Identification of genes required for excision of CTnDOT, a Bacteroides conjugative transposon:...

Identification of genes required for excision of CTnDOT,a Bacteroides conjugative transposon

Qi Cheng, Yuri Sutanto, Nadja B. Shoemaker,

Jeffrey F. Gardner and Abigail A. Salyers*

Department of Microbiology, 601 S. Goodwin Ave.,

University of Illinois, Urbana, IL 61801, USA.

Summary

Integrated self-transmissible elements called conju-

gative transposons have been found in many different

bacteria, but little is known about how they excise

from the chromosome to form the circular intermedi-

ate, which is then transferred by conjugation. We have

now identified a gene, exc, which is required for the

excision of the Bacteroides conjugative transposon,

CTnDOT. The int gene of CTnDOT is a member of the

lambda integrase family of recombinases, a family

that also contains the integrase of the Gram-positive

conjugative transposon Tn916. The exc gene was

located 15 kbp from the int gene, which is located at

one end of the 65 kbp element. The exc gene, together

with the regulatory genes, rteA, rteB and rteC, were

necessary to excise a miniature form of CTnDOT that

contained only the ends of the element and the int

gene. Another open reading frame (ORF) in the same

operon and upstream of exc, orf3, was not essential

for excision and had no significant amino acid

sequence similarity to any proteins in the databases.

The deduced amino acid sequence of the CTnDOT Exc

protein has significant similarity to topoisomerases. A

small ORF (orf2 ) that could encode a small, basic

protein comparable with lambda and Tn916 excision

proteins (Xis) was located immediately downstream of

the CTnDOT int gene. Although Xis proteins are

required for excision of lambda and Tn916, orf2 had

no effect on excision of the element. Excision of the

CTnDOT mini-element was not affected by the site in

which it was integrated, another difference from

Tn916. Our results demonstrate that the Bacteroides

CTnDOT excision system is tightly regulated and

appears to be different from that of any other known

integrated transmissible element, including those of

some Bacteroides mobilizable transposons that are

mobilized by CTnDOT.

Introduction

In Bacteroides species, antibiotic resistance genes are

being spread by conjugative transposons (CTns) and

mobilizable transposons (MTns) as well as by plasmids

(Salyers and Shoemaker, 1995; Salyers et al., 1995a;

Smith et al., 1998; Tribble et al., 1999). The transfer of one

type of Bacteroides CTn, CTnDOT and related CTns, has

been so efficient that, today, over 80% of natural isolates

from different Bacteroides species carry one of these

CTns (Shoemaker et al., 2001). CTns are integrated

elements that excise from the chromosome of a donor cell

to form a circular intermediate that transfers to a recipient

cell by conjugation. Previously, we showed that an

integrase gene located at one end of CTnDOT was able

to mediate integration of the joined ends of the CTnDOT

circular form into the Bacteroides thetaiotaomicron

chromosome but, once integrated, this construct was

unable to excise itself from the chromosome (Cheng et al.,

2000).

The deduced amino acid sequence of the CTnDOT

integrase gene (int ) has amino acid similarity to members

of the lambda integrase family (Cheng et al., 2000).

Excision of the integrated form of phage lambda requires

not only the integrase but also a small, basic protein

called Xis (Landy, 1989). Xis bends DNA and promotes

co-operative binding of Int, so that a higher order

nucleoprotein complex can form and catalyse the excision

reaction (Thompson and Landy, 1988; Numrych et al.,

1992; Wu et al., 1998). Similarly, excision of the Gram-

positive CTn, Tn916, which also has an integrase that is a

member of the lambda integrase family, requires both an

integrase and a xis-like gene (Rudy et al., 1997a,b). A

small open reading frame (ORF), orf2, was located next to

the int gene on CTnDOT, which could encode a small

basic protein such as Xis, but int and orf2 were not

sufficient for excision (Cheng et al., 2000). Using a mini-

element excision system described here, we identified

and characterized a gene (exc ) that, together with int,

mediates the excision of CTnDOT.

Transfer of CTnDOT is regulated by tetracycline.

Exposure of cells to tetracycline results in a 1000- to

10 000-fold increase in the transfer frequency of the CTn.

Tetracycline regulation of CTnDOT functions is mediated

by three regulatory genes, rteA, rteB and rteC (Stevens

et al., 1993; Salyers et al., 1995b). The rteA and rteB

genes are part of an operon that also contains the geneAccepted 4 May, 2001. *For correspondence. E-mail [email protected]; Tel. (11) 217 333 7378; Fax (11) 217 244 8485.

Molecular Microbiology (2001) 41(3), 625–632

Q 2001 Blackwell Science Ltd

tetQ, which encodes a ribosome protection-type tetra-

cycline resistance protein (Nikolich et al., 1992). Exposure

of cells to tetracycline stimulates expression of the tetQ–

rteA–rteB operon. RteA and RteB, in turn, activate the

expression of the rteC gene (Stevens et al., 1993). Thus,

RteA, RteB and RteC proteins are essential for transfer of

CTnDOT. We now show that excision is dependent on

tetracycline stimulation and that all three rte genes are

required for excision.

Results

Identification of genes involved in excision of CTnDOT

We showed previously that an integrated mini-element

that only carried the CTnDOT int and a small downstream

ORF, orf2, did not excise in Bacteroides thetaiotaomicron

4001 (Cheng et al., 2000). Thus, there must be at least one

other CTn gene involved in excision. Downstream of orf2

on CTnDOT is a 13 kbp DNA segment that contained an

ermF gene Fig. 1 (Shoemaker et al., 1989; Whittle et al.,

2001). As this 13 kbp segment was missing from a closely

related CTn, CTnERL, it seemed unlikely that any excision

genes would be located in the 13 kbp ermF region of

CTnDOT. A 7.6 kbp region downstream of the 13 kbp

insertion, which was present in both CTnDOT and

CTnERL, was cloned and sequenced.

Examination of the sequence of the 7.6 kbp cloned

region revealed two large ORFs, orf3 and orf4, that were

transcribed in the same direction (Fig. 1). Single cross-

over gene disruptions were made in each of these two

ORFs in both CTnDOT and CTnERL. Both insertions

completely eliminated transfer of the CTns; the transfer

frequencies were reduced from 1025 to 1026 per recipient

to less than 1028 per recipient. These disruptions also

eliminated excision, as determined by polymerase chain

reaction (PCR) amplification of the joined ends of the

circular intermediate (Fig. 2). Complementation of each

insertion with a plasmid carrying both ORFs (pKSO1)

restored excision for CTnERL as well as for CTnDOT. The

DNA cloned in pKSO1 came from CTnDOT. Thus, as

expected from the high sequence identity of the two

CTns outside the 13 kbp ermF region (Whittle et al.,

2001), genes from CTnDOT complemented functions of

CTnERL.

A disruption in rteC also abolished excision (Fig. 2). As

excision is a tetracycline-regulated step (Fig. 2), it seemed

reasonable that one or more of the rte genes would be

required. Normally, expression of rteC is controlled by

RteA and RteB, which in turn are produced only in

tetracycline-induced cells (Salyers et al., 1995b). Pre-

viously, we constructed a plasmid, pLYL52, which carries

a copy of rteC that is expressed constitutively (Li et al.,

1995). When pLYL52 was introduced into the strain that

had a disruption in rteC, the clone restored excision

(Fig. 2). Moreover, excision was now constitutive. Thus,

rteC controls excision. If RteA and RteB had been directly

involved in excision, except insofar as they are needed to

activate the wild-type rteC gene, excision would still have

been seen only when cells were exposed to tetracycline,

because RteA and RteB are only produced under these

conditions.

Minimal excision system for CTnDOT

To determine whether orf3 and orf4 could mediate

excision, the plasmid containing them (pKSO1) was

introduced into a strain of BT4001 that carried a copy

of the integrated CTnDOT mini-element containing int

(integrated pDJE2.3). No excision was detected. Thus,

orf3 and orf4 alone were not sufficient for excision of the

mini-element. One explanation for this was that orf3 and

orf4 were not expressed in the absence of rteC.

Accordingly, a compatible plasmid containing rteA, rteB

and rteC (pAMS9) was introduced into the strain contain-

ing the integrated mini-element, VpDJE2.3, and pKSO1

(Fig. 3). The rteA and rteB genes were included to induce

the expression of rteC. In the resulting strain, excision of

the mini-element could be detected by both PCR as in

Fig. 1. The functional regions on CTnDOT andCTnERL are shown at the top. The position ofthe 13 kbp ermF region, which is inserted at theend of orf2 on CTnDOT (Whittle et al., 2001), ismissing on CTnERL. The point of insertion ofthe ermF region is indicated. The locations ofthe 10 bp consensus regions on thechromosome and at one end of the integratedCTn are indicated by small horizontalarrowheads. The lower part contains anenlargement of the excision–integration regionof CTnERL/CTnDOT with the insertion site ofthe 13 kbp ermF-containing region of CTnDOTindicated. The 7.6 kbp region from CTnDOT thatis cloned on pKS01 is indicated.

626 Q. Cheng et al.

Q 2001 Blackwell Science Ltd, Molecular Microbiology, 41, 625–632

Fig. 2 and Southern blot analysis. pKSO1 had to be

present because no excision was detectable if only

pAMS9 and VpDJE2.3 or the larger element VpDJE2.6

were present. Thus, excision requires the central

regulatory genes, rteA, rteB and rteC, that are carried on

pAMS9 as well as the genes that are carried on pKSO1.

Excision of the mini-element was regulated by tetra-

cycline, as was excision of the wild-type CTnDOT. In the

case of the mini-element, however, even in the absence of

tetracycline, there was some detectable excision (Fig. 4),

whereas in the case of CTnDOT, no excision was detected

in the absence of tetracycline even by PCR (Fig. 2). The

low level of tetracycline-independent excision of the mini-

element could result from the fact that rteA–rteB and rteC

were present in multiple copies, and any basal expression

of these genes was thus magnified.

The orf2 gene has no role in excision of the mini-element

The orf2 gene is not essential for excision, because the

mini-element that only contained int, pDJE2.3, was

capable of excising if pKSO1 and pAMS9 were present.

Still, we could not rule out the possibility that the presence

of orf2 might stimulate excision. To test this possibility,

excision of two different versions of the mini-element

inserted at the same site were compared. One mini-

element contained int but not orf2 (VpDJE2.3). The

second contained both int and orf2 (VpDJE2.6). Strains

containing VpDJE2.3 or VpDJE2.6 plus the plasmids

pKSO1 and pAMS9 were constructed and tested for

excision. The results, shown in Fig. 4, demonstrated that,

although there was a slight experiment to experiment

variation in the amount of circular form seen in the same

strain, there was no significant increase in the amount of

excised circular form when orf2 was present. Thus, if orf2

has any role in excision, it was not evident in this particular

strain background.

The exc (orf4) gene is required for excision of the mini-

element

To determine whether both orf3 and orf4 were required for

excision, or whether only one was sufficient, a subclone of

pKS01, which contained orf3 but only part of orf4 (pKSO4;

Fig. 5A), was introduced into a strain containing VpDJE2.3

and pAMS9. No excision was observed (Fig. 5B). A

plasmid that had an internal deletion in orf3 (pKSO7;

Fig. 5A) did support excision when introduced into the

strain containing VpDJE2.3 and pAMS9 (Fig. 5B). pKS07

had 2.1 kbp of DNA upstream of the 50 end of orf3. A similar

construct (pKSO6), which had only 450 bp of upstream

DNA, did not support excision. A plasmid that contained all

of orf4 but only a portion of the 30 end of orf3 (pKSO5) did

not support excision. This result supports the hypothesis

that orf3 and orf4 are in an operon, with the promoter or

activator site . 300 bp upstream of orf3 in the region

deleted in pKS06. The fact that an insertion in orf3

eliminated excision was probably the result of a polar

effect of the insertion on orf4. This possibility was

confirmed by the observation that, when pKSO7 was

introduced into the Bacteroides strain that contained

CTnERL with a disruption in orf3, excision was restored,

as indicated by the fact that the mutant element now

transferred once again at wild-type levels (1025 transcon-

jugants per recipient) under tetracycline-induced con-

ditions. As orf4 was required for excision, we renamed it

exc.

Fig. 2. PCR assay for excision of wild type and insertional derivativesof CTnERL. BT4104 with CTnERL or with CTnERL containinginsertions in rteC (VrteC ), orf3 (Vorf3 ) or orf4 (Vorf4 ) were tested forexcision of CTnERL using PCR to detect the excised circularintermediate. The cells were grown with (1) or without (–) tetracycline(Tc) in the medium. The plasmids used to complement each mutationare indicated. pLYL52 contains rteC behind a constitutive promoter(Li et al., 1993). The region cloned on pKS01 is shown in Fig. 1.

Fig. 3. Components of the mini-element system. The suicide plasmid,pDJE2.3 (Cheng et al., 2000), which contains the ends of CTnDOTand the int gene, is integrated into the chromosome. pAMS9 (Stevenset al., 1992) contains tetQ, rteA, rteB and rteC. pKSO1 carries the7.6 kbp DNA segment depicted in Fig. 1. Growing cells in mediumcontaining tetracycline (Tc) causes the mini-element to excise andcircularize. Excision is monitored by PCR or by Southern blot to detectthe joined ends.

Excision system for CTnDOT 627


The site of integration does not affect excision

CTnDOT has at least seven integration sites in

B. thetaiotaomicron 4001 (Bedzyk et al., 1992; Cheng

et al., 2000). The chromosomal DNA sequences of these

sites are different from each other except for a conserved

10 bp sequence (Cheng et al., 2000). When CTnDOT

integrates, it integrates 5 bp 30 of this conserved sequence.

Thus, the chromosomal sequences flanking at least one

end of an integrated CTn can differ markedly from one

site to another. Transfer frequencies of Tn916, which

integrates nearly randomly, differ considerably depending

on the site in which Tn916 is located (Jaworski and

Clewell, 1994). These differences in transfer frequency are

presumed to result from differences in excision frequency.

In contrast, we did not observe any significant differences

in transfer frequency between copies of CTnDOT

integrated in different sites. However, this observation

does not rule out differences in excision frequency from

one site to another, because the transfer assay will only

detect differences that are . 10-fold. Also, if excision were

not the rate-limiting step in transfer, differences in excision

would be missed.

To determine whether the site of integration affected the

amount of circular intermediate formed, excision of

VpDJE2.3 from four different sites was compared.

Southern blot analysis showed that the strength of the

joined end signal was similar for all four inserted mini-

elements (Fig. 6). Thus, if there are differences in excision

levels associated with different sites, they are not very

great.

The exc gene product is related to topoisomerases

The deduced amino acid sequences of orf3 and exc were

compared with those of proteins in the protein databases.

The Orf3 protein was not significantly related to any of the

proteins in the databases. Exc protein was related most

closely to proteins encoded on Gram-positive plasmids,

the Gram-positive conjugative transposon Tn1549

(Garnier et al., 2000) and to known chromosomally

encoded topoisomerases. Although Exc was only 30–

32% identical to these topoisomerases, it had a similar

Fig. 4. Determination of the importance of orf2for excision. Two mini-element systems wereconstructed in B. thetaiotaomicron: VpDJE2.3(orf2–) and VpDJE2.6 (orf21). The two mini-elements were integrated into the chromosomeat the same site. Three separate cultures ofeach system were grown with (1) or without (–)tetracycline (Tc) and tested for excision bySouthern blot. Triplicate experiments wereperformed to assess the amount ofexperimental variation. The 1.1 kbp DJE probeused detects the integrated ends (attR, attL ) aswell as the joined ends of the excision product.The excision product seen with the orf21 strainis a little larger than the product seen with theorf2– strain, because the HindIII restrictionenzyme used to digest the DNA cut outside thecloned CTnDOT segment (in the vector andflanking the inserted element in thechromosome). pDJE2.6 contains 0.3 kbp morecloned DNA than pDJE2.3 (Cheng et al., 2000).

Fig. 5. A. Clones tested in the mini-element system to determinewhether orf3 and/or orf4 are essential for excision. The brackets withdashes between them indicate the location and extent of the 1308 bpin frame deletion in orf3, which leaves 216 bp of the 50 end fused to the45 bp 30 end. The excision results are summarized on the right.B. Southern blot analysis of the strains carrying vectors described in(A). The Bacteroides strains contained VpDJE2.3, pAMS9 and theindicated vector. The strains were grown with (1) and without (–)tetracycline (Tc) induction. The DNA from each culture was digestedwith HindIII, and the Southern blot was probed with the 1.1 kbp PCRproduct containing the CTnDOT joined ends (DJE) and part of intDOT.The positions of the attR, attL and pDJE2.3 excision products areindicated on the right, and the locations of the lambda standards areindicated on the left.

628 Q. Cheng et al.


length, and the regions in which identity was seen

were distributed throughout the entire protein sequence.

Moreover, Exc had high sequence identity to the

topoisomerases within a region known to be essential for

topoisomerase activity, and it had the tyrosine residue

that is conserved in all enzymes of this type (Fig. 7). Exc

had the putative decatenation loop that distinguishes

topoisomerase I from topoisomerase III (Li et al., 2000).

Discussion

The results of a previous study suggested that the

Bacteroides CTns integrate and excise similarly to the

Gram-positive CTn, Tn916 (Scott and Churchward, 1995).

That is, staggered cuts are made in the DNA usually

4–5 bp from the ends of the integrated form of the CTn.

These 4–5 bp chromosomal coupling sequences become

part of the circular form of the element (Bedzyk et al.,

1992; Cheng et al., 2000). Despite the apparent similarity

of the excision mechanisms of CTnDOT and Tn916,

the way this excision is accomplished appears to be

completely different. First, efficient excision of CTnDOT

occurred independently of the site in which the element

had inserted. The transfer efficiency of Tn916 differed

from one site to another by orders of magnitude (Jaworski

and Clewell, 1994). In the case of CTnDOT, there is a

conserved 10 bp sequence (GTA NA/T TTTGC) found at

one end of CTnDOT with high identity to a 10 bp sequence

in the target site (Cheng et al., 2000). As the element

integrates 5 bp away from the 30 end of the 10 bp sequence

in the chromosomal att site, a duplication of this 10 bp site

occurs, with one copy in the chromosome 5 bp from one

end of the element and one copy within the other end of the

element. If the excision complex bound to the 10 bp site,

there would be little or no effect of the non-conserved

chromosomal DNA around the integration site.

A second difference between CTnDOT and Tn916 is

that Tn916 excision requires int and a small xis-like gene,

whereas CTnDOT requires int and exc, a gene whose

predicted amino acid sequence indicates that it might

encode a topoisomerase. We found previously that

disruption of the int gene in the wild-type CTn abolished

excision (Cheng et al., 2000), so int is required for excision

as well as for integration. The orf2 gene that would encode

a small basic protein such as Xis of lambda and Tn916 was

not required for and did not stimulate excision. If exc does

indeed encode a topoisomerase, the excision system of

CTnDOT is completely different from all known excision

systems.

Topoisomerase genes have been found on plasmids,

conjugative transposons and bacteriophages as well as in

the chromosome. Little is known about the function of the

plasmid or conjugative transposon-encoded topoisome-

rases. The topoisomerase encoded on the Gram-positive

plasmid pAMb1 appears to play a role in the early steps of

Fig. 7. Alignment of Exc (Orf4) withtopoisomerases in the regions where thetopoisomerase sequences are highlyconserved. The sequences were aligned usingthe CLUSTALW multiple sequence alignmentprogram (Thompson et al., 1994). Bold lettersindicate amino acids that are identical to thosein the CTnDOT sequence. The boxed Y is thecatalytic tyrosine. Accession numbers are:TopB of E. coli (P14294), Orf24 of Tn1549(AAF72351), pAMb1 from Enterococcusfaecalis (AAC38606), pGO1 fromStaphylococcus aureus (A56976), TraE ofRP4(AAA26423) and TopA of E. coli (P06612).

Fig. 6. Excision of VpDJE2.3 from four different sites in theB. thetaiotaomicron chromosome. Each of the B. thetaiotaomicronstrains contained VpDJE2.3 in one of four different insertion sites aswell as pAMS9 and pKS01. The strains were grown with (1) or without(–) tetracycline (Tc) added to the medium. The DNA from each culturewas digested with HindIII, and the Southern blot was probed with the1.1 kbp DJE fragment. The location of the 3.9 kbp excision product isindicated on the right, and positions of the lambda standards areindicated on the left. On this blot, which is somewhat overexposed, thelow background excision seen in the absence of tetracycline is visible.



plasmid replication, although it is not essential for

replication (Bidnenko et al., 1998). The topoisomerase of

phage T4 also plays a role in the initiation of phage DNA

replication (Kreutzer and Morrical, 1994). CTnDOT appears

to excise in a way that does not involve replication. That is,

the CTn excises to form a circle, restoring the site from which

it has excised. Conjugal transfer of the circular form involves

a type of replication, in the sense that the single strand that is

transferred and the one that is left behind are both copied to

make a double-stranded intermediate, which can integrate

into the chromosome. However, exc appears not to be

essential for transfer, because it is not in the region

previously found to be necessary and sufficient for transfer

of the circular intermediate (Li et al., 1995; Bonheyo et al.,

2000). If Exc is a topoisomerase, a possible role for it might

be to change the supercoiling of the excised circular form

in a way that prevents reintegration, thus moving the

reaction in the direction of excision. A second possibility is

that, although Exc is not required for integration (Cheng

et al., 2000), it participates directly in the excision reaction.

For example, Int and Exc could form heterodimers on one

or both of the att sites used for excision. There is a

precedent for this mechanism. The Escherichia coli xer

system uses two related proteins, XerC and XerD, that

bind to discrete sites during recombination (Blakely et al.,

1993).

Experimental procedures

Bacterial strains, plasmids and DNA manipulation

For construction of B. thetaiotaomicron strains, DNA was firstcloned in a plasmid in E. coli S17-1 (Simon et al., 1983), thenintroduced by conjugation into B. thetaiotaomicron 4001. Themedium used for all experiments was trypticase–yeast–glucose (TYG) medium (Holdeman and Moore, 1975).Methods for DNA extraction, cloning and conjugation fromE. coli to B. thetaiotaomicron have been described previously(Saito and Miura, 1963; Shoemaker et al., 1986; Sambrooket al., 1989). The various cloned DNA fragments came fromCTnDOT but, in some experiments, CTnERL was used tomake the construction of mutations in the CTn easier.According to high-stringency Southern blots, these twoelements have a high sequence identity in this region. Thesequences (< 600 bp) flanking the 13 kbp ermF region thathas been inserted between orf2 and orf3 of CTnDOT are over99% identical to CTnERL (Whittle et al., 2001).

Gene disruptions in orf3 and orf4 (exc) and

complementation experiments

Because of the high sequence identity, gene disruptions usingCTnDOT fragments were readily made in CTnERL. UsingCTnERL allowed us to use ermF, which is already present onCTnDOT, as a selectable marker. Internal portions of orf3 andorf4 were cloned into pCQW1 (Feldhaus et al., 1991), a plasmidthat replicates in E. coli but not in B. thetaiotaomicron. The

resulting plasmids were mobilized into B. thetaiotaomicronusing E. coli S17-1 as the donor and were shown by Southernblot to have recombined into the correct region. Excision wasdetected by PCR amplification of the joined ends. Forcomplementation assays, various portions of the orf3/orf4region were cloned into pLYL05 (Li et al., 1993), shuttle vectorpFD160 (Smith et al., 1995) carrying a cefoxitin resistance gene(Smith and Parker, 1993) that replicates in B. thetaiotaomicronas well as in E. coli. pKSO1 was constructed by cloning the7.6 kbp Ssp I fragment (also contained on CTnERL) that liesdownstream of the 13 kbp ermF region of CTnDOT into theSma I site of pLYL05. pKSO4 was derived from pKSO1 bythe Erase-a-base (Promega) method performed according tothe manufacturer’s directions. The resulting plasmids weretransferred from E. coli S17-1 to B. thetaiotaomicron strainsthat contained CTnERL with a disruption in orf3 or orf4.Excision was detected by PCR using primers and conditionsdescribed previously (Cheng et al., 2000).

Construction of a strain carrying the minimal excision

system

A suicide plasmid (pDJE2.3; Cheng et al., 2000) wasintroduced into BT4001, a strain that contained no CTn.pDJE2.3 carries a 2.3 kbp portion of CTnDOT that consistedof the joined ends and the CTnDOT integrase gene (int ). Insome experiments, a slightly larger plasmid (pDJE2.6) wasused that carries not only the joined ends and int gene ofCTnDOT, but also orf2. The integrated forms of theseplasmids are called VpDJE2.3 and VpDJE2.6 respectively.Previously, we obtained the sequences of some of the sitesinto which CTnDOT integrates (Cheng et al., 2000). Todetermine which insertions of pDJE2.3 and pDJE2.6 went intothe same site, a primer seated in one end of CTnDOTsequences on the plasmids and a primer seated in adjacentchromosomal DNA was used in a PCR amplification reactionto determine whether the junction region was present ornot. Each strain used in the excision experiments waschecked on a Southern blot, using a probe from the CTnDOTint region, to ascertain that only one copy of the mini-elementhad integrated. The Southern blot procedure and probepreparation was the same as that described previously(Shoemaker et al., 2000).

Plasmids carrying rteA, rteB and rteC (pAMS9 chloram-phenicol resistance; Stevens et al., 1992) and orf3 and orf4(pKSO1 cefoxitin resistance) are compatible in Bacteroideshosts and were introduced sequentially by conjugation into thestrain carrying VpDJE2.3 or VpDJE2.6. In some experiments,derivatives of pKSO1 were used instead of pKSO1. Excisionwas detected by either PCR or Southern blot assays (Chenget al., 2000). The clones containing the in frame deletionof orf3 were made by ligating PCR products containing anupstream sequence of either 2.1 kbp (for pKS06) or 450 bp(for pKS07) plus 216 bp of the 50 end of orf3 to a PCR productcontaining 45 bp of the 30 end of orf3 and all of orf4. The PCRproducts were assembled in pGEM-T (Promega) usingrestriction sites encoded in the primers. Each construct wassequenced and then subcloned into the shuttle vector pLYL05to form pKS06 and pKS07. The resulting constructs containeda 1308 bp deletion of the 1569 bp orf3.

630 Q. Cheng et al.


DNA sequencing

Sequencing was performed by the University of IllinoisBiotechnology Genetic Engineering facility with an AppliedBiosystems model 373A version 2.0.1S dye terminatorautomated sequencer. The sequences of orf3 and orf4 havebeen submitted to EMBL Nucleotide Sequence Database.The accession number for the mini-element containing theCTnDOT joined ends, int, and orf2 (pDJE2.6) is AJ311171and the number for orf3 and exc is AJ319661.

Acknowledgements

This work was supported by grants AI22383 (to A.A.S.) andGM28717 (to J.F.G.) from the National Institutes of Health.

References

Bedzyk, L.A., Shoemaker, N.B., Young, K.E., and Salyers,A.A. (1992) Insertion and excision of Bacteroides con-jugative chromosomal elements. J Bacteriol 174: 166–172.

Bidnenko, V., Ehrlich, S.D., and Janniere, L. (1998) In vivorelationships between pAMb1-encoded type 1 topoisome-rase and plasmid replication. Mol Microbiol 28: 1005–1016.

Blakely, G., May, G., McCulloch, R., Arciszewska, L.K.,Burke, M., Lovett, S.T., and Sharratt, D.J. (1993) Tworelated recombinases are required for site-specific recom-bination at dif and cer in E. coli 12. Cell 75: 351–361.

Bonheyo, G., Graham, D., Shoemaker, N.B., and Salyers,A.A. (2001) Transfer region of a Bacteroides conjugativetransposon, CTnDOT. Plasmid 45: 41–51.

Cheng, Q., Paszkiet, B.J., Shoemaker, N.B., Gardner, J.F.,and Salyers, A.A. (2000) Integration and excision of aBacteroides conjugative transposon, CTnDOT. J Bacteriol182: 4035–4043.

Feldhaus, M.J., Hwa, V., Cheng, Q., and Salyers, A.A. (1991)Use of an Escherichia coli beta-glucuronidase gene asa reporter gene for investigation of Bacteroides promoters.J Bacteriol 173: 4540–4543.

Garnier, F., Taourit, S., Glaser, P., Courvalin, P., andGalimand, M. (2000) Characterization of transposonTn1549 conferring VanB-type resistance in Enterococcusspp. Microbiology 146: 1481–1489.

Holdeman, L.V., and Moore, W.E.C. (1975) AnaerobeLaboratory Manual. Blacksburg, VA: Virginia PolytechnicalInstitute and State University.

Jaworski, D.D., and Clewell, D.B. (1994) Evidence thatcoupling sequences play a frequency-determining role inconjugative transposition of Tn916 in Enterococcusfaecalis. J Bacteriol 176: 3328–3335.

Kreutzer, K.N., and Morrical, S.W. (1994) Initiation of DNAreplication. In Molecular Biology of Bacteriophage T4.Karam, J.D. (ed.). Washington, DC: American Society forMicrobiology Press, pp. 28–42.

Landy, A. (1989) Dynamic, structural and regulatory aspectsof site-specific recombination. Annu Rev Biochem 58:913–950.

Li, L.Y., Shoemaker, N.B., and Salyers, A.A. (1993)Characterization of the mobilization region of a Bacteroidesinsertion element (NBU1) that is excised and transferred by

Bacteroides conjugative transposons. J Bacteriol 175:6588–6598.

Li, L.Y., Shoemaker, N.B., and Salyers, A.A. (1995) Location

and characteristics of the transfer region of a Bacteroidesconjugative transposon and regulation of transfer genes.

J Bacteriol 177: 4992–4999.

Li, Z., Mondragon, A., Hiasa, H., Marians, K.J., and DiGate,R.J. (2000) Identification of a unique domain essential for

Escherichia coli DNA topoisomerase III-catalysed decate-

nation of replication intermediates. Mol Microbiol 35:888–895.

Nikolich, M.P., Shoemaker, N.B., and Salyers, A.A. (1992) A

Bacteroides tetracycline resistance gene represents a new

class of ribosome protection tetracycline resistance.Antimicrob Agents Chemother 36: 1005–1012.

Numrych, T.E., Gumport, R.I., and Gardner, J.F. (1992)

Characterization of the bacteriophage lambda excisionase(Xis) protein: the C-terminus is required for Xis-integrase

cooperativity but not DNA binding. EMBO J 11:

3797–3806.Rudy, C., Taylor, K.L., Hinerfeld, D., Scott, J.R., and

Churchward, G. (1997a) Excision of a conjugative

transposon in vitro by the Int and Xis proteins of Tn916.Nucleic Acids Res 25: 4061–4066.

Rudy, C.K., Scott, J.R., and Churchward, G. (1997b) DNA

binding by the Xis protein of the conjugative transposonTn916. J Bacteriol 179: 2567–2572.

Saito, H., and Miura, K.I. (1963) Preparation of transforming

deoxy-ribonucleic acid by phenol treatment. BiochimBiophys Acta 72: 619–629.

Salyers, A.A., and Shoemaker, N.B. (1995) Conjugative

transposons: the force behind the spread of antibioticresistance genes among Bacteroides clinical isolates.

Anaerobe 1: 143–150.

Salyers, A.A., Shoemaker, N.B., and Li, L.Y. (1995a) In thedriver’s seat: the Bacteroides conjugative transposons and

the elements they mobilize. J Bacteriol 177: 5727–5731.

Salyers, A.A., Shoemaker, N.B., and Stevens, A.M. (1995b)Tetracycline regulation of conjugal transfer genes. In Two-

Component Signal Transduction. Hoch, J.A., and Silhavy,

T.J. (eds). Washington, DC: American Society forMicrobiology Press, pp. 393–400.

Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989) Molecular

Cloning: a Laboratory Manual. Cold Spring Harbor, NY:Cold Spring Harbor Laboratory Press.

Scott, J.R., and Churchward, G.G. (1995) Conjugative

transposition. Annu Rev Microbiol 49: 367–397.Shoemaker, N.B., Getty, C., Guthrie, E.P., and Salyers, A.A.

(1986) Regions in Bacteroides plasmids pBFTM10 and

pB8-51 that allow Escherichia coli-Bacteroides shuttlevectors to be mobilized by IncP plasmids and by a

conjugative Bacteroides tetracycline resistance element.

J Bacteriol 166: 959–965.

Shoemaker, N.B., Barber, R.D., and Salyers, A.A. (1989)Cloning and characterization of a Bacteroides conjugal

tetracycline–erythromycin resistance element by using a

shuttle cosmid vector. J Bacteriol 171: 1294–1302.Shoemaker, N.B., Wang, G.R., and Salyers, A.A. (2000)

Multiple gene products and sequences required for excision

of the mobilizable integrated Bacteroides element NBU1.J Bacteriol 182: 928–936.



Shoemaker, N.B., Vlamakis, H., Hayes, K., and Salyers, A.A.(2001) Evidence for extensive resistance gene transferamong Bacteroides spp and between Bacteroides andother genera in the human colon. Appl Environ Microbiol 67:561–568.

Simon, R., Priefer, U., and Puhler, A. (1983) A broad hostrange mobilization system for in vivo genetic engineering:transposon mutagenesis in gram negative bacteria.Bio/Technology 1: 784–791.

Smith, C.J., and Parker, A.C. (1993) Identification of a circularintermediate in the transfer and transposition of Tn4555, amobilizable transposon from Bacteroides spp. J Bacteriol175: 2682–2691.

Smith, C.J., Rollins, L.A., and Parker, A.C. (1995) Nucleotidesequence determination and genetic analysis of theBacteroides plasmid, pBI143. Plasmid 34: 211–222.

Smith, C.J., Tribble, G.D., and Bayley, D.P. (1998) Geneticelements of Bacteroides species: a moving story. Plasmid40: 12–29.

Stevens, A.M., Sanders, J.M., Shoemaker, N.B., and Salyers,A.A. (1992) Genes involved in production of plasmidlikeforms by a Bacteroides conjugal chromosomal elementshare amino acid homology with two-component regulatorysystems. J Bacteriol 174: 2935–2942.

Stevens, A.M., Shoemaker, N.B., Li, L.Y., and Salyers, A.A.(1993) Tetracycline regulation of genes on Bacteroidesconjugative transposons. J Bacteriol 175: 6134–6141.

Thompson, J.F., and Landy, A. (1988) Empirical estimation ofprotein-induced DNA bending angles: applications tolambda site-specific recombination complexes. NucleicAcids Res 16: 9687–9705.

Thompson, J.D., Higgins, D.G., and Gibson, T.J. (1994)CLUSTAL W: improving the sensitivity of progressivemultiple sequence alignment through sequence weighting,position-specific gap penalties and weight matrix choice.Nucleic Acids Res 22: 4673–4680.

Tribble, G.D., Parker, A.C., and Smith, C.J. (1999) Geneticstructure and transcriptional analysis of a mobilizable,antibiotic resistance transposon from Bacteroides. Plasmid42: 1–12.

Whittle, G., Hund, B.D., Shoemaker, N.B., and Salyers, A.A.(2001) Characterization of the 13 kb ermF region ofBacteroides conjugative transposon CTnDOT. ApplEnviron Microbiol (in press).

Wu, Z., Gumport, R.I., and Gardner, J.F. (1998) Defining thestructural and functional roles of the carboxyl region of thebacteriophage lambda excisionase (Xis) protein. J Mol Biol281: 651–661.

632 Q. Cheng et al.


Identification of genes required for excision of CTnDOT, a Bacteroides conjugative transposon:...

Documents

Transcript of Identification of genes required for excision of CTnDOT, a Bacteroides conjugative transposon:...