THE JOURNAL OF BIOLWICAL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 269, No. 14, Issue of April 8, pp. 10940-10945, 1994 Printed in U.S.A.

Interaction of the NH2- and COOH-terminal Domains of the FLP Recombinase with the FLP Recognition Target Sequence*

(Received for publication, November 18, 1993, and in revised form, January 22, 1994)

Gagan B. Panigrahi and Paul D. SadowskiS

From the Department of Molecular and Medical Genetics, University of Toronto, llbronto M5S lA8, Canada

The FLP protein that is encoded by the 2-pm plasmid of yeast Saccharomyces cerevisicre is a 45-kDa site-spe- cific recombinase that belongs to the Int family of re- combination proteins. FLP catalyzes a recombination event within the plasmid by binding specifically to each of three 13-base pair (bp) symmetry elements of the FLP recognition target (FRT). We have shown previously that partial proteolysis of the FLP protein by proteinase K resulted in a COOH-terminal fragment of size 32 kDa (P32) and an NH2-terminal fragment of 13 kDa (P13). In this study we have used footprinting with dimethyl sul- fate to show that P32 binds specifically to the outer 9 bp of the 13 bp symmetry element. Binding of P13 alone to the FRT site was not detectable in this assay. However, when P13 and P32 were incubated together with the FRT site, protection of the remaining 4-bp region of the symmetry element was observed.

To confirm these results we used bromodeoxyuridine (BrdU)-dependent UV cross-linking. P32 became cross- linked to the substrate that contained BrdU substitu- tions in the outer 9 bp of a 13-bp symmetry element, but not to one with the BrdU substitutions in the inner 4 bp. Reciprocally P13 cross-linked to the latter substrate but not the former. Cross-linking was both BrdU and ultra- violet light-dependent. This study indicates that the COOH-terminal domain (P32) of FLP recognizes the outer 9 bp of the 13-bp symmetry element, whereas its NH,-terminal domain (P13) is needed for protection of the inner 4 bp of each symmetry element.

The FLP recombinase of 2 pm plasmid of yeast Saccharorny- ces cerevisiae carries out site-specific recombination between two identical target sequences (FRT site)’ that are situated within two 599-bp inverted repeats of the plasmid (Broach and Hicks, 1980; Hartley and Donelson, 1980). Each FRT site con- sists of three 13-bp symmetry elements (elements a , b, and c) (see Fig. 1). Binding of FLP to the FRT site is the initial event in the. process of recombination. Several footprinting studies

*This work was supported by the Medical Research Council of Canada and the National Sciences and Engineering Research Council of Canada. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

to the GenBankTMIEMBL Data Bank with accession number(s1 The nucleotide sequence(s1 reported in this paper has been submitted

M23380. $ To whom correspondence should be addressed: Tel.: 416-978-6061;

The abbreviations used are: FRT, FLP recognition target; bp, base pair(s); BrdU, bromodeoxyuridine; BrdUTP, bromodeoxyuridine tri- phosphate; DMS, dimethyl sulfate.

Fax: 416-978-6885.

have been carried out to elucidate the interaction of FLP with the FRT site (Andrews et al., 1985; Bruckner and Cox, 1986; Beatty and Sadowski, 1988; Panigrahi et al., 1992). However, little is known about the disposition of NH,- and COOH-termi- nal domains of FLP with respect to the symmetry elements. Recently FLP has been digested into three peptides by protein- ase K an amino-terminal 13-kDa peptide (P13), a 21-kDa in- ternal peptide (P21) whose NH, terminus begins at residue 148, and a 32-kDa carboxyl-terminal peptide whose NH, ter- minus begins at amino acid 124 (Pan et al., 1991; Pan and Sadowski, 1993). Both P21 and P32 have been shown to contain a DNA binding domain that recognizes the FRT site specifically.

In order to localize the interaction of the NH,- and COOH- terminal portions of the FLP recombinase within the symmetry elements of the FRT site, we have undertaken footprinting and cross-linking studies with the peptides derived from partial proteolysis of FLP. The large carboxyl-terminal fragment (P32) protects 9 of the 13 bp of each symmetry element against the methylation of DMS. However, P32 together with P13 protects the complete 13 bp. We have incorporated BrdU, a photoreac- tive analogue of thymine, within a single symmetry element, irradiated the protein-DNA complex with ultraviolet light and determined the regions of the DNA that are cross-linked to the protein. The cross-linking study confirmed the footprinting ob- servations. P32 cross-linked to the area of the symmetry ele- ment where the footprint of P32 occurred. The P13 did not cross-link to this area. On the other hand, P13 cross-linked to the region of the symmetry element corresponding to that de- tected by footprinting analysis. P32 did not cross-link to this footprinting area. The implication of this polarity of FLP inter- action and its relevance to the mechanism of binding and cata- lytic activities are discussed.

MATERIALS AND METHODS Plasmid Fragments

The plasmid pBA112 used in the present study contains a single FRT site and has been described elsewhere (Andrews et al., 1985). A 76-bp fragment from this plasmid was excised by digestion with BamHI and EcoRI and was dephosphorylated and labeled at the 5‘-end with [y3*P1ATP and polynucleotide kinase (Fig. 1).

FLP Protein and Proteolysis

The FLP protein used was a Sephacryl S300 fraction (85% pure) that contained 80 unitdpl (Pan et al., 1991). The proteolytic peptides P13 and P32 were prepared as described elsewhere (Pan and Sadowski, 1993). Protein concentrations were measured by using Bio-Rad protein assay kit (Bradford, 1976).

Synthesis of Single Symmetry Element Containing BrdU

The following oligonucleotides for construction of a single symmetry element were synthesized in an Applied Biosystems DNA synthesizer.

10940

Peptides of FLP and Target Site Interaction 10941 - I -

pBA112 mbp 5 5 Y c , b * d -

S ~ T T T G A A G T T C C T A T T C C G A ~ ~ G T T C ~ : T A T T C ) ~ C T ~ G A A A G T A T A G G A A C T T C A 9 "

3'AAACT TCAAGGATAAGGCTTCAAGGATAAGAGATCTTT,CATATCCTTGAAGT - "- C b a

FIG. 1. Schematic representation of the symmetry elements found in the FRT site. The symmetry elements are depicted by the horizontal arrows (a-c) and the core by an open box. The nucleotides are numbered above the sequence. The cleaved phosphodiester bonds are indicated by v and A.

a

b

used in the DNA protein cross-link- FIG. 2. Single symmetry element

ing analysis. Positions are numbered as in Fig. 1. The U represents bromode- c oxyuridine that replaces a thymine. The primers are underlined.

d

e

SUBSTRATE WITHOUT BRDU SUBSTlTUTlON 9

s C T A G A G G A T C C T A G A A A G T A T A G G A A C T T C A G A G G G " In z 51

3 3 ' G A T C T C C T A G G A T C T T T C A T A T C C T T G A A G T C T C C C T T A A S "

a

SUBSTRATE CONTAINING BRW IN THE BOTTOM STRAND 9 " In ?

s C T A G A G G A T C C T A G A A A G T A T A G G A A C T T C A G A G G G 3' 3' G A U C U C C U A G G U U C U U U C A U A U C C U U G A A G T C T C C C 5' "

a

SUBSTRATE CONTAINING BRDU IN THE TOP STRAND

s C T A G A G G A T C C T A G A A A G U A U A G G A A C U U C A G A G G G A A U U 3 9 " In z 51

3 G A T C T C C T A G G A T C T T T C A T A T C C T T G A A G T C T C C C T T A A 5' "

a

SUBSTRATE CONTAINING BRDU IN THE CORE DISTAL REGION 9 " In z 51

3 G A T C T C C T A G G A T C T T T C A T A T C C U U G A A G T C T C C C 5 s C T A G A G G A T C C T A G A A A G T A T A G G A A C T T C A G A G G G 3'

" a

SUBSTRATE CONTAINING BRW IN THE CORE PROXIMAL REGION 9 C Y In z 51

5' C T A G A G G A T C C T ~ ~ G A A A G U A U A G G A A C T T C A G A G G G A A T T 3 3 G A T C T C C T A G G A T C T T T C A T A T C C T T G A A G T C T C C C T T A A 5' "

a

Oligonucleotide 1: p and the incubation was continued for 2 h at 37 "C to complete the synthesis. The products were purified by polyacrylamide gel electro-

Oligonucleotide 2: phoresis. The control DNA fragments were synthesized in the presence

Oligonucleotide 3: (e) Synthesis of a DNA Fragment Containing BrdU in the Core-distal 5' CTAGAGGATCCTAGAAAG 3 ' Region (Fig. 2d)2-Oligonucleotides 1 (150 pmol) and 2 (300 pmol) were

mixed, annealed, and diluted into the synthesis buffer as described above. BrdUTP was added to a final concentration of 500 p. Three

3' GATCTCCTAGGATCTTTCATATCCTTGAAGTCTCCCTTAA 5' units of Henow fragment were added, and the reaction was incubated

5' CTAGAGGATCCTAGAAAGTATAGGAACTTCAGAGGG 3 '

3'GAAGTCTCCC 5' of dlTP instead of BrdUTP.

Oligonucleotide 4:

(a) Strand Hybridization-The single-stranded oligonucleotides 1 for 15 min at room-temperature. The mixture was passed through a (150 pmol) and (300 pmol) or (300 pmol) and (150 pmol) were SePhadex G50 Spun column (1-d bed volume) equilibrated with 30 mM annealed prior to enzymatic synthesis of the bottom or top strand, Tris-HC1, pH 8, 100 m~ NaCI, and 5 IMI MgCI, to remove BrdUTP. The

respectively (see b). The annealing solution contained 100 m~ NaCl and (20 ") was and bovine serum was added to a final concentration of 1 mg/ml. dlTP, dCTP, and dGTP were added to

cooled from 80 "C to room temperature over a period of 2 h. Md%' were heated at 95 "' for lo and then the mixture to a final concentration of 500 p followed by 50 pCi of

(b) Enzymatic Synthesis of Either Bottom or Top Strand DNA Con- (3000 ci/mmO1)' Three units Of fragment were

taining BrdU 2, b and c+Tris-HC1 buffer, pH 8.0, and bovine Cold u ~ p was added to a final concentration of500 ~, and the incu- added, and the reaction was incubated at room temperature for 15 min.

bation was continued at 37 "C for 2 h. The synthesized fragments were serum albumin were added to the hybridized oligonucleotides to a final concentrations of 30 m~ and 1 mg/ml, respectively. BrdUTP (Sigma), dCTP. and dGTP were added to a final concentration of 500 UM and 50 purified as described above' pCi of [(r-32PldATP (3000 Ci/mmol) were added. Three units of DNA polymerase I Klenow fragment were added, and the reaction was incu- We define the core-distal region as the outer 9 bp of the 1 3 - b ~ bated for 15 min. Then cold dATP was added to a concentration of 500 symmetry element and the core-proximal region as the inner 4 bp.

~ ~~~

10942 Peptides of FLP and Target Site Interaction W I- d

+ * e s 1 2 3

A

c Ill c II C I

., S

1 2 3 B

FIG. 3. Binding of FLP and proteolytic peptides to FRT site. Electrophoretic mobility shift assays were performed using approxi- mately 2 pmol of an 80-bp fragment containing the FRT site. A, lane 1, substrate alone; lane 2, substrate with 40 pmol of P32; lane 3, substrate with 4 pmol of intact FLP. B, lane 1, only substrate: lane 2, substrate with a mixture of 80 pmol of P13 and 40 pmol of P32; lane 3, substrate with 4 pmol of intact FLP. S represents substrate, GI, CII, and CIZZ represent complexes I, 11, and 111, respectively. The gel in A was run at room temperature, whereas that shown in B was run at 5 "C.

(dl Synthesis of a DNA Fragment Containing BrdU in Core-proximal Region Pig. 2e)-The oligonucleotides 3 (150 pmol) and 4 (300 pmol) were mixed, annealed, and suspended in the synthesis buffer as de- scribed above. In the first phase of synthesis, BrdUTP and dATP were added to a final concentration of 500 m. The synthesis was allowed to proceed for 15 min in the presence of 3 units of Klenow fragment a t room temperature. The whole mixture was passed through a Sephadex G50 spun column. The second phase of synthesis was done as described above.

W Cross-linking of BrdU-containing DNA Fragments

Experiments involving BrdU were performed under diffuse dim light. DNA fragments (-300,000 cpm) containing BrdU were dissolved in 50 pl of buffer containing 5 m Tris-HC1, pH 7.6, 125 m NaCI, 2.5 m MgCI,, and 1.5 pg of calf thymus DNA. The intact FLP protein or peptides were added to the mixture. The concentrations are described in the figure legends. The mixtures were incubated for 20 min at room temperature and then shifted to an ice-cold Petri dish. They were irra- diated for 5 min from a UV source generating 10 ergdmm2/s. Aliquots of irradiated or unirradiated samples were analyzed by electrophoresis on a 5% polyacrylamide gel. The efficiency of cross-linking to the DNA by U V irradiation was determined by adding heparin (200 pg/ml) to the irradiated samples prior to electrophoresis. This concentration of hep- arin completely abolishes the binding of FLP as determined in an un- irradiated control experiment.

FLP-DNA Electrophoretic Mobility Shift Assay

A gel mobility shift assay was performed to determine the degree of binding of intact FLP and the peptides. Briefly, DNA fragments labeled on either the top or bottom strands were dissolved in 60 pl of a buffer containing 5 m Tris-HCI, pH 7.6, 125 m NaCI, 2.5 m MgCI,, and 2 pg of calf thymus DNA. The concentrations of FLP or peptides were as described in the figure legends. The mixture was incubated for 20 min at room temperature, 6 pl of 10 x tracking dye were added, and the mixture was loaded on to a 5% polyacrylamide gel. The gel was elec- trophoresed as described previously (Panigrahi and Walker, 1991). When P13 was involved in the reaction the mobility shift gel was run in the cold room; otherwise the gels were run at room temperature.

DNase I Footprint Assay

The reaction conditions for DNase I were as described in the gel mobility shift section. After the incubation with P32, DNase 1(1 pg) was added, and the incubation was continued for 3 additional min. The

1 2 3 TOP BOTTOM

FIG. 4. Protection of FRT from DNase I by P32. Approximately 2 pmol of substrate were 5'-end-labeled either a t the BamHI end (top strand) or a t the EcoRI end (bottom strand), incubated with 40 pmol of P32 and treated with DNase I. The complexes were isolated and the DNA was analyzed on an 8% denaturing polyacrylamide gel. -P32, cleavage of unbound substrate; +P32, DNA isolated from complex 111. Symmetry elements are shown as arrows. The 8-base pair core is shown as an open box. The bracket shows the area of protection. The broken bracket in the bottom strand shows where protection is weak. Overex- posure of the same gel clearly shows the cleavage product in the 4'32 lane (lane 5) toward the bottom of the gel. The G>A lane shows the mobility of the sequencing markers.

reaction was stopped by adding 0.5 pl of 0.5 M EDTA. Finally 6 pl of 10 x loading dye were added, and the reaction was loaded on to a 5% gel. The peptide-bound and unbound fractions of DNA were located by au- toradiography and excised. The DNA was eluted and ethanol-precipi- tated and dissolved in 2 pl of dye solution for denaturing gel electro- phoresis.

DMS Methylation Protection Assay

The assay has been described in detail elsewhere (Panigrahi et al., 1992).

Quantitation

Autoradiograms were scanned with a LKB ultroscan XC laser den- sitometer or dried gels were analyzed on a Molecular Dynamics Phos- phorImager.

RESULTS

Mobility Shift Analysis of P32-FRT Site Interaction-Intact FLP protein and P32 bind to the three symmetry elements of the FRT site with the formation of three DNA-protein com- plexes that are resolved by gel electrophoresis (Fig. 3). The three complexes are thought to be the result from the binding of one molecule of FLP to one, two, or three symmetry elements of the FRT site (Andrews et al., 1987) The mobility of complexes generated by P32 is greater than that seen with intact FLP, consistent with the reduced molecular weight of the P32 pep- tide (Fig. 3A, lanes 2 and 3 ). P13 itself did not result in detect- able complexes with the FRT site under these conditions (data not shown but see below, Fig. 7, lane 18). However when P13

Peptides of FLP and Target Site Interaction 10943

C

+ + + I * * * I

.* qq m a

1 2 3 4 TOP

6 8

5 6 7 8 BOTTOM

FIG. 5. Protection by P32 against methylation by DMS at the N7 position of guanines and the N3 position of adenines. Approxi- mately 2 pmol of a 5'-end-labeled fragment containing an intact FRT site was incubated with 40 pmol of P32 and treated with DMS. The complexes were isolated, the DNA was depurinated, cleaved with alkali, and analyzed on an 8% polyacrylamide denaturing gel. -P32, cleavage of unbound substrate; +P32 I, +P32 I I , and +P32 III refer to the DNA from complexes I, 11, and 111, respectively. The nucleotides that are not protected by P32 are marked by an asterisk. The numbers correspond to the nucleotide numbers in Fig. 1. Note that these nucleotides are all protected by intact FLP (Panigrahi et al., 1992). Lanes 1 4 , top strand; lanes 5-3, bottom strand. The FRT site is shown schematically to the left of each panel.

was mixed with P32, the mobility of the complexes was com- parable with that produced by intact FLP (Fig. 3B, lanes 2 and 3 ).

Nuclease Footprinting Analysis with P32"Previous DNase I footprinting studies have reported that FLP protects 50 bp of DNA in a substrate containing the three symmetry elements of the FRT site (Andrews et al., 1985). In order to determine which nucleotides of the FRT site interact with the NH,- and COOH-terminal domains of FLP, FLP protein was partially proteolyzed by proteinase K, and the P13 and P32 peptides were isolated as described previously (Pan and Sadowski, 1993). P32 was incubated with a 5'-end-labeled fragment of DNA containing an FRT site, and the reaction was incubated with DNase I. The protein-DNAcomplexes were separated on a native polyacrylamide gel. Unbound DNA and DNA from com- plex I11 were isolated and analyzed on a denaturing polyacryl- amide gel (Fig. 4). Significant protection against DNase I cleav- age was observed in symmetry elements a, b, and c of the top strand as observed previously (Pan and Sadowski, 1993) (Fig. 4, lane 3 1. Like the intact FLP protein, protection conferred by P32 is -50 bp long. However, on the bottom strand, a few nucleotides of the symmetry elements a and b that were adja- cent to the core region were not protected (Fig. 4, lane 6) . Rather two nucleotides display enhanced reactivity when the DNA is bound to FLP. This hyperreactivity could be the result

+ I

I @a a / - d c

&r::

1 4-22

4 - 4 6 0 0

t t '3 4 8

8 0. 0. 0

0

-c 1 2 3 4 TOP BOrrOM

FIG. 6. Protection by mixture of P13 and P32 against methyla- tion by DMS at the N7 position of guanines and the N3 position of adenines. An 80-base pair fragment (approximately 2 pmol) con- taining an intact recognition site was 5'-end-labeled and incubated with the mixture of P32 (40 pmol) and P13 (80 pmol). The complexes were isolated in the cold room, depurinated, cleaved with alkali,and analyzed on an 8% polyacrylamide gel. -, cleavage of unbound substrates; +P13+P32 I l l , DNAisolated from complex 111. The arrows and numbers indicate the nucleotides which were protected here but not protected from methylation when only P32 was present (see Fig. 5). Lanes I and 2, top strand; lanes 3 and 4, bottom strand. The FRT sites are shown schematically.

of P32-induced rotation that causes the DNase I to cleave more efficiently.

Methylation Protection at Guanine and Adenine Bases in P32-FRT Site Complexes-The nuclease protection study of P32 suggested that P32 might not contact certain core-proxi- mal nucleotides of each of the symmetry elements a and b. To obtain better resolution of the binding of P32 to the symmetry elements we used DMS instead of DNase I. DMS methylates double-stranded DNA at the N7 position of guanine in the major groove and N3 position of adenine in the minor groove. Extensive footprinting studies with the intact FLP by DMS have been reported before (Panigrahi et al., 1992). In order to determine the nature of the interaction of P32 with the three symmetry elements of the FRT site, P32 was incubated with a 5'-end-labeled fragment of DNA containing a wild-type FRT site. The reaction was then treated with DMS. The mixture was loaded onto a native polyacrylamide gel and unbound DNA and the DNA from complexes I, 11, and I11 were isolated, depuri- nated at the modified G and Aresidues, cleaved with alkali, and analyzed on a denaturing polyacrylamide gel (Fig. 5). The pro- tection by P32 was different from that of intact FLP (Panigrahi et al., 1992). In the bottom strand, the purines at positions 8,6, -5, -6, -7, -19, -20, and -21 were not protected (Fig. 5, com- pare lanes 7 and 8, asterisks). In the top strand, the purines at positions 7, 5, -8, and -22 were not protected (Fig. 5, compare lanes 3 and 4, asterisks). In contrast, all the purines at the

10944

FIG. 7. Formation of cross-linked

BrdU, the thymines of the bottom strand FRT complexes. In the panel marked

were replaced with bromodeoxyuridine (Fig. 2b). The arrowheads show the cross- linked product generated by FLP, P32, and P13. The amount of FLP, P32, and P13 was 4, 40, and 40 pmol, respectively. The concentration of heparin ( H E P ) was 200 pg/ml. The W irradiation was for 5 min from a W source generating 10 ergs/ mm2/s.

Peptides of FLP and Target Site Interaction

CONTROL BRDU CONTROL BRDU CONTROL BRDU I n'

W X

I 3 4 f

n n n n + +

I + + + A J J J LLUUU

I nl W I

n + w > x 3 + +

I nI W I n T

I3 w >

+ + n n n n I + + + 22 22

1 2 3 4 5 6 7 8 9 1011 12 13141516 17 181920 21 222324

positions described above were protected by intact FLP (Pani- grahi et al., 1992). In other words, P32 protects a 9-bp region of each symmetry element away from the core (core-distal) but leaves a 4-bp region adjacent to the core (core-proximal) unpro- tected. The protected and unprotected areas were the same in all the three symmetry elements.

Methylation Protection at Guanine and Adenine Bases by a Mixture of P32 and P l 3 S i n c e P32 protected a 9-bp region of a single symmetry element but left 4 bp unprotected, it was of interest to determine whether addition of P13 would expand the region of protection of the symmetry elements against methylation by DMS. Accordingly, P32 and P13 were incubated together with a 5'-end-labeled fragment of DNA containing a wild-type FRT site and then DMS was added. The complexes were separated on a native polyacrylamide gel in the cold room. The bound and unbound material were isolated and processed as described above. The results are shown in Fig. 6. In the bottom strand, the purines at positions 8, 6, -5, -6, -7, -19, -20, and -21 which were not protected by P32 (Fig. 5, lane 7) were protected by the P13 and P32 mixture (Fig. 6, lane 3, arrowheads). Likewise in the case of the top strand, the purines at positions 7, 5, -8, and -22 that were unprotected by P32 alone (Fig. 5, lane 3) were protected by the mixture of P13 and P32 (Fig. 6, lane 1 ). Thus the addition of P13 expanded the footprint caused by P32 by 4 bp toward the core.

U V Cross-linking of Zntact FLP and Peptides to a BrdU- substituted Single Symmetry Element-The footprinting stud- ies have demonstrated that P32 binds to the distal 9 bp of the symmetry elements. Addition of P13 extended the footprinted area by four nucleotides. Since P13 binding to the symmetry element was not detectable in a band shift assay, it was unclear whether this expansion was due to the direct binding of P13 to that area or due to some conformational changes of P32 that occurred after addition of P13.

To answer this question directly, we have used a chemical cross-linking approach. Bromouracil and thymine are structur- ally similar, as the substituents at the 5-position differ little in size, with the van der Waals radii of bromine and methyl being 1.95 and 2.00 A, respectively (Hutchinson, 1973). W irradia- tion of BrdU-substituted DNA results in photodissociation of bromine with the generation of a reactive free radical at the C5 of uracil. A properly positioned substituent within a bound protein could donate a hydrogen to the uracil or deoxyribose radical or accept the free radical and directly cross-link to the DNA. The thymines present in either the bottom or top strand of a single symmetry element were replaced by 5-bromouracil (see Fig. 2, b and c) . This substrate was incubated with intact FLP, P32, or P13 and irradiated with UV light. After treatment with heparin, the DNA was analyzed on a 5% polyacrylamide gel to determine the amount of cross-linked product. Intact FLP and P32 bound to the single symmetry element substrate to

4

4

* FLP * P32

* P13

1 2 3 4 5 6 7 8

1-1 [ T I C A T R C C U U G R A G C R T R C C T T G R R C - -

A B FIG. 8. Cross-linking of FLP or ita peptides with a single sym-

metry element of the FRT site. Autoradiogram of a polyacrylamide gel used to isolate the cross-linked complexes from free DNA. A is core-distal substrate (Fig. 2 d ) , and B is core-proximal substrate (Fig. 2). Arrows show the protein-DNA complexes formed by intact FLP, P32, or P13. The sequences of the symmetry elements of the two sub- strates are shown a t the bottom along with the positions of BrdU sub- stitutions ( U ) . The rectangles over the sequence draw the boundary of p13 and P32 binding area.

make a single complex (Fig. 7, lanes 2 and 10). When all of the bottom strand thymines were replaced with BrdU, the degree of binding was comparable with that present in the unsubsti- tuted substrate control lane (compare lane 2 with lane 6 and lane 10 with lane 14). Heparin destroyed the complex in both control and BrdU-substituted substrates (lane 3, 7, 11, and 15). UV irradiation of substrate containing BrdU induced cross- linked DNA-protein products which were not destroyed by hep- arin (Fig. 7, lanes 8 and 16). P13 did not form a stable complex with the control or BrdU containing DNA (lanes 18 and 22). However a cross-linked product was observed when the reac- tion containing the BrdU-substituted substrate was UV-irradi- ated (Fig. 7, lane 24). In the absence of W cross-linking the P13-DNA interaction may be too weak to survive gel electro- phoresis. Similarly in the case of the substituted top strand, cross-linked products were observed with intact FLP, P32, and P13; the products were both UV- and BrdU-dependent (data not shown).

The footprinting data predicted that P32 should bind to the core-distal 9 bp of the symmetry element and that P13 should bind to the core-proximal 4 bp. In order to determine the speci- ficity of binding of P13 and P32 to a particular area, two dif- ferent substrates were synthesized. A substrate was synthe- sized in which the 2 thymines in the core-distal region were replaced with BrdU (Fig. 2d). Similarly, 2 thymines in the top strand core-proximal region were replaced with BrdU to create the core-proximal substrate (Fig. 2e) . Intact FLP formed cross-

Peptides of FLP and Target Site Interaction 10945

FIG. 9. Model showing the confor mational changes of the P32 induced by PW which allows its cross-linking with the core-proximal region of the symmetry element. P13 binds to the core-proximal region (top) providing a guide to the binding of P32 (middle). A -* conformational change in P32 allows it to = CORE-PROXIMAL AREA contact the core-proximal region (bottom).

CORE-DISTAL AREA

0 P13

links with both substrates (Fig. 8, lanes 1 and 5). As expected, P32 became cross-linked to the core-distal substrate (lane 2) but did not produce cross-links with the core proximal sub- strate (lune 6). Similarly P13 formed few cross-links with the core-distal substrate (lune 3) but could be cross-linked with the core-proximal substrate (lane 7). These results confirm that P13 binds to the core-proximal 4 bp of a symmetry element, whereas P32 is in close proximity to the core-distal 9 bp. The mixing of P13 and P32 together prior to cross-linking resulted in a doubling of the amount of P32 bound to the core-distal substrate (compare lanes 2 and 4) . P13 also stimulated the cross-linking of P32 to the core-proximal substrate (compare lanes 6 and 8).

DISCUSSION

In the present work, we have addressed the DNA binding properties of the FLP protein. We have shown that the NH,- terminal domain of FLP (P13) binds to the core-proximal 4 bp of each symmetry element, whereas the COOH-terminal do- main (P32) contacts the core-distal 9 bp. A similar mode of binding was reported in case of the ySresolvase (Abdel-Meguid et al., 1984). Resolvase can be cleaved by chymotrypsin into a large NH,-terminal fragment and a small COOH-terminal fragment. A DNase I protection study indicated that the COOH-terminal domain bound to the outer part of binding site, whereas the NH,-terminal bound to the middle of the site. The interaction of Cre with the lox site has been studied by Fe(I1)- EDTA footprinting (Hoess et al., 1990). A limited chymotryptic digest of Cre resulted in two fragments: a 25-kDa carboxyl- terminal fragment and a 13.5-kDa amino-terminal fragment. The 25-kDa peptide was found to protect sequences on the outer edges of the inverted repeats whereas the amino termi- nus of wild type Cre protected the DNA of the spacer region.

A model of the functional organization of the FLP protein must take into account the following observations. P32 binds specifically to the FRT site in presence of nonspecific DNA. P32 contacts the outer 9 bp of each symmetry element, whereas the 4 bp toward the core region were left unprotected. Addition of P13 restored the footprint of the entire 13 bp of each symmetry element. The cross-linking study indicates that P13 contacts the core-proximal 4 bp. This result predicts that the binding of FLP to a symmetry element should be sensitive to changes in the sequence of the first 4 (core proximal) bp. Mutations in these positions decrease recombination (Senecoff et al., 1988) perhaps by affecting DNA binding.

P13 has been shown previously to have some nonspecific DNA binding activity (Pan and Sadowski, 1993). This would account for the small amount of cross-linking of P13 to the core-distal substrate (Fig. 8A, lane 3). P13 also stimulated the

cross-linking of P32 to the core-distal substrate (Fig. SA, lane 4 1, a result which is consistent with previous findings (Pan and Sadowski, 1993) that P13 stimulated the binding of P32 to the FRT site. The mixing experiment in which P13 and P32 were mixed and cross-linked with either substrate provided unex- pected results in that P13 stimulated the cross-linking of P32 to both the core-distal (Fig. 8 A ) and to the core-proximal sub- strates (Fig. 8B).

We propose the following model for the action of P13 and P32 (Fig. 9). P13 binds to the core-proximal 4 bp of the symmetry element and facilitates the binding of P32 to the core-distal 9 bp. P13 induces a conformational change in P32 that allows it to contact the core-proximal 4 bp also apparently without dis- placement of P13 (Fig. 3B, lane 2). This accounts for the fact that P32 becomes cross-linked to the core-proximal substrate.

P32 is incapable of catalyzing cleavage of the FRT site unless P13 is also present (Pan and Sadowski, 1993). This may indi- cate that the conformational change is important for DNA cleavage, perhaps by bringing the required residues close to the scissile bond at the margin of the core and the symmetry ele- ment (Pan et al., 1993). Chen et al. (1992) have shown that FLP promotes cleavage in trans, that is a FLP molecule binds adja- cent to the scissile bond and another FLP molecule actually donates the nucleophilic tyrosine that breaks the phosphodi- ester backbone and covalently attaches the FLP molecule to the 3'-phosphoryl group. We cannot exclude the possibility that the cross-linking of P32 to the core-proximal substrate stimulated by P13 also takes place in trans.

Acknowledgments-We thank Helena Friesen, Barbara Funnell, Doug Kuntz, and Linda McBroom for critical reading of the manuscript.

REFERENCES

Abdel-Meguid, S . S. , Grindley, N. D. F., Templeton, N. S., and Steitz, T. A. (1984)

Andrews, B. J., Proteau, G. A,, Beatty, L. G., and Sadowski, P. D. (1985) Cell 40,

Andrews, B. J., Beatty, L. G, and Sadowski, P. D. (1987) J. Mol. Bwl. 193,345-358 Beatty, L. G., and Sadowski, P. D. (1988) J. Mol. Biol. 204,283-294 Bradford, M. M. (1976) Anal. Biochem. 72,248-254 Broach, J. R., and Hicks, J. B. (1980) Cell 21,501-508 Bruckner, R. C., and Cox, M. M. (1986) J. Biol. Chem. 261,11798-11807 Chen, J.-W., Lee, J., and Jayaram, M. (1992) Cell 69,647-658 Hartley, J. L., and Donelson, J. E. (1980) Nature 286,860-864 Hoess, R., Abremski, K., Irwin, S . , Kendall, M., and Mack, A. (1990) J. Mol. Biol.

Hutchinson, F. (1973) Q. Rev. Biophys. 6,201-246 Pan, G., and Sadowski, P. D. (1992) J. Biol. Chem. 267, 12397-12399 Pan, G., and Sadowski, P. D. (1993) J. Biol. Chem. 268,22546-22551

Pan, G., Luetke, K., and Sadowski, P. D. (1993) Mol. Cell. Biol. 13,3167-3175 Pan, G., Clary, D., and Sadowski, P. D. (1991) J. Bwl. Chem. 226,11347-11354

Panigrahi, G. B., Beatty, L. G., and Sadowski, P. D.(1992) Nucleic Acids Res. 20, Panigrahi, G. B., and Walker, I. G. (1991) Biochemistry 30,9761-9767

Senecoff, J. F., Rossmeissl, P. J., and Cox, M. M. (1988) J. Mol. Bid . 201, 405-421

Proc. Natl. Acad. Sci. U. S. A. 81, 2001-2005

795-803

216,873-882

5927-5935