Asymmetry in Flp-mediated cleavageKaren H. Luetke, Bao-ping Zhao+ and Paul D. Sadowski*
Department of Medical Genetics and Microbiology, University of Toronto, MSB, Room 4285, Toronto, Ontario M5S 1A8, Canada
Received July 30, 1997;Revised and Accepted September 10, 1997
ABSTRACT
Flp is a member of the integrase family of site-specific recombinases. Members of the integrase family mediate DNA strand cleavage via a transesterification reaction involving an active site tyrosine residue. The first step of the reaction results in covalent linkage of the protein to the 3'-phosphoryl DNA terminus, leaving a 5'-hydroxyl group at the site of the nick. We have used Flp recognition target (FRT) sites containing a 5'-bridging phosphorothioate linkage at the site of Flp cleavage to accumulate intermediates in which Flp is covalently bound at a cleavage site. We have probed these intermediates with dimethylsulfate using methylation protection and find that Flp-mediated cleavage is associated with protection of two adenine residues that are opposite the sites of cleavage and covalent attachment by Flp. Methylation interference studies showed that cleavage and covalent attachment are also accompanied by differences in the contacts of Flp with each of the two cleavage sites and with the surrounding symmetry elements. Therefore, we provide evidence that Flp-mediated cleavage and covalent attachment result in changes to the conformation of the Flp-FRT complex. These changes may be required for Flp-mediated strand exchange activity.
INTRODUCTION
Site-specific recombinases of the integrase family utilize a common catalytic mechanism of strand cleavage and ligation to mediate recombination (for reviews see 1 -5 ). These proteins bring about strand cleavage by employing an active site tyrosine to covalently link the protein to the 3'-DNA terminus and generate a free 5'-hydroxyl group at the nick. The covalent intermediate is dissociated by a second transesterification reaction in which the 5'-hydroxyl group from a second Flp recognition target (FRT) site that has been similarly cleaved attacks the covalent linkage and liberates the protein, resulting in intermolecular strand ligation and formation of a Holliday structure. The Holliday intermediate is subsequently resolved by a second pair of strand exchanges at the remaining sites to yield two recombinant products.
The Flp protein, which is a member of the integrase family of recombinases, is encoded by the Flp gene on the 2 µm plasmid of Saccharomyces cerevisiae. Flp-mediated recombination is thought to play a role in amplification of the 2 µm plasmid (6 ). Recombination is carried out between two FRT sites that are contained within two 599 bp inverted repeats of the plasmid. The minimal FRT site (Fig. 1 ) required for recombination consists of two inverted 13 bp symmetry elements (a and b) which flank an 8 bp core region. Binding of one monomer of Flp to each of the symmetry elements of the FRT site results in formation of an Flp-DNA complex (complex II) in which the DNA of the FRT site has undergone a protein-induced bend (7 ,8 ).
Flp assembles its active site from partial active sites contributed by two Flp monomers. This leads to cleavage of the DNA in trans (9 ). The Flp monomer which contributes the nucleophilic tyrosine is not bound to the symmetry element immediately adjacent to the site of cleavage. Recent evidence suggests that cleavage proceeds by a trans-horizontal mechanism (10 ), i.e. the Flp monomer contributing the nucleophilic tyrosine is positioned on the same substrate molecule as the phosphodiester bond that is to be cleaved. Other evidence suggests that synapsis between two substrate molecules occupied by Flp is not required for strand cleavage (11 ). This supports the view that the minimal catalytic unit of Flp is a dimer (12 ,13 ). In a dimeric Flp complex only one of two potential active sites is assembled at any one time, limiting strand cleavage to a single event (14 ,15 ).
The recombination events promoted by two other integrase family members, [lambda] integrase and P1 Cre, proceed by an ordered sequence of strand exchanges (16 -20 ). We have found previously that the top and bottom strand cleavage events mediated by Flp are associated with different positions of the DNA bends, suggesting that a preferred order to the initiation of strand exchange by Flp may exist (8 ).
In the present paper we examine covalent Flp-DNA intermediates to gain a better understanding of the events leading to Flp-mediated strand exchange activity. We used FRT sites containing a 5'-bridging phosphorothioate linkage at the site of Flp cleavage to accumulate such intermediates. We find that cleavage and covalent attachment by Flp are associated with unique changes in the dimeric complex formed by Flp with the FRT site as detected by chemical protection and interference experiments.
MATERIALS AND METHODS
Oligonucleotide substrates
Unmodified oligonucleotides were synthesized at the Hospital for Sick Children/Pharmacia Biotechnology Service Centre, Banting Institute, University of Toronto. Deprotected oligonucleotides used for methylation protection and interference experiments were purified by denaturing polyacrylamide gel electrophoresis [15% polyacrylamide, acrylamide:bis-acrylamide (19:1), 8 M urea, 1* TBE (90 mM boric acid, 90 mM Tris, 2 mM EDTA)].
Oligonucleotides used in this study (S indicates the position of the 5'-bridging phosphorothioate; sequences of the symmetry elements of the FRT site are indicated in bold; the core sequence of the FRT site is indicated in italics): KL-7, d(TGAAGTTCCTATTCTCTAGAAAGTATAGGAACTTCGACCT); KL-31, d(TTTCCAGGTCGAAGTTCCTATACTTTCTAGAGAATAGGAACTTCA- CTC); KL-7M(S), d(TTGTGAAGTTCCTATTCSTCTAGAAAGTATAGGAACTTCGA); KL-31M(S), d(TTTCCAGGTCGAAGTTCCTATACSTTTCTAGAGAATAGGAACTTC); HP-31, d(GGTCGAAGTTCCTATACTTTCTAGAGAATAGGAACTTCAC); KL-7V, d(TTGTGAAGTTCCTATTCTCTAGAAAGTATAGGAACTTCGA); HP-33V, d(TTGTGAAGTTCCTATTC); HP-14V, d(TCTAGAAAGTATAGGAACTTCGA); HP-2, d(TTTCTAGAGAATAGGAACTTCA); HP-5, d(AGGTCGAAGTTCCTATAC); HP-33, d(TGAAGTTCCTATTC); HP-14, d(TCTAG- AAAGTATAGGAACTTCGACCTGATC).
Oligonucleotides were 5'-end-labeled with [[gamma]-32P]ATP (NEN DuPont) using T4 polynucleotide kinase (New England Biolabs). Following extraction with phenol/chloroform and ethanol precipitation, the radioactively labeled oligonucleotide was annealed to the appropriate oligonucleotide(s) by heating and slow cooling in 100 mM NaCl, 5 mM MgCl2. Annealed DNA substrates were purified on 10% polyacrylamide gels [acrylamide:bis-acrylamide (30:1), 1* TBE]. Alternatively, annealed DNA substrates were 3'-end-labeled with [[alpha]-32P]dCTP (NEN DuPont) using AMV reverse transcriptase (Life Sciences) and purified as described above.
Preparation of 5'-bridging thiooligonucleotides
The thiooligonucleotides were synthesized on an ABI 394 DNA/RNA synthesizer and the standard protocol for oligonucleotide synthesis on a controlled pore glass support was followed except for the step in which the sulfur linkage was formed. For the latter linkage a protocol described previously (21 ) and modified by Burgin et al. (22 ) was adopted. This 5'-phosphorothioate link was created using 5'-(S-trityl)-mercapto- 5'-deoxythymidine-3'-O-(2-cyanomethyl)-N,N-isopropylphosphoramidite prepared as described previously (21 ) and as modified by Burgin et al. (22 ).
We also extended the time for detritylation of the sulfur by AgNO3 from 10 to 20 min and the coupling time of the phosphoramidite containing the 5'-mercapto group with the next phosphoramidite from 5 to 10 min. The thiooligonucleotides were deprotected by the standard procedure using ammonium hydroxide and purified by gel electrophoresis.
Flp preparations
Flp and Flp R191K were purified as described previously (23 ). Flp was >90% pure. Flp R191K was ~50% pure (15 ). The concentration of Flp was estimated by comparison with highly purified Flp standards on Coomassie blue stained SDS-PAGE. The Bradford assay (24 ) was used to determine the concentration of the homogeneous Flp standards.
Assay of Flp-mediated cleavage
Cleavage reactions containing 0.01 pmol DNA substrate radioactively labeled at both 5'-ends were incubated with 1.4, 2.8 or 14.1 pmol Flp or 58.7 pmol Flp R191K in a 20 µl volume containing 50 mM Tris-HCl, pH 7.4, 33 mM NaCl, 1 mM EDTA, 100 µl/ml calf thymus DNA. Reactions were incubated at 22oC for 90 min and subsequently treated with 10 µg proteinase K and 0.005% (w/v) SDS for 60 min at 37oC. Following phenol/chloroform extraction and ethanol precipitation the DNA was analyzed on a 15% denaturing polyacrylamide gel.
Gel mobility shift assays
Binding reactions containing 0.03 pmol radioactively labeled DNA substrate were incubated with 3.5 pmol Flp protein in a 20 µl volume containing 50 mM Tris-HCl, pH 7.4, 33 mM NaCl, 1 mM EDTA, 100 µg/ml calf thymus DNA. Reactions were incubated at 22oC for 60 min. Some reactions were shifted to higher temperatures for an additional 10 min (as indicated in the figure legends). A 2.7 µl quantity of dye mixture was added and reactions were subjected to electrophoresis on 5% polyacrylamide gels (1* TBE) at 4oC. Dried gels were subjected to autoradiography.
DMS methylation protection
DNA substrates (1.23-1.43 pmol), 5'-end-labeled on the top or bottom strands, were incubated with 38 pmol Flp protein for 90 min at 22oC in a 20 µl volume containing 50 mM Na cacodylate, pH 8.0, 5 mM Tris-HCl, pH 7.6, 12.5 mM NaCl, 2.5 mM MgCl2, 0.5 mM EDTA, 100 µg/ml calf thymus DNA, pH 7.7. Then, DMS was added to a final concentration of 24.5 mM and after 1 min at 22oC [beta]-mercaptoethanol was added to 200 mM. After 5 min at 22oC reactions in which Flp was bound to phosphorothioate-containing FRT sites were placed at 49oC for 15 min. Subsequently, 3 µl of a dye mixture were added and reactions subjected to electrophoresis on a 5% native polyacrylamide gel at 4oC. Both Flp-bound and unbound fractions of DNA were visualized by autoradiography, excised and eluted. The DNA was depurinated at 90oC for 10 min and cleaved by heating with NaOH at 90oC for 5 min (25 ). The DNA was ethanol precipitated with sodium acetate (0.3 M final concentration). In order to obtain DNA free of protein, the precipitated pellets were washed four or five times with 70% ethanol and the washes pooled and reprecipitated with 1.5 µg carrier DNA. Typically 6-12% of the DNA present in the Flp-DNA complexes was recovered by this method. The recovered DNA was analyzed on a 15% denaturing polyacrylamide gel.
Methylation interference
DNA substrates (8.8-11.1 pmol), 5'-end-labeled on the top or bottom strands, were methylated with DMS (30.1 mM final concentration) in a 50 µl volume as described above, except that Flp was omitted. The methylation reaction was terminated by addition of 12.5 µl stop solution (1.5 M Na acetate, pH 7.0, 1.0 M [beta]-mercaptoethanol) and the DNA precipitated with ethanol. Methylated DNA substrate (1.23 pmol) was incubated with Flp, as described above, in a reaction containing 50 mM Tris-HCl, pH 7.4, 33 mM NaCl, 1 mM EDTA, 100 µg/ml calf thymus DNA. Thereafter complexes and DNA were treated in the same way as described for the methylation protection protocol.
Quantitation
Dried gels or autoradiograms were scanned using a Molecular Dynamics Phosphorimager and analyzed using Imagequant software.
RESULTS
Accumulation of covalent Flp-DNA complexes
Flp mediates strand cleavage of the FRT site through a transesterification reaction which results in covalent linkage of the protein to the 3'-phosphoryl terminus through the active site tyrosine residue. The otherwise transient covalent intermediates resulting from Flp-mediated cleavage may be trapped by the use of ligation-defective mutant proteins (15 ,26 ,27 ) or by the use of suicide substrates containing nicks that are a few nucleotides away from the site of strand cleavage (16 ). Recently Burgin et al. (22 ) used a novel suicide substrate containing a 5'-bridging phosphorothioate linkage at the site of enzyme-mediated strand cleavage. Covalent enzyme-DNA intermediates accumulated after cleavage of the phosphorothioate linkage since the 5'-sulfhydryl group which was liberated was incompetent for the subsequent ligation reaction.
To determine the effect of a 5'-bridging phosphorothioate linkage on Flp-mediated cleavage activity we assayed strand cleavage on a denaturing gel (Fig. 2 ). Flp was incubated with an unmodified FRT site (Fig. 2 , lanes 2-4) or FRT sites containing single phosphorothioate modifications at the top (Fig. 2 , lanes 7-9) or bottom strand cleavage sites (Fig. 2 , lanes 12-14). Incubation of Flp with FRT sites containing phosphorothioate modifications results in accumulation of cleavage product at the site of modification (Fig. 2 , compare lanes 7-9, 12-14 and 2-4), consistent with published results for calf thymus topoisomerase I and [lambda] integrase (22 ). Incubation of Flp with an unmodified FRT site yields very little cleavage product (28 ; Fig. 2 , lanes 2-4). This is presumably due to the transient nature of the covalent Flp-DNA intermediate, which is quickly lost due to rejoining of the phosphodiester bond via the adjacent 5'-OH group. Cleavage of the phosphorothioate linkage by wild-type Flp was as efficient as by the ligation-defective protein Flp R191K (Fig. 2 , compare lanes 9 and 10 and 14 and 15), as one would expect if the phosphorothioate substrate is inhibiting Flp-mediated ligation. Cleavage of an unmodified FRT site by Flp R191K was also comparable in efficiency with cleavage of a phosphorothioate linkage by Flp (Fig. 2 , compare lanes 5 and 9 and 5 and 14). We also observed that the presence of the phosphorothioate in one strand seemed to inhibit cleavage of the non-substituted strand by Flp R191K (Fig. 2 , compare lanes 5, 10 and 15).
Methylation protection in Flp-DNA complexes
DMS methylates double-stranded DNA at the N7 position of guanine in the major groove and the N3 position of adenine in the minor groove. Protection from methylation by DMS can be used to probe the proximity of a protein to DNA (29 ,30 ).
Flp has been shown to confer significant protection of G and A residues in both strands of the symmetry elements (25 ,31 ,32 ). However, in DNA isolated from complexes II protection of bases in the core region was not observed (25 ,32 ). To determine whether Flp-mediated cleavage activity was associated with a detectable change in interaction of Flp with the FRT site we used the G>A specific cleavage reaction of methylated bases.
In order to accumulate covalent Flp-DNA complexes cleaved in only one strand we used FRT sites containing a 5'-bridging phosphorothioate at one or other of the Flp cleavage sites. To enable us to isolate the covalent Flp-DNA complexes we used heat to destroy the non-covalent Flp-DNA complexes. Flp was incubated with DNA substrates, the mixture was treated with DMS and the methylation reaction was terminated by addition of [beta]-mercaptoethanol. A control reaction in which Flp-DNA complexes were treated with [beta]-mercaptoethanol in the absence of DMS indicated that the Flp-DNA complexes were stable in 200 mM [beta]-mercaptoethanol (data not shown). A reaction in which Flp was bound to a phosphorothioate-containing FRT site was incubated at 49oC for 15 min to obtain a homogeneous population of covalent Flp-DNA complexes. Control reactions were also run in which Flp was bound to a phosphorothioate-containing FRT site, a wild-type FRT site or a FRT site with a nick at one cleavage site but which were not shifted to 49oC. An FRT site with a nick at one cleavage site was complexed with Flp as a control for the effect of a break in the DNA strand on the pattern of methylation protection. The complexes II, the substrate that was not bound by Flp and the untreated substrate DNAs were then separated and isolated from a preparative polyacrylamide gel. The DNAs were depurinated at modified G and A residues, cleaved with alkali and analyzed on a denaturing polyacrylamide gel (Figs 4 A and 5 A). The results are summarized in Figures 4 B and 5 B. DNA obtained from covalent Flp-DNA complexes was compared with substrate DNA, since the unbound fraction from heat-treated reactions consisted of DNA derived from multiple species. In these experiments we have examined the DNA strand opposite the site of Flp-mediated cleavage.
Methylation interference in Flp-mediated cleavage activity
Methylation protection provides information on the accessibility of DMS to G and A residues in a protein-DNA complex. A methylation interference experiment determines whether methylation of specific bases interferes with protein function. We were interested in determining whether methylation of G and A residues interferes with Flp-mediated cleavage and formation of covalent Flp-DNA complexes.
Our experimental approach was similar to that used for the methylation protection experiments except that DNA substrates were treated with DMS prior to incubation with Flp. A reaction in which Flp was bound to a phosphorothioate-containing FRT site was placed at 49oC for 15 min to obtain a homogeneous population of covalent Flp-DNA complexes cleaved in one strand. Control reactions were also done in which Flp was bound to a phosphorothioate-containing or a wild-type FRT site but the reactions were not shifted to 49oC. In order to examine the DNA strand opposite the site of Flp-mediated cleavage, DNA from isolated complexes II was analyzed on a denaturing gel to determine the pattern of methylation in the top (Fig. 6 A) or bottom strand of the FRT site (Fig. 7 A).
DISCUSSION
Phosphorothioate-substituted FRT sites trap covalent intermediates of Flp
We find that Flp-mediated cleavage of an FRT site containing a 5'-bridging phosphorothioate modification results in accumulation of cleavage product at the site of modification (Fig. 2 ), consistent with the results reported for calf thymus topoisomerase I and [lambda] integrase (22 ). Incorporation of a 5'-bridging phosphorothioate linkage at the Flp cleavage site in the FRT site has enabled us to trap covalent Flp-DNA intermediates because the 5'-sulfhydryl which is liberated is incompetent for subsequent ligation reactions (22 ). Use of phosphorothioate modifications, as opposed to ligation-defective mutants or nicked suicide substrates (16 ), more closely mimics a wild-type reaction, since covalent intermediates are accumulated by cleavage of an intact DNA strand using wild-type protein.
Surprisingly, methylation protection studies indicate that the phosphorothioate modification impedes binding of Flp and formation of a non-covalent complex II. Methylation protection experiments (Figs 4 and 5 ) suggested that the phosphorothioate substitution weakens the affinity of Flp for the symmetry element adjacent to the substitution. Covalent complexes II generated on a phosphorothioate-containing substrate, however, showed strong protection of the symmetry element. We suggest that the phosphorothioate modification may impose a structural restraint on flexibility of the DNA and may thereby inhibit Flp-induced DNA bending. This would weaken interaction of Flp with the binding element cis to the modification, leading to decreased protection of the FRT site. That this restraint is released by cleavage of the phosphorothioate linkage is demonstrated by our finding that protection of the FRT site is equally efficient in covalent complexes II as in non-covalent complexes II formed with an unsubstituted FRT site. Recent evidence shows that increased DNA flexibility due to tandem mismatches or due to replacement of thymine with 5-hydroxymethyluracil gives rise to tighter DNA binding by several DNA bending proteins (36 ,37 ). We have evidence suggesting that the core region of the FRT site contains a sequence-directed bend or point of flexure in the DNA (Luetke and Sadowski, in preparation). Bailly et al. (38 ) suggest that decreased flexibility of DNA due to replacement of adenine with 2,6-diaminopurine weakens interaction of the DNA bending protein Fis with its site. Phosphorothioate-containing sites are thus likely to be more effective suicide substrates for DNA binding proteins which mediate phosphoryl transfer reactions in the absence of DNA bending.
We find that cleavage of a phosphorothioate linkage by Flp is comparable in efficiency with cleavage of an unmodified FRT site by ligation-defective Flp R191K (Fig. 2 ; 15 ,27 ). This supports the postulate that the phosphorothioate substitution blocks ligation (22 ).
Flp-mediated cleavage causes changes in conformation of the Flp-FRT complex
We found that cleavage was associated with methylation protection of an adenine residue opposite each of the cleavage sites (+4A and -4A, Figs 4 and 5 ). These protections were not observed using unmodified or nicked substrates. They could be due to a conformational change in the Flp protein that occurs upon covalent attachment. This change could result in closer approximation of the protein to the A residues. Alternatively, they could be due to a conformational change in the FRT site itself, for example Flp-induced compression of the minor groove.
We have considered the possibility that protection of the adenine residues opposite the cleavage site is a result of the procedure used to isolate the covalent complexes. However, we have also observed partial protection of these A residues in complexes II generated by binding Flp to phosphorothioate-containing FRT sites which were not heated prior to electrophoresis. These complexes are thus a mixture of covalent and non-covalent complexes. In addition, a similar protection was seen by Bruckner and Cox (31 ) using an FRT site with a symmetrical core. We suggest that protection of these adenines may be indicative of a protein-controlled mechanism that prevents resealing of the nick and promotes the intermolecular ligation required for strand exchange.
A change of DNA structure associated with cleavage of the top strand
We observed that the -3G residue is not hypermethylated in complexes II that are nicked on the top strand. Hypermethylation of the -3G residue has previously been associated with a severe Flp-induced DNA bend (>140o) localized in the core (7 ,33 ). We attribute the difference in reactivity of the -3G residue to a localized change in DNA structure, resulting from breakage of the phosphodiester bond at the top strand cleavage site within the context of a Flp-DNA complex II. It is possible that the localized change in DNA structure extends to the neighboring -4A/T base pair at the site of the nick. It has been demonstrated that formation of the covalent vaccinia topoisomerase I-DNA intermediate results in unpairing of the T/A base pair 5' of the cleavage site (39 ). Type I DNA topoisomerases employ a similar cleavage mechanism to that of recombinases of the integrase family, involving transesterification by an active site tyrosine to create a covalent 3'-DNA intermediate. In addition, the recent report of the crystal structure of the covalent Cre-lox A complex shows that the first 3 nt of the core region, adjacent to the site of cleavage, are single stranded (40 ). The structure of the Cre-lox A complex also shows that a mobile [alpha]-helix of the cleaving Cre monomer is positioned to make three phosphate contacts in the concave surface of the ~100o DNA bend in the core region (40 ). This suggests that the change in DNA structure associated with breakage of the DNA strand at the cleavage site in Flp-DNA complexes may be indicative of unpairing of bases in covalent Flp-DNA complexes.
Although we have observed an absence of hypermethylation of the -3G residue, the +1G residue did not show a similar effect upon nicking of the bottom strand. This G remained hypermethylated in complexes II generated on phosphorothioate-substituted, nicked and unmodified FRT sites. This may mean that the structural change associated with cleavage is localized near the cleavage site and does not extend into the middle of the core. This conclusion is supported by the structure of the covalent Cre-lox A complex, which shows that unpairing of the core region is limited to the 3 bp of the 6 bp core region adjacent to the site of cleavage (40 ).
Flp contacts the two cleavage sites differently
We have examined interference of methylated residues with Flp-mediated cleavage activity. These studies showed asymmetrical interference of methylated core residues in Flp-mediated cleavage of the top versus bottom strands. Whereas methylation of the +2A, +3A and +4A residues in the top strand interfered with Flp-mediated cleavage of the bottom strand, methylation of the -2A, -3G and -4A residues in the bottom strand did not interfere with Flp-mediated cleavage of the top strand (Figs 6 and 7 ). The adenine contacts are in the minor groove, whereas methylation of the -G3 residue disrupts a potential contact in the major groove. We suggest that asymmetrical recognition of the cleavage sites may occur as a function of the asymmetry in the core sequence and possibly also the single base pair mismatch between the a and b elements. This suggestion is supported by the effect of several core mutations on Flp-mediated cleavage (Luetke and Sadowski, in preparation). These experiments showed that mutation of the -2 or -3 base pair had no effect on cleavage of the top or bottom strands, whereas mutation of the -1, +1, +2 or +3 base pair reduced cleavage of the bottom strand by 50%.
Asymmetrical recognition of the cleavage sites may represent one regulatory step leading to a bias in cleavage of the top versus bottom strands of the FRT site. It should follow that the subsequent sequence of strand exchanges would reflect this bias. However, to date there is no data to support a preferred order of initiation of strand exchange by Flp.
Covalent attachment is associated with changes in contacts of Flp with the symmetry elements
Cleavage by Flp is reported to proceed by a trans-horizontal mechanism (10 ). In a dimeric complex the Flp monomer which contributes the nucleophilic tyrosine and becomes covalently attached to the DNA is bound to the symmetry element on the other side of the core from the cleavage site. Thus the Flp monomer bound to the symmetry element cis to the site of cleavage is associated non-covalently with the DNA.
Our data show that methylation of the -7A, -6A or +7A base interfered specifically with formation of covalent complexes II. Thus the -7A, -6A and +7A residues do not appear to be essential for binding, but they do seem to be important for Flp-mediated cleavage. These data support a change in interaction of the Flp monomer bound non-covalently to the symmetry element cis to the site of cleavage which is associated with cleavage and covalent attachment. Consistent with our data, Beatty and Sadowski (32 ) observed that methylation of -6A and -7A interfered strongly with Flp-mediated recombination. This observation renders unlikely the possibility that the change in interaction of Flp with the symmetry element cis to the site of cleavage in covalent complexes II arises as an artifact due to the temperature shift used to isolate covalent complexes II. Changes in interaction of Flp with the FRT site upon formation of covalent complexes II may contribute to enhanced stability of covalent Flp-DNA complexes to temperatures of 46-55oC, as compared with non-covalently bound Flp complexes II (Fig. 3 ).
The crystal structure of Cre complexed with the lox A site reveals a difference in interaction of the non-cleaving monomer and the cleaving monomer with the DNA of each symmetry element (40 ). Guo et al. suggest that this difference may arise due to conformational changes associated with cleavage of the substrate by one of the two Cre monomers. This data is consistent with our finding that there is a change in interaction of the Flp monomer with the symmetry element cis to the site of cleavage which is associated with covalent attachment.
ACKNOWLEDGEMENTS
This work was supported by the Medical Research Council of Canada. K.H.L. held a University of Toronto Open Fellowship. We thank Drs A.Becker, H.Friesen and G.Pan for careful reading of the manuscript. We thank Donna Clary, Dr H.Friesen and Dr D.Kuntz for preparing Flp proteins.
36 Grove,A., Galeone,A., Mayol,L. and Geiduschek,E.P. (1996) J. Mol. Biol., 260, 120-125.MEDLINE Abstract
37 Grove,A., Galeone,A., Mayol,L. and Geiduschek,E.P. (1996) J. Mol. Biol., 260, 196-206.MEDLINE Abstract
38 Bailly,C., Waring,M.J. and Travers,A.A. (1995) J. Mol. Biol., 253, 1-7.MEDLINE Abstract
39 Sekiguchi,J. and Shuman,S. (1996) J. Biol. Chem., 271, 19436-19442.MEDLINE Abstract
40 Guo,F., Gopaul,D.N. and Van Duyne,G.D. (1997) Nature, 389, 40-46.MEDLINE Abstract
*To whom correspondence should be addressed. Tel: +1 416 978 6061; Fax: +1 416 971 2494; Email: p.sadowski@utoronto.ca +Present address: Department of Pharmaceutical Chemistry, State University of New Jersey-Rutgers, PO Box 789, Piscataway, NJ 08855-0789, USA
