Control of SXT integration and excision

Abstract

The Vibrio cholerae SXT element is a conjugative self-transmissible chromosomally integrating element that encodes resistance to multiple antibiotics. SXT integrates in a site-specific fashion at prfC and excises from the chromosome to form a circular but nonreplicative extrachromosomal form. Both chromosomal integration and excision depend on an SXT-encoded recombinase, Int. Here we found that Int is necessary and sufficient for SXT integration and that int expression in recipient cells requires the SXT activators SetC and SetD. Although no xis-like gene was annotated in the SXT genome, Int was not sufficient to mediate efficient SXT chromosomal excision. We identified a novel SXT Xis that seems to function as a recombination directionality factor (RDF), facilitating SXT excision and inhibiting SXT integration. Although unrelated to any previously characterized RDF, Xis is similar to five hypothetical proteins that together may constitute a new family of RDFs. Using real-time quantitative PCR assays to study SXT excision from the chromosome, we determined that while SXT excision is required for SXT transfer, the percentage of cells containing an excised circular SXT does not appear to be a major factor limiting SXT transfer; i.e., we found that most cells harboring an excised circular SXT molecule do not act as SXT donors. In the absence of prfC, SXT integrated into several secondary attachment sites but preferentially into the 5' end of pntB. SXT excision and transfer from a donor containing pntB::SXT were reduced, suggesting that the SXT integration site may also influence the element's transmissibility.

FIG. 1.

Genes required for SXT integration. Integration of replication-deficient plasmids containing the SXT attP site was assessed by a conjugation assay. The inserts in these plasmids, all of which were derived from the mobilizable suicide vector pGP704, are indicated. In the diagrams of the inserts, open reading frames are indicated by arrows, the SXT attP site is indicated by a black box, the putative promoter region upstream of the int-containing operon is indicated by a gray box with an arrow, and the frameshift mutations are indicated by x's. The frequency of exconjugant formation was obtained by dividing the number of exconjugants (Tc^r Ap^r CFU) by the number of recipients (Tc^r CFU). In all the cases, the donor strain was E. coli SM10λpir and the recipient strains were E. coli CAG18439 derivatives. For the plasmids in panel A, the recipient cells harbored pSetCD33, which contained setCD under control of P_BAD. For the plasmids in panel B, the recipient cells harbored pInt33, which contained int under control of P_BAD. To induce expression of setCD or int, the conjugation assays were carried out by using media supplemented with 0.02% arabinose. The bars indicate the mean values obtained from two independent experiments. The asterisk and the circle in lines 3 and 4 indicate that the frequencies of exconjugant formation were less than 10⁻⁵ exconjugant/recipient; in line 3, the frequency was 9.7 × 10⁻⁶, and in line 4, the frequency was 7.0 × 10⁻⁶.

FIG. 2.

Xis augments excision of integrated SXT-derived plasmids. The amount of an unoccupied attB site was quantified by real-time quantitative PCR. The means of triplicate measurements performed with each sample are shown. E. coli CAG18439, which contained one copy of attB per chromosome, was used as a control strain. DNA templates were extracted from strains containing the integrated plasmids described in Fig. 1 and pSetCD33 or pInt33 expressing setCD or int under control of P_BAD. Strains were grown overnight in LB medium supplemented either with 0.2% glucose (+glc) to repress expression or with 0.02% arabinose (+ara) to induce expression.

FIG. 3.

Xis promotes excision and transfer of SXT. Real-time quantitative PCR was used to determine the percentage of unoccupied attB sites and attP sequences resulting from circularization of SXT. Triplicate measurements were obtained for each sample, and the mean and standard deviation are shown for each assay. The control strain E. coli VI61 (CAG18439 ΔlacZ::attP-cat) contained a single chromosomal copy of attB and attP. E. coli HW220 (CAG18439 prfC::SXT) was used as the wild-type (wt) strain. The Δxis strain was E. coli VI100 (CAG18439 prfC::SXT Δxis). DNA templates were prepared from overnight LB broth cultures containing 0.2% glucose or 0.02% arabinose where indicated. The frequencies of exconjugant formation were obtained by dividing the number of exconjugants (Nal^r SXT^r CFU) by the number of donor cells (Tc^r CFU). In all the mating experiments, the recipient strain was E. coli BI533 (MG1655 Nal^r). The solid bars indicate the mean values obtained from two independent mating experiments. The expression of xis by pXis and the expression of setCD by pSetCD in the donor strains were repressed with 0.2% glucose or were induced with 0.02% arabinose.

FIG. 4.

ClustalW alignment of SXT Xis with five hypothetical proteins from V.cholerae (VC1785), M.leprae (ML2429), S.meliloti (SMc02201), M.loti (msr5461), and B.melitensis (BMEI1700). Amino acid residues that are identical in at least two-thirds of the sequences are indicated by a black background, and residues that are similar in at least two-thirds of the sequences are indicated by a grey background. The bar indicates the predicted helix-turn-helix (HTH) DNA-binding motif in SXT Xis, VC1785, ML2429, SMc02201, msr5461, and BMEI1700. These peptides have Dodd-Egan scores of 3.32, 4.13, 4.24, 2.57, 4.70, and 2.26, respectively.