymes, pML1, was constructed by inserting an oligonucleotide duplex of the requisite sequence between the EcoRI and HindIII sites of pBR322. c, pNEB193, which lacks PflMI, BstXI, and XcmI sites, was converted into a plasmid with one site for each of these enzymes, pAB2; an oligonucleotide duplex with the requisite sequence was inserted between the PstI and HindIII sites of pNEB193. The PvuII fragment spanning the insert (indicated in gray) was then cloned at the SspI site of pAB2 to yield pAB3, with two sites for PflMI, BstXI, and XcmI.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   BglI on plasmids with (a) one or (b) two recognition sites. The drawings indicate the various forms of DNA that can exist during the reaction of a restriction enzyme on circular DNA with either (a) one or (b) two recognition sites for the enzyme. SC, supercoiled substrate; OC, open circle DNA cleaved in one strand; FLL, full-length linear DNA cleaved in both strands at one site; L1 and L2 (only in b), the two linear DNA products from cleaving both strands at both sites. The reactions contained 0.07 nM BglI endonuclease and 5 nM 3H-labeled DNA (~95% supercoiled) in 50 mM Tris·HCl (pH 7.9), 100 mM NaCl, 10 mM MgCl2, and 1 mM DTT, at 37 °C. The DNA in (a) was pBGL1, which has one BglI site. The DNA in (b) was pUC19, which has two BglI sites. At timed intervals after adding the enzyme, samples from the reactions were quenched with stop mix prior to electrophoresis through agarose. Individual segments of the gels were then analyzed by scintillation counting to obtain concentrations of the following forms of the DNA at each time point sampled during the reaction: , supercoiled substrate (marked SC on both graphs); , open circle DNA (marked OC); , full-length linear DNA (marked FLL);  (only in b), total DNA in the two final products after cutting both sites (marked L1/2).

The above experiments with the SC form of pUC19 do not reveal which BglI site is cleaved to yield the FLL form. The two BglI sites in pUC19, noted as A and B (Fig. 1a), are flanked by different sequences, and they also have different sequences in the 5-bp spacer of nonspecific DNA in the middle of the site. Restriction activity is often affected by sequences flanking the recognition site (1), and the activity of SfiI is also modulated by the sequence of the spacer (32). Hence, one BglI site in pUC19 may be more susceptible than the other. To monitor the cleavage of each site, pUC19 was cut with AlwNI, and the linearized DNA was used as a substrate for BglI (Fig. 3). The cleavage of either BglI site on this substrate gives rise to two fragments but the pair of fragments produced by cutting site A are distinct from those from cutting B (Fig. 3a). A partial product of 1714 bp scores the fraction of the DNA cleaved only at A, because this fragment carries an intact site B. Likewise, a partial product of 2090 bp scores the fraction cleaved only at B. During the BglI reaction on the linearized pUC19, both the 1714-bp and 2090-bp partial products were formed transiently and were then devoured as their intact sites were cleaved (Fig. 3b). The singly cut DNA thus contains one fraction cleaved only at A and another fraction cleaved only at B. This shows that BglI must indeed cleave each site in a separate reaction. However, the transient for the 1714-bp product reached a higher maximum than that for the 2090-bp product. Hence, BglI site A is cleaved preferentially to site B, by a factor of ~2.5.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   BglI on linear DNA with two sites. a, the drawing illustrates the possible products from a BglI reaction on AlwNI-linearized pUC19, a 2686-bp DNA with two BglI sites marked A and B. Cleavage at B yields a final product of 596 bp and a partial product of 2090 bp; the latter is then cleaved at A to the final products of 1118 and 972 bp. Cleavage at A yields a final product of 972 bp and a partial product of 1714 bp; the latter is then cleaved at B to the final products of 596 bp and 1118 bp. b, the reaction contained 0.28 nM BglI and 10 nM AlwNI-linearized pUC19 in 50 mM Tris·HCl (pH 7.9), 100 mM NaCl, 10 mM MgCl2, and 1 mM DTT, at 37 °C. Samples were withdrawn from the reaction at timed intervals and added immediately to stop mix. Subsequent electrophoresis of each sample through agarose separated all six of the forms of DNA noted in a (not shown), and the concentrations of each were determined by analyzing individual segments of the gel in a scintillation counter. The concentrations of the partial products of 2090 bp () and 1714 bp () are shown.

DNA Binding by BglI-- Independent reactions at individual recognition sites are common to both the EcoRI- and the EcoRV-like enzymes, but in the absence of Mg²⁺, to prevent DNA cleavage, the enzymes in each group bind to DNA in distinctive ways (18). Gel-shift studies on the binding of EcoRV to a 381-bp DNA with one EcoRV site revealed multiple DNA-protein complexes, due to the association of 1, 2, 3···N molecules of protein/molecule of DNA, where N is the maximum that can fit onto the DNA (17). The same series of complexes were also formed at the same protein concentrations with an isogenic DNA lacking an EcoRV site. In contrast, gel shift studies with EcoRI had revealed a single DNA-protein complex due to specific binding at the recognition site (12). However, binding studies with EcoRV in the presence of Ca²⁺ as a noncatalytic analogue of Mg²⁺ gave a single DNA-protein complex with a DNA carrying an EcoRV site but no complex with a DNA lacking an EcoRV site (23). Ca²⁺ ions thus enhance the binding of EcoRV to its recognition site and diminish its binding to nonspecific DNA.

To determine whether BglI belongs to either group, its binding to DNA was examined by the gel shift method in the absence of divalent metal ions and in the presence of Ca²⁺; no DNA cleavage was detected in the presence of Ca²⁺ (see below). Two DNA fragments were used: a 465-bp fragment containing one BglI site and an isogenic 465-bp fragment from which the BglI site had been removed by changing one bp. The experiments were carried out initially in buffers containing 100 mM NaCl, the optimum for BglI activity (39). Under these conditions, the stoichiometric addition of 50 pM BglI protein to 50 pM DNA resulted in all of the DNA being converted to DNA-protein complexes, regardless of whether the DNA had a BglI site or whether Ca²⁺ was present (data not shown). Hence, no distinction could be made between the binding of BglI to specific or nonspecific DNA, with or without Ca²⁺. In contrast to EcoRV and to most other restriction enzymes, the affinity of BglI for DNA in 100 mM NaCl is too high to allow for a K_D value to be obtained from the titration of 50 pM DNA with increasing amounts of protein. However, many restriction enzymes behave as other DNA-binding proteins in binding to DNA more weakly at elevated salt concentrations (12, 25). By raising the NaCl concentration to 200 mM, the affinity of BglI for DNA was reduced sufficiently to permit such titrations (Fig. 4).

View larger version (107K): [in this window] [in a new window]   Fig. 4.   BglI binding to specific and nonspecific DNA in the absence and presence of Ca2+. The binding reactions contained 32P-labeled DNA (~0.05 nM), in either EDTA buffer with 200 mM NaCl (a and b) or Ca2+ buffer with 200 mM NaCl (c and d), and BglI at one of the following concentrations (from left to right across the gel, as indicated by the expanding scale); 0, 0.05, 0.1, 0.2, 0.4, 0.8, 1.6, 3.1, 6.3, and 12.5 nM. The DNA in a and c was the 465-bp AatII/EcoRI fragment from pUC19, which has one BglI site. The DNA in b and d was the 465-bp AatII/EcoRI fragment from a derivative of pUC19, which differs from the above by a single bp change in the recognition sequence for BglI. After 20 min at room temperature, the samples were subjected to electrophoresis through polyacrylamide, and the gels were subsequently analyzed in a PhosphorImager; the phosphorescence records are shown here. F marks the electrophoretic mobility of the free DNA.

The addition of progressively increasing amounts of BglI protein to the DNA with one BglI site, in a buffer containing 200 mM NaCl but no divalent metal ions, caused a progressive reduction in the amount of free DNA present in the binding mixture (Fig. 4a). However, instead of giving rise to the single retarded complex that would have been expected if the protein bound only to its recognition site, a sequential series of complexes were formed, with successively reduced mobilities. The same series of complexes were formed with the DNA lacking a BglI site, at the same concentrations of BglI protein (Fig. 4b). Hence, as with EcoRV (17), BglI binds to DNA in the absence of divalent metal ions with no detectable preference for its recognition site over the sum of the nonspecific sites on these DNA fragments. The successive complexes with both specific and nonspecific DNA are therefore the 465-bp fragments carrying 1, 2, 3···N molecules of BglI protein bound at random along the DNA. Because BglI covers ~15 bp of DNA (41), N  31.

When Ca²⁺ was present, the addition of increasing amounts of BglI to the DNA with a BglI site no longer caused the progressive loss of free DNA. Instead, even in 200 mM NaCl, the DNA was converted stoichiometrically to a single complex, though additional complexes were formed at higher concentrations of BglI (Fig. 4c). In contrast, Ca²⁺ had virtually no effect on the binding of BglI to the DNA without a BglI site; a progressive rather than a stoichiometric loss of free DNA was recorded, together with the formation of multiple complexes at similar protein concentrations to those in the absence of Ca²⁺ (Fig. 4d). The initial binding to the DNA with a BglI site, at a stoichiometric concentration of BglI protein, must therefore be at the recognition site and the additional complexes formed at higher concentrations must be due to the further binding of 1, 2, 3···N-1 molecules of BglI protein at nonspecific sites elsewhere on the DNA.

Metal Ion Specificity of BglI-- DNA cleavage by BglI was examined with other metal ions in place of Mg²⁺, to determine whether BglI interacts with metal ions in the same way as EcoRV. EcoRV has significant activity with either Co²⁺ or Mn²⁺ but no activity with Ca²⁺ (5, 49, 50). Its optimal turnover rates with Co²⁺ and Mn²⁺ are, respectively, about one third and one tenth of that with Mg²⁺ but it has higher affinities for Co²⁺ and Mn²⁺ than Mg²⁺; the maximal reaction rate achievable with Mg²⁺ requires a higher concentration of metal ion than is the case with Co²⁺ or Mn²⁺ (5). However, the substitution of Mg²⁺ with Mn²⁺ eradicates most of the specificity of EcoRV for its recognition sequence (50). With Mg²⁺, noncognate sites one bp different from the recognition site are cleaved at least 10⁶ times more slowly. With Mn²⁺, some noncognate sites are cleaved just six times more slowly than the recognition site. No such loss of specificity occurs with Co²⁺ (5).

The activity of BglI with alternative metal ions was examined first by adding logarithmically increasing concentrations of BglI protein to a plasmid with one BglI site in the presence of a fixed concentration of a divalent metal ion; the reactions were stopped after a fixed time, and the DNA was analyzed by electrophoresis through agarose (Fig. 5). With Mg²⁺ as the cofactor, the lowest concentration of BglI tested was sufficient to convert all of the SC substrate to FLL DNA, by cleaving the BglI site, and no further cleavages were detected even at a 1000-fold higher concentration of BglI. With Mn²⁺, the lowest enzyme concentration was insufficient to cleave all of the SC substrate; a 10-fold increase resulted in all of the SC substrate being cleaved at its BglI site, but a further 10-fold increase in the enzyme concentration then produced several smaller fragments of DNA, because of cleavages at a number of secondary sites (Fig. 5). Co²⁺ also failed to give complete cleavage of the SC substrate at the lowest BglI concentration, though a smaller fraction of the DNA remained intact than was the case with Mn²⁺, but, in contrast to Mn²⁺, Co²⁺ caused no additional cleavages at high enzyme concentrations. Exactly like EcoRV (5, 50), BglI cleaves DNA with a high degree of specificity for its recognition sequence in the presence of either Mg²⁺ or Co²⁺ but with a very much lower specificity in the presence of Mn²⁺.

View larger version (74K): [in this window] [in a new window]   Fig. 5.   Specificity of BglI with alternative metal ions. The reactions, in 50 mM Tris·HCl (pH 7.9), 100 mM NaCl, and 1 mM DTT, contained pBGL1 (10 nM), one of the concentrations of BglI noted below, and one of the following salts (5 mM): MgCl2, lanes 3-6; MnCl2, lanes 7-11; CoCl2, lanes 11-14. The concentrations of BglI in lanes 3-6 were 0.5, 5, 50, and 500 nM, respectively; likewise in lanes 7-10 and 11-14. Lane 2 had no BglI. After 1.5 h at 37 °C, the DNA was treated with SDS (0.2% w/v) and proteinase K (0.2 mg/ml) for 10 min at 37 °C, extracted with phenol/chloroform, and then analyzed by electrophoresis through agarose. The mobilities of the SC, OC, and FLL forms of pBGL1 are marked on the left of the gel. Lane 1 contained size markers ranging from 250 bp at the bottom of the gel to 10 kilobases at the top.

A quantitative analysis of BglI activity with various metal ions was carried out by measuring steady-state rates for the utilization of a SC substrate with one BglI site at varied concentrations of each metal ion (Fig. 6). In these reactions, with the enzyme at a lower concentration than the DNA, only the recognition site is cleaved (Fig. 5). The turnover rates with Mg²⁺ increased linearly as the concentration of MgCl₂ was increased from 1 to 5 mM but the rates with both Co²⁺ and Mn²⁺ remained virtually unaltered as the concentrations of these ions were increased above 1 mM (Fig. 6). Thus, as with EcoRV (5), BglI has a lower affinity for Mg²⁺ than for either Co²⁺ or Mn²⁺, but at high metal-ion concentrations, its turnover rate with Mg²⁺ exceeds that with Co²⁺, which in turn exceeds that with Mn²⁺. The relative rates with these three ions (Fig. 6) match the extents of DNA cleavage at the lowest BglI concentration used above (Fig. 5). No DNA cleavage was detected with Ca²⁺, at all concentrations tested (Fig. 6). However, in experiments similar to those described previously with EcoRV (49), Ca²⁺ was found to act as a competitive inhibitor to Mg²⁺; the addition of increasing concentrations of CaCl₂ to reactions containing a fixed concentration of MgCl₂ progressively reduced the rate of DNA cleavage (data not shown).

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Metal ion dependence of DNA cleavage by BglI. Reactions at 37 °C, in 50 mM Tris·HCl (pH 7.9), 100 mM NaCl, and 1 mM DTT, contained BglI (0.5 nM) and pBGL1 (5 nM), with one of the following salts at the concentration indicated on the x-axis: , MgCl2; , CoCl2; , MnCl2; , CaCl2. Samples were withdrawn from the reactions at varied times and analyzed as in Fig. 2a. Values for v1 were measured from the initial linear decline in the concentration of supercoiled pBGL1 with time.

Other Restriction Enzymes with Discontinuous Sites-- It has been suggested that the unusual reaction mechanism of SfiI is due to its elongated recognition site, which covers 13 bp of DNA (38). If so, then even though BglI acts independently at each copy of its recognition site in the orthodox manner for a type II restriction enzyme, other enzymes with elongated sites may act like SfiI. As noted above, a comparison of the reaction kinetics of a restriction enzyme on substrates with either one or two recognition sites provides a diagnostic test for concerted action at two recognition sites (11). This test was applied to several type II restriction enzymes with discontinuous recognition sequences. The enzymes selected for this analysis were chosen on the basis of similarities in their recognition sequences (Table I) and, in some instances, on account of unusual features in their genetic organization. The recognition sites for Tth111I, PshAI, AhdI, and DrdI all contain the same specified sequence, GAC···GTC, but the length of the spacer is different in each case. Likewise, the recognition sites for PflMI, BstXI, and XcmI are identical to each other except for the length of the spacer (Table I). If the mode of action of a restriction enzyme is determined by the length of the recognition site, then the enzymes with sites containing relatively short interruptions might be expected to act in the orthodox manner, whereas the enzymes recognizing extended sites covering 13 or more bp might act in unorthodox ways. The latter applies particularly to XcmI, whose recognition site contains a 9-bp interruption amid 6 specified bp and thus has an overall length of 15 bp, the longest known for a type II restriction enzyme (2). Moreover, three of the endonucleases tested here, PshAI, AhdI, and XcmI, are from restriction-modification systems with genetic organizations midway between the type I and the type II systems (13) and have been termed type 11/2 systems.²

The recognition sites for Tth111I, PshAI, and AhdI occur once in pBR322 (Fig. 1b), and this plasmid was used as the one-site substrate for these three enzymes. A two-site substrate for these enzymes, pML1, was constructed by cloning into pBR322 a duplex carrying each of these sites (Fig. 1b); in all three cases, the second copy had the same flanking and intervening sequences as the site present in pBR322. Similarly, two plasmids were constructed with either one or two copies of the recognition sequences for PflMI, BstXI and XcmI, pAB2 and pAB3, respectively, again with identical flanking and spacer sequences at each copy (Fig. 1c). For DrdI, pAT153 (46) and pUC19 (Fig. 1a) were used as the one-site and two-site substrates, respectively; the single DrdI site in pAT153 is identical to one of the DrdI sites in pUC19 but the other DrdI site in pUC19 has different flanking and spacer sequences. Because DrdI was the only enzyme of the seven examined here to be tested on substrates with nonidentical recognition sites, the data with DrdI are described separately from the other six enzymes.

The rates at which the six enzymes cleaved DNA/unit of enzyme varied considerably, but in all six cases, the rate for the utilization of the one-site substrate was virtually the same as that for the two-site substrate (Table II). Despite each having a discontinuous recognition site, none of these enzymes display the hallmark property of SfiI, faster cleavage of a two-site substrate than a one-site substrate (28). Five out of the six enzymes, PshAI, AhdI, PflMI, BstXI, and XcmI, cleaved their two-site substrates in sequential steps, with equal kinetics for the two steps; first at one site to give FLL DNA and then at the other site to give, after an initial lag phase, the two final products, L1 and L2 (representative data for XcmI shown in Fig. 7a, other data not shown). However, the reaction of Tth111I on its two-site substrate yielded less of the FLL DNA than expected for a sequential pathway with two kinetically equal steps; instead, the final products cut at both sites were generated directly from the start of the reaction without a lag phase (Fig. 7b). For both XcmI and Tth111I, the rates for converting the SC two-site substrate to FLL DNA (v_2A) were evaluated relative to those for the conversion of FLL DNA to L1 and L2 (v_2B), from the rise and fall in the concentration of FLL DNA during these reactions. For XcmI (Fig. 7a), the best fit was with a ratio of v_2B:v_2A at 0.9; for Tth111I (Fig. 7b), v_2B exceeded v_2A by a factor of 3. 

View larger version (20K): [in this window] [in a new window]   Fig. 7.   XcmI and Tth111I on two-site plasmids. a, the reaction at 37 °C contained 4 units/ml XcmI and 5 nM pAB3 in 10 mM Tris·HCl (pH 7.9), 50 mM NaCl, 10 mM MgCl2, 1 mM DTT, and 100 µg/ml bovine serum albumin. b, the reaction at 65 °C contained 12 units/ml Tth111I and 5 nM pUC19 in 20 mM Tris acetate (pH 7.9), 50 mM potassium acetate, 10 mM magnesium acetate, 1 mM DTT, and 100 µg/ml bovine serum albumin. Samples withdrawn from the reactions at various times were analyzed as in Fig. 2 to determine the concentrations of the following forms of the DNA: , supercoiled substrate (marked SC on both graphs); , open-circle DNA (marked OC); , full-length linear DNA (marked FLL); , total DNA in the two final products after cutting both sites (marked L1/2).

As with BglI (Fig. 2), PshAI, AhdI, PflMI, BstXI, and XcmI all catalyze independent reactions at each copy of their respective recognition sites, in the conventional manner for type II enzymes. For Tth111I, the diminished yield of FLL DNA during its reaction on a two-site substrate (Fig. 7b) could be due to a concerted reaction at two recognition sites. If so, then Tth111I must be able to bridge two sites on different DNA molecules as readily as two sites on the same DNA, because it cleaved its one-site and two-site substrates at the same rate (Table II). Even so, this is a possibility. Though SfiI and SgrAI act concertedly at two sites, the difference between their reaction rates on two-site and one-site substrates varies with ionic strength, and in the absence of added salts, they cleave one-site substrates almost as fast as two-site substrates (11, 29). Alternatively, the lack of FLL DNA from the Tth111I reaction on its two-site substrate could be due to a processive mechanism in which the enzyme first cleaves one site and then translocates to the second site without departing from the DNA. However, an enzyme that acts processively at low ionic strength is likely to act distributively at high ionic strength (10, 12). To distinguish these alternatives, the concentration of potassium acetate in the reaction buffer for Tth111I was raised from 50 mM to 200 mM. In contrast to SfiI at elevated ionic strength (29), Tth111I still cleaved the one-site substrate at the same rate as the two-site substrate, but its reaction on the two-site DNA now yielded, in successive stages, OC DNA cut in one strand, then FLL DNA cut in both strands at one site and finally the end products cut in both strands at both sites (data not shown). Tth111I thus shows a high degree of processivity at low ionic strength but, at high ionic strength, it cleaves each phosphodiester bond in a separate DNA-binding event.

In contrast, DrdI cleaved its two-site substrate, pUC19, faster than its one-site substrate, pAT153 (Table II). However, when DrdI was tested against a linear form of pUC19, as described above with BglI (Fig. 3), almost all of the DNA was cleaved first at the DrdI site that is distinct from the single site present in pAT153, and the DrdI site on pUC19 that is identical to the site on pAT153 was cleaved later on in a separate and much slower reaction (data not shown). Hence, the reason why DrdI cleaves pUC19 faster than pAT153 is not because its needs two recognition sites for its DNA cleavage reaction but rather because it has a marked preference for a particular site that is present on pUC19 but is absent from pAT153. The preference presumably arises from the flanking and/or spacer sequences at that site.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The recognition sites for many type II restriction endonucleases are discontinuous sequences, containing two symmetrically equivalent sets of specified bp separated by a defined length of nonspecific DNA (2; Table I). The enzymes that act at such sites must differ from the enzymes that recognize continuous sequences, such as BamHI and EcoRV, at least in the positioning of their DNA recognition functions. A crystal structure is currently available for only one type II enzyme with a discontinuous site, BglI (41). The recognition/catalytic functions of BglI are similar to those in EcoRV, an enzyme recognizing a continuous sequence, but are arranged differently because of a different subunit interface. However, very little was known about the mode of action of BglI in solution. The only known mechanism for a type II restriction enzyme with a discontinuous site had been that for SfiI (28). SfiI is a tetrameric protein that interacts with two recognition sites before being cleaving DNA (29-33) and thus differs radically from the conventional restriction enzymes such as BamHI and EcoRV, dimeric proteins that act at individual sites (1, 14).

It therefore seemed possible that BglI might behave similarly to SfiI, particularly given the precedence of Cfr10I and SgrAI, two enzymes that recognize sites that are related to each other in the same way as the BglI and SfiI sites and that both need two copies of their respective sites for their DNA cleavage reactions (11, 38). However, the experiments on BglI described here reveal no sign of the concerted mode of action of SfiI at two recognition sites. Unlike SfiI (29, 31), BglI does not cleave DNA with two sites faster than DNA with one site, nor does it convert DNA with two sites directly to the final products cut at both sites (Figs. 2 and 3). Instead, in accord with its crystal structure (41), the properties of BglI in solution are very similar to those of EcoRV.

Each turnover of BglI results in the cleavage of both strands of the DNA at an individual recognition site, without the liberation of any of the nicked OC form of the plasmid (Fig. 2). Hence, as with EcoRI (10) and EcoRV (9), the DNA cleavage steps in the reaction pathway of BglI must be faster than the subsequent dissociation of the cleaved DNA, and the latter is presumably rate-limiting for its turnover on plasmid substrates. However, individual recognition sites for BglI display varying susceptibilities to the enzyme; the initial reaction of BglI on pUC19 occurred mainly at site A rather than site B (Fig. 3). Surprisingly, the steady-state rate for the utilization of the one-site substrate, pBGL1 (Fig. 1a), is almost the same as that for pUC19, even though pBGL1 carries site B from pUC19. The selection of site A over B on pUC19 must therefore be due to a reduced K_m rather than an enhanced V_max.

Repressor proteins that recognize discontinuous sequences require the two half-sites to be oriented in a particular manner and thus often bind most strongly to sites where the intervening DNA is intrinsically twisted or bent as appropriate (51, 52). The same should apply here, because DNA bound to BglI is bent through ~20° and is undertwisted in the spacer region (41). The sequences at the BglI sites in pUC19 are as follows: Site A, TTCGCCATTCAGGCTGC and Site B, CCAGCCGGAAGGGCCGA (underlined bases denote the specified sequence). The spacer at the recalcitrant site, B, has only purines in one strand and only pyrimidines in the other and is thus likely to possess a less flexible structure than the mixed purine/pyrimidine sequence in the spacer at site A (53). Hence, one explanation for the elevated K_m at site B is that more energy is required to convert site B into the requisite structure for DNA cleavage than is the case at site A. The different flanking sequences either side of the 11-bp segments may also contribute to this effect.

The BglI endonuclease binds to DNA (Fig. 4) in a manner that is essentially identical to EcoRV (17, 23) and to some other enzymes such as TaqI (8, 18), but which is distinct from EcoRI (12). In the absence of divalent metal ions, BglI shows no detectable preference for its recognition site over the sum of the nonspecific sequences on the DNA fragments used here. In the presence of Ca²⁺, it has a marked preference for its recognition site over the remainder of the DNA. BglI and EcoRV are thus very similar to each other in terms of both DNA-binding behavior and three-dimensional structures. BamHI and EcoRI are also very similar to each other in both their DNA-binding properties and their three-dimensional structures but differ from EcoRV in both respects. This supports the proposal, first made by Barany and colleagues (22), that EcoRV and EcoRI represent distinct groups of type II restriction enzymes (10, 14, 20), in contrast to the view that no clear-cut distinction can be made between the groups (6, 19, 24-26).

The affinity of BglI for its recognition site in the presence of Ca²⁺ is, however, much higher than that of EcoRV for its site. Because 50 pM concentrations of both BglI and DNA resulted in the protein binding to all of the DNA with the BglI site (Fig. 4c), the K_D of BglI for its recognition site in 200 mM NaCl must be 50 pM. In contrast, the K_D of EcoRV for its recognition site in a buffer containing Ca²⁺ and 100 mM NaCl is ~200 pM (23). The difference in K_D values is probably not due to the overall length of the recognition site for BglI, 11 bp, compared with that for EcoRV, 6 bp. Though BglI interacts with the sugar-phosphate backbone of the DNA in both the 5-bp spacer and in the 2 bp of flanking DNA either side of the site, EcoRV makes more backbone contacts with the flanking DNA than BglI (41, 42). The total length of DNA contacted by EcoRV is similar to that for BglI, 14 versus 15 bp. On the other hand, BglI makes a larger number of direct interactions with the bases in its target sequence, contacting all 12 bases in its 6-bp target, than is the case with EcoRV, where only 8 out of the 12 bases are contacted directly. Alternatively, the relative affinities of BglI and EcoRV for their respective recognition sites may be due to the BglI site undergoing a modest deformation from B-form DNA on binding to the protein (41), whereas the EcoRV site undergoes a radical distortion (42). Hence, the fraction of the intrinsic binding energy used to deform the DNA may be larger with EcoRV than with BglI. The affinity of BglI for nonspecific DNA in both the absence and presence of Ca²⁺ is also higher than that of EcoRV. This might account for why Ca²⁺ ions fail to suppress BglI binding to nonspecific DNA (Fig. 4d), even though it suppresses nonspecific binding by EcoRV (23).

BglI responds to alternative metal ions as substitutes for Mg²⁺ (Figs. 5 and 6) identically to EcoRV (5, 49, 50). Hence, the roles of the metal ions in DNA recognition and catalysis by BglI are likely to be the same as in EcoRV. The crystal structures of EcoRV and BglI bound to their respective recognition sites in the presence of divalent metal ions both reveal two metal ions adjacent to the scissile phosphodiester bond at each active site but not in identical positions (16, 41). In BglI, the metal ions are aligned parallel to the axis of the DNA, as in a general mechanism for phosphodiester hydrolysis with two metal ions (54), whereas the metal ions in EcoRV are aligned more or less perpendicular to the axis of the DNA. However, neither structure denotes the active complex. The structure for BglI used Ca²⁺ as a noncatalytic cofactor. Because Ca²⁺ acts as a competitive inhibitor of Mg²⁺ in DNA cleavage by BglI, it is likely to occupy the same sites as Mg²⁺ but in a distorted geometry, due to its larger ionic radius. The structures for EcoRV used Mg²⁺, Mn²⁺, Co²⁺, and Ca²⁺, but even when catalytically competent metals were soaked into the DNA-protein crystals, no DNA cleavage ensued. Moreover, in a structure of the enzyme-product complex for EcoRV generated by crystallizing the complex after DNA cleavage in solution, the two metal ions and the phosphate from the scissile bond are positioned differently from their locations in the enzyme-substrate complex (16). Catalysis by EcoRV may take place within a structure more like the crystal structure of the enzyme-product complex than the crystal structure of the enzyme-substrate complex (5, 55, 56). If this scheme for phosphodiester hydrolysis by EcoRV is to be applied to BglI, the active site in BglI would also have to be reorganized prior to catalysis.

In addition to BglI, seven other restriction enzymes with discontinuous recognition sites were tested against plasmids with either one or two copies of their sites (Table II). None of the enzymes needed two copies of their recognition sites for their DNA cleavage reactions. Hence, enzymes with unusual recognition sites, in which the specified sequence is interrupted by a defined length of nonspecific DNA, do not necessarily follow unusual reaction mechanisms, even when the site is elongated to the extent of the XcmI site. For six of these enzymes, PshAI, AhdI, DrdI, PflMI, BstXI, and XcmI, each recognition site was cleaved in a separate reaction in the orthodox fashion for type II restriction enzymes. By analogy with the crystal structure of BglI (41), perhaps restriction enzymes whose target sites differ only in the length of the spacer DNA, such as PflMI, BstXI, and XcmI (Table I), possess common structures for DNA recognition but different dimer interfaces, so that the recognition elements are separated in each case by the appropriate distance. The seventh, Tth111I, utilized a substrate with one site at the same rate as the two-site substrate, but it acted processively on the latter and generally cleaved both sites before departing from the DNA, at least at the ionic strength of its standard reaction buffer (Fig. 7b). In a survey of restriction enzymes with 8-bp recognition sites (11), AscI was the only one out of 12 tested, across sites separated by ~700 bp, that acted in a processive manner. The processivity of Tth111I is, however, remarkable because the two Tth111I sites on its two-site substrate are separated by over 2100 bp, almost as far as is possible given the circumference of pBR322 (Fig. 1b). The ability of Tth111I to translocate over considerable distances of DNA might be due to the fact that it is assayed at 65 °C but comes from an organism that is grown at 75 °C (57). At temperatures below its optimum for DNA cleavage, SfiI dissociates from the cleaved DNA at extremely slow rates (31). Hence, Tth111I may also dissociate from DNA very slowly at its subphysiological temperature of 65 °C.

The enzymes examined here represent only a small fraction of the type II restriction enzymes with discontinuous sites that have been identified to date (2). Hence, even though concerted action at two sites is clearly not obligatory for such enzymes, other type II enzymes with discontinuous sites may still behave like SfiI. Indeed, other sorts of restriction-modification systems encode endonucleases that utilize two copies of a discontinuous site. For example, BcgI typifies a group of restriction enzymes that cleave DNA on both sides of a discontinuous recognition site (58). Efficient cleavage by BcgI requires two such sites on the same DNA (59). The recognition sites for the type I restriction-modification systems are also bipartite sequences (60). The type I endonucleases cleave DNA at loci distant from their recognition sites, but although the tracking from a single site on a circular DNA can result in DNA cleavage, virtually no cleavage occurs on a linear DNA with a single site (61). Three of the endonucleases tested here, PshAI, AhdI, and XcmI, are from type 11/2 restriction-modification systems.² The modification genes of these systems are like those from type I systems, whereas their restriction genes are like those from type II systems. In all three cases, the endonucleases from the type 11/2 systems acted at individual sites in independent reactions, in exactly the same manner as the conventional type II restriction enzymes such as BglI and EcoRV.

	ACKNOWLEDGEMENTS

We thank Michael Livingstone, Susan Milsom, Matthew Newman, Ira Schildkraut, Mark Watson, and Geoff Wilson for aid and advice.

	FOOTNOTES

^* This work was supported by Grants 7/G10881 and 046178 from the Biotechnology and Biological Sciences Research Council and the Wellcome Trust.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed. Tel.: 44-117-928-7429; Fax: 44-117-928-8274; E-mail: s.halford@bris.ac.uk.

² G. Wilson, personal communication.

	ABBREVIATIONS

The abbreviations used are: bp, base pair(s); DTT, dithiothreitol; SC, supercoiled; OC, open circle; FLL, full-length linear; L1 and L2, linear DNA fragments from the cleavage of a circular DNA with two sites at both sites.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				S. R. Coffin and N. O. Reich Escherichia coli DNA Adenine Methyltransferase: THE STRUCTURAL BASIS OF PROCESSIVE CATALYSIS AND INDIRECT READ-OUT J. Biol. Chem., July 3, 2009; 284(27): 18390 - 18400. [Abstract] [Full Text] [PDF]

				T.-J. Su, M. R. Tock, S. U. Egelhaaf, W. C. K. Poon, and D. T. F. Dryden DNA bending by M.EcoKI methyltransferase is coupled to nucleotide flipping Nucleic Acids Res., June 7, 2005; 33(10): 3235 - 3244. [Abstract] [Full Text] [PDF]

				S. Chandrashekaran, M. Saravanan, D. R. Radha, and V. Nagaraja Ca2+-mediated Site-specific DNA Cleavage and Suppression of Promiscuous Activity of KpnI Restriction Endonuclease J. Biol. Chem., November 26, 2004; 279(48): 49736 - 49740. [Abstract] [Full Text] [PDF]

				D. M. Gowers, S. R.W. Bellamy, and S. E. Halford One recognition sequence, seven restriction enzymes, five reaction mechanisms Nucleic Acids Res., June 29, 2004; 32(11): 3469 - 3479. [Abstract] [Full Text] [PDF]

				A. J. Bath, S. E. Milsom, N. A. Gormley, and S. E. Halford Many Type IIs Restriction Endonucleases Interact with Two Recognition Sites before Cleaving DNA J. Biol. Chem., February 1, 2002; 277(6): 4024 - 4033. [Abstract] [Full Text] [PDF]

				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]

				S. A. Williams and S. E. Halford SfiI endonuclease activity is strongly influenced by the non-specific sequence in the middle of its recognition site Nucleic Acids Res., April 1, 2001; 29(7): 1476 - 1483. [Abstract] [Full Text] [PDF]

				M. Soundararajan, Z. Chang, R. D. Morgan, P. Heslop, and B. A. Connolly DNA Binding and Recognition by the IIs Restriction Endonuclease MboII J. Biol. Chem., January 4, 2002; 277(2): 887 - 895. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted 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 Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gormley, N. A. Articles by Halford, S. E. Search for Related Content PubMed PubMed Citation Articles by Gormley, N. A. Articles by Halford, S. E. Social Bookmarking                      What's this?