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

J. Biol. Chem., Vol. 277, Issue 6, 4034-4041, February 8, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/6/4034    most recent M108442200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager 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?

The Type IIs Restriction Endonuclease BspMI Is a Tetramer That Acts Concertedly at Two Copies of an Asymmetric DNA Sequence^*

Niall A. Gormley, Anna L. Hillberg, and Stephen E. Halford

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

Received for publication, August 31, 2001, and in revised form, November 26, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Type IIs endonucleases recognize asymmetric DNA sequences and cleave both strands at fixed positions downstream of the sequence. Many type IIs enzymes, including BspMI, cleave substrates with two sites more rapidly than those with one site. They usually act sequentially on DNA with two sites, but BspMI converted such a substrate directly to the final products cut at both sites. The BspMI endonuclease was found to be a tetramer, in contrast to the monomeric structures for many type IIs enzymes. No change in subunit association occurred during the BspMI reaction. Plasmids with two BspMI sites were cleaved in cis, in reactions spanning sites in the same DNA, even when the sites were separated by just 38 bp. Plasmids with one BspMI site were cleaved in trans, with the enzyme bridging sites in separate DNA molecules: these slow reactions could be accelerated by adding a second DNA with the recognition sequence. Thus, whereas many type IIs enzymes dimerize before cleaving DNA, a process facilitated by two recognition sites in cis, the BspMI tetramer binds two copies of its recognition sequence before cleaving the DNA in both strands at both sites.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The endonucleases from type II restriction-modification systems recognize short DNA sequences, 4-8-bp long, and cleave both strands of the DNA at fixed locations (1). Their recognition sites are often palindromic sequences that are contacted symmetrically by a dimeric protein, with the result that both strands are cleaved at equivalent positions (2). However, the endonucleases from the type IIs systems (3) recognize asymmetric sequences and cleave the DNA in both strands several bp downstream of the site, but not at equivalent positions (4). An asymmetric sequence cannot be recognized symmetrically by a homodimeric protein, and the type IIs endonucleases are generally monomers with separate domains for DNA recognition and catalysis, at least in the case of the best-known type IIs enzyme, FokI (5, 6). The catalytic domain of FokI has the functions for cleaving one DNA strand; to cut both strands at one site, two monomers associate to form a dimer (7-9).

In the preceding study (10), FokI and several other type IIs enzymes were found to cleave DNA substrates with two recognition sites more rapidly than those with one site, in a manner indicative of DNA looping (2, 11), although they cleaved each site in their two-site substrates in separate reactions. The formation of an active dimer at one DNA site thus seems to be facilitated by a second site in cis, in the same molecule of DNA, presumably because of the effect that such a site must have on the equilibrium constant for the monomer-dimer association (12). The reaction profiles of these type IIs enzymes on two-site substrates are similar to those of the type IIe endonucleases, such as EcoRII and NaeI, that are activated to cleave one recognition site by a second copy of the same sequence (13). However, the preceding study (10) also identified one type IIs endonuclease, BspMI, that acts in a radically different manner from FokI.

As with many other type IIs systems (14), the BspMI restriction-modification system consists of three proteins: a single endonuclease of 492 amino acids, and two methyltransferases that each modify one strand of the asymmetric recognition site.¹ Its recognition sequence is ACCTGC, and the restriction enzyme cuts both strands of the DNA downstream of this site, between the fourth and fifth bases in the strand shown and between the eight and ninth bases in the complementary strand (4). It had been noted previously that the BspMI endonuclease is recalcitrant to cleave its single recognition site in pBR322 but is activated to cleave this site by DNA molecules with multiple BspMI sites (15). When tested on plasmids with one or two BspMI sites, it cleaved the two-site substrate more rapidly than that with one site (10). Moreover, during its reaction on the two-site plasmid, only a small fraction of the DNA was liberated from the enzyme after cutting just one site. Instead, BspMI cleaved almost all of the two-site substrate directly to the final products cut at four phosphodiester bonds, two at each site. BspMI thus seems to act like the tetrameric type IIf endonucleases, such as SfiI, Cfr10I, and NgoMIV, that interact concurrently with two DNA sites and cut both strands at both sites before dissociating from the DNA (13, 16-18). This study reports on the subunit structure of BspMI and its reactions with DNA carrying one or two BspMI sites.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Proteins-- A strain of Escherichia coli that overproduces the BspMI endonuclease was a gift from I. Schildkraut (New England Biolabs). The strain was grown at 37 °C in 2× YT broth (19) to an A₆₀₀ of ~0.8 before the addition of isopropyl--D-thiogalactopyranoside to a final concentration of 0.5 mM. The culture was continued for 5 h before harvesting the cells by centrifugation. All subsequent steps were performed at 4 °C. The cells were resuspended in Buffer P (20 mM Tris-HCl (pH 8.0), 1 mM EDTA, 1 mM DTT,² and 10% (v/v) glycerol) supplemented with 300 mM NaCl, 50 µM phenylmethylsulfonyl fluoride, and 100 µM benzamidine. The resuspended cells were disrupted by sonication, and the debris was removed by centrifugation. The supernatant was diluted with 2 volumes of Buffer P and applied to a phosphocellulose column (P11; Whatman) that had been prepared as described by Greene et al. (20) and pre-equilibrated with Buffer P containing 100 mM NaCl. The column was developed with a linear gradient from 100 to 500 mM NaCl in Buffer P, and the fractions were assayed for BspMI activity and by SDS-PAGE. The BspMI endonuclease eluted at ~230 mM NaCl. The peak fractions were mixed with a slurry of hydroxylapatite (BioGel HTP; Bio-Rad) that had been equilibrated in Buffer P with 230 mM NaCl. After gentle agitation for 1 h, the hydroxylapatite resin was allowed to settle, and the supernatant was recovered by low-speed centrifugation. The supernatant contained the majority of the BspMI that had been added to the slurry, whereas most of the impurities were removed, presumably by adsorption onto the hydroxylapatite. This left preparations in which 95% of the protein was BspMI endonuclease, as judged by SDS-PAGE. The preparations were stored at 20 °C. BspMI concentrations were determined by absorbance at 280 nm, using an extinction coefficient of 76,810 M¹ cm¹ (per subunit) calculated from its amino acid composition, and are cited in terms of the tetrameric form of the protein. All other enzymes used for manipulating DNA were purchased from commercial suppliers and used as recommended by the supplier.

Ultracentrifugation-- Samples (110 µl) of BspMI were examined by centrifugation to equilibrium at 20 °C in 3-mm columns in a Beckman XLA analytical ultracentrifuge. After centrifugation for the requisite times, typically 16 h and 20 h at 8000 rpm followed by the same time intervals at 11,000 rpm, the radial distribution of the protein was measured by absorption at 230 nm (for samples of 0.2 mg/ml) or 280 nm (for samples of >0.1 mg/ml). The distributions recorded 4 h apart duplicated each other, thus confirming that the sedimentation was at equilibrium. Another absorption record was taken after extending the centrifugation for 6 h at 40,000 rpm. The absorption at the meniscus in the latter record was used as a fixed offset in the subsequent analysis with the ORIGIN software from Beckman (21). The analysis fitted by nonlinear regression either single or multiple absorption records to obtain the best fit for molecular mass (M) in the equation

(Eq. 1)

where A_r and A₀ are the absorbancies at radius r and at the reference radius r₀,  is the angular velocity, R is the gas constant, T is the temperature, is the partial specific volume (0.735 ml/g for BspMI),  is the buffer density (1.039 g/ml for the buffer used here), and B is the baseline offset (determined as described above). Partial specific volumes and buffer densities were calculated with SEDNTERP (Ref. 22; alpha.bbri.org/rasmb/spin/ms_dos/sednterp-philo/).

DNA-- The plasmids pAT153 and pNAG1 are described elsewhere (10). Further derivatives of pNAG1 with insertions or deletions of nonspecific sequences between the two BspMI recognition sites were generated by standard procedures (19): each derivative retained from pNAG1 the head-to-tail orientation of the two sites. The plasmids were ³H-labeled and purified as recorded previously (11). The preparations contained largely the supercoiled form of the monomeric plasmid, with typically 10% dimeric plasmid and/or nicked open-circle DNA. Complementary pairs of high pressure liquid chromatography-purified oligodeoxyribonucleotides were purchased from MWG Biotech Ltd. (Milton Keynes, United Kingdom) and annealed to form duplexes by heating to 95 °C before slow cooling overnight.

Reactions-- The reactions were generally carried out in 200-µl volumes at 37 °C. An aliquot (typically 20 µl) of the BspMI endonuclease, diluted to the requisite concentration in the appropriate dilution buffer (10), was added to the ³H-labeled plasmid (usually 5 or 10 nM) in 20 mM HEPES (pH 8.0), 1 mM DTT, 10 mM MgCl₂ and the concentration of NaCl as specified for each experiment (usually 40 or 80 mM). At various times after adding the enzyme, 15-µl samples were removed from the reaction, mixed immediately with 10 µl of stop-mix, and analyzed as described previously (10).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Analytical Ultracentrifugation-- The DNA sequence of the gene for the BspMI endonuclease predicts a molecular weight (M_r) of 55,435 for the monomer.¹ Several type IIs endonucleases have been shown by gel filtration to exist in solution as monomers (3), but the elution volume of protein from size-exclusion columns depends not only on the mass of the protein but also on its shape (23). Hence, to determine the oligomerization state of BspMI in solution, the M_r of the enzyme was evaluated by sedimentation equilibrium (Fig. 1), a procedure that is unaffected by the shape of the protein (23). In the low-ionic-strength buffers used for the DNA cleavage reactions described below, the BspMI protein was insufficiently soluble to permit its sedimentation to be monitored by uv absorption: even at the minimal concentrations needed in the analytical ultracentrifuge, some of the protein precipitated from solution. However, the enzyme remained in solution in the buffer from the final stage of the enzyme preparation, and this buffer, of relatively high ionic strength, was used for the sedimentation studies.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Analytical ultracentrifugation. The data points () in the bottom panel show the absorption at 230 nm as a function of the radius of centrifugation, following the centrifugation of a sample of BspMI endonuclease (0.15 mg/ml) in Buffer P with 230 mM NaCl for 20 h at 8000 rpm. The line drawn through the data points corresponds to a global fit to a single value for Mr, using both the data shown and additional data sets recorded at other protein concentrations and at other centrifugal speeds: the best fit was with Mr 215,000. The top panel shows the residuals between the data shown and the globally fitted curve.

Sedimentation equilibrium experiments were carried out with a range of concentrations of BspMI (0.02-0.5 mg/ml) and at various rotor speeds (8000-12,000 rpm). In each individual experiment at a particular protein concentration and rotor speed, the radial distribution of the protein at sedimentation equilibrium matched that expected for a single monodisperse species: the fits to this model (Eq. 1) yielded random variations in the residuals between the experimental data and the theoretical curves (Fig. 1, top panel). The M_r values from the individual experiments all fell within the range of 196,000-225,000. No systematic variation in the M_r values was observed at different protein concentrations, nor was any significant improvement obtained by fitting the data to a model for a self-associating system. When the data at each protein concentration were simultaneously fitted to Eq. 1, the global fit yielded a M_r of 215,000 (Fig. 1). This value lies close to that expected for four subunits of the BspMI protein, 221,740. Thus, in contrast to the monomeric structures of other type IIs enzymes, BspMI is a tetramer.

Steady-state Kinetics-- The reactions of BspMI on one-site and two-site substrates were analyzed by steady-state kinetics (Fig. 2). The substrates were the plasmids pAT153, which has a single site for BspMI, and pNAG1, a derivative of pAT153 with a second site for BspMI in directly repeated orientation 700-bp distant from the original site (10). The second site in pNAG1 ought to be equivalent to the site in pAT153 because it is flanked by the same sequences for 15 bp downstream of the recognition sequence, so as to encompass the sites of cleavage 4 and 8 bp away from the recognition site, and also for 9 bp upstream of the sequence. The progress of the reactions was monitored by using ³H-labeled DNA: samples withdrawn from the reactions at various times were analyzed by electrophoresis through agarose to separate the supercoiled substrate and the cleaved products, and the concentrations of each form of the DNA were measured by scintillation counting (11). The reaction velocities were determined from the initial linear decline in the concentration of supercoiled substrate with time rather than from the appearance of the reaction products because the latter include both the DNA cut at one site and the DNA cut at two sites.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Steady-state kinetics on plasmids with one or two BspMI sites. The reactions at 37 °C in 20 mM HEPES (pH 8.0), 10 mM MgCl2, 1 mM DTT, and 80 mM NaCl contained BspMI endonuclease and one of the following plasmids (>90% supercoiled): pAT153 (one BspMI site,  in both a and b) or pNAG1 (two BspMI sites,  in both a and b). The reactions in a were with 0.5 nM BspMI and varied substrate concentrations (indicated on the x axis). The reactions in b were with 10 nM substrate and varied BspMI concentrations (indicated on the x axis). Samples were removed from the reaction at timed intervals and analyzed as described under "Experimental Procedures" to determine the concentrations of the substrate and each of the reaction products. The reaction rates shown on the ordinates were evaluated from the initial linear decline in the concentration of the supercoiled substrate with time. In a, these are plotted on a discontinuous scale; in b, the rates on pNAG1 (two sites, ) and pAT153 (one site, ) are plotted on the left and right ordinates, respectively.

In one set of experiments, the concentrations of the substrates were varied from 2.5 to 20 nM while the enzyme concentration was held constant at a level below the lowest substrate concentration (Fig. 2a). The objective of these experiments was to determine the V_max and the K_1/2 values of the one-site and two-site substrates for BspMI (where K_1/2 denotes the substrate concentration that yields half the V_max value, regardless of whether the reaction velocities vary with substrate concentrations in hyperbolic or sigmoidal fashions). The reaction velocities on pNAG1 were invariant across this range of substrate concentrations (Fig. 2a). The K_1/2 for the two-site substrate must therefore be < 2.5 nM, the lowest concentration that could be tested, and the uniform reaction velocity observed across this range of concentrations corresponds to the V_max value, ~0.5 nM DNA consumed/nM enzyme tetramer/min. In contrast, the reaction velocities on pAT153 increased more or less linearly with increasing concentrations of this substrate (Fig. 2a). The K_1/2 for the one-site substrate must therefore be > 20 nM, the highest concentration tested, and the reaction velocity observed at 20 nM pAT153, 0.1 nM DNA consumed/nM enzyme tetramer/min, must be well below its V_max value. (If BspMI cleaves a one-site DNA only after interacting with two molecules of the DNA, as indicated below, its reaction velocities on the one-site DNA should show a sigmoidal dependence on the substrate concentration (24). However, the characteristic S shape of a sigmoidal plot becomes apparent only at substrate concentrations approaching the K_1/2 value.)

With another type IIs enzyme, FokI, the steady-state velocities on DNA substrates with either one or two FokI sites increase disproportionately with the enzyme concentration; a 2-fold increase in the FokI concentration causes a >2-fold increase in reaction rate (8, 10). To examine whether this is a general property of the type IIs enzymes, an additional set of steady-state experiments was conducted with varied concentrations of BspMI and fixed concentrations of either the one-site or the two-site substrate (Fig. 2b). With both the one- and two-site plasmids, the reaction velocities increased in direct proportion to the increases in the concentrations of BspMI. Thus, in contrast to FokI (8, 9), it is highly unlikely that the reaction mechanism of BspMI on either one- or two-site substrates involves an association of protein subunits during the catalytic cycle.

Pre-steady-state Kinetics-- Many enzymes exhibit a pre-steady-state "burst" phase of product formation, with an amplitude proportional to the enzyme concentration, due to the rapid formation of an enzyme-product complex followed by the slow dissociation of the product from the enzyme (25). The rate-limiting step in the turnover of many restriction enzymes is product dissociation (26-29). Hence, if product dissociation is rate-limiting for BspMI, the higher BspMI concentrations used for the reactions in Fig. 2b should have given a burst phase: for example, at 4 nM BspMI and 10 nM DNA, 40% of the DNA could have been consumed in a rapid exponential phase and 60% of the DNA could have been consumed in the zero-order steady-state phase. However, a burst phase of product formation was not detected in any of the reactions described above (Fig. 2). Instead, the amount of product formed during these reactions increased with time as a linear zero-order process from the start of the reaction, which could be extrapolated back to zero product at zero time (data not shown).

The individual DNA cleavage steps during the catalytic cycle of a restriction enzyme can be monitored in single-turnover reactions, with the enzyme in excess of the DNA. For example, single turnovers of the SfiI endonuclease on a plasmid with two SfiI sites revealed a sequential series of transient intermediates due to the cleavage of the enzyme-bound DNA in 1  4 phosphodiester bonds, all at the same intrinsic rate, before the release of the DNA cut in both strands at both sites (28). Single-turnover reactions of BspMI were conducted with a 2-fold molar excess of enzyme over pNAG1: higher enzyme concentrations resulted in diminished rates (data not show), presumably due to the binding of BspMI molecules to each site rather than the binding of one molecule to both sites (30, 31). Whereas single turnovers of most restriction enzymes can be monitored only with rapid reaction instrumentation (26-29), those with BspMI were sufficiently slow to monitor by hand-mixing the reagents (Fig. 3).

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Single-turnover kinetics on a plasmid with two BspMI sites. A reaction with 10 nM pNAG1 (85% supercoiled) at 37 °C in 20 mM HEPES (pH 8.0), 10 mM MgCl2, 1 mM DTT, and 80 mM NaCl was initiated by adding BspMI to a final concentration of 20 nM. At timed intervals after the addition of the enzyme, samples were removed from the reaction and analyzed to determine the concentrations of the following forms of the DNA: , supercoiled substrate; , nicked open-circle DNA; , full-length linear DNA cut in both strands at one site; , total DNA in the two final products after cutting both sites.

The reactions were initiated by either adding the enzyme to a solution containing pNAG1 and MgCl₂ (Fig. 3) or adding MgCl₂ to a mixture of the enzyme and pNAG1. No significant difference was observed between the two procedures (data not shown). During these reactions, all of the supercoiled DNA was utilized in a single-exponential process (Fig. 3), with a first-order rate constant of 0.47 min¹. This rate constant is similar to the steady-state turnover rate of ~0.5 nM DNA consumed/nM enzyme tetramer/min for BspMI on pNAG1 under the same reaction conditions (Fig. 2). Hence, both the rate of substrate utilization in the single-turnover reactions and the absence of a pre-steady-state burst phase indicate that the rate-limiting step in the catalytic cycle of BspMI occurs at or before the cleavage of the first phosphodiester bond in the substrate. Furthermore, the rate-limiting step is unlikely to be the initial binding of the enzyme to the DNA: if this had been the case, the single turnovers initiated by adding MgCl₂ to the enzyme-DNA mixture would have given faster rates of substrate utilization than those initiated by adding the enzyme.

During the single turnover, the concentration of the open-circle DNA present in the plasmid preparation declined in parallel with the supercoiled DNA. Thus, virtually none of the intermediate containing the DNA cleaved in one phosphodiester bond accumulated during the course of the reaction. The rate constant for forming this intermediate must therefore be smaller than that for its progression to the intermediate with the DNA cut in two bonds. Some accumulation of the intermediate carrying the DNA cleaved in both strands at one recognition site was observed, but, after a short lag phase, the majority of the DNA in this intermediate was cleaved at the second site to give the final products. (A fraction of the DNA in the intermediate with the DNA cut at one site failed to progress to that cut at both sites. This is probably due to the dissociation of this DNA from the enzyme before the next cleavage step (28), so as to liberate the DNA cut at one site; the latter cannot readily be re-utilized as a substrate for BspMI (Fig. 2).)

The kinetics of DNA cleavage by BspMI on a substrate with two BspMI sites are consistent with a scheme in which the enzyme binds to both sites and then proceeds to cleave the four scissile phosphodiester bonds in the substrate in sequential steps, but the rates of the successive reactions are limited by a step at or before the cleavage of the first phosphodiester bond.

Varied Separations of BspMI Sites-- To examine the ability of the BspMI endonuclease to interact with two sites in cis, a series of derivatives of pNAG1 were generated by either deleting DNA from the 700-bp segment separating its two BspMI sites or inserting additional DNA into this region. The new plasmids differed from each other only in the length of the DNA in one of the two arcs between the BspMI sites (the lengths cited below are for the shorter arc). Steady-state reactions were carried out with each derivative: the rates at which BspMI cleaved each plasmid, and also that with a single BspMI site, were measured from the decline in the concentration of the supercoiled substrate with time (Fig. 4). No change in reaction rate was observed as the inter-site spacing was increased from 700 bp (the separation in pNAG1) to either 1605 bp (Fig. 4) or 2644 bp (data not shown). On the other hand, the plasmids with the short inter-site spacings (94 or 38 bp) were cleaved more slowly than that with the 700-bp spacing, particularly in the case of the 38-bp separation (Fig. 4)

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Varied lengths of DNA between two BspMI sites. Reactions at 37 °C in 20 mM HEPES (pH 8.0), 10 mM MgCl2 1 mM DTT, and 80 mM NaCl contained 0.5 nM BspMI and 10 nM DNA (~90% supercoiled). The DNA was one of the following: , pAT153 (one BspMI site); , a derivative of pNAG1 with two BspMI sites separated by 38 bp; , a derivative of pNAG1 with two sites 94 bp apart; , pNAG1 (two sites 700 bp apart); , a derivative of pNAG1 with two sites 1605 bp apart. At timed intervals after the addition of the enzyme, samples were removed from the reactions and analyzed as described previously to separate the supercoiled substrate from the reaction products. The graph shows the decline in the concentration of supercoiled DNA during each reaction.

This behavior is largely as expected for a looping interaction across two sites in the same molecule of DNA. The probability for the juxtaposition of two sites in a supercoiled circle of DNA varies little as the length of DNA between the sites is increased from one-fifth to one-half of its overall length (31, 32). With short inter-site spacings of <300 bp, looping may be disfavored by the need to bend and/or twist the intervening DNA to position the requisite surfaces of the DNA at both sites against the corresponding DNA-binding surfaces in the protein (33). The energy required to bend and/or twist an intervening segment of just 38 bp into the requisite geometry ought to severely impede the looping interaction (34). Yet the plasmid with two BspMI sites 38 bp apart was still cleaved more rapidly than the plasmid with one BspMI site (Fig. 4). The cleavage of this two-site plasmid must therefore occur, at least in part, by reactions in cis involving the two sites 38 bp apart, rather than by reactions in trans involving sites on separate DNA molecules.

Interactions in trans-- The slow reactions on substrates with one cognate site were examined by comparing the activities of BspMI on the supercoiled form of pAT153 and on a linear form that had been generated by prior cleavage of pAT153 with another restriction enzyme (Fig. 5). The linear one-site substrate was cleaved more rapidly than the same DNA in its supercoiled form, although not quite as rapidly as the supercoiled DNA with two sites.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Linear and supercoiled DNA with one site. The reactions at 37 °C contained 0.5 nM BspMI endonuclease in 20 mM HEPES (pH 8.0), 10 mM MgCl2, 1 mM DTT, and 80 mM NaCl and DNA as noted below. At timed intervals after the addition of the enzyme, samples were removed from the reactions and analyzed as described previously to determine the residual concentration of the DNA substrate. In one reaction (), the DNA was 10 nM pAT153 (~85% supercoiled); the utilization of the supercoiled DNA is shown. In a second reaction (), the DNA was 10 nM pAT153 that had been previously linearized with AlwNI; the utilization of the linear DNA is shown. In a third reaction (), the DNA was 10 nM pAT153 (85% supercoiled) and 10 nM linearized pAT153; only the utilization of the supercoiled DNA is shown.

Reactions were also conducted with equal concentrations of both the linear and the supercoiled forms of pAT153. If the linear and the supercoiled forms of the one-site DNA act independently and compete for binding to the enzyme, the rate of cleavage of, for example, 10 nM supercoiled DNA in the presence of 10 nM linear DNA should be slower than that in a reaction containing 10 nM supercoiled DNA and no linear DNA. In contrast to this expectation, the supercoiled DNA was cleaved more rapidly in reactions containing both forms of the DNA than in reactions containing only the supercoiled form (Fig. 5). Thus, a linear DNA with one BspMI site is not only more susceptible to cleavage by this enzyme than a supercoiled DNA, but it can also confer susceptibility to the supercoiled DNA. In these respects, BspMI behaves like SfiI (16, 35).

As with SfiI (24), BspMI must therefore cleave DNA molecules with one recognition site by means of interactions in trans, in which the enzyme bridges two sites on separate DNA molecules. The juxtaposition of two sites on separate DNA molecules is likely to be facilitated if at least one of the two DNA molecules is linear, as opposed to both being supercoiled (16). A linear DNA may be able to penetrate the volume occupied by another DNA molecule by reptation (36), with one end of the linear molecule threading its way through the other DNA.

Activation by Oligoduplexes-- To characterize how a linear DNA with one BspMI site can affect the activity of BspMI against a plasmid with one cognate site, the cleavage of pAT153 by BspMI site was examined in the presence of a series of DNA duplexes made from synthetic oligonucleotides (Fig. 6). If the addition of the duplex results in the formation of a trans complex in which the enzyme is bound to both the plasmid and the duplex, then the extent of cleavage of the plasmid should first increase with increasing amounts of the duplex but then decline as the concentration of duplex is increased further; the decline is due to the enzyme engaging two molecules of the duplex rather than one duplex and one plasmid (16, 37). The concentration of the duplex needed for maximal activation and the extent of the decline in plasmid cleavage at higher duplex concentrations indicate the relative affinity of BspMI for that duplex.

View larger version (24K): [in this window] [in a new window]   Fig. 6.   Activation by oligoduplexes. The reactions at 37 °C contained 2 nM BspMI endonuclease and 10 nM pAT153 (~85% supercoiled) in 20 mM HEPES (pH 8.0), 10 mM MgCl2, 1 mM DTT, and 40 mM NaCl and an oligoduplex at the concentration indicated on the x axes. Ten min after the addition of the enzyme, the reactions were stopped, and the samples were analyzed as described above to determine the residual concentration of supercoiled pAT153; the concentrations of pAT153 consumed during the 10-min reactions are given on the y axes. Both a and b show reactions with three different oligoduplexes. In both panels, the sequence of the top strand of the duplex is noted alongside the corresponding data symbol (the bottom strands had the same lengths and exactly complementary sequences). The duplex marked  in a lacks the BspMI recognition sequence; in all other cases, the recognition sequence is underlined, and the sites of cleavage are marked by vertical hatch marks (the left-hand mark is for top-strand cleavage, and the right-hand mark is for bottom-strand cleavage). Each data point is the average of four repeats.

One of the oligoduplexes lacked the recognition sequence for BspMI, and at all concentrations tested, this neither activated nor inhibited the cleavage of pAT153 by BspMI (Fig. 6a, ). All of the other duplexes possessed the recognition site, and these were all capable of activating the cleavage of the one-site plasmid, albeit at different concentrations (Fig. 6). The activation thus seems to be a specific response to the recognition sequence. The specific duplexes all carried the same flanking sequences as those around the BspMI site in pAT153, but they differed from each other in terms of the lengths of flanking DNA either upstream and/or downstream of the site. Two duplexes had 6 bp of DNA upstream of the site and either 4 bp or 10 bp of downstream sequence, so that the former lacks the sites of DNA cleavage, whereas the latter possesses these sites (Fig. 6a,  and , respectively). At each concentration examined, these two duplexes gave the same enhancement in pAT153 cleavage, but in both cases, comparatively high concentrations of the duplex were needed to activate BspMI, and neither caused a decline in the cleavage of pAT153 at the highest concentrations used here. Hence, the two duplexes with 6 bp of upstream DNA seem to have relatively low affinities for BspMI, and the presence of the downstream site for DNA cleavage has no effect on their interactions with BspMI.

Three more duplexes were made: 1) a duplex with 10 bp of upstream DNA instead of the 6 bp used above and with 4 bp of downstream DNA, thus lacking the cleavage sites (Fig. 6b, ), 2) a duplex with 10 bp of downstream DNA, which includes the cleavage sites and 2 bp beyond (Fig. 6b, ), or 3) a duplex with 15 bp of downstream DNA, which extends 7 bp beyond the distal site of cleavage (Fig. 6b, ). All three of these longer duplexes (Fig. 6b) produced their maximal activation of BspMI at lower concentrations than the shorter duplexes noted above (Fig. 6a), and in all cases, they caused a decline in the extent of pAT153 cleavage at higher concentrations. However, the extent of the decline was most marked with the duplex carrying 15 bp of downstream sequence and least marked with the duplex possessing 4 bp of downstream DNA. The increase in the length of the upstream DNA from 6 to 10 bp seems to markedly enhance the affinity of BspMI for these duplexes. Moreover, in contrast to the duplexes with 6 bp of upstream DNA, the affinity for the duplexes with 10 bp of upstream DNA is enhanced further by the presence of the cleavage sites 4-8 bp downstream of the recognition site and enhanced further still by the DNA 2-7 bp beyond the cleavage site.

Salt Dependence-- Increasing salt concentrations attenuate DNA-protein interactions (38). For an enzyme that interacts with two DNA sites, its activity on sites in trans, relative to that on sites in cis, can be affected by the ionic strength of the reaction buffer. Interactions spanning two sites on separate molecules of DNA are generally weaker than those across two sites in the same molecule (12, 33), so the minimal salt concentration needed to block an interaction in trans will be lower than that needed to block the equivalent interaction in cis (35). For example, in reactions at high ionic strength, both the SfiI and SgrAI endonucleases cleave substrates with two recognition sites more rapidly than substrates with one site, but at low ionic strength, they cleave their one-site and two-site substrates at similar rates (11, 35). To further distinguish the cis and the trans reactions of BspMI, the extent of cleavage of one-site and two-site substrates for BspMI were measured under steady-state conditions in reactions containing varied NaCl concentrations (Fig. 7a). The reactions were also carried out in the presence of an oligonucleotide duplex that carries the recognition sequence for BspMI (Fig. 7b) to examine the enhancement of BspMI activity (Fig. 6) at varied salt concentrations. The oligonucleotide was the 31-bp duplex used above (Fig. 6b, ).

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Salt dependence. a, the reactions at 37 °C contained 2 nM BspMI and 10 nM plasmid (~90% supercoiled), either pAT153 (one BspMI site, ) or pNAG1 (two BspMI sites, ), in 20 mM HEPES (pH 8.0), 10 mM MgCl2, 1 mM DTT, and the concentrations of NaCl indicated on the x axis. b, the reactions were as described in a except for the inclusion of a 31-bp oligoduplex ( in Fig. 6b) that has the BspMI recognition sequence at a concentration of 25 nM. Ten min after the addition of the enzyme, the reactions were stopped and analyzed as described previously to determine the utilization of the supercoiled substrate within the 10-min period. Each data point is the average of four repeats.

In the absence of the oligoduplex (Fig. 7a), the activity of BspMI on the one-site substrate changed with the salt concentration in a manner different from that on its two-site substrate. Instead of the large difference in cleavage rates noted previously (Fig. 2; Ref. 10), in reactions at 80 mM NaCl, the one-site and two-site substrates were cleaved equally in reactions without added salt. As the concentration of NaCl was raised, its activity on the one-site substrate fell progressively, whereas its activity on the two-site substrate rose to a maximum at 60 mM and was only attenuated at salt concentrations > 80 mM. However, in BspMI reactions containing <80 mM NaCl, a fraction of the DNA was cleaved not only at the recognition site but also at an additional site: the additional site mapped to a locus where the sequence differs from the recognition sequence for BspMI by 1 bp (data not shown). As a consequence, a NaCl concentration of 80 mM was used in all BspMI reactions, unless stated otherwise. The different salt profiles for the reactions on the one-site and two-site substrates match the expectation for the former being cleaved in trans and the latter being cleaved in cis: the interactions in trans will be weaker than those in cis, so that the former are disrupted at lower salt concentrations than the latter (35).

The addition of the 31-bp duplex had a major effect on the salt profiles (Fig. 7b). At all NaCl concentrations < 100 mM, the one-site plasmid was now cleaved to a greater extent than the two-site plasmid. This switch in the relative activities of BspMI is due in part to the enhanced cleavage of the one-site plasmid caused by the duplex (viz. Fig. 6b) but is also due to a reduction in the extent of cleavage of the two-site plasmid. The duplex thus acts as an inhibitor rather than an activator of the reaction on the two-site substrate, presumably by disrupting the interaction in cis in favor of an interaction in trans. (The maximal cleavage of the one-site plasmid in the presence of the duplex was seen at 40 mM NaCl. Hence, the activation by oligoduplexes (Fig. 6) was studied at 40 mM NaCl rather than at 80 mM.)

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The type IIs restriction endonucleases that have been characterized previously are monomeric proteins that recognize asymmetric sequences and cleave the DNA in both strands at fixed distances away from the recognition site (3). In the case of FokI, the monomer contacts the entire recognition sequence, but it has the functions for cleaving only one phosphodiester bond (6). To cut both strands of the DNA, two monomers associate on the DNA to form a dimer (7-10). BspMI clearly deviates from this paradigm.

Sedimentation equilibrium studies showed that BspMI exists in solution as a tetramer of identical subunits (Fig. 1). However, the sedimentation analysis was carried out in a different buffer and at higher protein concentrations than those used for DNA cleavage assays. Hence, it is possible that BspMI is a monomer or a dimer under assay conditions. Nonetheless, the kinetics of DNA cleavage by BspMI can be readily reconciled to a tetrameric structure. During one DNA-binding event, BspMI can convert a plasmid with two BspMI sites directly to the final products cut at both sites in both strands (10). The cleavage of four phosphodiester bonds in a single turnover (Fig. 3) implies a protein with four active sites. Whereas some endonucleases have two active sites in one subunit (39), most contain the same number of active sites as protein subunits. For example, the dimeric restriction enzymes such as EcoRI, EcoRV, and MunI have two active sites and can cut two phosphodiester bonds per turnover to make a double-strand break at one DNA site (26, 27, 29). On the other hand, the tetrameric enzymes such as SfiI and Cfr10I have four active sites and can cut four phosphodiester bonds per turnover to make double-strand breaks at two sites (16-19, 28). Moreover, unlike the monomeric restriction enzymes FokI and Sau3AI, whose reaction rates increase disproportionately to the enzyme concentration (8, 10, 40), the linear relationship between the steady-state rates for DNA cleavage by BspMI and the enzyme concentration (Fig. 2b) suggests that BspMI remains in the same oligomerization state throughout its reaction. The assembly of BspMI in its DNA cleavage reactions is thus likely to be the tetramer seen in the ultracentrifuge.

Under its regular reaction conditions, BspMI cleaves DNA with two target sites more rapidly than DNA with one site (Fig. 2), although the ratio of its rates on the one- and the two-site plasmids is affected by a number of factors: 1) the substrate concentration (Fig. 2a), 2) the length of DNA between the sites in the two-site substrate (Fig. 4), 3) the topological state of the DNA (Fig. 5), and 4) the ionic strength of the reaction buffer (Fig. 7a). The way in which these factors affect the ratio of the rates on the two substrates is consistent with the view that BspMI cleaves DNA with one recognition site by means of reactions in trans, in which the tetramer bridges two sites in separate molecules, and cleaves DNA with two sites by means of reactions in cis, spanning sites in the same molecule of DNA.

The steady-state kinetics (Fig. 2a) show that at least part and perhaps all of the difference in the reaction velocities on the one- and two-site plasmids stems from the former having a much higher value for K_1/2 than the latter. The difference in K_1/2 values can be assigned to a local concentration effect: the concentration of one DNA site in the vicinity of another is higher when both sites are in the same molecule than when they are in separate molecules (33). Nevertheless, the reaction velocities on the one- and two-site plasmids may reach the same V_max at saturating substrate concentrations. Indeed, in the presence of the optimal concentration of an oligoduplex carrying the BspMI site, the reaction velocity on the plasmid with one BspMI site approached that for the two-site plasmid in the absence of the duplex (Figs. 6 and 7).

The different concentrations at which the oligoduplexes of varied lengths activated and/or inhibited the cleavage of the one-site plasmid (Fig. 6) indicate that the BspMI endonuclease contacts a considerable length of DNA (at least 22 bp) at each recognition site: not only its 6-bp recognition sequence but also >6 bp upstream and >10 bp downstream of the site. Strikingly, in duplexes with 6 bp of upstream DNA, the length of the downstream DNA had no effect on the interaction with the enzyme (Fig. 6a), but in duplexes with 10 bp of upstream DNA, the downstream DNA at and beyond the sites of DNA cleavage had a marked effect (Fig. 6b). Hence, it appears that to interact with its sites for DNA cleavage downstream of its recognition site, BspMI also needs to interact with the DNA >6 bp upstream of the recognition site.

On substrates with two BspMI sites, the concurrent action of the enzyme at the sites in cis presumably sequesters the intervening DNA in a loop. Whereas a plasmid with two BspMI sites 38 bp apart is not cleaved as rapidly as one with two sites separated by 700 bp, it is still cleaved more rapidly than a plasmid with one site (Fig. 4). DNA-looping interactions generally require inter-site separations of >38 bp, on account of the energy needed to bend and/or twist the DNA into the requisite configuration (33). For example, the SfiI endonuclease is less effective on plasmids with two SfiI sites separated by ~110 bp than on plasmids with sites >150 bp apart (41). One explanation for the ability of BspMI to mediate communications between very closely spaced sites in cis is that the intervening DNA is not held as a loop away from the protein but is instead wrapped around the surface of the protein, in a manner akin to a nucleosome (42). DNA wrapping would also account for the unusually long length of DNA needed for the optimal interaction with BspMI, as revealed by the oligoduplexes discussed above. This suggestion is similar to previous models for the interactions of the lac repressor with operator sites 60 bp apart, where the intervening DNA is wrapped right around the protein (43, 44).

In terms of its mode of action, BspMI, a type IIs enzyme, is similar to the tetrameric type IIf endonucleases, such as SfiI, Cfr10I, and NgoMIV, which also act concurrently at two DNA sites (13). However, BspMI recognizes an asymmetric DNA sequence, whereas all of the other tetrameric restriction enzymes characterized to date recognize palindromic sequences (16-18). The BspMI-DNA complex must therefore be organized differently from the other tetramers. Both Cfr10I and NgoMIV possess two DNA-binding clefts located on the opposite sides of the protein, each of which is made from two subunits (17, 18). The pairs of subunits that constitute each DNA-binding cleft interact symmetrically with the palindromic recognition sequence, in much the same manner as a dimeric restriction enzyme, and the two active sites within each pair are positioned to cleave the two strands of the DNA bound at that cleft.

With BspMI, the active sites from two subunits might be juxtaposed at the subunit interface to create a structural unit capable of inflicting a double-strand break in DNA, as in the dimeric form of FokI (7), and likewise the active sites of the other two subunits. But the recognition sequence for BspMI is asymmetric, therefore, unlike Cfr10I and NgoMIV (17, 18), the two subunits that make up one unit for inflicting double-strand breaks cannot interact symmetrically with this sequence. Instead, each copy of this asymmetric sequence may be recognized in its entirety by just one subunit of the BspMI tetramer. If so, then the complex of BspMI bound to two copies of its recognition site would involve the DNA recognition functions from two of the four subunits, but DNA cleavage within this complex, in both strands at both sites, would involve the catalytic functions of all four subunits. This scheme might allow BspMI to bind to more than two copies of its DNA, but the observation that its reaction on a plasmid with two BspMI sites is inhibited rather than activated by an oligoduplex with the cognate sequence (Fig. 7b) suggests otherwise. If the binding of the oligoduplex to one of the four subunits in BspMI still permitted two other subunits to bind to the two recognition sites in the two-site plasmid, then the duplex would not have inhibited the cleavage of this plasmid. However, an excess of DNA recognition functions over DNA sites is not unique to BspMI. The type III restriction enzymes recognize two target sites in DNA via their Mod subunits but cleave DNA within a complex in which only two of the four Mod subunits are likely to contact the DNA (45).

	ACKNOWLEDGEMENTS

We thank Ira Schildkraut, Rick Morgan, and Huimin Kong at New England Biolabs for clones and sequences and Abigail Bath, Isabel Kingston, Susan Milsom, and Mark Szczelkun for aid and advice.

	FOOTNOTES

^* This work was supported by Grant 7/G10881 from the Biotechnology and Biological Sciences Research Council and Grant 046178 from 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.

Published, JBC Papers in Press, November 29, 2001, DOI 10.1074/jbc.M108442200

¹ R. Morgan and H. Kong, personal communication.

	ABBREVIATIONS

The abbreviation used is: DTT, dithiothreitol.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit