Download What makes the bacteriophage λ Red system useful for genetic

FEMS Microbiology Letters 201 (2001) 9^14
www.fems-microbiology.org
MiniReview
What makes the bacteriophage V Red system useful for genetic
engineering: molecular mechanism and biological function
Anthony R. Poteete *
Department of Molecular Genetics and Microbiology, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA 01655, USA
Received 20 March 2001; received in revised form 14 May 2001; accepted 15 May 2001
First published online 14 June 2001
Abstract
Recent studies have generated interest in the use of the homologous recombination system of bacteriophage V for genetic engineering. The
system, called Red, consists primarily of three proteins: V exonuclease, which processively digests the 5P-ended strand of a dsDNA end ;
L protein, which binds to ssDNA and promotes strand annealing; and Q protein, which binds to the bacterial RecBCD enzyme and inhibits
its activities. These proteins induce a `hyper-rec' state in Escherichia coli and other bacteria, in which recombination events between DNA
species with as little as 40 bp of shared sequence occur at high frequency. Red-mediated recombination in the hyper-rec bacterium proceeds
via a number of different pathways, and with the involvement of different sets of bacterial proteins, depending in part on the nature of the
recombining DNA species. The role of high-frequency double-strand break repair/recombination in the life cycle of the lambdoid phages is
discussed. ß 2001 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
Keywords : Bacteriophage V; Homologous recombination ; Genetic engineering; V Red system
1. Introduction
Escherichia coli normally engages in homologous recombination cautiously. The characteristics of its recombination system suggest that genetic exchange may be almost a
side e¡ect of its main functions, namely, restoring collapsed replication forks, repairing damage-induced double-strand breaks, and maintaining the genetic integrity
of its chromosome [1,2].
The static and conservative picture of homologous recombination in E. coli changes radically when the bacterium is host to lytically replicating bacteriophage V. In the
last minutes of its existence, the infected cell is a hotbed of
genetic exchange. In a single round of lytic growth, each
progeny phage is estimated to undergo approximately one
exchange in its lineage [3], despite a chromosome length of
less than 50 kb.
The V-induced `hyper-rec' state is not, in itself, harmful
to the bacterium. It can be induced, either permanently or
transiently, in the absence of other phage processes. Re-
combination between DNA molecules with as little as
40 bp of shared sequence is so e¤cient in such a cell
that it has been used in place of restriction enzymes and
ligase for genetic engineering [4,5].
The parts of V most responsible for the hyper-rec state
are proteins encoded by three genes, which together constitute the Red recombination system. The purpose of this
communication is to review the development of our
present understanding of the V Red system, focusing on
those aspects most directly related to the use of the Red
system in genetic engineering. An attempt is made to rationalize the properties of the system in terms of its biological function.
Recombination of phage V has been intensively studied
over several decades; indeed, such studies are the original
sources of much of our understanding of the molecular
mechanisms of genetic recombination. Readers interested
in the broader topic, or more of the original references, are
referred to previous reviews [2,3,6].
2. Components of the Red recombination system
* Tel. : +1 (508) 856^3708;
E-mail : [email protected]
The genes of the V Red system were among the earliest
0378-1097 / 01 / $20.00 ß 2001 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 1 0 9 7 ( 0 1 ) 0 0 2 4 2 - 7
FEMSLE 9996 29-6-01
10
A.R. Poteete / FEMS Microbiology Letters 201 (2001) 9^14
recombination genes to be described. Following the isolation of recombination-de¢cient recA, recB, and recC mutants of E. coli by Clark and coworkers, it was found that
V could recombine normally in these strains. Reasoning
that V must therefore encode its own recombination functions, researchers isolated red3 (recombination-defective)
V mutants, which were partially defective in homologous
recombination in a wild-type host, and grossly defective in
a recA3 host [3]. The red3 alleles were found to a¡ect two
genes, which were called redK and redL. When redK was
found to encode the previously characterized V exonuclease, and redL to encode a protein previously named L,
the names of these genes were changed to exo and L. A
third gene subsequently found to have a role in Red-mediated recombination was named Q. Despite an attempt to
standardize V genetic nomenclature by renaming the genes
exo, bet, and gam, all of the aforementioned names are
still used in the recent literature.
The genes of the Red system are clustered in the PL
operon, one of the two V operons which are expressed
early in the phage's transcriptional program (see Fig. 1).
The exo gene product, V exonuclease, monomer Mr 24 000,
processively degrades the 5P-ended strand of dsDNA, generating 3P-ended overhangs [7]. Crystallographic studies of
V exonuclease provided a rare instance in which a protein's
structure revealed much about how it works. The exonuclease is a ring-shaped trimer in solution, with a passage
through the center which can accommodate dsDNA at
one end, but only ssDNA at the other [8]. L protein,
monomer Mr 28 000, binds to ssDNA, promotes renaturation of complementary strands, and is capable of mediating strand annealing and exchange reactions in vitro [9]. It
Fig. 1. Genomic organization of the lambdoid bacteriophages. The order of genes shown is that of the integrated prophage, with `up' corresponding to `left' in the usual map representation. The major promoters
of the phage lytic cycle are indicated by arrows. Two types of homologous recombination systems, consisting of orthologs of the V Red genes,
or of the P22 erf-abc genes, have been identi¢ed [30,32].
forms rings when it is free in solution, larger rings when
bound to ssDNA, and helical ¢laments when bound to
dsDNA. These structural properties suggest that L protein
belongs to a family of recombination proteins which includes the Erf protein of Salmonella phage P22, the RecT
protein of the cryptic E. coli prophage Rac, and the Rad52
protein of eukaryotes [10]. L protein forms complexes with
V exonuclease, and modulates both its nucleolytic and recombination-promoting activities. The gam gene product
is a polypeptide of Mr 16 000, which, in the form of a
dimer, binds to the host RecBCD protein, and inhibits
all of its known activities ([11], and references therein).
3. Mechanisms of Red-mediated V recombination
In early studies of the genetics of recombination in
E. coli, Clark and coworkers found evidence of multiple
recombination pathways. Conjugational and transductional recombination normally proceeds through the RecBCD
pathway, in which the functions of recB and recC are
essential (the recD-encoded subunit, discovered much later, is not essential). If the RecBCD pathway was blocked
by mutation of either gene, further mutation of the bacterium could induce either the RecE or the RecF pathway,
restoring recombination pro¢ciency. V strains lacking the
Red system could engage in homologous recombination
via any of these pathways. Conversely, V strains possessing
the Red system could engage in homologous recombination independent of any single bacterial recombination
function. Red recombination was thus considered to be
a separate pathway.
Studies of the Red pathway in its proper context, a cell
undergoing the V lytic cycle, are experimentally challenging, for a number of reasons. (i) It is necessary to distinguish Red-mediated recombination from other recombination mechanisms available to the phage. (ii) In the V lytic
cycle, DNA replication and recombination are interdependent-each stimulates the other. (iii) Tracing the rearrangement of DNA atoms in crosses is practical only
under conditions of restrained, or blocked, DNA replication. However, under these conditions, in which the `rolling circle' mode of synthesis is blocked, production of a
maturable phage chromosome precursor ^ hence, of progeny phage to analyze ^ depends upon recombination. Experimentally disentangling replication and the several
modes of recombination in a V infection, Stahl and coworkers established key mechanistic aspects of the Red
pathway: the involvement of dsDNA ends, made partially
single-stranded by the action of V exonuclease; parts of
the relationship between replication, chromosome packaging, and recombination; and the involvement of RecA
protein [6].
Red-mediated recombination in the context of the V lytic
cycle is best understood as proceeding via two di¡erent
mechanisms: invasion and annealing (see Fig. 2). In
FEMSLE 9996 29-6-01
A.R. Poteete / FEMS Microbiology Letters 201 (2001) 9^14
Fig. 2. Pathways of Red-mediated double-strand break repair. Invasion
(left) is e¤cient only in the presence of RecA, whereas annealing is e¤cient in the absence of RecA [12].
both mechanisms, recombination is initiated by a dsDNA
end, which is resected by L-modulated V exonuclease, producing a 3P overhang. If the only partner available for
recombination with the dsDNA end is an unbroken circular homologous duplex, then recombination proceeds via
strand invasion, and is dependent upon the bacterial RecA
protein. If the partner is replicating, or has a doublestrand end at a genetically di¡erent location, then recombination can proceed by L protein-mediated annealing,
which is RecA-independent [12].
4. Red-mediated recombination out of the phage context
Even before the red genes were identi¢ed, there were
indications that a V-encoded recombination function could
promote bacterial recombination (see discussion in reference [13]). Various studies of V-infected cells suggested
that Red could substitute for RecBCD in promoting bacterial recombination, and that this recombination was
RecA- and RecF-dependent [14,15].
Studies of Red-mediated bacterial recombination were
simpli¢ed with the cloning of the V red genes. The genes
gam, bet, and exo, expressed at moderate levels under
control of the lac promoter from a multicopy plasmid,
were found to be non-toxic to E. coli. The plasmid-encoded Red system could restore recombination pro¢ciency
in conjugational crosses to a recB recC E. coli strain;
however, as this Red-mediated recombination was dependent upon RecF function (unlike recombination in wild-
11
type E. coli), it was clear that the Red system did not
simply substitute for RecBCD [16].
The RecF dependence of Red-mediated conjugational
recombination was consistent with a well-known pattern
of similarities between the Red and RecE pathways. The
genes of the RecE pathway, recE and recT, are encoded by
a cryptic prophage, called Rac, which is found in certain
strains of E. coli. Like exo, recE encodes an exonuclease ;
like bet, recT encodes a ssDNA-binding strand-exchange
protein [17].
Plasmid-encoded red functions were found to promote
recombination events between non-replicating homologs
at a high e¤ciency (up to 20% of the input parental chromosomes converted to recombinants). Consistent with the
Stahl model for Red-mediated recombination by invasion
[18], this high-e¤ciency recombination occurred only
when (1) one of the recombining partners had a doublestrand break ; (2) V exonuclease and L protein were both
present; (3) RecA protein was present; and (4) RecBCD
was inactivated, either by mutation or by Q protein [19].
Another technical advance in the study of Red-mediated
bacterial recombination came from replacing the E. coli
recC-ptr-recB-recD gene cluster with the red genes [20].
In the resulting strain, unlike a cell bearing a multicopy
plasmid with red genes, it was possible to knock out most
of the known recombination genes of E. coli, while maintaining strain viability. The contributions of the encoded
recombination proteins to Red-mediated recombination
could then be readily assessed. A surprising ¢nding from
this line of experiment was that knocking out the RecE
(and RecF) pathway genes recJ or recG increased the e¤ciency of Red-mediated recombination. The degree of dependence of Red-mediated recombination on other bacterial genes was found to vary with the nature of the
recombining DNA species, as well as with the presence
or absence of RecG [20^22]. These observations, and
others described below, indicated clearly that the two
Red pathways invoked by Stahl and coworkers ^ invasion
and annealing ^ are themselves multiply branched downstream.
The Red-mediated break-join recombination event for
which we have, at present, the clearest description, occurs
between a linear dsDNA species and the bacterial chromosome, in a strain in which the recC-ptr-recB-recD gene
cluster has been replaced with the genes of the Red system,
and recG has been deleted. Variants of this strain are infected with a suicide vector version of phage V. The
phage's chromosome, which is unable to transcribe its lytic
genes or replicate, is cut by a cellular restriction endonuclease, releasing a linear dsDNA. The linear molecule,
bearing the gene for chloramphenicol acetyltransferase
(cat) £anked on either side by 1.3 kb of lac operon sequences, can recombine with the bacterial chromosome,
generating a chloramphenicol-resistant, Lac3 recombinant
(Fig. 3).
The production of recombinants in these crosses be-
FEMSLE 9996 29-6-01
12
A.R. Poteete / FEMS Microbiology Letters 201 (2001) 9^14
tween a phage-delivered linear dsDNA and the bacterial
chromosome depends on recA, recF, recO, recR, recQ,
ruvAB, and ruvC; it is independent of recN ([22]; unpublished results). Fig. 4 shows a recombination mechanism
based on these observations, as well as on extensive studies
of E. coli recombination functions. This mechanism accounts fairly well for events in the absence of RecG; in
the presence of RecG, this particular pathway is partly
impeded, but, evidently, others become available [21].
5. Use of Red in genetic engineering
Murphy [20] found that the Red system, plasmid- or
chromosome-encoded, would promote recombination between the bacterial chromosome and linear dsDNA molecules introduced into the cell via electroporation. The
Red system was far more e¤cient than systems used previously for making gene replacements in E. coli. PCR-generated DNA species could be used in this reaction.
The Red system works in other bacteria as well. It has
been used to make gene replacements in Salmonella
(E. Kofoid and J. Roth, personal communication), as
well as enteropathogenic E. coli and enterohemorrhagic
E. coli (K. Murphy, personal communication).
All of the research described above involved recombination between DNA segments with identical sequences of
several hundred bases or more. For purposes of genetic
engineering, recombination reactions based on short DNA
sequences, of a size readily synthesized, is far more useful.
Stewart and coworkers [4] showed that the RecET system,
as well as the Red system, would promote such events.
Fig. 3. Chromosomal gene replacement by a linear dsDNA fragment, either released from an infecting phage by the action of an endogenous
restriction endonuclease, or introduced directly into the cell by electroporation.
Fig. 4. Red-mediated replacement of lac with cat in the chromosome of
E. coli recGv. The depicted molecular events would have to take place
on both sides of the cat gene to generate a recombinant ; for clarity,
only one side is shown. The double-stranded end initiates recombination. V exonuclease processively digests the 5P-ended strand, leaving a
3P-ended single-stranded tail. The combined action of RecA and the V L
protein mediates invasion of the 3P-ended strand into an unbroken homologous duplex. RecFOR is a key participant in the overall reaction
pathway ; its properties suggest that it is involved either in loading or
unloading RecA (and L?) from the joint molecule. Once a threestranded junction is formed, the crossed strands are subject to RuvAB
and/or RecQ helicase-driven branch migration, resulting in a Holliday
junction, which can be resolved by RuvC into a recombinant molecule.
Their experimental approach involved co-electroporation
of an E. coli strain expressing the RecET system, or the
Red system, with a plasmid and a linear DNA species. The
linear species was PCR-generated, and encoded a selectable drug resistance marker. The PCR primers included up
to 60 bases of plasmid sequences, which, as homologous
£anks in the PCR products, promoted recombination with
the plasmid, generating a new plasmid in which the drug
resistance marker had replaced speci¢ed plasmid sequences. In further studies of this system, it was found that the
functionally analogous RecE/RecT and RedK/RedL pairs
interact speci¢cally with their partners, that is, the RecE
exonuclease would not work with RedL strand exchanger,
nor would V exonuclease (RedK) work with RecT [23].
Two groups of researchers developed e¤cient methods,
based on the use of the Red system, for replacing genes in
the E. coli chromosome. The Court group's method involved a brief heat induction, followed by rapid cooling,
of a defective V prophage under control of the thermosensitive repressor encoded by the cI857 allele. Electroporation of the transiently induced V lysogen with a linear
FEMSLE 9996 29-6-01
A.R. Poteete / FEMS Microbiology Letters 201 (2001) 9^14
dsDNA bearing 30^50-bp homologous £anks resulted in
e¤cient recombination with the E. coli chromosome. This
recombination was found to be dependent upon the three
genes of the Red system, but not other V genes [24]. The
Wanner group obtained high-e¤ciency recombination between the E. coli chromosome and electroporated short£ank linear DNA species by placing gam, bet, and exo
under control of the ParaB promoter on a low-copy plasmid, and inducing with arabinose [25].
The Red system also promotes high-frequency recombination between the E. coli chromosome and short singlestrand oligonucleotides, resulting in gene conversion (H.
Ellis and D. Court, personal communication; F. Stewart,
personal communication). The only V function required
for this activity is bet. Investigators hypothesize that
L protein promotes annealing between the oligonucleotide
and chromosomal DNA made transiently single-stranded
during replication. Mismatch repair, or actual incorporation of the base-paired oligonucleotide into one of the
replication products, could then lead to gene conversion.
A new mechanistic question emerges from studies of the
Red system in genetic engineering : How do short-homology (e.g. 40 bp) and long-homology (1000 bp) break-join
recombination di¡er? Red-promoted gene replacement by
linear dsDNAs with long homologous £anks is e¤cient in
recBCD3 cells expressing V exonuclease and L protein.
Three additional factors appear to increase the relative
frequency of chromosomal gene replacement with short
homologous £anks: Q protein, heat shock, or the presence
in the cell of a low-copy plasmid ([20,24,25]; K. Murphy,
personal communication). This stimulation by Q protein,
even in a cell in which RecBCD has been eliminated, suggests an involvement of SbcCD protein, the other known
target of Q [26]. SbcCD is an endonuclease which cleaves
DNA palindromes [27]. Perhaps it attacks a structure
which is formed in the course of short-homology recombination, but not in long-homology recombination. The
stimulation of Red-mediated short-homology recombination by heat shock or by the presence of certain plasmids
in the cell is mechanistically obscure at present.
6. Modulation of Red-mediated recombination by other V
proteins
The red genes are the only known V functions required
for recombination in wild-type E. coli or in mutants lacking any of the bacterium's known recombination genes.
However, functions encoded by other V genes may have
signi¢cant roles in Red-mediated recombination.
The orf gene encodes a protein which is able to substitute for the RecFOR complex in promoting phage (but
not bacterial) recombination in the absence of the Red
system [28]. The rap gene encodes an endonuclease with
resolvase-like properties [30]. Studies by Tarkowski et al.
(cited in reference [6]) suggest that the orf and rap func-
13
tions may focus Red-mediated recombination events, causing crossovers to occur near the initiating double-strand
break. One or both of these functions may be able to substitute for RuvC and, partially, for RecFOR in promoting
Red-mediated gene replacement reactions in E. coli [22].
Whether yet other V gene products in£uence Red-mediated recombination is unknown. The observation that
Red-mediated bacterial gene replacement recombination
is enhanced in recJ and recG mutants [20,21] might be
interpreted as a hint that some V function(s) alter the
activities of RecJ and RecG proteins.
7. Biological function of Red
Is the Red system important to V? The importance is
not immediately obvious: a V mutant lacking exo and bet
is not much impaired for lytic growth or lysogeny in wildtype E. coli. However, several features of V biology point
to a central role for the highly e¤cient double-strand
break repair/recombination that possession of the Red system provides.
1. Recombination stimulates V DNA replication. Redand RecA-mediated invasion of a duplex DNA by 3Pended strands can directly prime DNA synthesis, in
theory, but this has not been demonstrated [6]. Kuzminov has suggested that recombination may stimulate
replication by joining greater-than-unit-chromosomelength linear V DNA pieces into circles, which serve
as more e¤cient templates for replication initiation ;
and that Red-mediated annealing-type recombination
between linear V pieces may generate very long multimers, which would be more e¤ciently packaged into
phage capsids [2].
2. A phage chromosome, in moving from one host cell to
another, is likely to encounter restriction systems, and
experience double strand breaks. An e¤cient system for
repairing double strand breaks might have greater utility for V than for E. coli.
3. Prophage V is launched into its lytic mode, which includes expression of the red genes, by agents which
damage DNA. It is therefore likely that most non-passive replication of V in nature takes place under conditions in which rates of DNA damage are high. In
these circumstances, an e¤cient recombinational repair
system like Red could be especially valuable.
4. V is a member of a large and diverse group of phages
which exchange blocks of genes by genetic recombination [29]. DNA sequence data indicate that recombination is quite active in the evolution of these phages.
Each di¡erent lambdoid phage can be considered a
mosaic of rapidly reshu¥ing genetic modules [30]. A
description of the lambdoid phages as a sexual species,
and of V itself as merely an individual with some parthenogenetic capability, is an exaggeration, but it sug-
FEMSLE 9996 29-6-01
14
A.R. Poteete / FEMS Microbiology Letters 201 (2001) 9^14
gests the importance of recombination to this form of
life.
5. The lambdoid phages include P22, as well as other generalized transducing phages. In this type of phage, unlike V and the other specialized transducing lambdoid
phages, homologous recombination is required for circularizing the linear chromosome which is injected into
the host bacterium upon infection [29]. P22 encodes a
recombination system functionally analogous to the
Red system. Mutants lacking this system are greatly
impaired for lytic growth and lysogeny in a wild-type
host, and completely defective in a recA3 mutant, but
are completely restored by the V Red system [31]. If the
functional modules of the lambdoid phages, rather than
individual phages, are the primary units operated on by
natural selection, then one function of the Red system
is to circularize the chromosomes of the P22-like lambdoid phages.
Acknowledgements
I thank Kenan Murphy for helpful discussions; thank
Kenan Murphy, Donald Court, Hilary Ellis, Barry Wanner, and John Roth for communicating unpublished results ; and apologize to the authors of key references which
were eliminated to ¢t the format constraints of this journal. Research on V recombination in my laboratory was
supported by Grant R01 GM 51609 from the United
States National Institutes of Health.
References
[1] Myers, R.S. and Stahl, F.W. (1994) Chi and the RecBCD enzyme of
Escherichia coli. Annu. Rev. Genet. 28, 49^70.
[2] Kuzminov, A. (1999) Recombinational repair of DNA damage in
Escherichia coli and bacteriophage V. Microbiol. Mol. Biol. Rev.
63, 751^813.
[3] Signer, E. (1971) General recombination. In: The Bacteriophage
Lambda (Hershey, A.D., Ed.), pp. 139^174. Cold Spring Harbor
Laboratory, Cold Spring Harbor, NY.
[4] Zhang, Y., Buchholz, F., Muyrers, J.P.P. and Stewart, A.F. (1998) A
new logic for DNA engineering using recombination in Escherichia
coli. Nature Genet. 20, 123^128.
[5] Muyrers, J.P.P., Zhang, Y., Testa, G. and Stewart, A.F. (1999) Rapid
modi¢cation of bacterial arti¢cial chromosomes by ET-recombination. Nucleic Acids Res. 27, 1555^1557.
[6] Stahl, F.W. (1998) Recombination in phage V: one geneticist's historical perspective. Gene 223, 95^102.
[7] Little, J.W. (1967) An exonuclease induced by bacteriophage V. II.
Nature of the enzymic reaction. J. Biol. Chem. 242, 679^686.
[8] Kovall, R. and Matthews, B.W. (1997) Toroidal structure of lambdaexonuclease. Science 277, 1824^1827.
[9] Li, Z., Karakousis, G., Chiu, S.K., Reddy, G. and Radding, C.M.
(1998) The beta protein of phage lambda promotes strand exchange.
J. Mol. Biol. 276, 733^744.
[10] Passy, S.I., Yu, X., Li, Z., Radding, C.M. and Egelman, E.H. (1999)
Rings and ¢laments of L protein from bacteriophage V suggest a
superfamily of recombination proteins. Proc. Natl. Acad. Sci. USA
96, 4279^4284.
[11] Murphy, K.C. (1991) V gam protein inhibits the helicase and M-stimulated recombination activities of Escherichia coli RecBCD enzyme.
J. Bacteriol. 173, 5808^5821.
[12] Stahl, M.M., Thomason, L., Poteete, A.R., Tarkowski, T., Kuzminov, A. and Stahl, F.W. (1997) Annealing vs. invasion in phage V
recombination. Genetics 147, 961^977.
[13] Signer, E.R. and Weil, J. (1968) Recombination in bacteriophage V. I.
Mutants de¢cient in general recombination. J. Mol. Biol. 34, 261^
271.
[14] Weisberg, R.A. and Sternberg, N. (1974) Transduction of recB3
hosts is promoted by V red+ function. In: Mechanisms in Recombination (Grell, R.F., Ed.), pp. 107^109. Plenum Press, New York.
[15] Armengod, M.E. (1981) Role of the recF gene of Escherichia coli K12 in V recombination. Mol. Gen. Genet. 181, 497^504.
[16] Poteete, A.R. and Volkert, M.R. (1988) Activation of recF-dependent
recombination in Escherichia coli by bacteriophage V and P22-encoded functions. J. Bacteriol. 170, 4379^4381.
[17] Kowalczykowski, S.C., Dixon, D.A., Eggleston, A.K., Lauder, S.D.
and Rehrauer, W.M. (1994) Biochemistry of homologous recombination in Escherichia coli. Microbiol. Rev. 58, 401^465.
[18] Stahl, F.W., Kobayashi, I. and Stahl, M.M. (1985) In phage V, cos is
a recombinator in the Red pathway. J. Mol. Biol. 181, 199^209.
[19] Poteete, A.R. and Fenton, A.C. (1993) E¤cient double-strand breakstimulated recombination promoted by the general recombination
systems of phages V and P22. Genetics 134, 1013^1021.
[20] Murphy, K.C. (1998) Use of bacteriophage V recombination functions to promote gene replacement in Escherichia coli. J. Bacteriol.
180, 2063^2071.
[21] Poteete, A.R., Fenton, A.C. and Murphy, K.C. (1999) Roles of
RuvC and RecG in phage V Red-mediated recombination. J. Bacteriol. 181, 5402^5408.
[22] Poteete, A.R. and Fenton, A.C. (2000) Genetic requirements of
phage V Red-mediated gene replacement in Escherichia coli K-12.
J. Bacteriol. 182, 2336^2340.
[23] Muyrers, J.P.P., Zhang, Y., Buchholz, F. and Stewart, A.F. (2000)
RecE/RecT and RedK/RedL initiate double-stranded break repair by
speci¢cally interacting with their respective partners. Genes Dev. 14,
1971^1982.
[24] Yu, D., Ellis, H.M., Lee, E.-C., Jenkins, N.A., Copeland, N.G. and
Court, D.L. (2000) An e¤cient recombination system for chromosome engineering in Escherichia coli. Proc. Natl. Acad. Sci. USA 97,
5978^5983.
[25] Datsenko, K.A. and Wanner, B.L. (2000) One-step inactivation of
chromosomal genes in Escherichia coli K-12 using PCR products.
Proc. Natl. Acad. Sci. USA 97, 6640^6645.
[26] Kulkarni, S.K. and Stahl, F.W. (1989) Interaction between the sbcC
gene of Escherichia coli and the gam gene of phage V. Genetics 123,
249^253.
[27] Leach, D.R.F., Okely, E.A. and Pinder, D.J. (1997) Repair by recombination of DNA containing a palindromic sequence. Mol. Microbiol. 26, 597^606.
[28] Sawitzke, J.A. and Stahl, F.W. (1994) The phage V orf gene encodes a
trans-acting factor that suppresses Escherichia coli recO, recR, and
recF mutations for recombination of V but not of E. coli. J. Bacteriol.
176, 6730^6737.
[29] Campbell, A. and Botstein, D. (1983) Evolution of the lambdoid
phages. In: Lambda II (Hendrix, R.W. et al., Eds.), pp. 365^380.
Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
[30] Juhala, R.J., Ford, M.E., Duda, R.L., Youlton, A., Hatfull, G.F. and
Hendrix, R.W. (2000) Genomic sequences of bacteriophages HK97
and HK022: pervasive genetic mosaicism in the lambdoid bacteriophages. J. Mol. Biol. 299, 27^51.
[31] Poteete, A.R. and Fenton, A.C. (1984) Lambda red-dependent
growth and recombination of phage P22. Virology 134, 161^167.
[32] Plunkett, G., Durfee, D.J. and Blattner, F.R. (1999) Sequence of
Shiga toxin 2 phage 933W from Escherichia coli O157:H7: Shiga
toxin as a phage late-gene product. J. Bacteriol. 181, 1767^1778.
FEMSLE 9996 29-6-01