An Unpaired 3 ’ Terminus Stimulates recA Protein-promoted DNA Strand Exchange

In the presence of 100 mM NaCl, the efficient exchange of strands between a circular single strand and an homologous DNA duplex promoted by the recA and single-stranded DNA binding proteins of Escherichia coli requires an unpaired 3’ terminus. Of the duplex DNAs tested, only those with 4 unpaired bases at the 3’ termini are effective. Without added NaCl, strand exchange proceeds efficiently with all duplex DNA termini examined including a nicked circular duplex. Thus, at approximately physiological salt concentrations,  factors in addition to the recA and singlestranded DNA binding proteins are needed to promote efficient strand exchange. One such factor may be a DNA helicase(s).

In the presence of 100 m M NaCl, the efficient exchange of strands between a circular single strand and an homologous DNA duplex promoted by the recA and single-stranded DNA binding proteins of Escherichia coli requires an unpaired 3' terminus. Of the duplex DNAs tested, only those with 4 unpaired bases at the 3' termini are effective. Without added NaCl, strand exchange proceeds efficiently with all duplex DNA termini examined including a nicked circular duplex. Thus, at approximately physiological salt concentrations, factors in addition to the recA and singlestranded DNA binding proteins are needed to promote efficient strand exchange. One such factor may be a DNA helicase(s).
The recA protein of Escherichia coli can promote the exchange of strands between homologous circular SS' and fulllength linear duplex DNA molecules. The products are a circular duplex with a single strand interruption and a linear single-strand (1).
The reaction proceeds in two distinct phases. In the first, recA protein and circular SS DNA form a stable complex in the presence of ATP and SSB which is necessary for efficient strand exchange (2, 3). The recA protein-SS DNA complex then interacts with the linear duplex resulting in strand assimilation (2). In the second phase, the heteroduplex region formed during strand assimilation is expanded through recA protein-promoted branch migration which is polar and proceeds in a 3' to 5' direction relative to the minus strand of the linear duplex (4-7). In the experiments described here, we have examined the effect of duplex DNA termini on strand exchange, and have found that at 100 mM NaCl there is a strong preference for duplex DNAs with unpaired bases a t the 3' termini.

Materials
RecA protein (8) and SSB (3) were purified as described previously. E. coli helicase I1 and rep protein were the generous gifts of N. Dixon and N. Arai (Stanford University), respectively. Labeled and unlabeled $X circular single-stranded, closed circular, and linear duplex * This work was supported by Grant GM 06196 from the National Institutes of Health and Grant PCM 79-04638 from the National Science Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
DNAs were prepared and their concentrations determined as described (1,2). All DNA concentrations are given in nucleotides.
Restriction endonucleases PstI, StuI, and XhoI were purchased from New England Biolabs. DNase I was obtained from Worthington. 3H-labeled nucleotides were purchased from Amersham and ENRHANCE was obtained from New England Nuclear. Phosphoenolpyruvate, pyruvate kinase, and S1 nuclease were purchased from Sigma.
Singly nicked circular duplex $X DNA was prepared by treating the closed circular duplex with DNase I in the presence of ethidium bromide (9). The reaction mixture contained 10 mM Tris.HC1 (pH 8.0), 1 mM MgCl,, 1 mM EDTA, 0.1 mg/ml of bovine serum albumin, 66 pg/ml of ethidium bromide, 190 mM closed circular duplex $X DNA, and 1 pg/ml of DNase I. The reaction was incubated a t 25 "C for 30 min and then stopped by phenol extraction.

Methods
All reactions were performed in 25 mM Tris.HC1 (pH 7.2), 10 mM MgCl,, 1 mM dithiothreitol, 5% glycerol, and where indicated, 100 mM NaC1. An ATP regeneration system was added as indicated and contained a final concentration of from 1.7 to 4.2 mM phosphoenolpyruvate and from 14 to 33 pg/ml of pyruvate kinase.
Reactions performed with linear duplex DNA were started by the addition of a mixture of ATP and SSB after preincubation of all other components at 37 "C. Reactions containing nicked circular duplex DNA were initiated by combining two separately preincubated mixtures. In one, nicked circular duplex DNA was incubated with SSB, ATP, and an ATP regenerating system a t 37 "C for 30 min, and helicase I1 and rep protein were added as indicated. In the second preincubation, recA protein was incubated a t 37 "C with circular SS DNA, and SSB and ATP were added to form the recA protein-SS DNA complex (2, 3). Thus, the reaction was started by mixing a preformed recA protein-SS DNA complex with the nicked circular duplex.
Assay for Heteroduplex Formation-This assay measures the incorporation of 3H-labeled SS DNA into S1 nuclease resistant heteroduplex. The procedure used is the same as described previously (z), except that the S1 nuclease digestions were performed a t 45 "C. Since the nicked circular duplex was made by random nicking of closed circular duplex DNA, a procedure that should produce equal numbers of molecules with single nicks in either the plus or minus strand, only half of the nicked circular duplex DNAs will be capable of reacting. Agarose Gel Electrophoresk and Fluorography-Electrophoresis was performed in 0.8% agarose as previously described (1). After soaked in EN3HANCE for approximately 1 h with shaking. The gel electrophoresis, the gel was blotted to remove excess water and then was then soaked in water for 1 h, after which it was dried down onto Whatmann 3" paper. The labeled bands were visualized by exposure of a sheet of Kodak XAR-5 x-ray film a t -80 "C.

RESULTS
Effect of Duplex DNA Termini on Strand Exchange-The effect of linear duplex DNA termini on strand exchange in the presence of 100 mM NaCl and SSB is shown in Fig. 1. Closed circular @X DNA cleaved with the PstI enzyme which generates 4 unpaired bases at the 3' termini was significantly more reactive than DNAs cleaved with either XhoI which produces 5' termini with 4 unpaired bases, or StuI which generates blunt ends. Similar results were obtained for strand exchange reactions performed in the presence of 100 mM NaCl, but without SSB.' We had reported earlier that the rate and extent of heteroduplex formation in the presence of SSB and in the absence of added 100 mM NaCl were nearly the same for all three duplex DNA termini (10). Since branch migration proceeds in a 3' to 5' direction on the minus strand D. A. Soltis, and I. R. Lehman, manuscript in preparation.
of the linear duplex (4-7), these findings suggest that an unpaired 3' terminus is required for efficient strand exchange. Presumably, the added NaCl stabilizes the duplex DNA termini, so that only the PstI-cleaved DNA possesses the requisite unpaired 3' terminus.

PsrI ' A C G T C T C A ' '
,

5C C T C C A C Tr
, Xhol 'CT T C 5 I  cular Duplex DNAs-RecA protein can promote the exchange of strands between "H-labeled circular SS and nicked circular duplex DNAs (Fig. 2). In the presence of SSB, this reaction proceeded nearly to completion although a t a somewhat lower rate than strand exchange reactions performed with a linear duplex (not shown). The products were a nicked circular duplex, into which the "H-labeled circular SS DNA had been incorporated, and a displaced circular single strand (Fig. 3). However, when 100 mM NaCl was added, the rate and extent of heteroduplex formation was substantially reduced (Fig. 2) and very little labeled nicked circular duplex DNA was formed (Fig. 3). These results again suggest that an unpaired 3' terminus is necessary for efficient strand exchange in the presence of 100 mM NaCI, and that addition of NaCl stabilizes the duplex structure of the nicked circular duplex DNA, thereby preventing interaction with the circular SS DNA.

. T C G A G A A G G 3 .
Effect of DNA Helicases on Strand Exchange-When E. coli DNA helicase I1 and rep protein were added to a strand exchange reaction performed in the presence of 100 mM NaCl, there was a %fold stimulation in the rate and extent of heteroduplex formation (Fig. 2). An approximate &fold increase was also observed in the amount of labeled nicked circular duplex formed (Fig. 3). The stimulation of heteroduplex formation by helicase I1 and rep protein was additive and may result from the unwinding of the 3' terminus at the nick in the circular duplex DNA by these proteins. The addition of more helicase I1 and/or rep protein, however, did not substantially increase the rate or extent of reaction (data not shown). Additional recA protein and purified recBC nuclease, which can also unwind duplex DNA (11)(12)(13)(14), had no effect (data not shown). We do not know what factors limit the extent of heteroduplex formation, but the effect of helicase I1 and rep protein does appear to be dependent on the added NaCl since they did not influence strand exchange in the absence of added NaCl (data not shown). Thus, although helicase I1 and rep protein are able to stimulate heteroduplex formation between circular SS and duplex DNAs in the presence of 100 mM NaCI, it would appear that other factors are necessary to promote this reaction fully.

REFERENCES
Our principal conclusion is that an unpaired 3' terminus on the minus strand of the duplex DNA is required for recA protein to promote the efficient exchange of strands between circular SS and duplex DNAs. This conclusion is based on the assumption that addition of 100 mM NaCl stabilizes basepairing at the 3' termini, thereby inhibiting initiation of strand exchange. In the absence of NaCl, the 3' termini may exist in an equilibrium between paired and unpaired forms, such that the unpaired termini can react during strand exchange. Since an increase in ionic strength is known to raise the melting point of duplex DNA, this assumption is not unreasonable.
The finding that an unpaired 3' terminus is required for strand exchange would suggest that strand assimilation involves the initial pairing of two complementary SS DNAs over a rather short region. Inasmuch as strand exchange proceeds efficiently in the presence of 100 mM NaCl on duplex DNAs with 4 unpaired bases at the 3' termini, but not on DNAs with 4 unpaired bases at the 5' termini, or on DNA with blunt ends, 4 unpaired bases at the 3' end of the duplex DNA would appear to be sufficient to initiate productive strand exchange. However, a more detailed study is clearly required to evaluate the effect of length of unpaired 3' termini and the role of helicases in the initiation of strand exchange.