Transposition and site-specific recombination

Transposition and site-specific recombination Transposition and site-specific recombination

05.06.2014 Views

cut-and-paste mechanisms to a variety of genetic rearrangements Bernard Hallet *, David J. Sherratt Received 28 May 1997 ; revised 1 July 1997 ; accepted 1 July 1997 In bacteria, two categories of specialised recombination promote a variety of DNA rearrangements. Transposition is the process by which genetic elements move between different locations of the genome, whereas site-specific recombination is a reaction in which DNA strands are broken and exchanged at precise positions of two target DNA loci to achieve determined biological function. Both types of recombination are represented by diverse genetic systems which generally encode their own recombination enzymes. These enzymes, generically called transposases and site-specific recombinases, can be grouped into several families on the basis of amino acid sequence similarities, which, in some cases, are limited to a signature of a few residues involved in catalysis. The well characterised site-specific recombinases are found to belong to two distinct groups, whereas the transposases form a large super-family of enzymes encompassing recombinases from both prokaryotes and eukaryotes. In spite of important differences in the catalytic mechanisms used by these three classes of enzymes to cut and rejoin DNA molecules, similar strategies are used to coordinate the biochemical steps of the recombination reaction and to control its outcome. This review summarises our current understanding of transposition and site-specific recombination, attempting to illustrate how relatively conserved DNA cut-and-paste mechanisms can be used to bring about a variety of complex DNA rearrangements. ; Site- speci¢c recombinatio n ; Recomb ination mechanism ; Nucleoprote in comple x 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 2. Transpositional and site-speci¢c recombination : overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 2.1. A plethora of transposable elements with common transposition motifs . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 2.2. Site-speci¢c recombination : integration, resolution, inversion and transposition . . . . . . . . . . . . . . . . . . . . . 160 3. Three mechanisms of DNA cut-and-paste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 3.1. The DDE transposases : a universal one-step transesteri¢cation reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . 162 3.2. Site-speci¢c recombinases : two-step transesteri¢cations by distinct mechanisms . . . . . . . . . . . . . . . . . . . . . . 165 3.2.1. The resolvase/invertase family : concerted breakage and rejoining of four DNA strands . . . . . . . . . . . 166 3.2.2. The V family : sequential pairs of DNA strand exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168 0168-6445 / 97 / $32.00 ß 1997 Federation of Europ ean Microbiolog ical Societies. Published by Elsevier Science B.V. 0 1 6 8 - 6 4 4 5 ( 9 7 ) 0 0 0 5 5 - 7 FEMS Microbiology Reviews 21 (1997) 157^178 * Corresponding author. Tel. : +44 (1865) 275-290 ; fax : +44 (1865) 275-297 ; e-mail : Hallet@bioch.ox.ac.uk Transposition and site-speci¢c recombination : adapting DNA Microbiology Unit, Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU, UK Abstract Keywords : Transposon Contents Int PII S

cut-<strong>and</strong>-paste mechanisms to a variety of genetic rearrangements<br />

Bernard Hallet *, David J. Sherratt<br />

Received 28 May 1997 ; revised 1 July 1997 ; accepted 1 July 1997<br />

In bacteria, two categories of specialised <strong>recombination</strong> promote a variety of DNA rearrangements. <strong>Transposition</strong> is the<br />

process by which genetic elements move between different locations of the genome, whereas <strong>site</strong>-<strong>specific</strong> <strong>recombination</strong> is a<br />

reaction in which DNA str<strong>and</strong>s are broken <strong>and</strong> exchanged at precise positions of two target DNA loci to achieve determined<br />

biological function. Both types of <strong>recombination</strong> are represented by diverse genetic systems which generally encode their own<br />

<strong>recombination</strong> enzymes. These enzymes, generically called transposases <strong>and</strong> <strong>site</strong>-<strong>specific</strong> recombinases, can be grouped into<br />

several families on the basis of amino acid sequence similarities, which, in some cases, are limited to a signature of a few<br />

residues involved in catalysis. The well characterised <strong>site</strong>-<strong>specific</strong> recombinases are found to belong to two distinct groups,<br />

whereas the transposases form a large super-family of enzymes encompassing recombinases from both prokaryotes <strong>and</strong><br />

eukaryotes. In spite of important differences in the catalytic mechanisms used by these three classes of enzymes to cut <strong>and</strong><br />

rejoin DNA molecules, similar strategies are used to coordinate the biochemical steps of the <strong>recombination</strong> reaction <strong>and</strong> to<br />

control its outcome. This review summarises our current underst<strong>and</strong>ing of transposition <strong>and</strong> <strong>site</strong>-<strong>specific</strong> <strong>recombination</strong>,<br />

attempting to illustrate how relatively conserved DNA cut-<strong>and</strong>-paste mechanisms can be used to bring about a variety of<br />

complex DNA rearrangements.<br />

; Site- speci¢c recombinatio n ; Recomb ination mechanism ; Nucleoprote in comple x<br />

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158<br />

2. <strong>Transposition</strong>al <strong>and</strong> <strong>site</strong>-speci¢c <strong>recombination</strong> : overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158<br />

2.1. A plethora of transposable elements with common transposition motifs . . . . . . . . . . . . . . . . . . . . . . . . . . . 158<br />

2.2. Site-speci¢c <strong>recombination</strong> : integration, resolution, inversion <strong>and</strong> transposition . . . . . . . . . . . . . . . . . . . . . 160<br />

3. Three mechanisms of DNA cut-<strong>and</strong>-paste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161<br />

3.1. The DDE transposases : a universal one-step transesteri¢cation reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . 162<br />

3.2. Site-speci¢c recombinases : two-step transesteri¢cations by distinct mechanisms . . . . . . . . . . . . . . . . . . . . . . 165<br />

3.2.1. The resolvase/invertase family : concerted breakage <strong>and</strong> rejoining of four DNA str<strong>and</strong>s . . . . . . . . . . . 166<br />

3.2.2. The V<br />

family : sequential pairs of DNA str<strong>and</strong> exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168<br />

0168-6445 / 97 / $32.00 ß 1997 Federation of Europ ean Microbiolog ical Societies. Published by Elsevier Science B.V.<br />

0 1 6 8 - 6 4 4 5 ( 9 7 ) 0 0 0 5 5 - 7<br />

FEMS Microbiology Reviews 21 (1997) 157^178<br />

* Corresponding author. Tel. : +44 (1865) 275-290 ; fax : +44 (1865) 275-297 ; e-mail : Hallet@bioch.ox.ac.uk<br />

<strong>Transposition</strong> <strong>and</strong> <strong>site</strong>-speci¢c <strong>recombination</strong> : adapting DNA<br />

Microbiology Unit, Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU, UK<br />

Abstract<br />

Keywords : Transposon<br />

Contents<br />

Int<br />

PII S


158<br />

5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173<br />

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174<br />

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178<br />

4. Higher order nucleoprotein complexes : an additional level of control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170<br />

<strong>recombination</strong> by distinct biochemical mechanisms<br />

1. Introduction<br />

may show very similar outcomes.<br />

The present review attempts to give a general survey<br />

of the specialised <strong>recombination</strong> repertoire in<br />

It is now well documented that genetic information<br />

can be reshu¥ed by inversion, insertion, duplication,<br />

deletion, or translocation of DNA segments.<br />

bacteria, trying to outline the di¡erences, but also<br />

discussing some aspects which in the light of recent<br />

Whereas some rearrangements ful¢l speci¢c physiological<br />

functions or are involved in programmed<br />

studies, sound very much like `variations on a<br />

processes, others occur `spontaneously' with respect<br />

theme'. Although this review is focused on bacterial<br />

to time <strong>and</strong> position in the genome <strong>and</strong> contribute to<br />

<strong>recombination</strong> systems, those in eukaryotes show<br />

the genetic diversity of a population. In bacteria, as<br />

many similarities. Other discussion on <strong>site</strong>-speci¢c<br />

in most organisms, DNA rearrangements can be<br />

<strong>recombination</strong> <strong>and</strong> transposition can be found in<br />

recent reviews [2^10].<br />

mediated by general <strong>recombination</strong> between homologous<br />

DNA sequences, but also by a variety of specialised<br />

<strong>recombination</strong> mechanisms commonly<br />

grouped into two categories : genetic transposition<br />

2. <strong>Transposition</strong>al <strong>and</strong> <strong>site</strong>-speci¢c <strong>recombination</strong>:<br />

<strong>and</strong> conservative <strong>site</strong>-speci¢c <strong>recombination</strong>.<br />

overview<br />

Although transposition <strong>and</strong> <strong>site</strong>-speci¢c <strong>recombination</strong><br />

seem to be fundamentally distinct processes<br />

2.1. A plethora of transposable elements with common<br />

that often have very di¡erent biological outcomes,<br />

motifs<br />

transposition<br />

increasing evidence indicates that they are related<br />

Transposable elements or transposons can be described<br />

as discrete DNA segments that are able to<br />

move between di¡erent, non homologous, genomic<br />

in many ways. In general <strong>recombination</strong>, genetic material<br />

is exchanged through a cascade of events involving<br />

multiple proteins assembled in di¡erent enzymatic<br />

complexes [1]. In contrast, both categories of<br />

specialised <strong>recombination</strong> systems utilise relatively<br />

loci. <strong>Transposition</strong> of an element is thus a <strong>recombination</strong><br />

reaction involving three separate <strong>site</strong>s : the<br />

simple <strong>recombination</strong> machineries in which one (or<br />

two transposon ends <strong>and</strong> the new target locus. In<br />

sometimes two) enzyme, generically referred to as the<br />

autonomous transposable elements, the transposase<br />

transposase <strong>and</strong> <strong>site</strong>-speci¢c recombinase, catalyses<br />

is encoded by a gene located within the element.<br />

the essential DNA breakage <strong>and</strong> joining reactions.<br />

These proteins act at speci¢c DNA sequences which<br />

Transposon ends often contain inverted repeated sequences<br />

on which the transposase binds speci¢cally.<br />

are characteristic for each genetic element.<br />

Although most transposons can transpose into many<br />

di¡erent places, insertion is never totally r<strong>and</strong>om.<br />

For all systems analysed, coordination of the <strong>recombination</strong><br />

reactions requires the formation of an<br />

While a speci¢c target consensus sequence has been<br />

enzymatically competent nucleoprotein complex in<br />

found for some elements, the insertion of many<br />

which the recombinase <strong>and</strong> at least two distant<br />

transposons is also in£uenced by regional features<br />

of the target locus such as a local DNA structure,<br />

the presence host protein binding <strong>site</strong>s, or DNA<br />

DNA <strong>site</strong>s are brought together. Whereas the molecular<br />

transactions controlling the assembly of this active<br />

complex can vary from one system to another<br />

supercoiling (reviewed in Ref. [10]). The transposon<br />

Tn7 is distinguished by its ability to transpose either<br />

within the same family, unrelated systems also appear<br />

to use convergent strategies. As a consequence,<br />

to a speci¢c <strong>site</strong> of the bacterial chromosome<br />

a single <strong>recombination</strong> mechanism can be adapted to<br />

(attTn7) at a high frequency or to many di¡erent<br />

many di¡erent <strong>recombination</strong> functions. Conversely,<br />

<strong>site</strong>s at a low frequency (see below) [11].


Fig. 1. Two major modes of transposition. In non-replicative<br />

transposition, the element (rectangle) is excised from the donor<br />

locus by double-str<strong>and</strong> breaks at both ends (solid triangles) <strong>and</strong><br />

transferred to the target <strong>site</strong> (cut-<strong>and</strong>-paste mechanism). The<br />

product of intermolecular transposition by the replicative mechanism<br />

is a cointegrate molecule in which the donor <strong>and</strong> target<br />

backbones are fused. The element is duplicated during the process<br />

<strong>and</strong> each copy remains attached to the donor backbone at<br />

one end, whereas the other end is joined to the target. The cointegrate<br />

may be resolved by homologous or <strong>site</strong>-speci¢c <strong>recombination</strong><br />

between the two transposon copies to restore the initial<br />

donor molecule <strong>and</strong> yield a target copy in which the element is<br />

inserted. Both modes of transposition usually generate short target<br />

duplications £anking the element in the new target locus<br />

(open<br />

triangles).<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178 159<br />

transposition. In a non-replicative or `cut-<strong>and</strong>-paste'<br />

mechanism, the element is excised from its original<br />

location <strong>and</strong> inserted into the new target locus. This<br />

mode of transposition results in the simple insertion<br />

of the element in the target DNA (Fig. 1). The<br />

gapped donor molecule resulting from the transposon<br />

excision is either degraded or repaired by homologous<br />

<strong>recombination</strong> using a second copy of the donor<br />

as a template [12,13]. Rarely, the ends of the<br />

donor molecule may be resealed. In replicative transposition,<br />

the element is copied so that insertion of<br />

the element into the same DNA molecule leads to<br />

deletion <strong>and</strong>/or replicative inversion, while transposition<br />

from one circular molecule to another generates<br />

a cointegrate structure in which the donor <strong>and</strong><br />

target backbones are joined by directly repeated copies<br />

of the transposon at each junction. This cointegrate<br />

may be subsequently resolved by <strong>recombination</strong><br />

between the two copies of the element, giving<br />

rise to a restored donor molecule <strong>and</strong> a target molecule<br />

in which one copy of the transposon has inserted<br />

(Fig. 1) [9]. Both modes of transposition generate<br />

short target sequence duplications of a<br />

characteristic length (2^14 bp) that form direct repeats<br />

£anking the element in its new locus. These<br />

duplications are created by repair of the gaps arising<br />

from staggered cuts of the target DNA (see below).<br />

Beyond this overall description hides a broad diversity<br />

of transposable elements in bacteria. More<br />

than 400 di¡erent elements have been identi¢ed<br />

<strong>and</strong> characterised from a wide range of archaebacteria<br />

<strong>and</strong> eubacteria (J. Mahillon <strong>and</strong> M. Ch<strong>and</strong>ler,<br />

personal communication). The size <strong>and</strong> genetic organisation<br />

of bacterial transposons is highly variable.<br />

They range from relatively small <strong>and</strong> compact (0.8^2<br />

kb) insertion sequences (IS), which in the simplest<br />

form contain a single transposase gene [14], to larger<br />

(<strong>and</strong> more complex) multi-drug resistance transposons<br />

such as Tn7 (14 kb) [11], or the bacteriophage<br />

Mu (37.5 kb) which uses transposition as part of its<br />

life cycle strategy [6,15].<br />

Historically, bacterial transposable elements have<br />

been grouped into several classes on the basis of<br />

di¡erent criteria. However, it is quite clear that the<br />

largest class, the IS elements, de¢ned by the sole<br />

feature that they only encode functions required<br />

for transposition, actually represents a very heterogeneous<br />

group of genetically <strong>and</strong> functionally distinct<br />

transposable elements. The identi¢cation <strong>and</strong><br />

comparison of an increased number of transposons<br />

has made an alternative classi¢cation possible based<br />

on sequence similarities ([14] ; J. Mahillon <strong>and</strong> M.<br />

Ch<strong>and</strong>ler, personal communication). These studies<br />

not only revealed the existence of several bacterial<br />

families of related elements, but more importantly,<br />

they also identi¢ed conserved transposase domains<br />

amongst a broad diversity of transposons from prokaryotes<br />

<strong>and</strong> eukaryotes. This was ¢rst shown for<br />

the IS3 family of transposases which exhibit striking<br />

similarities with the integrase proteins of retroviruses<br />

<strong>and</strong> retrotransposons, notably by the presence of a<br />

invariant triad of acidic residues termed the DDE<br />

motif [16^18]. Subsequently, a DDE-like signature<br />

has been identi¢ed within the transposase of many<br />

di¡erent transposons including IS10, Tn7 <strong>and</strong> the<br />

bacteriophage Mu, which are amongst the best characterised<br />

bacterial elements [19^22]. The widespread<br />

Bacterial transposons utilise two major modes of


160<br />

Fig. 2. Outcomes from <strong>site</strong>-speci¢c <strong>recombination</strong>. Triangles<br />

show the orientation of the <strong>recombination</strong> <strong>site</strong>s. a <strong>and</strong> b indicate<br />

the position of distinct genetic markers around the <strong>recombination</strong><br />

loci. `Excision' <strong>and</strong> `integration' refer to <strong>recombination</strong> events involving<br />

genetic entities of di¡erent size <strong>and</strong>/or function (e.g., the<br />

bacterial chromosome <strong>and</strong> a phage genome), whereas `resolution'<br />

<strong>and</strong> `fusion' apply to equivalent DNA molecules, (e.g., two plasmids).<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178<br />

occurrence of this signature indicates that transposases<br />

<strong>and</strong> retroviral integrases from very di¡erent<br />

sources form a super-family of <strong>recombination</strong> enzymes<br />

which share a common catalytic mechanism.<br />

It is worth noting that not all the proteins involved<br />

in transposition contain a DDE motif, <strong>and</strong> that some<br />

of them clearly belong to completely distinct classes<br />

of enzymes. For example, the recombinases involved<br />

in the translocation of di¡erent kinds of conjugative<br />

transposons <strong>and</strong> members of the IS110/IS492 family,<br />

belong to three di¡erent groups of <strong>site</strong>-speci¢c recombinases<br />

[23^25]. IS1 transposase also shares similarities<br />

with the restriction endonuclease EcoRII<br />

[26]. In contrast, the IS91 family transposase is related<br />

to proteins involved in replication by rollingcircle<br />

mechanism [27,28]. Therefore, it is not surprising<br />

that these elements appear to use quite di¡erent<br />

<strong>recombination</strong><br />

mechanisms.<br />

Transposable elements are often regarded as selfish<br />

genetic entities that have evolved a set of mechanisms<br />

to replicate faster than their host [29]. An<br />

extreme case is bacteriophage Mu which transposes<br />

into the host chromosome for the establishment of<br />

segment. Recombination between <strong>site</strong>s on separate<br />

lysogeny <strong>and</strong>, during lytic growth, uses the host machinery<br />

to replicate its own DNA by undergoing<br />

DNA molecules will integrate one molecule into<br />

multiple rounds of replicative transposition [6,15].<br />

By promoting various DNA rearrangements such<br />

as deletions, inversions, <strong>and</strong> replicon fusions, transposable<br />

elements can also be viewed as naturally<br />

the other (Fig. 2). Although these DNA rearrangements<br />

are reminiscent of those promoted by transposable<br />

elements, one main di¡erence is that <strong>site</strong>speci¢c<br />

<strong>recombination</strong> occurs between predetermined<br />

occurring `genetic engineers' which by modifying<br />

loci, whereas rearrangements mediated by transposons<br />

largely rely on their target-<strong>site</strong> speci¢city. A<br />

the genome of their host may contribute to the genetic<br />

diversity of a population [30]. Through their<br />

second major di¡erence lies in the conservative nature<br />

of <strong>site</strong>-speci¢c <strong>recombination</strong>, i.e., DNA str<strong>and</strong><br />

exchange is completed without involving any DNA<br />

ability to translocate genes, transposons are also potent<br />

agents for horizontal transfer <strong>and</strong> in that respect,<br />

they play a crucial role in the spread of antibiotic<br />

resistance factors amongst bacteria.<br />

synthesis or degradation. That allows <strong>site</strong>-speci¢c <strong>recombination</strong><br />

to be reciprocal, i.e., a second round of<br />

2.2. Site-speci¢c <strong>recombination</strong> : integration,<br />

<strong>recombination</strong> can restore the initial DNA con¢guration.<br />

resolution, inversion <strong>and</strong> transposition<br />

As a consequence of these distinguishing features,<br />

<strong>site</strong>-speci¢c <strong>recombination</strong> is exploited for a range of<br />

In DNA rearrangements mediated by <strong>site</strong>-speci¢c<br />

biological functions [5]. In a number of examples,<br />

<strong>site</strong>-speci¢c inversion is used to switch the expression<br />

<strong>recombination</strong>, four DNA str<strong>and</strong>s are broken, exchanged<br />

<strong>and</strong> resealed at speci¢c positions of two<br />

of genes between alternative patterns [32]. As the<br />

separate <strong>recombination</strong> <strong>site</strong>s [3^5,31]. The outcome<br />

of a <strong>recombination</strong> event depends on the relative<br />

genes controlled by these systems often code for surface<br />

structures proteins, e.g., bacteriophage tail ¢-<br />

disposition of the two <strong>site</strong>s. Intramolecular <strong>recombination</strong><br />

between inverted or directly repeated <strong>site</strong>s will<br />

bres, £agellar antigens or pilin, <strong>recombination</strong> provides<br />

a source of genetic variability which helps the<br />

invert or excise respectively the intervening DNA<br />

host organism to adapt to environmental £uctua-


B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178 161<br />

tions. For these di¡erent systems, alternative gene<br />

expression units located on various replicons or<br />

within transposons [43]. Conjugative transposons<br />

expression is achieved, either by £ipping the orientation<br />

of a promoter between two structural genes or<br />

by connecting variable C-terminal portions of a gene<br />

have combined <strong>site</strong>-speci¢c <strong>recombination</strong> <strong>and</strong> conjugation<br />

mechanisms to transfer a circular excised<br />

to the constant N-terminal region located outside the<br />

form of the element between a donor <strong>and</strong> a recipient<br />

invertible segment [32]. An extreme example of the<br />

cell [23]. A remarkable feature of integrons <strong>and</strong> conjugative<br />

transposons is their low level of target <strong>site</strong><br />

latter is provided by the shu¥on system of R64 plasmid<br />

in which a set of four invertible fragments controls<br />

the selection of seven C-terminal segments for<br />

selectivity when compared to other <strong>site</strong>-speci¢c <strong>recombination</strong><br />

systems, illustrating the potential for a<br />

<strong>recombination</strong> mechanism to accommodate distinct<br />

the pilV gene [33]. DNA splicing reactions by <strong>site</strong>speci¢c<br />

<strong>recombination</strong> between directly repeated <strong>site</strong>s<br />

functions.<br />

has also been reported as a mechanism of gene activation<br />

in developmentally regulated prokaryotic<br />

processes, such as Bacillus subtilis sporulation [34]<br />

The functional £exibility of <strong>site</strong>-speci¢c <strong>recombination</strong><br />

is also demonstrated by the fact that the various<br />

biological functions listed above are ful¢lled by<br />

<strong>and</strong> Anabaena sp. heterocyst di¡erentiation [35].<br />

recombinases belonging to two major families of enzymes<br />

using distinct biochemical mechanisms. The<br />

Transposons of the Tn3 family encode a <strong>site</strong>-speci¢c<br />

resolution system to reduce the cointegrate intermediates<br />

generated by their replicative mode of<br />

resolvase/invertase family, of which there are currently<br />

approximately 40 di¡erent members, forms a<br />

transposition [36]. Another widespread function of<br />

rather homogenous group of related proteins in<br />

<strong>site</strong>-speci¢c <strong>recombination</strong> is to resolve replicon<br />

which a conserved serine residue plays a key catalytic<br />

dimers arising from homologous <strong>recombination</strong> in<br />

role [44,45]. Recombinases of the V integrase (V Int)<br />

order to facilitate their segregation to daughter cells<br />

family (V60 members) are much more divergent,<br />

at the time of cell division. Whereas several di¡erent<br />

replicons encode their own resolution systems (e.g.,<br />

with only four completely invariant residues intimately<br />

involved in catalysis : the RHRY tetrad [46^<br />

Refs. [37^39]), plasmids of the ColE1 family utilise<br />

48]. There is no functional segregation between the<br />

two families <strong>and</strong> systems of both groups are used to<br />

the multifunctional <strong>recombination</strong> system Xer encoded<br />

by the Escherichia coli chromosome [40]. By<br />

perform virtually all kinds of rearrangements [5]. As<br />

acting at a speci¢c <strong>recombination</strong> <strong>site</strong> (dif) located in<br />

for DDE transposases, these two families are not<br />

the terminus region of replication, Xer <strong>recombination</strong><br />

also appears to play an active role in the partitioning<br />

of the bacterial chromosome itself. Bacteria<br />

exclusive <strong>and</strong> there are a few recombinases that appear<br />

to belong to neither group. An interesting example<br />

is the Piv invertase of Moraxella sp. which<br />

de¢cient in Xer <strong>recombination</strong> form ¢laments with<br />

together with the transposases of unusual IS elements<br />

(the IS110/IS492 family) form what can be<br />

aberrant nucleoids, presumably as a result the accumulation<br />

of chromosome dimers [41]. This seems to<br />

viewed as a third family of related recombinases<br />

be an important physiological role of <strong>site</strong>-speci¢c<br />

<strong>recombination</strong> that is highly conserved among bacteria.<br />

[25]. Like conjugative transposons, several IS elements<br />

of this family fail to generate target duplications<br />

at the insertion <strong>site</strong> <strong>and</strong>/or produce an excised<br />

circular form as a transposition intermediate.<br />

Finally, the paradigm of <strong>site</strong>-speci¢c <strong>recombination</strong><br />

is the reaction by which bacteriophage V <strong>and</strong><br />

related phages choose whether they prefer to remain<br />

3. Three mechanisms of DNA cut-<strong>and</strong>-paste<br />

dormant as a prophage or to embark on the lytic<br />

cycle of phage growth by integrating into <strong>and</strong> excising<br />

from the host genome [4,42]. A similar excision/<br />

integration reaction is used by di¡erent systems to<br />

The establishment of a cell-free system for di¡erent<br />

types of specialised <strong>recombination</strong> has allowed<br />

translocate non-replicating genetic material, e.g.,<br />

elucidation of the biochemical steps of the DNA<br />

antibiotic resistance genes, between di¡erent genomic<br />

str<strong>and</strong> exchange reactions. These in vitro studies<br />

loci. Integrons are compo<strong>site</strong> <strong>recombination</strong> systems<br />

have identi¢ed three di¡erent mechanisms for cutting<br />

that exchange mobile gene cassettes between speci¢c<br />

<strong>and</strong> rejoining DNA molecules, corresponding to the


162<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178<br />

Fig. 3. Chemistry of transposition <strong>and</strong> <strong>site</strong>-speci¢c <strong>recombination</strong> reactions. Recombination DNA str<strong>and</strong> breakage <strong>and</strong> joining occur by<br />

transesteri¢cation reactions in which the phosphate of the scissile phosphodiester bond is subject to nucleophile attack by a hydroxyl<br />

group (arrows). Endonucleolytic cleavage at the transposon ends (A) <strong>and</strong> the str<strong>and</strong>-transfer reaction that join the ends to the target<br />

DNA (B) are one-step transesteri¢cations in which the nucleophile is a water molecule <strong>and</strong> the 3P-OH end of the element, respectively.<br />

Str<strong>and</strong> exchange catalysed by <strong>site</strong> speci¢c recombinases (C <strong>and</strong> D) occurs by two steps of transesteri¢cation (cleavage <strong>and</strong> rejoining) involving<br />

a covalent protein-DNA intermediate. The nature of the catalytic residue <strong>and</strong> the line of entry of the nucleophile is di¡erent between<br />

the two recombinase families. For cleavage catalysed by the invertase/resolvase family (C), the nucleophile hydroxyl is derived from<br />

a serine <strong>and</strong> the leaving group is the 3P-OH of the deoxyribose. For the V integrase family (D), the catalytic residue is a tyrosine <strong>and</strong> the<br />

leaving group is the 5P-OH. For both recombinase families, the rejoining step is the reverse of the cleavage step. Phosphate backbones are<br />

drawn in thick <strong>and</strong> thin lines to distinguish the donor <strong>and</strong> target DNA (panel B) or the two <strong>recombination</strong> partner DNA str<strong>and</strong>s (panels<br />

C <strong>and</strong> D).<br />

three major classes of proteins described above : the<br />

DDE transposases, the resolvase/invertase, <strong>and</strong> the V<br />

level of resolution by having obtained crystal structure<br />

data for each of the three families of proteins.<br />

Int families of <strong>site</strong>-speci¢c recombinases. A common<br />

feature of the three mechanisms is that they proceed<br />

3.1. The DDE transposases : a universal one-step<br />

by transesteri¢cation reactions without requiring<br />

reaction<br />

transesteri¢cation<br />

high-energy cofactors such as ATP (Fig. 3)<br />

[3,31,49]. The energy of the broken phosphodiester<br />

The biochemistry of reactions catalysed by the<br />

bonds is conserved for the formation of new bonds.<br />

DDE recombinases has been examined in detail for<br />

several di¡erent systems including three bacterial<br />

The DDE transposases use a one-step transesteri¢cation<br />

mechanism, whereas the two distinct families of<br />

transposable elements : the bacteriophage Mu, IS10<br />

<strong>site</strong>-speci¢c recombinases use contrasting two-step<br />

<strong>and</strong> Tn7 (for recent reviews, see [6,11,15,50]. In all<br />

transesteri¢cation mechanisms involving di¡erent<br />

amino acid residues in the formation of a covalent<br />

cases, the transposase executes a set of critical reactions<br />

that join the 3P ends of the element to the target<br />

DNA-enzyme intermediate (Fig. 3). Underst<strong>and</strong>ing<br />

of these mechanisms has now reached the atomic<br />

DNA, whereas connection of the 5P ends <strong>and</strong> completion<br />

of transposition requires processing events


Fig. 4. Biochemical steps underlying the non-replicative or replicative<br />

transposition of three bacterial elements. The shaded rectangles<br />

represent the DNA str<strong>and</strong>s of the transposable elements<br />

ends. For IS10 <strong>and</strong> Tn7, reactions occurring at a single end are<br />

shown. Black <strong>and</strong> white rectangles are the £anking donor sequences<br />

<strong>and</strong> the target DNA, respectively. The black <strong>and</strong> white<br />

arrowheads show the 3P-end <strong>and</strong> 5P-end cleavage, respectively.<br />

Curved arrows indicate the nucleophilic attack transferring the<br />

3P-OH ends on staggered phosphates of the target DNA (black<br />

dots). Crenellated lines represent the few target nucleotides that<br />

are duplicated during the transposition process. For the three elements,<br />

the biochemical steps are catalysed by the transposase in<br />

a complex where the transposon ends are in synapsis. For IS10,<br />

the target is captured after completion of the double-str<strong>and</strong><br />

breaks at the transposon ends, whereas for Mu <strong>and</strong> Tn7, the<br />

presence of the target within the complex is required to activate<br />

the cleavage reactions. The cross-hatching represents replication<br />

events that complete transposition after complex dissociation.<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178 163<br />

str<strong>and</strong> is also cleaved to reveal the transposon 5P<br />

ends <strong>and</strong> to excise the element from its initial genomic<br />

locus. IS10 excision yields £ush transposons<br />

ends [51], whereas the double-str<strong>and</strong> breaks at the<br />

ends of Tn7 are staggered by three nucleotides,<br />

with the 5P end cleavages occurring in the adjacent<br />

donor backbone [52]. In sharp contrast, the second<br />

str<strong>and</strong> cleavage does not occur at this stage of bacteriophage<br />

Mu transposition which remains connected<br />

at its 5P ends to the £anking DNA sequences.<br />

In a second step, the 3P-OH ends of the excised<br />

transposon (IS10 <strong>and</strong> Tn7) or the nicks in the donor<br />

molecule (Mu) participate in a concerted str<strong>and</strong><br />

transfer reaction that join both ends of the element<br />

to staggered phosphates of the two target DNA<br />

str<strong>and</strong>s (Fig. 4 ; see also Fig. 3B).<br />

In the case of IS10 <strong>and</strong> Tn7, host processing of the<br />

resulting str<strong>and</strong> transfer intermediate results in the<br />

¢lling-in of the short single-str<strong>and</strong> gaps lying at<br />

each transposon-target DNA junction. This repair,<br />

which in the case of Tn7 is also presumed to remove<br />

the overhanging three nucleotides at the 5P ends,<br />

generates the short target duplications that £ank<br />

the element in its new locus (Fig. 4). After eventual<br />

cleavage of the donor backbone by an undetermined<br />

mechanism, the str<strong>and</strong> transfer product of bacteriophage<br />

Mu can be processed in a similar way during<br />

the non-replicative integration of the phage. Alternatively,<br />

in the absence of donor backbone dissociation,<br />

complete replication of the element gives rise<br />

to a cointegrate, which is the ¢nal product of replicative<br />

transposition (Fig. 4).<br />

Variations of the transposition pathway have been<br />

reported for both prokaryotic <strong>and</strong> eukaryotic transposons.<br />

In particular, the transposase of IS911, a<br />

member of the IS3 family, has been found to carry<br />

out a distinctive single-str<strong>and</strong> circularisation reaction<br />

performed by host repair <strong>and</strong> replication functions<br />

in which one transposon end is transferred to a target<br />

<strong>site</strong> three nucleotides distant from the other end<br />

(Fig. 4). An important variation between the three<br />

[53,54]. In vivo, the resulting `¢gure-eight' molecule<br />

bacterial systems is in the number <strong>and</strong> nature of cuts<br />

is processed, giving rise to a circular excised form of<br />

that sever the transposon from the £anking donor<br />

the transposon thought to be a transposition intermediate<br />

that can e¤ciently insert in a new locus.<br />

Although mechanistically distinct, this `<strong>site</strong>-speci¢c'<br />

DNA, resulting in di¡erent transposition end-products.<br />

For the three transposons, transposition initiates<br />

with a pair of speci¢c single str<strong>and</strong> cleavages<br />

circularisation of IS911 is reminiscent of the excision<br />

exposing the 3P-OH ends of the element (Fig. 4 ; see<br />

reaction performed by <strong>site</strong>-speci¢c recombinases. Interestingly,<br />

the formation of excised transposon<br />

also Fig. 3A). For IS10 <strong>and</strong> Tn7 which use a nonreplicative<br />

cut-<strong>and</strong>-paste mechanism, the other<br />

circles has been observed for other members of the


164<br />

IS3 family [54,55], <strong>and</strong> also for IS1 [56] <strong>and</strong> for diverse<br />

eukaryotic transposons, suggesting that it may<br />

represent a major mode of transposition [54].<br />

As mentioned above, the initial cleavage that generates<br />

the 3P-OH ends <strong>and</strong> the str<strong>and</strong> transfer reaction<br />

catalysed by bacterial DDE transposases are<br />

chemically equivalent to the reactions catalysed by<br />

the retroviral integrase proteins (IN) for the integration<br />

of the retrovirus cDNA into the genome of an<br />

infected cell [6,7,57]. IN-mediated cleavages at the<br />

ends of HIV cDNA <strong>and</strong> the subsequent str<strong>and</strong> transfer<br />

step proceed with inversion of chirality at the<br />

target DNA phosphates. The same result was obtained<br />

for the str<strong>and</strong> transfer reaction catalysed by<br />

the bacteriophage MuA transposase [58,59]. This<br />

analysis indicates that the reactions catalysed by<br />

the DDE recombinases occur by a one-step transesteri¢cation<br />

mechanism in which the scissile phosphodiester<br />

bond is directly attacked, either by a<br />

water molecule (end cleavage), or by the 3P-OH<br />

end of the element (str<strong>and</strong> transfer) without the formation<br />

of a protein-DNA covalent intermediate<br />

(Fig. 3). Both transesteri¢cation steps require diva-<br />

cations (Mg<br />

2‡<br />

lent<br />

or<br />

2‡<br />

) <strong>and</strong> it was postulated<br />

Mn<br />

that the role of the DDE residues was to form a<br />

metal ion binding pocket in the active <strong>site</strong> [49].<br />

This suggestion is strongly supported by the recent<br />

¢nding that the catalytic domain of IN, MuA <strong>and</strong><br />

other enzymes involved in metal ion-dependent phosphoryl<br />

transfers exhibit very similar structures.<br />

The crystal structure of the DDE-containing catalytic<br />

core of MuA shows that this region of the protein<br />

contains two sub-domains [60]. The 70 C-terminal<br />

residues form a L-barrel, on one face of which is<br />

a large region of positive potential that could play a<br />

role in the non speci¢c DNA binding activity associated<br />

with the whole catalytic core. It was proposed<br />

that this activity might be important for interactions<br />

either with the DNA sequences surrounding the<br />

cleavage points at the Mu ends or with the target<br />

DNA, in order to position the substrate into the<br />

enzyme active <strong>site</strong> [60]. Remarkably, in spite of a<br />

very low level of sequence homology, the structure<br />

of the N-terminal sub-domain containing the DDE<br />

triad shows striking similarities, not only with the<br />

catalytic domain of the IN proteins of HIV <strong>and</strong><br />

ASV retroviruses, but also with the structure of<br />

more functionally distant enzymes such as the E.<br />

HIV ribonucleases H (RNase H) <strong>and</strong> the<br />

junction resolving enzyme RuvC<br />

(reviewed in Refs. [61^63]). In the structure shared<br />

by these di¡erent proteins (a central ¢ve-str<strong>and</strong>ed L-<br />

sheet surrounded by K-helices on either side) three or<br />

four catalytically important acidic residues, i.e., the<br />

DDE motif in the case of the MuA <strong>and</strong> IN proteins,<br />

are clustered to form a possible two metal ion binding<br />

pocket as actually observed in the structure of<br />

HIV RNase H active <strong>site</strong> [64]. This structural similarity<br />

between unrelated proteins strongly supports<br />

the view that the DDE recombinases belong to a<br />

larger group of enzymes, including polymerases <strong>and</strong><br />

ribozymes, that catalyse transesteri¢cation reactions<br />

using a common `carboxylate-chelated two-metal<br />

ion' catalytic mechanism [61,63]. In this mechanism,<br />

initially proposed for the 3P-5P exonuclease activity<br />

of DNA polymerase I, the two chelated metal ions<br />

participate in the activation of the nucleophile hydroxyl<br />

<strong>and</strong> the scissile phosphodiester bond by stabilising<br />

the phosphate in a penta-coordinated form<br />

[65].<br />

However, DDE recombinases are not simple nucleases<br />

or polynucleotidyl transferases. These enzymes<br />

are distinguished by the fact that they consecutively<br />

catalyse two (or three) transesteri¢cation<br />

reactions at both transposons ends. Successful transposition<br />

requires that these two sets of reactions be<br />

temporally <strong>and</strong> spatially coordinated in order to prevent<br />

incomplete <strong>recombination</strong> events which are<br />

likely to be deleterious to the host <strong>and</strong> to the transposable<br />

element itself. Recent work from di¡erent<br />

laboratories has provided new interesting insight<br />

into how this control is e¡ected in the case of Mu<br />

(reviewed in Refs. [66,67]). A key step in Mu transposition<br />

is the assembly of a synaptic complex into<br />

which catalytically inert monomers of MuA become<br />

activated by the formation of a tetrameric core<br />

tightly bound to the two transposon ends. Assembly<br />

of the tetramer thus implies a structural transition by<br />

which the enzyme <strong>and</strong> the DNA cleavage <strong>site</strong>s become<br />

engaged for catalysis ([6,15] ; see below). Complementation<br />

experiments performed by mixing distinct<br />

catalytic mutants of MuA have shown that the<br />

tetramer active <strong>site</strong>s are built by interlocking separate<br />

domains of distinct transposase subunits. In<br />

each active <strong>site</strong>, one monomer provides the DDE<br />

catalytic core (domain II), whereas a di¡erent mono-<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178<br />

coli <strong>and</strong><br />

E. coli Holliday


C-terminal part of the protein [68]. Although the<br />

exact function of this second catalytic domain (domain<br />

IIIA) is not known, it appears to play a role in<br />

the activation <strong>and</strong>/or positioning of the DNA [69].<br />

For the str<strong>and</strong> transfer reactions, the active <strong>site</strong> domain<br />

IIIA is provided by MuA monomers occupying<br />

the inner part of the Mu ends whereas the DEEcontaining<br />

domain II is provided by transposase subunits<br />

bound on the external <strong>site</strong>s [70,71]. By reciprocality<br />

of domain sharing, this spatial arrangement<br />

appears to be reversed for the ends cleavage reactions<br />

[71]. Furthermore, in both reaction steps, the<br />

DDE domain II operates in trans, i.e., the subunit<br />

bound to one end catalyses the cleavage <strong>and</strong> joining<br />

of the oppo<strong>site</strong> end [70,72]. This intricate architecture<br />

of the MuA enzymatic complex is viewed as a<br />

mechanism which ensures that substrate synapsis<br />

<strong>and</strong> catalysis are tightly coupled. The involvement<br />

of a structural transition upon tetramerisation <strong>and</strong><br />

substrate binding is also consistent with the fact that<br />

the DDE residues of the MuA catalytic core appear<br />

to be in an inactive con¢guration [60]. Taking these<br />

data into account, Yang et al. propose a pathway in<br />

which the Mu ends are ¢rst nicked within the two<br />

cleavage active <strong>site</strong>s <strong>and</strong> then swing onto two other<br />

catalytic pockets to be transferred on the target<br />

DNA [71].<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178 165<br />

As for Mu, a single transposase protein carries out<br />

the three reaction steps underlying IS10 transposition<br />

<strong>and</strong> the assembly of a precleaved synaptic complex<br />

is required to ensure the coordination of the<br />

processing events occurring at both ends [50,73,74].<br />

IS10 transposition pathway is also highly ordered.<br />

Cleavage of the transferred str<strong>and</strong> (3P end) always<br />

precedes cleavage of the non-transferred str<strong>and</strong> (5P<br />

end), <strong>and</strong> complete excision of the transposon is required<br />

before a target DNA molecule is captured by<br />

the complex for the str<strong>and</strong> transfer reactions [75,76].<br />

In sharp contrast to Mu, a single transposase monomer<br />

catalyses the three chemical steps of IS10 transposition<br />

at each of the two transposon ends [77].<br />

Thus, two monomers of transposase appear to carry<br />

out the entire transposition reaction without sharing<br />

domains. The mechanism which allows the repeated<br />

use of a single DDE-containing catalytic pocket is<br />

not known. Current models propose that consecutive<br />

structural rearrangements must occur between each<br />

reaction step to re-adjust the active <strong>site</strong> to novel<br />

substrate con¢gurations [50,77].<br />

A third <strong>and</strong> distinctive example of transposase architecture<br />

is provided by Tn7 [11]. In contrast to Mu<br />

<strong>and</strong> IS10, Tn7 transposition reactions are catalysed<br />

by two DDE-containing enzymes with clearly separate<br />

activities. TnsA mediates cleavage at the 5P ends<br />

of the transposon whereas TnsB, which is also responsible<br />

for the recognition <strong>and</strong> binding of Tn7<br />

ends, promotes the 3P end cleavage <strong>and</strong> carries out<br />

the str<strong>and</strong> transfer reaction [22]. Although the respective<br />

functions of TnsA <strong>and</strong> TnsB can be selectively<br />

blocked by mutation of their DDE residues,<br />

the presence of both proteins is required for the formation<br />

of an active complex. Tn7 transposase is thus<br />

a heteromeric enzyme containing two di¡erent catalytic<br />

subunits [22]. The possibility that the active <strong>site</strong>s<br />

of the TnsA+B transposase core could be assembled<br />

by contribution of distinct TnsA or TnsB protomers<br />

has not been examined thus far. As with IS10, there<br />

is currently no evidence for cis or trans activity by<br />

TnsA <strong>and</strong>/or TnsB. Nevertheless, the biochemical<br />

separation of the 5P <strong>and</strong> 3P ends processing reactions<br />

has a remarkable implication. Inactivation of TnsA<br />

converts the normal cut-<strong>and</strong>-paste transposition<br />

pathway of Tn7 into a Mu-like replicative mechanism,<br />

the 3P ends breakage <strong>and</strong> joining reactions catalysed<br />

by TnsB within the transposase core being<br />

functionally equivalent to those performed by the<br />

MuA tetramer in Mu transposition [78].<br />

These three di¡erent examples of `division of labour'<br />

within a transposition complex illustrate the<br />

remarkable £exibility by which a common chemical<br />

mechanism can be adapted in di¡erent ways to accomplish<br />

complex <strong>and</strong> highly controlled <strong>recombination</strong><br />

reactions.<br />

Unlike transposases, <strong>site</strong>-speci¢c recombinases, at<br />

least in principle, execute all <strong>recombination</strong> DNA<br />

breakage <strong>and</strong> joining reactions without involving<br />

host repair or replication functions [3^5,31]. These<br />

reactions are biochemically related to those catalysed<br />

by the topoisomerase enzymes that regulate the intracellular<br />

level of DNA supercoiling [79]. Indeed,<br />

<strong>site</strong>-speci¢c recombinases often exhibit type I topo-<br />

mer donates another catalytic region located in the<br />

3.2. Site-speci¢c recombinases : two-step<br />

transesteri¢cations by distinct mechanisms


166<br />

proteins. In the E. coli inversion system Fim, the two<br />

Fig. 5. Concerted DNA breakage <strong>and</strong> rejoining reactions catalysed<br />

by resolvase/invertase family enzymes. The subunit rotation<br />

model is shown. The ovals represent recombinase subunits with<br />

the conserved catalytic serine `S'. Thick <strong>and</strong> thin lines are the<br />

top <strong>and</strong> bottom str<strong>and</strong>s of the <strong>recombination</strong> <strong>site</strong>s, respectively.<br />

The short vertical bars are the 2 bp of the overlap region between<br />

the two cleavage points. Black arrows represent the nucleophilic<br />

attacks of phosphates (black dots) by hydroxyl groups<br />

(arrowheads). The four DNA str<strong>and</strong>s are cleaved (a), exchanged<br />

by 180³ rotation of the half-<strong>site</strong> bound subunits (b) <strong>and</strong> religated<br />

in the recombinant con¢guration (c).<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178<br />

isomerase activity by which they can relax supercoiled<br />

DNA substrates. However, in a <strong>recombination</strong><br />

reaction between two <strong>site</strong>s, DNA str<strong>and</strong>s are not<br />

only broken <strong>and</strong> rejoined, they are also exchanged.<br />

Therefore, as discussed above for the DDE recombinases,<br />

<strong>site</strong>-speci¢c <strong>recombination</strong> requires the assembly<br />

of a synaptic complex containing multiple recombinase<br />

subunits <strong>and</strong> the relevant DNA recombining<br />

partners.<br />

For most of the systems of the two major families<br />

(i.e., the resolvase/invertase <strong>and</strong> the V Int families),<br />

<strong>recombination</strong> takes place within a short (V30 bp)<br />

DNA segment called the `core' or `crossover' <strong>site</strong><br />

onto which two recombinase subunits bind, usually<br />

by recognising speci¢c sequences with dyad symmetry<br />

[3,80,81]. In the synaptic complex, the two core<br />

<strong>site</strong>s are brought in close proximity. It is generally<br />

admitted that the four recombinase subunits bound<br />

on the two duplexes participate in the <strong>recombination</strong><br />

reaction (Figs. 5 <strong>and</strong> 6). A few systems, all belonging<br />

to the V Int family, involve two distinct recombinase<br />

recombinases, FimB <strong>and</strong> FimE, act independently on<br />

the same <strong>recombination</strong> <strong>site</strong>s to control the `on' or<br />

`o¡' position of this particular genetic switch [82]. By<br />

contrast, the XerC <strong>and</strong> XerD recombinases cooperate<br />

in all Xer-mediated DNA rearrangements <strong>and</strong><br />

the two proteins bind with distinct speci¢cities to<br />

separate regions of the di¡erent <strong>recombination</strong> <strong>site</strong>s<br />

of this system [83,84].<br />

The <strong>recombination</strong> locus of all systems, even those<br />

working with a single recombinase, contain a certain<br />

degree of asymmetry so that `top' <strong>and</strong> `bottom'<br />

str<strong>and</strong>s can be distinguished. In the simplest cases,<br />

this asymmetry is entirely encoded within the core<br />

<strong>site</strong> whereas in other systems, external elements<br />

may contribute to the <strong>recombination</strong> <strong>site</strong> polarity<br />

by imposing a speci¢c geometry on the synaptic<br />

complex ([80,81] ; see also below). The catalytic<br />

mechanism used by the two families of <strong>site</strong>-speci¢c<br />

recombinases is di¡erent as is the structural organisation<br />

of the enzymes.<br />

<strong>and</strong> QN transposons [3,32,36,80,81,85]. Although the<br />

3.2.1. The resolvase/invertase family : concerted<br />

breakage <strong>and</strong> rejoining of four DNA str<strong>and</strong>s<br />

The best characterised recombinases of this family<br />

chemistry of str<strong>and</strong> exchange used by these recombinases<br />

appears to be highly conserved, important variation<br />

is found in the assembly of the <strong>recombination</strong><br />

are the invertases Gin from bacteriophage Mu <strong>and</strong><br />

synapse determining the selectivity of the reaction,<br />

Hin from Salmonella sp. <strong>and</strong> the resolvases of Tn3<br />

i.e., inversion or resolution (see below).


B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178 167<br />

arm region that connects the two domains. The small<br />

In a <strong>recombination</strong> reaction catalysed by resolvases<br />

or invertases, double str<strong>and</strong> breaks staggered<br />

C-terminal DNA binding domain is involved in the<br />

by 2 bp occur at the middle of the two paired core<br />

recognition of the outer part of the core <strong>site</strong> consensus<br />

sequence by making speci¢c contacts in both the<br />

<strong>site</strong>s, giving rise to recessed 5P ends <strong>and</strong> 3P-OH overhangs<br />

(Fig. 5 ; see also Fig. 3C). One recombinase<br />

major <strong>and</strong> minor grooves. Its structure, containing a<br />

subunit is linked to each of the 5P ends through the<br />

helix-turn-helix DNA binding motif, is similar to<br />

conserved serine residue of the family [86,87]. This<br />

that previously found for the DNA binding domain<br />

serine presumably provides the primary nucleophile<br />

of Hin invertase [96]. The arm region that joins the<br />

hydroxyl group in the cleavage reaction [45]. The<br />

two globular domains also contributes to DNA<br />

ligation step that follows str<strong>and</strong> exchange can be<br />

binding through interactions in the minor groove.<br />

viewed as the converse of the cleavage : the protein-<br />

The large N-terminal catalytic domain contains the<br />

DNA phosphoseryl bond of one str<strong>and</strong> is attacked<br />

active <strong>site</strong> serine <strong>and</strong> other catalytic residues, as well<br />

as a set of residues forming a hydrophobic core at<br />

the dimer interface. This domain also appears to be<br />

by the 3P-OH end of the partner to release the enzyme<br />

<strong>and</strong> reseal the DNA backbone in the recombinant<br />

con¢guration (Fig. 5 ; see also Fig. 3C).<br />

Although cleavages of the four DNA str<strong>and</strong>s or<br />

important for higher order interactions between resolvase<br />

dimers in the <strong>recombination</strong> complex [97^99].<br />

the religation steps can be experimentally uncoupled<br />

The DNA in the co-crystal is bent by 60³ away from<br />

the enzyme catalytic domain.<br />

by using <strong>recombination</strong> <strong>site</strong> variants or mutated recombinases,<br />

both types of reaction are normally<br />

The QN resolvase-DNA complex appears to be in<br />

highly coordinated [88,89]. The cleaved complex in<br />

an inactive con¢guration, the catalytic serines of the<br />

which four enzyme-linked <strong>recombination</strong> half-<strong>site</strong>s<br />

two resolvase subunits being too far away from the<br />

scissile phosphates to cleave the DNA [95]. However,<br />

are held together by recombinase subunit interactions<br />

seems to be an obligate intermediate [88].<br />

the active serine of either monomer is closest to the<br />

DNA cleavage position proximal to the half-<strong>site</strong><br />

Thus, <strong>recombination</strong> by the resolvase/invertase family<br />

occurs by a mechanism in which four DNA<br />

bound by the recombinase subunit. This correlates<br />

with the fact that the active serine <strong>and</strong> several other<br />

str<strong>and</strong>s are broken <strong>and</strong> rejoined in a concerted manner.<br />

catalytic residues of QN resolvase act in cis on the<br />

nearest scissile phosphodiester bond [88]. The crystal<br />

The st<strong>and</strong>ard substrates for invertases <strong>and</strong> resolvases<br />

are supercoiled molecules (see below) <strong>and</strong> the<br />

structure also suggests that, as in MuA, catalytic<br />

topological change induced by <strong>recombination</strong> has<br />

residues may be shared between the two monomers<br />

been found to be equivalent to a right-h<strong>and</strong>ed 180³<br />

to form the active <strong>site</strong>. While there is currently no<br />

rotation of one pair of cleaved half-<strong>site</strong>s relative to<br />

experimental evidence to support the possibility for a<br />

the other prior the rejoining step (Fig. 5) [90^92]. In<br />

compo<strong>site</strong> catalytic <strong>site</strong>, mutations in one resolvase<br />

subunit seem to activate the adjacent subunit in<br />

this mechanism, the 2 bp `overlap' regions that separate<br />

the top <strong>and</strong> bottom str<strong>and</strong> cleavage positions<br />

need to be identical in the two core <strong>site</strong>s to stably<br />

trans, presumably by altering inter-subunit interactions<br />

at the dimer interface ([88] ; M.R. Boocock<br />

reconnect the recombinant DNA duplexes [93,94]. It<br />

has been shown recently that in reactions involving<br />

<strong>and</strong> N.D.F. Grindley, unpublished). Likewise, mutations<br />

in the equivalent dimerisation domain of the<br />

non-homologous overlap regions, str<strong>and</strong> exchange<br />

DNA invertases Gin, Hin <strong>and</strong> Cin, make the recombinases<br />

more reactive <strong>and</strong> independent of structural<br />

mediated by Tn3 resolvase proceeds through apparent<br />

360³ (2U180³) rotation of the half-<strong>site</strong>s without<br />

rejoining the mis-paired str<strong>and</strong>s in the recombinant<br />

elements controlling the formation of the <strong>recombination</strong><br />

synaptic complex ([92,100^103] ; see also below).<br />

(180³ rotation) con¢guration [89].<br />

These observations indicate that activation of the<br />

The crystal structure of the QN resolvase dimer<br />

active <strong>site</strong> serine recombinases appears to require<br />

complexed to the core <strong>recombination</strong> <strong>site</strong> has been<br />

conformational changes during the assembly of an<br />

determined recently [95]. The DNA-bound resolvase<br />

enzymatically competent <strong>recombination</strong> structure as<br />

monomer contains two globular domains lying on<br />

discussed above for the DDE transposases. Indeed,<br />

oppo<strong>site</strong> faces of the DNA helix <strong>and</strong> an extended<br />

some level of structural £exibility as revealed by the


168<br />

Fig. 6. Sequential str<strong>and</strong> exchange by the V Int family <strong>site</strong>-speci¢c recombinases. The DNA str<strong>and</strong> swapping/isomerisation model is presented.<br />

The letter `Y' refers to the conserved catalytic tyrosine. Other symbols are as in Fig. 5. The top str<strong>and</strong>s (thick lines) are cleaved<br />

¢rst (a), swapped between the two partners (b), <strong>and</strong> then religated (c). The branch point of the generated Holliday junction intermediate<br />

is positioned at the middle of the (6-bp) overlap region <strong>and</strong> the top str<strong>and</strong>s are crossed. Isomerisation of the Holliday junction to a recombinant<br />

con¢guration in which the bottom str<strong>and</strong>s are crossed requires the reorganisation of the DNA helices <strong>and</strong> the four half-<strong>site</strong>sbound<br />

recombinase subunits within the complex (d). The resulting Holliday junction isoform is resolved by repeating steps a to c in order<br />

to exchange the bottom str<strong>and</strong>s (e).<br />

QN resolvase-DNA co-complex could be compatible<br />

with limited distortions [95].<br />

The conformational changes that take place during<br />

str<strong>and</strong> exchange appear less straightforward.<br />

Since resolvase cleaves the DNA in cis <strong>and</strong> seems<br />

not to dissociate from its binding <strong>site</strong> during <strong>recombination</strong><br />

[104], two models have been proposed to<br />

account for the apparent 180³ half-<strong>site</strong> rotation in<br />

the reaction. In the `subunit rotation' model, str<strong>and</strong><br />

exchange is coupled to a rotational rearrangement of<br />

the DNA-linked recombinase subunits within the tetramer<br />

[91]. A major di¤culty with this mechanism is<br />

that the recombinase dimer interface holding the<br />

cleaved ends in the complex must be disrupted transiently<br />

during the dissociation/reassociation process<br />

(Fig. 5). An alternative `static subunit' model postulates<br />

that the recombinase tetramer does not dissociate<br />

<strong>and</strong> that <strong>recombination</strong> occurs by localised<br />

conformational <strong>and</strong> topological rearrangement of<br />

the DNA molecules within the complex [95,99].<br />

This second model is di¤cult to reconcile with the<br />

recent observation that Tn3 resolvase brings about<br />

multiple rounds of rotations without rejoining the<br />

single 180³ rotation intermediate. It is argued that<br />

in a static subunit mechanism, such an iteration of<br />

the reaction would entangle the DNA within the<br />

catalytic complex to an unacceptable level [89,105].<br />

Unlike the recombinases of the resolvase/invertase<br />

family, <strong>site</strong>-speci¢c recombinase related to V Int such<br />

as the Cre recombinase of phage P1, E. coli XerC<br />

<strong>and</strong> XerD <strong>and</strong> the Flp protein from the yeast 2W<br />

plasmid, exchange the two pairs of DNA str<strong>and</strong>s<br />

separately <strong>and</strong> sequentially (Fig. 6) [3,4,31,106].<br />

To initiate the ¢rst str<strong>and</strong> exchange, the tyrosine<br />

residue of the conserved catalytic motif RHRY<br />

attacks a speci¢c scissile phosphate in one str<strong>and</strong><br />

(de¢ned here after as the top str<strong>and</strong>) of each <strong>recombination</strong><br />

core <strong>site</strong>s, thereby forming a 3P phosphotyrosyl-linked<br />

recombinase-DNA complex <strong>and</strong><br />

generating a free 5P-OH end (Fig. 6). The polarity<br />

of this cleavage reaction is thus reversed when compared<br />

to that of the resolvase/invertase-mediated<br />

cleavages (compare also Figs. 3C <strong>and</strong> 3D). In a sec-<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178<br />

3.2.2. The V Int family : sequential pairs of DNA<br />

exchange<br />

str<strong>and</strong>


B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178 169<br />

ond step, the recombinase-DNA phosphotyrosyl<br />

bond is attacked by the 5P-OH end from the partner<br />

bottom str<strong>and</strong> exchange [111,112]. Also, in the recombinase-Holliday<br />

junction complex, the centre of<br />

duplex to generate a four-way branched structure, or<br />

the overlap region is partially unstacked [113].<br />

`Holliday junction' intermediate, in which only two<br />

Although variations of both models can be envisioned,<br />

the str<strong>and</strong> swapping mechanism seems more<br />

DNA str<strong>and</strong>s have recombined. To resolve this intermediate<br />

<strong>and</strong> complete the <strong>recombination</strong> reaction,<br />

suitable for other <strong>recombination</strong> systems of the V Int<br />

the two other (bottom) str<strong>and</strong>s are exchanged by<br />

repeating the cleavage/religation process 6^8 bp<br />

family, such as conjugative transposons <strong>and</strong> integrons,<br />

which are less strict in their requirement for<br />

downstream of the ¢rst str<strong>and</strong> cleavage position<br />

sequence homology between partner <strong>recombination</strong><br />

(Fig. 6).<br />

<strong>site</strong>s [23,43]. These systems may be less sensitive to<br />

DNA mispairing by catalysing both str<strong>and</strong> exchanges<br />

without annealing recombinant str<strong>and</strong>s. Alternatively,<br />

the Holliday junction iso-form generated<br />

As in the resolvase/invertase-mediated <strong>recombination</strong>,<br />

sequence homology in the 6^8 bp overlap region<br />

that separates the top <strong>and</strong> bottom str<strong>and</strong> cleavage<br />

positions in the two partner <strong>recombination</strong> core<br />

<strong>site</strong>s is also essential for most (but not all) of the<br />

by a single str<strong>and</strong> exchange may be a stable <strong>recombination</strong><br />

product that is not processed back by the<br />

systems belonging to the V Int family. Although<br />

recombinases but could be well resolved by some<br />

other host function (as discussed in Ref. [113]).<br />

the homology appears to play a role after the synapsis<br />

of the two core <strong>site</strong>s, the exact mechanism by<br />

Another matter of divergence within the V Int<br />

which it in£uences the reaction remains unclear. The<br />

family concerns the catalytic role of the di¡erent<br />

recombinase subunits within the synaptic complex.<br />

classical model supposes that sequence identity between<br />

<strong>recombination</strong> <strong>site</strong>s is required for a reversible<br />

A wealth of data indicate that the yeast recombinase<br />

process called `branch migration' that moves the<br />

Flp uses a trans cleavage mechanism in which the<br />

branch point of the Holliday junction from its <strong>site</strong><br />

catalytic RHR triad of one monomer activates the<br />

of formation at one end of the overlap region to the<br />

scissile phosphodiester adjacent to its binding <strong>site</strong>,<br />

<strong>site</strong> of resolution at the oppo<strong>site</strong> end. This is<br />

while the tyrosine nucleophile is contributed by a<br />

achieved by stepwise melting <strong>and</strong> reannealing of<br />

di¡erent Flp monomer bound on a di¡erent half<strong>site</strong>.<br />

As for MuA, this active <strong>site</strong> sharing observed<br />

for Flp was proposed as a mechanism for coupling<br />

catalysis <strong>and</strong> <strong>recombination</strong> <strong>site</strong> pairing [31,114].<br />

the complementary str<strong>and</strong>s of the parental <strong>and</strong> partner<br />

DNA duplexes, [5,107]. This view is now challenged<br />

by an alternative `str<strong>and</strong> swapping/isomerisation'<br />

mechanism (Fig. 6) [108]. The model proposes<br />

However, this view was weakened by experiments<br />

that, after cleavage, two or three nucleotides from<br />

showing that the two collaborating monomers are<br />

the parental overlap sequences are melted <strong>and</strong> then<br />

bound on the same core <strong>site</strong> <strong>and</strong> that cleavage may<br />

actually precede synapsis [115,116]. An alternative<br />

swapped between the partner duplexes. In this mechanism,<br />

sequence homology is required in the reannealing<br />

reaction that orients the 5P-OH end of the<br />

model based on an asymmetrical (head to tail) assembly<br />

of Flp monomers has been proposed, again<br />

suggesting that active <strong>site</strong>-dependent molecular<br />

invading str<strong>and</strong> for ligation. Movement of the Holliday<br />

junction is limited to the 1^3 central bp of the<br />

bridges may be required [117]. In contrast, all current<br />

data indicate that XerC <strong>and</strong> XerD recombinases<br />

overlap region. This movement would simply be unstacking-restacking<br />

events whereby the Holliday<br />

act on the closest cleavage <strong>site</strong> by providing all catalytic<br />

residues in cis [118]. For V Int, experiments<br />

supporting both cis <strong>and</strong> trans mechanisms have<br />

junction DNA helices are reorganised from a `parental'<br />

(top str<strong>and</strong>s crossed) con¢guration to a `recombinant'<br />

(bottom str<strong>and</strong>s crossed) con¢guration<br />

been reported [119,120]. These apparently con£icting<br />

results have raised a number questions that have<br />

in which the bottom str<strong>and</strong> cleavage <strong>site</strong>s are adequately<br />

positioned for str<strong>and</strong> exchange (Fig. 6)<br />

been debated in the recent literature [48,121^123].<br />

[108^110]. This view is supported by recent work<br />

Although the details of the molecular interactions<br />

between recombinases subunits are not known, some<br />

showing that one Holliday junction isomer preferentially<br />

undergoes top str<strong>and</strong> exchange, whereas resolution<br />

of the other iso-form predominantly occurs by<br />

new insight has been provided by the recent acquisition<br />

of structural data for the C-terminal catalytic


170<br />

attP is wrapped in an `intasome' complex in which<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178<br />

moieties of V Int <strong>and</strong> the Haemophilus phage 1 integrase<br />

(HP1 Int) <strong>and</strong> for the complete 298 aa XerD<br />

These systems contain additional accessory DNA sequences,<br />

to which further recombinase subunits <strong>and</strong>/<br />

or other proteins bind thereby forming part of the<br />

recombinase [124^126]. The C-terminal catalytic domain<br />

of these three recombinases exhibits a similar<br />

functional nucleoprotein complex. In many cases,<br />

fold where the conserved residues RHR are exposed<br />

DNA supercoiling is also required for the assembly<br />

in a basic groove likely to contact the DNA for the<br />

<strong>and</strong>/or activity of the <strong>recombination</strong> complex. The<br />

activation of the scissile phosphodiester bond. However,<br />

the structure of the region containing the active<br />

tyrosine (the extreme C-terminal part of the domain)<br />

is very di¡erent in the three structures. This region<br />

precise architecture of the complex provides a mechanism<br />

for controlling the outcome of a <strong>recombination</strong><br />

reaction, while a requirement for cellular proteins<br />

may provide a means of relating the frequency<br />

of a <strong>recombination</strong> reaction to host metabolism.<br />

also seems to be involved in recombinase inter-subunit<br />

interactions. In V Int, the tyrosine is positioned<br />

Thus, the requirement for higher order interactions<br />

on a disordered £exible loop in a con¢guration<br />

represents an additional level of regulation ensuring<br />

which, by allowing conformational changes, could<br />

that <strong>recombination</strong> only occurs at the correct time,<br />

be consistent with either a cis or trans cleavage<br />

between the correct <strong>site</strong>s [5,9].<br />

[124]. In contrast, the structures of both HP1 Int<br />

<strong>and</strong> XerD support a cis mechanism, the tyrosine<br />

For example, bacteriophage V integration <strong>and</strong> excision<br />

need to be irreversible <strong>and</strong> separate steps of<br />

being docked at a ¢xed position within the catalytic<br />

phage development to permit a committed decision<br />

pocket. In the HP1 Int catalytic domain, which crystallised<br />

as a dimer, the tyrosine is in a position ready<br />

between lysogeny <strong>and</strong> lytic growth [42]. Both <strong>recombination</strong><br />

reactions are mediated by the formation of<br />

for in line nucleophilic attack of the scissile phosphate<br />

[125], whereas in the model proposed for<br />

nucleoprotein structures that arrange the <strong>recombination</strong><br />

partners in a complementary con¢guration for<br />

XerD-DNA complex, some local alteration of the<br />

synapsis [4,42,127]. In addition to the core <strong>site</strong> DNA<br />

structure would be required to bring the residue in<br />

binding activity, V Int has a second DNA binding<br />

an active con¢guration [126]. As for QN resolvase, it is<br />

domain that recognises accessory `arm <strong>site</strong>s' in the<br />

thought that this conformational activation is induced<br />

by protein interactions between XerD <strong>and</strong><br />

£anking regions of the <strong>recombination</strong> loci. This enables<br />

a single recombinase protomer to form a molecular<br />

bridge between adjacent <strong>site</strong>s on the same<br />

the partner recombinase XerC (our unpublished results).<br />

Another major conformational change would<br />

DNA molecules or between the two recombining<br />

be required to move the globular N-terminal domain<br />

of XerD (absent from the V <strong>and</strong> HP1 Int structures)<br />

duplexes in the synaptic complex. For both integration<br />

<strong>and</strong> excision, Int subunits bound to the arms of<br />

that blocks the access to the active <strong>site</strong>. The fact that<br />

one partner are delivered in trans to the core <strong>site</strong> of<br />

the other partner.<br />

the peptide connecting the two XerD domains is disordered<br />

in the crystal is again indicative of structural<br />

However, V Int-mediated integration <strong>and</strong> excision<br />

£exibility [126].<br />

employ distinct subsets of proteins acting on di¡erent<br />

DNA binding <strong>site</strong>s <strong>and</strong>, therefore, the two reactions<br />

are not the simple reverse of each other. For<br />

integration, the 240-bp phage <strong>recombination</strong> locus<br />

4. Higher order nucleoprotein complexes: an<br />

additional level of control<br />

the relevant Int subunits can capture the naked 25-<br />

The strategies used by transposable elements <strong>and</strong><br />

bp core region of the bacterial attachment <strong>site</strong>, attB<br />

<strong>site</strong>-speci¢c <strong>recombination</strong> systems to assemble their<br />

<strong>recombination</strong> machinery share many similarities. In<br />

[127,128]. Formation of the intasome involves precisely<br />

positioned DNA bends induced by the host<br />

particular, for the majority of elements, the minimal<br />

number of recombinase molecules bound to their<br />

protein IHF <strong>and</strong> a set of Int-mediated bridges between<br />

the core <strong>site</strong> <strong>and</strong> speci¢c arm binding <strong>site</strong>s in<br />

binding <strong>site</strong>s on DNA dictated by the chemistry of<br />

the £anking DNA segments of attP. Excision utilises<br />

two other proteins in addition to Int <strong>and</strong> IHF,<br />

the reaction is usually insu¤cient to mediate a `normal'<br />

<strong>recombination</strong> reaction at a `normal' frequency.<br />

namely the phage-encoded Xis <strong>and</strong> the host factor


B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178 171<br />

7. Models for the assembly of a synaptic complex for di¡erent specialised <strong>recombination</strong> systems showing topological selectivity. a:<br />

Fig.<br />

in resolvase/res system. b: Invertasome of DNA inversion systems. c: Synaptic complex assembled by Xer <strong>recombination</strong><br />

Synaptosome<br />

at the ColE1 plasmid resolution <strong>site</strong> cer. d: Possible structure of the LER synaptic complex intermediate during bacteriophage Mu<br />

system<br />

formation. Grey arrows represent the directly repeated (DR) or inverted repeated (IR) <strong>recombination</strong> core <strong>site</strong>s, or the<br />

transpososome<br />

as well as a di¡erent set of arm <strong>site</strong>/core <strong>site</strong>s<br />

FIS,<br />

to assemble a synaptic complex between<br />

interactions,<br />

<strong>and</strong> attR.<br />

this complex, the four Int subunits are now contributed<br />

In<br />

by both partners (three from attL <strong>and</strong> one<br />

distinct <strong>and</strong> mutually exclusive `lock <strong>and</strong> key'<br />

use<br />

which allow the phage to become com-<br />

interactions<br />

to either lysogeny or lytic growth [5,42]. Furthermoremitted<br />

the higher order interactions responsible<br />

the assembly of the integrative <strong>and</strong> excisive synaptic<br />

for<br />

structures also in£uence the resolution of the<br />

Mu left (L) <strong>and</strong> right (R) ends. Open ribbons are the res <strong>site</strong> accessory sub<strong>site</strong>s II <strong>and</strong> III, the cer <strong>site</strong> accessory sequences<br />

bacteriophage<br />

or the <strong>recombination</strong>al <strong>and</strong> transpositional enhancers (E). In the Mu LER complex, the thick <strong>and</strong> thin lines represent the phage ge-<br />

(AS),<br />

<strong>and</strong> the donor DNA molecule, respectively. The wavy line is the target DNA. Only the MuA subunits that will form the active tetramenome<br />

after conversion of the LER complex are represented. Additional molecules of MuA <strong>and</strong> of the proteins HU <strong>and</strong> IHF which participate<br />

in the formation of the LER complex are not shown.<br />

the ends of the integrated prophage attL<br />

from attR) [129]. Thus, V integration <strong>and</strong> excision


172<br />

maintained (reviewed in Refs. [6,15,66,67]). Like V<br />

be<br />

MuA has two distinct DNA binding domains<br />

Int,<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178<br />

to determine the right-h<strong>and</strong>ed rotational direction of<br />

Holliday junction intermediates, thereby reinforcing<br />

the directionality of both <strong>recombination</strong> reactions<br />

str<strong>and</strong> exchange [90^92,132]. As mentioned above,<br />

[130].<br />

single amino acid changes in resolvase <strong>and</strong> invertase<br />

enzymes allow the requirement for supercoiling <strong>and</strong><br />

During V integration <strong>and</strong> excision, DNA supercoiling<br />

acts as an additional architectural element<br />

facilitating directional wrapping of the DNA around<br />

higher order nucleoprotein interactions to be overcome,<br />

underlining the regulatory function of these<br />

interactions ([92,100,101,103] ; M.R. Boockock <strong>and</strong><br />

the complex, but <strong>recombination</strong> then occurs by r<strong>and</strong>om<br />

collision between <strong>site</strong>s in various con¢gurations.<br />

In contrast, it is crucial for the in vivo function of<br />

N.D.F. Grindley, personal communication). By promoting<br />

deletions <strong>and</strong> inversions as well as fusion<br />

<strong>site</strong>-speci¢c resolution <strong>and</strong> DNA inversion systems<br />

products, the behaviour of these relaxed recombinase<br />

to recombine selectively <strong>site</strong>s which are in direct or<br />

variants is similar to that of recombinases such as<br />

Cre <strong>and</strong> Flp which mediate <strong>recombination</strong> with little<br />

inverted repeat on the same DNA molecule, respectively.<br />

For these systems, such selectivity is achieved<br />

selectivity for a particular <strong>site</strong> organisation <strong>and</strong> with<br />

through a <strong>recombination</strong>al synapse of a precise local<br />

no requirement for a speci¢c DNA topology [4,106].<br />

geometry [80].<br />

The Xer <strong>recombination</strong> system provides an intriguing<br />

example of functional £exibility. Recombination<br />

at the E. coli chromosome <strong>site</strong> dif which can<br />

The 120-bp resolvase <strong>recombination</strong> <strong>site</strong> res contains<br />

three sub<strong>site</strong>s, each of which binds two resolvase<br />

subunits. In the <strong>recombination</strong> complex, str<strong>and</strong><br />

occur intermolecularly <strong>and</strong> intramolecularly appears<br />

to be achieved solely by the action of the XerC <strong>and</strong><br />

exchange occurs at sub<strong>site</strong> I, while the eight resolvase<br />

molecules bound to the other two sub<strong>site</strong>s<br />

XerD recombinases on a 28-bp core <strong>recombination</strong><br />

of the two synapsed res <strong>site</strong>s form a protein sca¡old<br />

<strong>site</strong>. In contrast, Xer <strong>recombination</strong> at <strong>site</strong>s present<br />

around which three plectonemic DNA supercoils are<br />

in natural multicopy plasmids (e.g., cer in ColE1<br />

interwrapped (Fig. 7a) [36]. Such a 33 synapse, in<br />

plasmid) is preferentially intramolecular <strong>and</strong> has<br />

which three negative supercoils are trapped, can only<br />

the requirement for<br />

V200 bp of accessory DNA<br />

form readily on directly repeated <strong>site</strong>s on the same<br />

sequence <strong>and</strong> for additional host proteins (ArgR,<br />

DNA molecule, thereby limiting <strong>recombination</strong> to<br />

the repressor of the arginine regulon ; <strong>and</strong> PepA,<br />

intramolecular resolution. To form a 33 synapse<br />

the aminopeptidase A) [40]. These are involved in<br />

between inverted repeated <strong>site</strong>s, or between <strong>site</strong>s on<br />

forming a synapse of de¢ned geometry in which, as<br />

separate circular molecules, unfavourable compensating<br />

positive supercoils would have to be introduced<br />

elsewhere. This system therefore constitutes a<br />

in the resolvase synaptosome, three negative supercoils<br />

are trapped (Fig. 7c) [133]. It is likely that functionally<br />

equivalent topological ¢lters have evolved<br />

so-called `topological ¢lter' [80]. A similar topological<br />

¢lter is used in the inversion reaction catalysed by<br />

independently in a number of di¡erent <strong>recombination</strong><br />

systems by multiplying the recombinase binding<br />

the DNA invertases like Gin <strong>and</strong> Hin. In this case<br />

however, two negative supercoils are trapped in the<br />

<strong>site</strong>s within the <strong>recombination</strong> locus (as in the resolvase<br />

system) <strong>and</strong>/or by recruiting di¡erent host DNA<br />

three-branched `invertasome' complex in which the<br />

binding proteins (as in Xer <strong>recombination</strong> or in Gin<br />

two inverted repeated <strong>recombination</strong> <strong>site</strong>s bound by<br />

<strong>and</strong> Hin inversion systems).<br />

the recombinase are aligned on a third <strong>and</strong> distant<br />

element, the <strong>recombination</strong>al enhancer, which is<br />

bound by the host factor FIS (Fig. 7b) [32,80,85].<br />

In transposition, the higher order interactions controlling<br />

the assembly of the active complex, or `transpososome',<br />

not only dictate the con¢guration with<br />

The FIS-bound enhancer is required for both for<br />

which the transposon ends can interact, but they<br />

the synaptic complex assembly <strong>and</strong> the activation of<br />

also in£uence target DNA <strong>site</strong> selection. Multiple<br />

the recombinase subunits. The latter may be<br />

achieved by inducing a conformational change at<br />

checkpoints in the assembly <strong>and</strong> progression of bacteriophage<br />

Mu transpososome ensure that each cycle<br />

the dimer interface [92,103,131]. Likewise, in both<br />

resolvase <strong>and</strong> invertase-mediated <strong>recombination</strong>,<br />

of replicative transposition will be successfully executed<br />

<strong>and</strong> that the integrity of the phage genome will<br />

DNA supercoiling energy is used after initial synapsis<br />

to provide the driving force for the reaction <strong>and</strong>


B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178 173<br />

to the transposase TnsA+B core, three other proteins<br />

enabling a single transposase subunit to interact with<br />

speci¢c sequence elements at the ends of Mu, as well<br />

(TnsC, D <strong>and</strong> E) participate in Tn7 transposition<br />

as with a transposition enhancer located within the<br />

[11]. The use of either TnsD or TnsE in the complex<br />

Mu genome, about 1kb from the left end. At an<br />

directs the transposon to di¡erent types of target<br />

early stage of the transposition reaction, a complex<br />

<strong>site</strong>s. TnsD is responsible for the recognition of the<br />

circuit of interactions between MuA subunits bound<br />

speci¢c chromosome integration <strong>site</strong> attTn7, whereas<br />

to the left end (L), the enhancer (E), <strong>and</strong> the right<br />

end (R) promotes the formation of a transient three<strong>site</strong><br />

synaptic complex called `LER' (Fig. 7d) [134].<br />

TnsE allows Tn7 to transpose with little target speci¢city<br />

into di¡erent <strong>site</strong>s. TnsC mediates communication<br />

between the transposon ends <strong>and</strong> the target<br />

Assembly of this complex requires DNA supercoiling<br />

as well as precisely-positioned DNA bends induced<br />

by the host proteins HU <strong>and</strong> IHF. Upon assembly,<br />

DNA complex, <strong>and</strong> has an immunity function analogous<br />

to that of MuB. In the TnsD-dependent pathway,<br />

no cleavage is observed in the absence of the<br />

target DNA, suggesting that all participants of the<br />

the catalytically inert components of the LER complex<br />

undergo a conformational transition converting<br />

reaction, i.e., TnsA, B, C <strong>and</strong> D, the transposon ends<br />

<strong>and</strong> the target <strong>site</strong> attTn7, must be assembled in the<br />

the inactive pre-transpososome into the stable synaptic<br />

complex in which the two Mu ends <strong>and</strong> the active<br />

transpososome to initiate <strong>recombination</strong> [22,139].<br />

<strong>site</strong>s of the MuA tetramer are engaged for catalysis.<br />

As for other <strong>recombination</strong> systems, the geometry<br />

By contrast, in IS10 transposition, the doublestr<strong>and</strong><br />

breaks severing the element from the donor<br />

of the LER complex may act as a topological ¢lter<br />

restricting synapsis to two adjacent inverted ends<br />

locus must occur before the transpososome can interact<br />

with the target DNA [76]. IS10 transposase is<br />

[135]. DNA supercoiling may also provide the free<br />

also able to use transposon ends located on distinct<br />

molecules or in directly repeated con¢guration rather<br />

energy required for the DNA <strong>and</strong>/or protein conformational<br />

changes likely to occur during conversion<br />

than the canonical inverted orientation, to carry-out<br />

transposition [140]. Finally, when compared to Mu<br />

of the LER complex [136]. In transpososome assembly,<br />

the enhancer acts as a platform from which<br />

<strong>and</strong> Tn7, IS10 is also more promiscuous with respect<br />

MuA monomers are delivered to form the active<br />

to target <strong>site</strong> selection. Intra-transposon insertion<br />

events are observed <strong>and</strong> may be favoured through<br />

tetramer at the Mu ends [68,137,138]. These MuApromoted<br />

bridges between the enhancer <strong>and</strong> the two<br />

a regulation pathway involving the host factor IHF<br />

ends are reminiscent of those formed by V Int for the<br />

synapse of two <strong>recombination</strong> <strong>site</strong>s.<br />

A similar role as a sca¡old has been proposed for<br />

the Mu transposition cofactor protein, MuB [137].<br />

([141] ; R. Chalmers <strong>and</strong> N. Kleckner, personal communication).<br />

This relative £exibility of IS10 transposition<br />

accounts for how IS10-promoted adjacent deletions<br />

<strong>and</strong> replicon fusions occur, as well as how<br />

MuB is an allosteric activator of MuA that is also<br />

new transposable element based on IS10 can form<br />

involved in target capture <strong>and</strong> in the mechanism of<br />

[50].<br />

target immunity whereby self-integration by Mu is<br />

prevented. Such immunity is conferred by a set of<br />

interactions between MuA <strong>and</strong> MuB that precludes<br />

5. Conclusion<br />

Recent advances in our underst<strong>and</strong>ing of transposition<br />

<strong>and</strong> <strong>site</strong>-speci¢c <strong>recombination</strong> have provided<br />

the formation of a MuB-DNA complex in the vicinity<br />

of a transposase-bound end. Although association<br />

of a MuB-bound target DNA to the transpososome<br />

is not absolutely required to initiate<br />

new pictures of specialised <strong>recombination</strong> machines.<br />

transposition, its presence strongly stimulates both<br />

Some of these snapshots must now be viewed with<br />

the end cleavage <strong>and</strong> str<strong>and</strong> transfer reactions<br />

[6,15]. This ensures that Mu transposition will only<br />

stereoglasses ! The emerging theme is that recombinases<br />

are built by combining simple <strong>and</strong> conserved<br />

chemical mechanisms of DNA cut <strong>and</strong> paste with<br />

occur in an appropriate target, thus avoiding selfdestruction.<br />

speci¢c DNA binding activities <strong>and</strong> the ability to<br />

The coordination of the transposition reactions<br />

form a complex in which the DNA substrates are<br />

seems even tighter in the case of Tn7. In addition<br />

brought together in order to trigger the recombina-


174<br />

[1] Eggleston, A.K. <strong>and</strong> West, S.C. (1996) Exchanging partners :<br />

<strong>recombination</strong> in E. coli. Trends Genet. 12, 20^26.<br />

[2] Sadowski, P.D. (1993) Site-speci¢c genetic <strong>recombination</strong> :<br />

hops, £ips, <strong>and</strong> £ops. FASEB J. 7, 760^767.<br />

[3] Stark, W.M., Boocock, M.R. <strong>and</strong> Sherratt, D.J. (1992) Catalysis<br />

by <strong>site</strong>-speci¢c recombinases. Trends Genet. 8, 432^439.<br />

[4] L<strong>and</strong>y, A. (1993) Mechanistic <strong>and</strong> structural complexity in the<br />

<strong>site</strong>-speci¢c <strong>recombination</strong> pathways of Int <strong>and</strong> FLP. Curr.<br />

Opin. Genet. Dev. 3, 699^707.<br />

[5] Nash, H.A. (1996) Site-speci¢c <strong>recombination</strong> : integration,<br />

excision, resolution, <strong>and</strong> inversion of de¢ned DNA segments.<br />

In : Escherichia coli <strong>and</strong> Salmonella typhimurium : Cellular <strong>and</strong><br />

Molecular Biology (Neidhart, F.C. et al., Eds.), 2nd edn., Vol.<br />

2, pp. 2363^2376. American Society for Microbiology, Washington,<br />

DC.<br />

[6] Mizuuchi, K. (1992) <strong>Transposition</strong>al <strong>recombination</strong> : mechanistic<br />

insights from studies of Mu <strong>and</strong> other elements. Annu.<br />

Rev. Biochem. 61, 1011^1051.<br />

[7] Polard, P. <strong>and</strong> Ch<strong>and</strong>ler, M. (1995) Bacterial transposases <strong>and</strong><br />

retroviral integrases. Mol. Microbiol. 15, 13^23.<br />

[8] Plasterk, R.H.A. (1995) Mechanisms of DNA transposition.<br />

In : Mobile Genetic Elements (Sherratt, D.J., Ed.), pp. 18^37.<br />

IRL,<br />

Oxford.<br />

[9] Craig, N.L. (1996) <strong>Transposition</strong>. In : Escherichia coli <strong>and</strong><br />

Salmonella typhimurium : Cellular <strong>and</strong> Molecular Biology<br />

(Neidhart, F.C. et al., Eds.), 2nd edn., Vol. 2, pp. 2339^<br />

2362. American Society for Microbiology, Washington, DC.<br />

[10] Craig, N.L. (1997) Target <strong>site</strong> selection in transposition.<br />

Annu. Rev. Biochem. 66, 437^474.<br />

[11] Craig, N.L. (1996) Transposon Tn7. Curr. Top. Microbiol.<br />

Immunol. 204, 27^48.<br />

[12] Bender, J., Kuo, J. <strong>and</strong> Kleckner, N. (1991) Genetic evidence<br />

against intramolecular rejoining of the donor DNA molecule<br />

following IS10 transposition. Genetics 128, 687^694.<br />

[13] Hagemann, A.T. <strong>and</strong> Craig, N.L. (1993) Tn7 transposition<br />

creates a hotspot for homologous <strong>recombination</strong> at the transposon<br />

donor <strong>site</strong>. Genetics 133, 9^16.<br />

[14] Ohtsubo, F. <strong>and</strong> Sekine, Y. (1996) Bacterial insertion sequences.<br />

Curr. Top. Microbiol. Immunol. 204, 1^26.<br />

[15] Lavoie, B.D. <strong>and</strong> Chaconas, G. (1996) <strong>Transposition</strong> of phage<br />

Mu DNA. Curr. Top. Microbiol. Immunol. 204, 83^102.<br />

[16] Fayet, O., Ramond, P., Polard, P., Prere, M.F. <strong>and</strong> Ch<strong>and</strong>ler,<br />

M. (1990) Functional similarities between retroviruses <strong>and</strong> the<br />

IS3 family of bacterial insertion sequences ? Mol. Microbiol. 4,<br />

1771^1777.<br />

[17] Khan, E., Mack, J.P., Katz, R.A., Kulkosky, J. <strong>and</strong> Skalka,<br />

A.M. (1991) Retroviral integrase domains : DNA binding <strong>and</strong><br />

the recognition of LTR sequences. Nucleic Acids Res. 19,<br />

851^860.<br />

[18] Kulkosky, J., Jones, K.S., Katz, R.A., Mack, J.P. <strong>and</strong> Skalka,<br />

A.M. (1992) Residues critical for retroviral integrative <strong>recombination</strong><br />

in a region that is highly conserved among retroviral/<br />

retrotransposon integrases <strong>and</strong> bacterial insertion sequence<br />

transposases. Mol. Cell. Biol. 12, 2331^2338.<br />

[19] Rezsohazy, R., Hallet, B., Delcour, J. <strong>and</strong> Mahillon, J. (1993)<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178<br />

References<br />

tion reactions. Adding more requirements to the system<br />

allows the outcome of <strong>recombination</strong> to be controlled<br />

at a higher level to prevent undesired DNA<br />

rearrangements.<br />

Although the £exibility of the conservative <strong>recombination</strong><br />

mechanism used by <strong>site</strong>-speci¢c recombinases<br />

to achieve di¡erent outcomes is obvious, the<br />

mechanism of one-step transesteri¢cation used by<br />

the DDE transposases seems restricted to the movement<br />

of transposable elements. Nevertheless, the<br />

DNA double-str<strong>and</strong> break reactions that initiate<br />

V(D)J <strong>recombination</strong>, which in vertebrates serves<br />

to assemble the immunoglobulin <strong>and</strong> T-cell receptor<br />

genes, are chemical equivalents of those catalysed by<br />

the DDE recombinases [142]. Further characterisation<br />

of the bacterial systems that are not related to<br />

those reviewed here, such as IS1, IS91 <strong>and</strong> the Piv<br />

family of recombinases, will undoubtedly provide<br />

new examples on how di¡erent DNA breakage <strong>and</strong><br />

joining mechanisms are adapted to promote speci¢c<br />

DNA rearrangements [25,26,28].<br />

For the di¡erent systems, the next challenge is to<br />

obtain a dynamic view describing how the <strong>recombination</strong><br />

partners come together <strong>and</strong> then, how the<br />

structure of the <strong>recombination</strong> complex changes during<br />

the di¡erent reaction steps. These studies are not<br />

only crucial for the underst<strong>and</strong>ing of fundamental<br />

<strong>recombination</strong> mechanisms, but should also assist<br />

in the improvement or development of new tools<br />

based on transposition <strong>and</strong> <strong>site</strong>-speci¢c <strong>recombination</strong><br />

systems for use in a variety of in vivo <strong>and</strong> in<br />

vitro<br />

applications.<br />

Acknowledgments<br />

We are grateful to Sean Colloms, Francois Cornet<br />

<strong>and</strong> Finbarr Hayes for critical reading of the manuscript<br />

<strong>and</strong> for helpful comments. Work in D.J.S.'s<br />

laboratory was supported by grants from the Wellcome<br />

Trust <strong>and</strong> the Medical Research Council. B.H.<br />

was the recipient of a postdoctoral fellowship from<br />

the European Molecular Biology Organization <strong>and</strong><br />

from the European Communities BIOTECH program.


served transposase motif. Mol. Microbiol. 9, 1283^1295.<br />

[20] Doak, T.G., Doerder, F.P., Jahn, C.L. <strong>and</strong> Herrick, G. (1994)<br />

A proposed superfamily of transposase genes : transposon-like<br />

elements in ciliated protozoa <strong>and</strong> a common `D35E' motif.<br />

Proc. Natl. Acad. Sci. USA 91, 942^946.<br />

[21] Baker, T.A. <strong>and</strong> Luo, L. (1994) Identi¢cation of residues in<br />

the Mu transposase essential for catalysis. Proc. Natl. Acad.<br />

Sci. USA 91, 6654^6658.<br />

[22] Sarnovsky, R.J., May, E.W. <strong>and</strong> Craig, N.L. (1996) The Tn7<br />

transposase is a heteromeric complex in which DNA breakage<br />

<strong>and</strong> joining activities are distributed between di¡erent gene<br />

products. EMBO J. 15, 6348^6361.<br />

[23] Scott, J.R. <strong>and</strong> Churchward, G.G. (1995) Conjugative transposition.<br />

Annu. Rev. Microbiol. 49, 367^397.<br />

[24] Bannam, T.L., Crellin, P.K. <strong>and</strong> Rood, J.I. (1995) Molecular<br />

genetics of the chloramphenicol-resistance transposon Tn4451<br />

from Clostridium perfringens : the TnpX <strong>site</strong>-speci¢c recombinase<br />

excises a circular transposon molecule. Mol. Microbiol.<br />

16, 535^551.<br />

[25] Lenich, A.G. <strong>and</strong> Glasgow, A.C. (1994) Amino acid sequence<br />

homology between Piv, an essential protein in <strong>site</strong>-speci¢c<br />

DNA inversion in Moraxella lacunata, <strong>and</strong> transposases of<br />

an unusual family of insertion elements. J. Bacteriol. 176,<br />

4160^4164.<br />

[26] Serre, M.C., Turlan, C., Bortolin, M. <strong>and</strong> Ch<strong>and</strong>ler, M.<br />

(1995) Mutagenesis of the IS1 transposase : importance of a<br />

His-Arg-Tyr triad for activity. J. Bacteriol. 177, 5070^5077.<br />

[27] Mendiola, M.V. <strong>and</strong> de la Cruz, F. (1992) IS91 transposase is<br />

related to the rolling-circle-type replication proteins of the<br />

pUB110 family of plasmids. Nucleic Acids Res. 20, 3521.<br />

[28] Mendiola, M.V., Bernales, I. <strong>and</strong> de la Cruz, F. (1994) Di¡erential<br />

roles of the transposon termini in IS91 transposition.<br />

Proc. Natl. Acad. Sci. USA 91, 1922^1926.<br />

[29] Brook¢eld, J.F.Y. (1995) Transposable elements as sel¢sh<br />

DNA. In : Mobile Genetic Elements (Sherratt, D.J., Ed.),<br />

IRL,<br />

Oxford.<br />

[30] Shapiro, J.A. (1993) Natural genetic engineering of the bacterial<br />

genome. Curr. Opin. Genet. Dev. 3, 845^848.<br />

[31] Jayaram, M. (1994) Phosphoryl transfer in Flp <strong>recombination</strong><br />

: a template for str<strong>and</strong> transfer mechanisms. Trends Biochem.<br />

Sci. 19, 78^82.<br />

[32] van de Putte, P. <strong>and</strong> Goosen, N. (1992) DNA inversions in<br />

phages <strong>and</strong> bacteria. Trends Genet. 8, 457^462.<br />

[33] Komano, T., Kim, S.R., Yoshida, T. <strong>and</strong> Nisioka, T. (1994)<br />

DNA rearrangement of the shu¥on determines recipient speci¢city<br />

in liquid mating of IncI1 plasmid R64. J. Mol. Biol.<br />

243, 6^9.<br />

[34] Sato, T., Samori, Y. <strong>and</strong> Kobayashi, Y. (1990) The cisA cistron<br />

of Bacillus subtilis sporulation gene spoIVC encodes a<br />

protein homologous to a <strong>site</strong>-speci¢c recombinase. J. Bacteriol.<br />

172, 1092^1098.<br />

[35] Carrasco, C.D., Buettner, J.A. <strong>and</strong> Golden, J.W. (1995) Programmed<br />

DNA rearrangement of a cyanobacterial hupL gene<br />

in heterocysts. Proc. Natl. Acad. Sci. USA 92, 791^795.<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178 175<br />

[36] Grindley, N.D.F. (1994) Resolvase-mediated <strong>site</strong>-speci¢c <strong>recombination</strong>.<br />

In : Nucleic Acids <strong>and</strong> Molecular Biology (Eckstein,<br />

F. <strong>and</strong> Lilley, D.M.J., Eds.), pp. 236^267. Springer-Verlag,<br />

Berlin.<br />

[37] Austin, S., Ziese, M. <strong>and</strong> Sternberg, N. (1981) A novel role for<br />

<strong>site</strong>-speci¢c <strong>recombination</strong> in maintenance of bacterial replicons.<br />

Cell 25, 729^736.<br />

[38] Lane, D., de-Feyter, R., Kennedy, M., Phua, S.H. <strong>and</strong> Semon,<br />

D. (1986) D protein of miniF plasmid acts as a repressor of<br />

transcription <strong>and</strong> as a <strong>site</strong>-speci¢c resolvase. Nucleic Acids<br />

Res. 14, 9713^9728.<br />

[39] Alonso, J.C., Ayora, S., Canosa, I., Weise, F. <strong>and</strong> Rojo, F.<br />

(1996) Site-speci¢c <strong>recombination</strong> in gram-positive theta-replicating<br />

plasmids. FEMS Microbiol. Lett. 142, 1^10.<br />

[40] Sherratt, D.J., Arciszewska, L.K., Blakely, G., Colloms, S.,<br />

Grant, K., Leslie, N. <strong>and</strong> McCulloch, R. (1995) Site-speci¢c<br />

<strong>recombination</strong> <strong>and</strong> circular chromosome segregation. Phil.<br />

Trans. R. Soc. Lond. B. Biol. Sci. 347, 37^42.<br />

[41] Lobner-Olesen, A. <strong>and</strong> Kuempel, P.L. (1992) Chromosome<br />

partitioning in Escherichia coli. J. Bacteriol. 174, 7883^7889.<br />

[42] L<strong>and</strong>y, A. (1989) Dynamic, structural, <strong>and</strong> regulatory aspects<br />

of lambda <strong>site</strong>-speci¢c <strong>recombination</strong>. Annu. Rev. Biochem.<br />

58, 913^949.<br />

[43] Hall, R.M. <strong>and</strong> Collis, C.M. (1995) Mobile gene cassettes <strong>and</strong><br />

integrons : capture <strong>and</strong> spread of genes by <strong>site</strong>-speci¢c <strong>recombination</strong>.<br />

Mol. Microbiol. 15, 593^600.<br />

[44] Hatfull, G.F. <strong>and</strong> Grindley, N.D.F. (1988) Resolvases <strong>and</strong><br />

DNA-invertases : a family of enzymes active in <strong>site</strong>-speci¢c<br />

<strong>recombination</strong>. In : Genetic Recombination (Kucherlapati,<br />

R. <strong>and</strong> Smith, G., Eds.), pp. 357^396. American Society for<br />

Microbiology, Washington, DC.<br />

[45] Leschziner, A.E., Boocock, M.R. <strong>and</strong> Grindley, N.D. (1995)<br />

The tyrosine-6 hydroxyl of gamma delta resolvase is not required<br />

for the DNA cleavage <strong>and</strong> rejoining reactions. Mol.<br />

Microbiol. 15, 865^870.<br />

[46] Argos, P., L<strong>and</strong>y, A., Abremski, K., Egan, J.B., Haggard-<br />

Ljungquist, E., Hoess, R.H., Kahn, M.L., Kalionis, B., Narayana,<br />

S.V., Pierson, L.S., Sternberg, N. <strong>and</strong> Leong, J.M.<br />

(1986) The integrase family of <strong>site</strong>-speci¢c recombinases :<br />

regional similarities <strong>and</strong> global diversity. EMBO J. 5, 433^<br />

440.<br />

[47] Abremski, K.E. <strong>and</strong> Hoess, R.H. (1992) Evidence for a second<br />

conserved arginine residue in the integrase family of <strong>recombination</strong><br />

proteins. Protein Eng. 5, 87^91.<br />

[48] Blakely, G.W. <strong>and</strong> Sherratt, D.J. (1996) Cis <strong>and</strong> trans in <strong>site</strong>speci¢c<br />

<strong>recombination</strong>. Mol. Microbiol. 20, 233^238.<br />

[49] Mizuuchi, K. (1992) Polynucleotidyl transfer reactions in<br />

transpositional DNA <strong>recombination</strong>. J. Biol. Chem. 267,<br />

21273^21276.<br />

[50] Kleckner, N., Chalmers, R.M., Kwon, D., Sakai, J. <strong>and</strong> Boll<strong>and</strong>,<br />

S. (1996) Tn10 <strong>and</strong> IS10 transposition <strong>and</strong> chromosome<br />

rearrangements : mechanism <strong>and</strong> regulation in vivo <strong>and</strong> in<br />

vitro. Curr. Top. Microbiol. Immunol. 204, 49^82.<br />

[51] Benjamin, H.W. <strong>and</strong> Kleckner, N. (1992) Excision of Tn10<br />

from the donor <strong>site</strong> during transposition occurs by £ush double-str<strong>and</strong><br />

cleavages at the transposon termini. Proc. Natl.<br />

Acad. Sci. USA 89, 4648^4652.<br />

[52] Bainton, R., Gamas, P. <strong>and</strong> Craig, N.L. (1991) Tn7 transposition<br />

in vitro proceeds through an excised transposon inter-<br />

The IS4 family of insertion sequences : evidence for a con-


176<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178<br />

information within the Mu transposase tetramer : catalytic<br />

mediate generated by staggered breaks in DNA. Cell 65, 805^<br />

816.<br />

contributions of individual monomers. Cell 85, 447^455.<br />

[53] Polard, P. <strong>and</strong> Ch<strong>and</strong>ler, M. (1995) An in vivo transposasecatalyzed<br />

single-str<strong>and</strong>ed DNA circularization reaction. Genes<br />

[72] Savilahti, H. <strong>and</strong> Mizuuchi, K. (1996) Mu transpositional <strong>recombination</strong><br />

: donor DNA cleavage <strong>and</strong> str<strong>and</strong> transfer in<br />

Dev. 9, 2846^2858.<br />

trans by the Mu transposase. Cell 85, 271^280.<br />

[73] Haniford, D. <strong>and</strong> Kleckner, N. (1994) Tn10 transposition in<br />

vivo : temporal separation of cleavages at the two transposon<br />

[54] Polard, P., Ton-Hoang, B., Haren, L., Betermier, M., Walczak,<br />

R. <strong>and</strong> Ch<strong>and</strong>ler, M. (1996) IS911-mediated transpositional<br />

<strong>recombination</strong> in vitro. J. Mol. Biol. 264, 68^81.<br />

ends <strong>and</strong> roles of terminal basepairs subsequent to interaction<br />

[55] Sekine, Y., Eisaki, N. <strong>and</strong> Ohtsubo, E. (1996) Identi¢cation<br />

of ends. EMBO J. 13, 3401^3411.<br />

<strong>and</strong> characterization of the linear IS3 molecules generated by<br />

staggered breaks. J. Biol. Chem. 271, 197^202.<br />

[74] Sakai, J., Chalmers, R.M. <strong>and</strong> Kleckner, N. (1995) Identi¢cation<br />

<strong>and</strong> characterization of a pre-cleavage synaptic complex<br />

that is an early intermediate in Tn10 transposition. EMBO J.<br />

[56] Turlan, C. <strong>and</strong> Ch<strong>and</strong>ler, M. (1995) IS1-mediated intramolecular<br />

rearrangements : formation of excised transposon circles<br />

14, 4374^4383.<br />

<strong>and</strong> replicative deletions. EMBO J. 14, 5410^5421.<br />

[75] Boll<strong>and</strong>, S. <strong>and</strong> Kleckner, N. (1995) The two single-str<strong>and</strong><br />

[57] Craig, N.L. (1995) Unity in transposition reactions. Science<br />

cleavages at each end of Tn10 occur in a speci¢c order during<br />

270, 253^254.<br />

transposition. Proc. Natl. Acad. Sci. USA 92, 7814^7818.<br />

[76] Sakai, J. <strong>and</strong> Kleckner, N. (1997) The Tn10 synaptic complex<br />

[58] Mizuuchi, K. <strong>and</strong> Adzuma, K. (1991) Inversion of the phosphate<br />

chirality at the target <strong>site</strong> of Mu DNA str<strong>and</strong> transfer :<br />

can capture a target DNA only after transposon excision. Cell<br />

evidence for a one-step transesteri¢cation mechanism. Cell 66,<br />

89, 205^214.<br />

129^140.<br />

[77] Boll<strong>and</strong>, S. <strong>and</strong> Kleckner, N. (1996) The three chemical steps<br />

[59] Engelman, A., Mizuuchi, K. <strong>and</strong> Craigie, R. (1991) HIV-1<br />

of Tn10/IS10 transposition involve repeated utilization of a<br />

DNA integration : mechanism of viral DNA cleavage <strong>and</strong><br />

single active <strong>site</strong>. Cell 84, 223^233.<br />

DNA str<strong>and</strong> transfer. Cell 67, 1211^1221.<br />

[78] May, E.W. <strong>and</strong> Craig, N.L. (1996) Switching from cut-<strong>and</strong>paste<br />

to replicative Tn7 transposition. Science 272, 401^404.<br />

[60] Rice, P. <strong>and</strong> Mizuuchi, K. (1995) Structure of the bacteriophage<br />

Mu transposase core : a common structural motif for<br />

DNA transposition <strong>and</strong> retroviral integration. Cell 82, 209^<br />

[79] Wang, J.C. (1996) DNA topoisomerases. Annu. Rev. Biochem.<br />

65, 635^692.<br />

220.<br />

[61] Yang, W. <strong>and</strong> Steitz, T.A. (1995) Recombining the structures<br />

[80] Stark, W.M. <strong>and</strong> Boocock, M.R. (1995) Topological selectivity<br />

in <strong>site</strong>-speci¢c <strong>recombination</strong>. In : Mobile Genetic Elements<br />

of HIV integrase, RuvC <strong>and</strong> RNase H. Structure 3, 131^134.<br />

(Sherratt, D.J., Ed.), pp. 101^129. IRL, Oxford.<br />

[62] Grindley, N.D. <strong>and</strong> Leschziner, A.E. (1995) DNA transposition<br />

: from a black box to a color monitor. Cell 83, 1063^1066.<br />

[81] Johnson, R.C. (1995) Site-speci¢c recombinases an their interaction<br />

with DNA. In : DNA-Protein : Structural Interactions<br />

(Lilley, D.M.J., Ed.), pp. 141^176. IRL, Oxford.<br />

[63] Rice, P., Craigie, R. <strong>and</strong> Davies, D.R. (1996) Retroviral integrases<br />

<strong>and</strong> their cousins. Curr. Opin. Struct. Biol. 6, 76^<br />

83.<br />

[82] Gally, D.L., Leathart, J. <strong>and</strong> Blom¢eld, I.C. (1996) Interaction<br />

of FimB <strong>and</strong> FimE with the ¢m switch that controls the<br />

[64] Davies, J.d., Hostomska, Z., Hostomsky, Z., Jordan, S.R. <strong>and</strong><br />

phase variation of type 1 ¢mbriae in Escherichia coli K-12.<br />

Matthews, D.A. (1991) Crystal structure of the ribonuclease H<br />

Mol. Microbiol. 21, 725^738.<br />

domain of HIV-1 reverse transcriptase. Science 252, 88^95.<br />

[83] Blakely, G., May, G., McCulloch, R., Arciszewska, L.K.,<br />

[65] Beese, L.S. <strong>and</strong> Steitz, T.A. (1991) Structural basis for the 3P-<br />

Burke, M., Lovett, S.T. <strong>and</strong> Sherratt, D.J. (1993) Two related<br />

5P exonuclease activity of Escherichia coli DNA polymerase I :<br />

recombinases are required for <strong>site</strong>-speci¢c <strong>recombination</strong> at dif<br />

a two metal ion mechanism. EMBO J. 10, 25^33.<br />

<strong>and</strong> cer in E. coli K12. Cell 75, 351^361.<br />

[66] Craigie, R. (1996) Quality control in Mu DNA transposition.<br />

[84] Colloms, S.D., McCulloch, R., Grant, K., Neilson, L. <strong>and</strong><br />

Cell 85, 137^140.<br />

Sherratt, D.J. (1996) Xer-mediated <strong>site</strong>-speci¢c <strong>recombination</strong><br />

[67] Chaconas, G., Lavoie, B.D. <strong>and</strong> Watson, M.A. (1996) DNA<br />

in vitro. EMBO J. 15, 1172^1181.<br />

transposition : jumping gene machine, some assembly required.<br />

Curr. Biol. 6, 817^820.<br />

[85] Johnson, R.C. (1991) Mechanism of <strong>site</strong>-speci¢c DNA inversion<br />

in bacteria. Curr. Opin. Genet. Dev. 1, 404^411.<br />

[68] Yang, J.Y., Kim, K., Jayaram, M. <strong>and</strong> Harshey, R.M. (1995)<br />

A domain sharing model for active <strong>site</strong> assembly within the<br />

[86] Reed, R.R. <strong>and</strong> Moser, C.D. (1984) Resolvase-mediated <strong>recombination</strong><br />

intermediates contain a serine residue covalently<br />

linked to the DNA. Cold Spring Harbor Symp. Quant. Biol.<br />

Mu A tetramer during transposition : the enhancer may specify<br />

domain contributions. EMBO J. 14, 2374^2384.<br />

49, 245^249.<br />

[69] Wu, Z. <strong>and</strong> Chaconas, G. (1995) A novel DNA binding <strong>and</strong><br />

[87] Klippel, A., Mertens, G., Patschinsky, T. <strong>and</strong> Kahmann, R.<br />

nuclease activity in domain III of Mu transposase : evidence<br />

(1988) The DNA invertase Gin of phage Mu : formation of a<br />

for a catalytic region involved in donor cleavage. EMBO J. 14,<br />

covalent complex with DNA via a phosphoserine at amino<br />

3835^3843.<br />

acid position 9. EMBO J. 7, 1229^1237.<br />

[88] Boocock, M.R., Zhu, X. <strong>and</strong> Grindley, N.D. (1995) Catalytic<br />

[70] Aldaz, H., Schuster, E. <strong>and</strong> Baker, T.A. (1996) The interwoven<br />

architecture of the Mu transposase couples DNA synapsis<br />

residues of gamma delta resolvase act in cis. EMBO J. 14,<br />

to catalysis. Cell 85, 257^269.<br />

5129^5140.<br />

[71] Yang, J.Y., Jayaram, M. <strong>and</strong> Harshey, R.M. (1996) Positional<br />

[89] McIlwraith, M.J., Boocock, M.R. <strong>and</strong> Stark, W.M. (1997)


B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178 177<br />

of a knotting reaction catalysed by Tn3 resolvase. J. Mol.<br />

Tn3 resolvase catalyses multiple <strong>recombination</strong> events without<br />

intermediate rejoining of DNA ends. J. Mol. Biol. 266, 108^<br />

Biol. 239, 25^36.<br />

121.<br />

[106] Sadowski, P.D. (1995) The Flp recombinase of the 2-microns<br />

plasmid of Saccharomyces cerevisiae. Prog. Nucleic Acid Res.<br />

Mol. Biol. 51, 53^91.<br />

[90] Kanaar, R., van de Putte, P. <strong>and</strong> Cozzarelli, N.R. (1988) Ginmediated<br />

DNA inversion : product structure <strong>and</strong> the mechanism<br />

of str<strong>and</strong> exchange. Proc. Natl. Acad. Sci. USA 85, 752^<br />

[107] Burgin, A., Jr. <strong>and</strong> Nash, H.A. (1995) Suicide substrates<br />

756.<br />

reveal properties of the homology-dependent steps during<br />

integrative <strong>recombination</strong> of bacteriophage lambda. Curr.<br />

[91] Stark, W.M., Sherratt, D.J. <strong>and</strong> Boocock, M.R. (1989) Sitespeci¢c<br />

<strong>recombination</strong> by Tn3 resolvase : topological changes<br />

Biol. 5, 1312^1321.<br />

in the forward <strong>and</strong> reverse reactions. Cell 58, 779^790.<br />

[92] Klippel, A., Kanaar, R., Kahmann, R. <strong>and</strong> Cozzarelli, N.R.<br />

[108] Nunes-Duby, S.E., Azaro, M.A. <strong>and</strong> L<strong>and</strong>y, A. (1995) Swapping<br />

DNA str<strong>and</strong>s <strong>and</strong> sensing homology without branch<br />

migration in lambda <strong>site</strong>-speci¢c <strong>recombination</strong>. Curr. Biol.<br />

(1993) Analysis of str<strong>and</strong> exchange <strong>and</strong> DNA binding of enhancer-independent<br />

Gin recombinase mutants. EMBO J. 12,<br />

5, 139^148.<br />

1047^1057.<br />

[93] Moskowitz, I.P., Heichman, K.A. <strong>and</strong> Johnson, R.C. (1991)<br />

[109] Dixon, J.E. <strong>and</strong> Sadowski, P.D. (1994) Resolution of immobile<br />

chi structures by the FLP recombinase of 2 microns<br />

Alignment of <strong>recombination</strong> <strong>site</strong>s in Hin-mediated <strong>site</strong>-speci¢c<br />

plasmid. J. Mol. Biol. 243, 199^207.<br />

DNA <strong>recombination</strong>. Genes Dev. 5, 1635^1645.<br />

[110] Arciszewska, L., Grainge, I. <strong>and</strong> Sherratt, D. (1995) E¡ects<br />

[94] Stark, W.M., Parker, C.N., Halford, S.E. <strong>and</strong> Boocock, M.R.<br />

(1994) Stereoselectivity of DNA catenane fusion by resolvase.<br />

of Holliday junction position on Xer-mediated <strong>recombination</strong><br />

in vitro. EMBO J. 14, 2651^2660.<br />

Nature 368, 76^78.<br />

[111] Azaro, M.A. <strong>and</strong> L<strong>and</strong>y, A. (1997) The isomeric preference<br />

[95] Yang, W. <strong>and</strong> Steitz, T.A. (1995) Crystal structure of the <strong>site</strong>speci¢c<br />

recombinase gamma delta resolvase complexed with a<br />

of holliday junctions in£uence resolution bias by lambda integrase<br />

<strong>and</strong> also shows a predictable dependence on the<br />

34 bp cleavage <strong>site</strong>. Cell 82, 193^207.<br />

branch point sequence. EMBO J. 16, 3744^3755.<br />

[96] Feng, J.A., Johnson, R.C. <strong>and</strong> Dickerson, R.E. (1994) Hin<br />

[112] Arciszewska, L., Grainge, I. <strong>and</strong> Sherratt, D. (1997) Action<br />

recombinase bound to DNA : the origin of speci¢city in major<br />

of <strong>site</strong>-speci¢c recombinases XerC <strong>and</strong> XerD on tethered<br />

<strong>and</strong> minor groove interactions. Science 263, 348^355.<br />

holliday junctions. EMBO J. 16, 3731^3743.<br />

[97] Hughes, R.E., Rice, P.A., Steitz, T.A. <strong>and</strong> Grindley, N.D.<br />

[113] McCulloch, R., Coggins, L.W., Colloms, S.D. <strong>and</strong> Sherratt,<br />

D.J. (1994) Xer-mediated <strong>site</strong>-speci¢c <strong>recombination</strong> at cer<br />

(1993) Protein-protein interactions directing resolvase <strong>site</strong>-speci¢c<br />

<strong>recombination</strong> : a structure-function analysis. EMBO J.<br />

generates Holliday junctions in vivo. EMBO J. 13, 1844^<br />

12, 1447^1458.<br />

1855.<br />

[98] Rice, P.A. <strong>and</strong> Steitz, T.A. (1994) Re¢nement of gamma delta<br />

[114] Chen, J.W., Lee, J. <strong>and</strong> Jayaram, M. (1992) DNA cleavage in<br />

resolvase reveals a strikingly £exible molecule. Structure 2,<br />

trans by the active <strong>site</strong> tyrosine during Flp <strong>recombination</strong> :<br />

371^384.<br />

switching protein partners before exchanging str<strong>and</strong>s. Cell<br />

[99] Rice, P.A. <strong>and</strong> Steitz, T.A. (1994) Model for a DNA-mediated<br />

69, 647^658.<br />

synaptic complex suggested by crystal packing of gamma delta<br />

[115] Lee, J., Whang, I., Lee, J. <strong>and</strong> Jayaram, M. (1994) Directed<br />

resolvase subunits. EMBO J. 13, 1514^1524.<br />

[100] Ha¡ter, P. <strong>and</strong> Bickle, T.A. (1988) Enhancer-independent<br />

protein replacement in <strong>recombination</strong> full <strong>site</strong>s reveals transhorizontal<br />

DNA cleavage by Flp recombinase. EMBO J. 13,<br />

mutants of the Cin recombinase have a relaxed topological<br />

5346^5354.<br />

speci¢city. EMBO J. 7, 3991^3996.<br />

[116] Voziyanov, Y., Lee, J., Whang, I., Lee, J. <strong>and</strong> Jayaram, M.<br />

(1996) Analyses of the ¢rst chemical step in Flp <strong>site</strong>-speci¢c<br />

[101] Klippel, A., Cloppenborg, K. <strong>and</strong> Kahmann, R. (1988) Isolation<br />

<strong>and</strong> characterization of unusual Gin mutants. EMBO<br />

<strong>recombination</strong> : Synapsis may not be a pre-requi<strong>site</strong> for<br />

J. 7, 3983^3989.<br />

str<strong>and</strong> cleavage. J. Mol. Biol. 256, 720^735.<br />

[102] Crisona, N.J., Kanaar, R., Gonzalez, T.N., Zechiedrich,<br />

[117] Qian, X.H. <strong>and</strong> Cox, M.M. (1995) Asymmetry in active complexes<br />

of FLP recombinase. Genes Dev. 9, 2053^2064.<br />

E.L., Klippel, A. <strong>and</strong> Cozzarelli, N.R. (1994) Processive <strong>recombination</strong><br />

by wild-type Gin <strong>and</strong> an enhancer-independent<br />

[118] Blakely, G.W., Davidson, A.O. <strong>and</strong> Sherratt, D.J. (1997)<br />

Binding <strong>and</strong> cleavage of nicked substrates by <strong>site</strong>-speci¢c<br />

mutant. Insight into the mechanisms of <strong>recombination</strong> selectivity<br />

<strong>and</strong> str<strong>and</strong> exchange. J. Mol. Biol. 243, 437^457.<br />

recombinases XerC <strong>and</strong> XerD. J. Mol. Biol. 265, 30^39.<br />

[103] Haykinson, M.J., Johnson, L.M., Soong, J. <strong>and</strong> Johnson,<br />

[119] Han, Y.W., Gumport, R.I. <strong>and</strong> Gardner, J.F. (1993) Complementation<br />

of bacteriophage lambda integrase mutants :<br />

R.C. (1996) The Hin dimer interface is critical for Fis-mediated<br />

activation of the catalytic steps of <strong>site</strong>-speci¢c DNA<br />

evidence for an intersubunit active <strong>site</strong>. EMBO J. 12, 4577^<br />

inversion. Curr. Biol. 6, 163^177.<br />

4584.<br />

[104] McIlwraith, M.J., Boocock, M.R. <strong>and</strong> Stark, W.M. (1996)<br />

[120] Nunes-Duby, S.E., Tirumalai, R.S., Dorgai, L., Yagil, E.,<br />

Weisberg, R.A. <strong>and</strong> L<strong>and</strong>y, A. (1994) Lambda integrase<br />

Site-speci¢c <strong>recombination</strong> by Tn3 resolvase, photocrosslinked<br />

to its supercoiled DNA substrate. J. Mol. Biol. 260,<br />

cleaves DNA in cis. EMBO J. 13, 4421^4430.<br />

299^303.<br />

[121] Stark, W.M. <strong>and</strong> Boocock, M.R. (1995) Gatecrashers at the<br />

[105] Stark, W.M. <strong>and</strong> Boocock, M.R. (1994) The linkage change<br />

catalytic party. Trends Genet. 11, 121^123.


178<br />

catalytic party. Trends Genet. 11, 432^433.<br />

[123] Jayaram, M. (1997) The Cis-trans paradox of integrase. Science<br />

276, 49^50.<br />

[124] Kwon, H.J., Tirumalai, R., L<strong>and</strong>y, L. <strong>and</strong> Ellenberger, T.<br />

(1997) Flexibility in DNA <strong>recombination</strong> : structure of the<br />

lambda integrase catalytic core. Science 276, 126^131.<br />

[125] Hickman, A.B., Waninger, S., Scocca, J.J. <strong>and</strong> Dyda, F.<br />

(1997) Molecular organization in <strong>site</strong>-speci¢c <strong>recombination</strong> :<br />

the catalytic domain of bacteriophage HP1 integrase at 2.7 A<br />

î<br />

resolution. Cell 89, 205^214.<br />

[126] Subramanya, H.S., Arciszewska, L.K., Baker, R.A., Bird,<br />

L.E., Sherratt, D.J. <strong>and</strong> Wigley, D.B. Crystal structure of<br />

the <strong>site</strong>-speci¢c recombinase, XerD. Submitted.<br />

[127] Nash, H.A. (1990) Bending <strong>and</strong> supercoiling of DNA at the<br />

attachment <strong>site</strong> of bacteriophage lambda. Trends Biochem.<br />

Sci. 15, 222^227.<br />

[128] Richet, E., Abcarian, P. <strong>and</strong> Nash, H.A. (1988) Synapsis of<br />

attachment <strong>site</strong>s during lambda integrative <strong>recombination</strong><br />

involves capture of a naked DNA by a protein-DNA complex.<br />

Cell 52, 9^17.<br />

[129] Kim, S. <strong>and</strong> L<strong>and</strong>y, A. (1992) Lambda Int protein bridges<br />

between higher order complexes at two distant chromosomal<br />

loci attL <strong>and</strong> attR. Science 256, 198^203.<br />

[130] Franz, B. <strong>and</strong> L<strong>and</strong>y, A. (1995) The Holliday junction intermediates<br />

of lambda integrative <strong>and</strong> excisive <strong>recombination</strong><br />

respond di¡erently to the bending proteins integration host<br />

factor <strong>and</strong> excisionase. EMBO J. 14, 397^406.<br />

[131] Lim, H.M. (1994) Analysis of subunit interaction by introducing<br />

disul¢de bonds at the dimerization domain of Hin<br />

recombinase. J. Biol. Chem. 269, 31134^31142.<br />

[132] Benjamin, K.R., Abola, A.P., Kanaar, R. <strong>and</strong> Cozzarelli,<br />

N.R. (1996) Contributions of supercoiling to Tn3 resolvase<br />

<strong>and</strong> phage Mu Gin <strong>site</strong>-speci¢c <strong>recombination</strong>. J. Mol. Biol.<br />

256, 50^65.<br />

[133] Colloms, S.D., Bath, J. <strong>and</strong> Sherratt, D.J. (1997) Topological<br />

selectivity in Xer <strong>site</strong>-speci¢c <strong>recombination</strong>. Cell 88, 855^<br />

864.<br />

[134] Watson, M.A. <strong>and</strong> Chaconas, G. (1996) Three-<strong>site</strong> synapsis<br />

during Mu DNA transposition : a critical intermediate preceding<br />

engagement of the active <strong>site</strong>. Cell 85, 435^445.<br />

[135] Craigie, R. <strong>and</strong> Mizuuchi, K. (1986) Role of DNA topology<br />

in Mu transposition : mechanism of sensing the relative orientation<br />

of two DNA segments. Cell 45, 793^800.<br />

[136] Wang, Z. <strong>and</strong> Harshey, R.M. (1994) Crucial role for DNA<br />

supercoiling in Mu transposition : a kinetic study. Proc. Natl.<br />

Acad. Sci. USA 91, 699^703.<br />

[137] Mizuuchi, M., Baker, T.A. <strong>and</strong> Mizuuchi, K. (1995) Assembly<br />

of phage Mu transpososomes : cooperative transitions<br />

assisted by protein <strong>and</strong> DNA sca¡olds. Cell 83, 375^385.<br />

[138] Yang, J.Y., Jayaram, M. <strong>and</strong> Harshey, R.M. (1995) Enhancer-independent<br />

variants of phage Mu transposase : enhancer-speci¢c<br />

stimulation of catalytic activity by a partner<br />

transposase. Genes Dev. 9, 2545^2555.<br />

[139] Bainton, R.J., Kubo, K.M., Feng, J.N. <strong>and</strong> Craig, N.L.<br />

(1993) Tn7 transposition : target DNA recognition is mediated<br />

by multiple Tn7-encoded proteins in a puri¢ed in vitro<br />

system. Cell 72, 931^943.<br />

[140] Chalmers, R.M. <strong>and</strong> Kleckner, N. (1996) IS10/Tn10 transposition<br />

e¤ciently accommodates diverse transposon end<br />

con¢gurations. EMBO J. 15, 5112^5122.<br />

[141] Signon, L. <strong>and</strong> Kleckner, N. (1995) Negative <strong>and</strong> positive<br />

regulation of Tn10/IS10-promoted <strong>recombination</strong> by IHF :<br />

two distinguishable processes inhibit transposition o¡ of<br />

multicopy plasmid replicons <strong>and</strong> activate chromosomal<br />

events that favor evolution of new transposons. Genes<br />

Dev. 9, 1123^1136.<br />

[142] van Gent, D.C., Mizuuchi, K. <strong>and</strong> Gellert, M. (1996) Similarities<br />

between initiation of V(D)J <strong>recombination</strong> <strong>and</strong> retroviral<br />

integration. Science 271, 1592^1594.<br />

B. Hallet, D.J. Sherratt / FEMS Microbiology Reviews 21 (1997) 157^178<br />

[122] Jayaram, M. <strong>and</strong> Lee, J. (1995) Return to sobriety after the

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!