Crossing Over Between Regions of Limited Homology in
Escherichia coli:
RecA-Dependent and RecA-Independent Pathways
Susan T. Lovett,
1Rebecca L. Hurley, Vincent A. Sutera, Jr.,
Rachel H. Aubuchon
2and Maria A. Lebedeva
Rosenstiel Basic Medical Sciences Research Center and the Department of Biology, Brandeis University, Waltham, Massachusetts 02454-9110
Manuscript received September 17, 2001 Accepted for publication December 10, 2001
ABSTRACT
We have developed an assay for intermolecular crossing over between circular plasmids carrying variable amounts of homology. Screens ofEscherichia colimutants demonstrated that known recombination functions can only partially account for the observed recombination. Recombination rates increased three to four orders of magnitude as homology rose from 25 to 411 bp. Loss of recAblocked most recombination; however, RecA-independent crossing over predominated at 25 bp and could be detected at all homology lengths. Products ofrecA-independent recombination were reciprocal in nature. This suggests that RecA-independentrecombination may involve a true break-and-join mechanism, but the genetic basis for this mechanism remains unknown. RecA-dependentcrossing over occurred primarily by the RecF pathway but considerable recombination occurred independent of both RecF and RecBCD. In many respects, the genetic dependence of RecA-dependent crossing over resembled that reported for single-strand gap repair. Surprisingly,ruvCmutants, in bothrecA⫹andrecAmutant backgrounds, scored as hyperrecombinational. This may occur because RuvC preferentially resolves Holliday junction intermediates, critical to both RecA-dependent and RecA-inRecA-dependent mechanisms, to the noncrossover configuration.Levels of crossing over were increased by defects in DnaB helicase and by oxidative damage, showing that damaged DNA or stalled replication can initiate genetic recombination.
R
ECOMBINATION can occur between exogenous or by exonucleolytic degradation or unwinding of bro-ken DNA. The RecA protein initiates recombination by DNA introduced into bacteria by conjugation, phagebinding to ssDNA and catalyzing strand pairing and transduction, or DNA transformation and the bacterial
strand transfer. Loading of RecA is facilitated by specific genome. Recombination plays a critical role in shaping
interactions with the RecBCD nuclease (Andersonand bacterial chromosomes: DNA can be integrated or
de-Kowalczykowski1997;Churchillet al.1999) or by leted from the chromosome by recombination, and
du-action of the RecFOR complex (UmezuandKolodner plicated gene segments can be substituted for one
an-1994;Shan et al. 1997; Webb et al. 1997). Because of other. Recombinational interactions between sister
RecBCD’s specificity for dsDNA ends, recombination chromosomes also repair DNA damage and restore the
involving double-strand breaks (DSBs) in DNA initia-integrity of chromosomes that are broken during the
tion employs RecBCD in the initiation step. In contrast, process of DNA replication.
gap repair appears to be initiated by RecFOR proteins
The steps of recombination: Models for genetic
re-(WangandSmith1984), which enhance the efficiency combination have been proposed to explain the
proper-of RecA loading onto ssDNA gaps (Umezu et al.1993; ties of genetic exchange during fungal meiosis
(Holli-Shan et al.1997;Webbet al.1997). day1964;MeselsonandRadding1975;Szostaket al.
After strand transfer from either type of substrate, 1983). The combined genetic and biochemical analysis
branched intermediates (Holliday junctions) are pro-of recombination functions from the bacteriumE. coli
cessed by one of several enzymes that can branch mi-has confirmed many elements of these recombinational
grate such structures (RecA, RuvAB, or RecG) and can models (reviewed inClark1991;West1992;
Kowalc-be cleaved by resolvases such as RuvC (West 1997). zykowski2000; and summarized below).
Mature recombinants are then formed by DNA ligase. Single-strand DNA (ssDNA) is the recombinogenic
The genetics of recombination is assay specific: Ge-substrate. It may be produced by incomplete replication
netic effects on the efficiency of recombination in bacte-ria depend on the type of recombination assay employed. This presumably reflects differences in the
recombina-1Corresponding author:Rosenstiel Ctr. MS029, Brandeis University,
tion substrates, intermediates, or products involved in
Waltham, MA 02454-9110. E-mail: [email protected]
these reactions. Most early studies of recombination
2Present address:Academic Press, 200 Wheeler Rd., Burlington, MA
01803. genes ofE. coliemployed assays of conjugational
MATERIALS AND METHODS bination (Clark1973). Since that time, studies of genetic
rearrangements in bacteria have revealed that there are Strains:The strains used in this study are listed in Table 1. differences between the genetic dependence of these pro- Strains were constructed by P1virA transduction using the selections noted in Table 1 (Miller 1992). Growth in the cesses as compared to conjugation. For example, although
presence of trimethoprim (Tp, 10g/ml) was used to select conjugation is mediated primarily by the RecBCD
path-thyAmutant derivatives as described (Miller1992; Viswana-way and only in a minor Viswana-way by the RecF pathViswana-way inE.
thanet al.2000) and Thy⫹transductants were subsequently
coli(HoriiandClark 1973), recombination between selected on 56/2 minimal medium (Willetts et al. 1969) direct repeats in Salmonella is catalyzed equally well by supplemented with 0.4% glucose; 0.001% thiamine; and either the RecF or the RecBCD pathway (Galitskiand 0.01% arginine, proline, histidine, threonine, and leucine. All other experiments employed Luria-Bertani (LB) medium Roth1997).
supplemented with the antibiotics ampicillin (Ap, 30g/ml),
What initiates recombination?Many of the most
com-chloramphenicol (Cm, 20g/ml), or tetracycline (Tc, 15g/ mon genetic recombination assays involve introduction
ml) as indicated. Growth was at 37⬚unless otherwise noted. of foreign DNA (by conjugation, transduction, or trans- Plasmids:Plasmid pSTL330 was generated from plasmid formation) in which the DNA is broken. In these cases, pBR322 template DNA by PCR with primers 5⬘GGTCTAGATC TCTGAAATCT AACAATGCGC TC and 5⬘ GGTCTAGAAT recombination is initiated by the provision of a DNA
TCTTGAAGAC GAAAGGGCC. The resultant PCR product end and so occurs at a conveniently high frequency.
was cut withXbaI, ligated, and introduced by electroporation We know much less about what initiates recombination
(Doweret al.1988) into strain HMS174, selecting Ap resis-events between bacterial genesin situ. If recombination tance on LB⫹Ap medium and screening for Tc sensitivity. is initiated by strand breaks, the cellular processes that Plasmids pSTL331, -332, -333, -334, -335, and -336 were pro-duced similarly from pACYC184 with a common primer (se-provide such breaks are unknown.
quence listed 5⬘ to 3⬘): GGGAGATCTC GAGGTACCAT TTG
Critical recombination intermediates:After
recombi-AGAAGCA CACGG and specific primers GGGGTACCAC GAT nation has been initiated, branched DNA structures
in-GAGCGCA TTGTTAGA (pSTL331); GGGGTACCAT CCAGG cluding Holliday junctions are formed. The branch mi- GTGAC GGTGCCGAG (pSTL332); GGGGTACCGA TATCCC gration of these structures can drive the recombination GCAA GAGGCCC (pSTL333); GGGGTACCGC ATATAGCGCT reaction forward or reverse it (West1997). Instability AGCAGCACGC CATAGTGACT GGCGATGCTG (pSTL334); GGGGTACCCC AAAGCGGTCG GACAGTGCTC CGAGAAC of these initial joint molecules may act as a proofreading
GGG TGCGCATAGA (pSTL335); GGGGTACCGC TCATGAG step for recombination, to prevent recombination
be-CCC GAAGTGGCGA Gbe-CCCGATCTT be-CCCCATCGGT (pSTL-tween spurious regions of short homologies. Alternative 336). The PCR products were digested with
Acc65I, recircular-resolution by cleavage of the Holliday junctions can ized by ligation, and transformed into HMS174 (F-recA1 rif produce either crossover or noncrossover products. hsdR). These plasmids confer a Cm-resistant but Tc-sensitive phenotype. Plasmids were introduced sequentially into the
Assay design:We wished to design an assay with
prop-test strains by electroporation (Doweret al.1988), selecting erties that would allow several unknown features of
bac-either ampicillin or chloramphenicol resistance on LB me-terial recombination to be explored. Our two-plasmid
in-dium. The appropriate construction was confirmed for each tegration assay detects recombination between resident plasmid derivative by sequence analysis of the entiretetA re-replicons; therefore, this assay should respond to factors gion (the Molecular Biology Core Facilities at Dana-Farber
Cancer Institute). that initiate recombination by production of strand
To insure that the various mutant strain backgrounds did breaks or other recombinogenic lesions. Our selection
not grossly influence plasmid copy number, plasmid DNA was detects only a subset of recombination events—those
extracted from strains carrying both pSTL330 and pSTL336 involving reciprocal crossing over—and is insensitive to using QIAGEN (Valencia, CA) miniprep kits and subjected other types of recombination such as nonreciprocal gene to agarose gel electrophoresis. From digital photographs of conversions. Since gene conversions and crossing over are the ethidium bromide-stained gels, the relative intensities of plasmid DNA bands were determined using the Molecular believed to be outcomes of differential processing of a
Analyst program (Bio-Rad, Richmond, CA). Duplicate plasmid common intermediate, the Holliday junction, genes
isolates from AB1157, N2096, N2731, STL230, STL970, that bias the resolution of recombination would also be STL3607, STL3919, and STL4959 were analyzed. When nor-revealed by our assay. Our system also includes recombi- malized to the optical density of the cultures, the relative nation substrates with very short sequence homologies, amount of plasmid DNA in any of the mutants assayed was at
least 60% of that in a wild-type strain. whose rate of recombination may be limited by factors
Assays:Fluctuation assays for rate determination were per-different from those involving large homologies. In this
formed by inoculation of 8–24 independent entire single colo-way, we hoped to reveal functions that stabilize or
desta-nies and aerobic growth for 2 hr in LB broth⫹Ap⫹Cm. bilize critical recombination intermediates. Our initial Subsequent serial dilution and plating on LB ⫹ Ap⫹ Cm experiments reported here show that the recombina- and LB⫹Ap⫹Cm⫹Tc determined the number of plasmid-bearing cells and plasmid recombinants, respectively, in each tion genes known forE. colican only partially account
culture. Using these values, recombination rates were calcu-for the recombination we observe. In addition,
difficul-lated from either the method of the median or the maximum-ties in DNA replication and oxidative damage to DNA
likelihood method as described (Leaand Coulson 1949). both can elevate levels of recombination between resi- The maximum-likelihood method was used in cases where dent genes of E. coli, suggesting that they may play a 25% or more of the cultures contained zero Tcrcolonies.
TABLE 1
E. coliK-12 strains
Strain Relevant genotype Derivation
AB1157 rec⫹ Bachmann(1996)
CS140 ruvC53 Shurvintonet al.(1984)
JC10287 (recA-srl⌬)304 CsonkaandClark(1979)
N2096 ruvA⌬63 R. Lloyd
N2731 recG258::miniTn10kan R. Lloyd
STL230 recO1504::Tn5 KmrUVstransductant P1 RDK1504 (R. Kolodner)⫻AB1157
STL970 ruvC53 recG258::miniTn10kan KmrUVstransductant P1 N2731⫻CS140
STL1131 ruvC53 cysC95::Tn10 TcrCys-transductant P1 CAG12173 (Singeret al.1989)⫻CS140
STL1143 ruvC53(recA-srl⌬)304 Cys⫹UVstransductant P1 JC10287⫻STL1131
STL1270 dnaB107(ts) malE::Tn10kan (SavesonandLovett1997)
STL3607 recF400::kan KmrUVstransductant P1 JC18970⫻AB1157
STL3919 recR252::miniTn10kan KmrUVstransductant P1 AM207 (R. Lloyd)⫻AB1157
STL3998 recB2053::Tn10kan KmrUVstransductant P1 STL3839⫻AB1157
STL4955 recF400::kan thyA Tprspontaneous mutant of STL3998
STL4959 recB2053::Tn10kan recF400::kan Thy⫹UVstransductant P1 STL3839⫻STL4955
All strains carry the additional markers F⫺Rac⫺⫺thi-1 hisG4⌬(gpt-proA)62 argE3 thr-1 leuB6 kdgK51 rfbD1 ara-14 lacY1 galK2 xyl-5 mtl-1 tsx-33 supE44 rpsL31(Bachmann1996).
were grown to an OD660of 0.3 and split into four 2-ml cultures. tetAhomologies between pSTL330 and any one of the
Two of the split cultures were treated with 5 mmH2O2and pSTL331–336 derivatives produces a dimeric integrant
two remained untreated. After 20 min of growth, 50g/ml
plasmid. This recombination event also restores a func-of catalase (Sigma, St. Louis) was added to the cultures to
tional tetA and hence can be selected by tetracycline inactivate the H2O2. Each culture was diluted and plated on
the appropriate media in duplicate to determine the number resistance. This assay, as designed, is specific to integ-of recombinants and the total colony-forming units integ-of surviv- rative recombinational events within a limited region ing plasmid-containing cells. Average cell survival at this dose and will not detect “gene-conversion” type recombina-was 16%. Determinations were made for four independent
tion in which a patch of genetic information has been cultures and the averages and range of values are presented.
transferred from one replicon to the other. Strains of
Product analysis: Minipreparations of plasmid DNA were
purified (QIAGEN) from independent tetracycline-resistant interest were transformed with pSTL330 and then one isolates. This DNA was subjected to agarose gel electrophoresis of the pSTL331–336 series plasmids, producing a set of with and without prior treatment with restriction endonucle- six strains with two plasmids that share variable regions aseEcoRI. Reciprocal products ran as heterodimers or higher
of sequence homology. Fluctuation analysis of frequen-multimers when uncut and exhibited characteristic restriction
products of 5.8 kb (tetA⫹⫹ColE1ori⫹bla⫹p15aori) and 0.7–1.1 kb (doubly deletedtetAallele⫹cat).
RESULTS
Assay design: A set of moderate-copy-number plas-mids was designed to detect crossing over between ho-mologies of variable lengths (Figure 1). One plasmid, pSTL330, was derived from pBR322 (Bolivar et al.
1977), carrying a ColE1 origin of replication; this plas-mid encodes ampicillin resistance and carries a mutant form of thetetAgene lacking promoter, ribosome-bind-ing site, and initiation codon. Another set of plasmids,
Figure 1.—Design of the two-plasmid crossing over assay. pSTL331–336, were derived from pACYC184 (Chang One plasmid, pSTL330, confers ampicillin resistance and car-andCohen1978), carrying the p15a origin of replica- ries a 5⬘-truncated allele of tetA. The second plasmid in the series pSTL331–336 confers chloramphenicol resistance and tion; these plasmids encode chloramphenicol resistance
carries a 3⬘-truncated allele oftetA. Crossing over in the region and carry various 3⬘ deletions of the tetAgene. These
of homology shared between the two plasmids, which varies plasmids were designed to share between 25 and 411
chromosomes (reviewed inBzymekandLovett2001), probably during the process of DNA replication. The results reported here also suggest that RecA-indepen-dent crossing over between two separate replicons can indeed occur, albeit at low rates in the population. The genetic basis of this RecA-independent recombination remains unknown.
RecBCD vs.RecFOR pathways: Loss ofrecF (Figure 3A) produced a 6- to 40-fold reduction of recombination at every homology length except the smallest one. Re-combination at 25 bp of homology was independent of both RecF and RecA; therefore, the lack of effect of RecF at this homology confirms that the reduction of recombi-nation at higher homologies was not due to artifacts of Figure2.—Recombination rates as a function of homology
plasmid maintenance or copy number problems. Mu-length: RecA dependence. By selection fortet⫹recombinants
tants inrecOandrecRalso exhibited reduction of recom-between plasmids sharing different lengths of homologous
sequences, rates of recombination were determined by fluctu- bination (five- and threefold, respectively) when assayed ation analysis for isogenic wild-type (open squares, dashed with the pSTL330 and STL336 plasmids sharing 411 line) andrecA⌬mutant (solid circles) strains. bp of homology. The RecA-dependent crossing over
measured in this assay, therefore, primarily involves the RecF pathway, as do other assays of recombination cies of Tc resistance in these plasmid-bearing
popula-measured with plasmids (Cohen and Laban 1983; tions determined individual rates of recombination at
Kolodneret al.1985;McFarlaneandSaunders1996). each length of homology.
The plasmids employed in our assay for recombination
Homology and RecA dependence:In a wild-type strain
naturally lack Chi sites, which attenuate the nuclease background, rates of recombination were quite low,
activity of RecBCD and promote its ability to activate
⬍10⫺8, at 25 bp and rose dramatically, over three orders
recombination (Dabertet al.1992;Dixonand Kowal-of magnitude, with increasing length Kowal-of homology
(Fig-czykowski1993;Kuzminovet al.1994). Therefore, it ure 2). Rates appeared to approach a plateau at a rate
was not surprising that the recB mutant exhibited no of 2⫻10⫺5for homologies⬎200 bp. Restriction
diges-reduction but, rather, especially high levels of recombi-tion of 6–10 independenttet⫹recombination products
nation (Figure 3B). Broken plasmid DNA would be rap-at each homology length confirmed the reciprocal
na-idly degraded by the RecBCD nuclease and therefore ture of the selected recombination event: All products
inactivation of this enzyme should stabilize broken were heteromultimeric and contained both an intact
DNA, permitting it to recombine with its homologous
tetA gene and the doubly deleted tetA gene with both
partner. This enhanced recombination in therecB mu-3⬘- and 5⬘-deleted segments (Figure 1).
tant is again mostly due to the RecF-dependent pathway, In arecAmutant strain, recombination rates remained
since rates are reduced in the double recB recFmutant at a low level ofⵑ10⫺8–10⫺7/cell generation (Figure 2).
relative to the singlerecBmutant (Figure 3B). However, There was a 10-fold elevation of recombination rates
substantial RecF-independent recombination is evident. with increasing homology, suggesting that
RecA-inde-Recombination rates in recF are considerably higher pendent recombination is also more efficient with
in-than those inrecAmutants; recombination rates inrecB
creasing homology, although not to the same extent
recFstrains are similar to those of wild type. Since the as evident for dependent recombination.
RecA-RecBCD and RecFOR pathways differ in the initiation independent recombination was the major contributor
step of recombination, this may reflect an alternative to recombination between very limited homologies of
way in which ssDNA is created or by which RecA is 25 bp in length. Restriction analysis of 6–10tet⫹plasmid
loaded on ssDNA. This could involve an unknown factor products formed independently of RecA, at each
homol-or, alternatively, a stochastic process in which a subset ogy length, showed that they, like the RecA-dependent
of substrates is generated that does not require RecFOR products, had experienced an apparent reciprocal
ex-or other additional factex-ors to suppex-ort initiation. change reaction. All products were plasmid
heterodim-Resolution factors: Since these reactions require ers with onetetA⫹ and one doubly deleted tetA allele.
crossing over, which may involve cleavage of a Holliday (Although thetet⫹allele is demanded by the selection,
junction, it was of interest to determine the effect of the reciprocal doubly deletedtetjoint is not; therefore
functions that are known to either branch migrate or “illegitimate” joining at other sites could have occurred.
cleave Holliday junctions. A mutation in ruvC, which Nevertheless, no such illegitimate joints were observed,
encodes the only known Holliday junction cleavage pro-even when only 25 bp of homology existed between the
tein normally expressed in E. coli, promoted hyperre-two plasmids.) Previous studies show that efficient
Figure3.—Effects of recombination initia-tion pathways on recombinainitia-tion rates at differ-ent length homologies. (A) Wild-type (open squares, dashed line) andrecF mutant (solid circles). (B) Mutants inrecB(solid squares) or recB recF(solid diamonds).
ofruvAB, encoding the Holliday junction helicase asso- At lowest length of homology, 25 bp, crossing over in wild-type strains is independent of RecA. We wished ciated with ruvCresolvase, did not significantly affect
recombination rates, except for a modest elevation at to know whether the hyperrecombination of ruvC mu-tants at 25 bp of homology was likewise independent of the highest homology, 411 bp (Figure 4A). A mutant in
recG, which has been proposed to provide an alternative recA. With 25 bp of homology shared between plasmids STL330 and pSTL331, the rate of crossing over in the mode of Holliday junction processing (Whitby et al.
1993), showed hyperrecombination at the lowest homol- recA ruvCmutant was 3.9⫻10⫺8(16 assays) compared
to the rate of 4.0⫻10⫺8(16 assays) seen with theruvC
ogy (25 bp) and a reduction, two- to fivefold, or no
effect at higher homologies (Figure 4B). A ruvC recG single mutant. Therefore, the absence of RuvC stimu-lates RecA-independent recombination at short homol-double mutant had a modest reduction, two- to fourfold
in rates at all homology lengths, except the smallest ogies.
Recombinogenic factors:A number of cellular events and the largest, which showed no effect (Figure 4B). If
Holliday junction resolution is required for these inte- may trigger recombination between intact DNA mole-cules. First, difficulties in replication may stall forks and grative reciprocal reactions, our experiments suggest
that it is not mediated by RuvC. This may be because elicit subsequent recombinational repair. Second, re-pair of DNA damage such as oxidative or other lesions RuvABC resolution is biased to noncrossovers in these
reactions, as has recently been proposed (Cromieand may produce recombinogenic ssDNA that provokes crossing over between replicons. We investigated two Leach 2000; Michel et al. 2000). A RecG-dependent
pathway may contribute partially, but considerable situations in which recombination may be elevated: in a dnaBts mutant with defects in replication and with crossing over remains after the inactivation of both
Ru-vABC and RecG. This crossing over is probably not due treatment of hydrogen peroxide to induce oxidative lesions. At its permissive temperature for growth, 30⬚, to Rus, a cryptic bacteriophage-encoded resolvase
(Mandal et al. 1993;Sharples et al. 1994), since it is introduction of a temperature-sensitive mutation in
dnaB replicative helicase elevated recombination rates not normally expressed in E. coli K12. Furthermore,
recombination rates ofrus ruvCdouble mutants at 411 substantially relative to wild-type rates at 30⬚. This effect was pronounced at all but the largest homology (Figure bp of homology were not lower than rates inruvC
mu-tants (data not shown). 5). At shorter homology lengths recombination rates
provide substrates with broken ends. The selected events can be produced only by crossing over and not by gene conversion-type reactions. Because of these features, such an assay may be useful to define the factors that initiate recombinationin situand those that affect the stability or processing of critical recombination intermediates. Indeed, our analysis shows that, depending on the length of sequence homology between recombining partners, factors affect recombination differentially. Re-combination measured by our assay shows that the re-combination genes identified by defects in repair or conjugation at present can only partially define the mo-lecular events we observe.
No single recombination assay can reflect the diversity Figure 5.—Elevation of recombination rates by dnaB-ts.
of genetic exchanges that occurin vivo. The genes that Recombination rates at different homology lengths for
wild-type (open squares, dashed line) or dnaB107strains (solid influence recombination can be assay specific, depending triangles) at 30⬚are shown. on the constraints of the particular assays. For example,
our system may be insensitive to double-strand-break-initiated recombination because the recombining part-may be limited by the probability of an initiating lesion
ners lack Chi sites that promote RecBCD-mediated ex-occurring in the region of homology shared between
change and attenuate the destruction of linear DNA by the two replicons. At the largest homology, other factors
RecBCD. In addition, because our assay demands cross-may become rate limiting. Difficulties in replication can
ing over between independent replicons, factors that trigger RecA-dependent recombinational repair and
encourage recombination strictly between sister chro-lead to genetic rearrangements, as supported by other
mosomes would not lead to recombination scored by analyses (HoriuchiandFujimura1995;Savesonand
our assay and might, rather, inhibit interreplicon cross-Lovett1997;Michel2000). Hydrogen peroxide
stimu-ing over. lated recombination at all homologies tested,
approxi-Homology dependence of dependent and
RecA-mately two- to fivefold (Table 2). As with the stimulation
independent crossing over:The length of the interacting conferred bydnaBts, stimulation of crossing over by H2O2
homologies was a strong determinant of recombina-was relatively greater at the lower homology lengths.
tion rates. Between 50 and 400 bases, integrative recom-bination was dependent on RecA and strongly sensitive DISCUSSION to the homology length. In this sense our results agree
with previous studies of integrative recombination
be-Advantages and constraints of our assay for crossing
tween bacteriophage and plasmid DNA (King and
over:We have developed a new assay to investigate
cross-Richardson1986;Shen andHuang 1986), although ing over between independent replicons that share
lim-the absolute frequencies of recombination vary in lim-the ited regions of homology. Unlike many other assays
three analyses. At 25 bp of homology, recombination for recombination used in bacteria, this one does not
was independent of RecA, although it occurred at a low rate in the population. RecA-independent crossing over
TABLE 2 was weakly dependent on the length of sequence homol-Stimulation of crossing over by hydrogen peroxide treatment ogy in contrast to RecA-dependent recombination, which showed a strong homology dependence. The apparent Fold elevation in recombinant homology threshold for maximally efficient crossing Homology (bp) frequency by 5 mmH2O2(range) over also was lower for RecA-independent than for
RecA-dependent recombination:ⵑ50vs.150 bp,
respec-25 5.1 (1.8–9.0)
tively.
51 3.0 (2.2–4.1)
104 1.9 (1.5–2.3) RecA-dependent crossing over occurs primarily via
158 1.9 (1.1–3.6) the RecFOR pathway and may be initiated by gap repair: 211 1.9 (0.8–2.8) RecA-dependent recombination contributes to most
411 1.6 (1.5–1.8)
crossing over we observed with homologies⬎50 bp in Four independent cultures in exponential growth phase length. This recombination appeared to be initiated via were split and treated with H2O2 or not and recombinant the RecFOR pathway and not via the RecBCD pathway.
frequency was determined in the cultures. All dilutions and This is probably because our substrate plasmids lack Chi plating were performed in duplicate. The average of the
sites that potentiate the recombinogenic activity and treated cultures is expressed relative to the untreated cultures.
diminish the degradative activity of RecBCD. The hyper-The range of values from lowest to highest is also given in
RecBCD suggests that linear DNA is generatedin vivo
but restrained from recombinational interactions by degradation by RecBCD if the molecules lack Chi sites. The RecFOR pathway is presumed to be a mechanism for recombination initiated at ssDNA gaps. The RecFOR pathway contributes to approximately one-half of the crossing over detected between E. coli sister chromo-somes (SteinerandKuempel1998) and to about one-half of the recombination between direct repeats in Salmonella (Galitski and Roth 1997). In the latter case, dependence onrecFwas diminished with the intro-duction of a DSB (GalitskiandRoth1997), suggesting that the RecFOR pathway recombination was primarily occurring via a gap-repair mechanism. In our two-plas-mid crossing over assay,ⵑ10% of RecA-dependent re-combination was independent of the RecFOR pathway, as well as the alternate RecBCD pathway. With respect to its dependence on RecF, our two-plasmid crossing over assay resembles chromosomal gap repair observed by
Figure6.—Resolution of Holliday junctions and crossing physical analysis after UV-irradiation: likewise, its major
over. The branched Holliday junction can be resolved in two component is RecF dependent, RecBCD independent
ways. (A) Cleavage of the outside strands and religation pro-and its minor component is both RecF pro-and RecBCD duce a crossover product. (B) Cleavage of the inside strands independent (Wang and Smith 1984). Whether the and religation produce noncrossover products. The hyperre-combinational phenotype ofruvCmutants may be explained RecF- and RecBCD-independent recombination utilizes
by the preferential RuvC cleavage as in B. an unknown function or merely represents the
propor-tion of initiating DNA molecules that do not require further processing is unknown.
found to be stimulated by inactivation of RuvC,
sug-Holliday junction resolution and crossing over:
Mo-gesting that both modes of recombination share a criti-lecular models of recombination (Holliday1964;
Mes-cal branched intermediate. elsonandRadding1975;Szostaket al.1983) presume
The possibility remains that an unknown activity in E.
that appropriate endonucleolytic resolution of Holliday
coliresolves Holliday junctions and gives rise to recombi-junctions generates reciprocal exchanges, seen as
ge-nants in our assay. If so, this activity is at least partially netic crossovers between chromosomes (Figure 6A).
dependent on RecG branch migration helicase since mu-Alternatively, apparent reciprocal exchanges can be
tations inrecGproduce a modest reduction in crossing generated by sequential nonreciprocal exchanges.
Al-over both inruvC⫹andruvCmutant backgrounds. With though the selection imposed by our system demands a
respect to its peculiar response toruv,recG, orruv recG
crossover and does involve reciprocal exchange between
mutations as well as a dependence on recF, our two-two DNA molecules, the only Holliday junction
resolv-plasmid crossing over assay shares a striking resem-ing activity expressed normally inE. coli, RuvC,was not
blance to the recombinational repair pathway reported required. In fact,ruvCmutants were
hyperrecombino-for etheno-adducts in DNA (Pandyaet al.2000). genic. The hyperrecombinational phenotype of ruvC
To explain our observations we favor a gap-filling mutants suggests that RuvC normally cleaves critical
in-mechanism for crossing over similar to that shown in termediates, removing their potential to give rise to
Figure 7 that generates a Holliday junction intermediate recombinants scored by our assay. The simplest
explana-concomitant with a template switch. Since one DNA tion is that RuvC biases resolution of Holliday junctions
strand of this structure is already recombinant, replica-toward the noncrossover configuration in certain
recombi-tion of the intermediate without resolurecombi-tion could gener-nation reactions (CromieandLeach2000;Michelet al.
ate both crossover and noncrossover products. Or an 2000; as in Figure 6B). Another possibility is that stalled
unknown resolvase may generate these products by the replication initiates crossing over by a gap-filling
mecha-appropriate strand cleavage. nism; however, a competing reaction, which does not
An alternative model is that seemingly reciprocal yield recombinants, could be retrograde fork movement
products detected by our assay may be generated by to produce a Holliday-junction-like structure. The
cleav-exchanges that do not involve formation of Holliday age of this structure by RuvC, as demonstrated by Michel
junctions. For example, as shown in Figure 8, linear and collaborators (reviewed inMichel2000), and the
DNA generated by cleavage or by rolling circle replica-destruction of the resultant linear DNA by RecBCD
tion could invade the second partner, producing the could preclude recombination in our assay. Both
Figure 8.—Two rounds of nonreciprocal recombination can generate a reciprocal product. (A) Linear plasmid DNA may be generated by cleavage or by rolling circle replication Figure7.—Initiation of crossing over by gap-filling
as shown here. (B) This linear end invades the second recom-bination. The genetic dependence of crossing over as detected
bination partner. (C) Cleavage of the invaded molecule can in our assay is reminiscent of gap-filling repair. (A) The initial
produce a second linear end in the region of homology be-gapped substrate. (B) Strand invasion between the be-gapped
tween the two plasmids, as shown. An alternative, not shown, DNA and homologies present on the second plasmid. (C) A
is that the invading DNA strand primes DNA synthesis whose template switch allows the gapped material to be synthesized
termination generates the new end. (D) A second recombina-from the other partner and the displaced strand is cleaved.
tion event occurs by invasion of the linear end into a region (D) This forms a Holliday junction intermediate. (E)
Replica-of homology. This second event could also arise by single-tion of this intermediate without resolusingle-tion can produce both
strand annealing of the two ends if the plasmid DNA marked crossover and noncrossover products. Alternatively, Holliday
in boldface has been liberated from the circle, providing a junction resolution produces either product.
resected second linear end. (E) Strand cleavage and ligation generate a crossover product.
cleavage or by fork collapse could produce a second
directly by reactive oxygen species (Imlay et al. 1988) linear end that invades or anneals at homology to form
or by subsequent enzymatic processing of oxidative le-the circular crossover product. We think this type of
sions in DNA (Ondaet al.1999). Whether or not intrin-mechanism is less likely to explain our observed
recom-sic replication arrest or oxidative damage contributes bination since it invokes two linear intermediates that
substantially to spontaneous recombination rates re-should be unstable to RecBCD degradation. However,
mains to be investigated. Mutants or treatments that such a mechanism may explain the very high levels of
produce lower spontaneous crossing over rates may re-crossovers observed inrecB mutants.
veal which processes initiate recombination during
nor-Initiation of recombination:Defects in replication or
mal cell growth. oxidative damage elevate crossing over detected with
our two-plasmid assay. A temperature-sensitive mutation We thank Robert Lloyd (University of Nottingham), Richard Ko-lodner (University of California, San Diego), and Mary Berlyn (E. coli of the DnaB helicase, at its permissive temperature of
Genetic Stock Center, Yale University) for bacterial strains. This work
growth, and hydrogen peroxide treatment promoted
was supported by a grant from the General Medical Institute of the
exchange that was particularly evident at the lower ho- National Institutes of Health (GM-51753) and a summer fellowship mologies tested. This result emphasizes the importance from the Howard Hughes Foundation to M.A.L.
of examining recombination between substrates of lim-ited homology, since at larger homologies initiation may
LITERATURE CITED not be rate limiting. Mutations indnaBhave been found
to promote both RecA-dependent and RecA-indepen- Anderson, D. G., andS. C. Kowalczykowski, 1997 The translocat-ing RecBCD enzyme stimulates recombination by directtranslocat-ing RecA
dent genetic rearrangements (Saveson and Lovett
protein onto ssDNA in a chi-regulated manner. Cell90:77–86.
1997, 1999) and can result in chromosome breakage Bachmann, B. J., 1996 Derivations and genotypes of some mutant (Michelet al.1997). Presumably, in this mutant replica- derivatives ofEscherichia coliK-12, pp. 2460–2488 inEscherichia coli and Salmonella: Cellular and Molecular Biology, edited by F. C.
tion frequently arrests and replication gaps initiate
cross-Neidhardt. ASM Press, Washington, DC.
ing over between replicons. It has been noted previously Bolivar, F., R. Rodriquez, P. Greene, M. Betlag, H. Heyneker that oxidative damage stimulates recombination within et al., 1977 Construction and characterization of new cloning
vehicles. II. A multipurpose cloning system. Gene2:95–113.
bacteria (Chen and Bernstein 1987; DeRose and
Bzymek, M., andS. T. Lovett, 2001 Instability of repetitive DNA
Claycamp1991;SteinerandKuempel1998;Ondaet sequences: the role of replication in multiple mechanisms. Proc.
Natl. Acad. Sci. USA98:8319–8325.
Chang, A. Y. C., andS. N. Cohen, 1978 Construction and character- Michel, B., 2000 Replication fork arrest and DNA recombination.
Trends Biochem. Sci.25:173–178.
ization of amplifiable multicopy DNA cloning vehicles derived
Michel, B., S. D. EhrlichandM. Uzest, 1997 DNA double-strand from the P15A cryptic miniplasmid. J. Bacteriol.134:1141–1156.
breaks caused by replication arrest. EMBO J.16:430–438.
Chen, D., andC. Bernstein, 1987 Recombinational repair of
hydro-Michel, B., G. D. Recchia, M. Penel-Colin, S. D. EhrlichandD. J.
gen peroxide-induced damages in DNA of phage T4. Mutat. Res.
Sherratt, 2000 Resolution of Holliday junctions by RuvABC
184:87–98.
prevents dimer formation inrepmutants and UV-irradiated cells.
Churchill, J. J., D. G. AndersonandS. C. Kowalczykowski, 1999
Mol. Microbiol.37:180–191. The RecBC enzyme loads RecA protein onto ssDNA
asymmetri-Miller, J. H., 1992 A Short Course in Bacterial Genetics. Cold Spring cally and independently of chi, resulting in constitutive
recombi-Harbor Laboratory Press, Cold Spring recombi-Harbor, NY. nation activation. Genes Dev.13:901–911.
Onda, M., K. Hanada, H. KawachiandH. Ikeda, 1999 Escherichia Clark, A. J., 1973 Recombination deficient mutants ofE. coliand
coli mutMsuppresses illegitimate recombination induced by oxi-other bacteria. Annu. Rev. Genet.7:67–86.
dative stress. Genetics151:439–446.
Clark, A. J., 1991 recgenes and homologous recombination
pro-Pandya, G. A., I. Y. Yang, A. P. GrollmanandM. Moriya, 2000 teins inEscherichia coli.Biochimie73:523–532.
Escherichia coli responses to a single DNA adduct. J. Bacteriol.
Cohen, A., andA. Laban, 1983 Plasmidic recombination in
Esche-182:6598–6604.
richia coliK-12: the role ofrecFgene function. Mol. Gen. Genet.
Saveson, C. J., andS. T. Lovett, 1997 Enhanced deletion formation 189:471–474.
by aberrant DNA replication in Escherichia coli. Genetics 146: Cromie, A. G., andR. D. Leach, 2000 Control of crossing over.
457–470. Mol. Cell6:815–826.
Saveson, C. J., andS. T. Lovett, 1999 Tandem repeat
recombina-Csonka, L., and A. J. Clark, 1979 Deletions generated by the
tion induced by replication fork defects inEscherichia colirequires transposon Tn10in thesrl recAregion of theEscherichia coliK-12
a novel factor, RadC. Genetics152:5–13.
chromosome. Genetics93:321–343.
Shan, Q., J. M. Bork, B. L. Webb, R. B. InmanandM. M. Cox, 1997
Dabert, P., S. D. EhrlichandA. Gruss, 1992 Chi sequence protects
RecA protein filaments: end-dependent dissociation from ssDNA
against RecBCD degradation of DNAin vivo.Proc. Natl. Acad.
and stabilization by RecO and RecR proteins. J. Mol. Biol.265:
Sci. USA89:12073–12077.
519–540.
DeRose, C. M., andH. G. Claycamp, 1991 Oxidative stress effects on
Sharples, G. J., S. N. Chan, A. A. Mahdi, M. C. WhitbyandR. G.
conjugational recombination and mutation in catalase-deficient
Lloyd, 1994 Processing of intermediates in recombination and
Escherichia coli.Mutat. Res.255:193–200. DNA repair: identification of a new endonuclease that specifically
Dixon, D. A., andS. C. Kowalczykowski, 1993 The recombination cleaves Holliday junctions. EMBO J.13:6133–6142.
hotspot chi is a regulatory sequence that acts by attenuating the Shen, P., andH. V. Huang, 1986 Homologous recombination in
nuclease activity of theE. coliRecBCD enzyme. Cell73:87–97. Escherichia coli:dependence on substrate length and homology.
Dower, W. J., J. F. MillerandC. W. Ragsdale, 1988 High efficiency Genetics112:441–457.
transformation ofE. coliby high voltage electroporation. Nucleic Shurvinton, C. E., R. G. Lloyd, F. E. BensonandP. V. Attfield,
Acids Res.16:6127–6145. 1984 Genetic analysis and molecular cloning of theEscherichia
Galitski, T., andJ. R. Roth, 1997 Pathways for homologous recom- coli ruvgene. Mol. Gen. Genet.194:322–329.
bination between chromosomal direct repeats inSalmonella typhi- Singer, M., T. A. Baker, G. Schnitzler, S. M. Deischel, M. Goel
murium.Genetics146:751–767. et al., 1989 A collection of strains containing genetically linked
Holliday, R., 1964 A mechanism for gene conversion in fungi. alternating antibiotic resistance elements for genetic mapping
Genet. Res.5:282–304. ofEscherichi coli.Microbiol. Rev.53:1–24.
Horii, Z. I., and A. J. Clark, 1973 Genetic analysis of the RecF Steiner, W. W., andP. L. Kuempel, 1998 Sister chromatid exchange
pathway of genetic recombination inEscherchia coliK12: isolation frequencies inEscherichia colianalyzed by recombination at the
and characterization of mutants. J. Mol. Biol.80:327–344. dif resolvase site. J. Bacteriol.180:6269–6275.
Horiuchi, T., andY. Fujimura, 1995 Recombinational rescue of Szostak, J. W., T. L. Orr-Weaver, R. J. RothsteinandF. W. Stahl,
the stalled DNA replication fork: a model based on analysis of 1983 The double-strand-break repair model for recombination.
anEscherichia colistrain with a chromosome region difficult to Cell33:25–35.
replicate. J. Bacteriol.177:783–791. Umezu, K., andR. D. Kolodner, 1994 Protein interactions in
ge-Imlay, J. A., S. M. ChinandS. Linn, 1988 Toxic DNA damage by netic recombination in Escherichia coli. Interactions involving RecO and RecR overcome the inhibition of RecA by
single-hydrogen peroxide through the Fenton reactionin vivoandin
stranded DNA-binding protein. J. Biol. Chem.269:30005–30013.
vitro.Science240:640–642.
Umezu, K., N. W. ChiandR. D. Kolodner, 1993 Biochemical
inter-King, S. R., andJ. P. Richardson, 1986 Role of homology and
action of theEscherichia coliRecF, RecO, and RecR proteins with pathway specificity for recombination between plasmids and
bac-RecA protein and single-stranded DNA binding protein. Proc.
teriophage lambda. Mol. Gen. Genet.204:141–147.
Natl. Acad. Sci. USA90:3875–3879.
Kolodner, R., R. A. FishelandM. Howard, 1985 Genetic
recombi-Viswanathan, M., J. J. Lacirignola, R. L. HurleyandS. T. Lovett, nation of bacterial plasmid DNA: effect of RecF pathway
muta-2000 A novel mutational hotspot in a natural quasipalindrome
tions on plasmid recombination inEscherichia coli.J. Bacteriol.
inEscherichia coli.J. Mol. Biol.302:553–564.
163:1060–1066.
Wang, T. V., andK. C. Smith, 1984 recF-dependent andrecF recB
-Kowalczykowski, S. C., 2000 Initiation of genetic recombination
independent DNA gap-filling repair processes transfer dimer-and recombination-dependent replication. Trends Biochem. Sci.
containing parental strands to daughter strands inEscherichia coli
25:156–165.
K-12uvrB.J. Bacteriol.158:727–729.
Kuzminov, A., E. SchabtachandF. W. Stahl, 1994 ⫺sites in
Webb, B. L., M. M. CoxandR. B. Inman, 1997 Recombinational combination with RecA protein increase the survival of linear
DNA repair: the RecF and RecR proteins limit the extension of DNA inE. coliby inactivating exoV activity of RecBCD nuclease.
RecA filaments beyond single-strand DNA gaps. Cell91:347–356.
EMBO J.3:2764–2776.
West, S. C., 1992 Enzymes and molecular mechanisms of genetic
Lea, D. E., andC. A. Coulson, 1949 The distribution of the numbers
recombination. Annu. Rev. Biochem.61:603–640.
of mutants in bacterial populations. J. Genet.49:264–285.
West, S. C., 1997 Processing of recombination intermediates by the
Mandal, T. N., A. A. Mahdi, G. J. SharplesandR. G. Lloyd, 1993
RuvABC proteins. Annu. Rev. Genet.31:213–244.
Resolution of Holliday intermediates in recombination and DNA Whitby, M. C., L. RyderandR. G. Lloyd, 1993 Reverse branch
repair: indirect suppression ofruvA,ruvB, andruvCmutations. migration of Holliday junctions by RecG protein: a new
mecha-J. Bacteriol.175:4325–4334. nism for resolution of intermediates in recombination and DNA
McFarlane, R. J., andJ. R. Saunders, 1996 Molecular mechanisms repair. Cell75:341–350.
of intramolecular recombination-dependent recircularization of Willetts, N. S., A. J. ClarkandB. Low, 1969 Genetic location of
linearized plasmid DNA inEscherichia coli:requirements for the certain mutations conferring recombination deficiency in
Esche-ruvA,ruvB,recG,recFandrecRgene products. Gene177:209–216. richia coli.J. Bacteriol.97:244–249.
Meselson, M., andC. Radding, 1975 A general model for genetic
