An accessory protein is required for relaxosome formation by small staphylococcal plasmids

Abstract

Mobilization of the staphylococcal plasmid pC221 requires at least one plasmid-encoded protein, MobA, in order to form a relaxosome. pC221 and closely related plasmids also possess an overlapping reading frame encoding a protein of 15 kDa, termed MobC. By completing the nucleotide sequence of plasmid pC223, we have found a further example of this small protein, and gene knockouts have shown that MobC is essential for relaxosome formation and plasmid mobilization in both pC221 and pC223. Primer extension analysis has been used to identify the nic site in both of these plasmids, located upstream of the mobC gene in the sense strand. Although the sequence surrounding the nic site is highly conserved between pC221 and pC223, exchange of the oriT sequence between plasmids significantly reduces the extent of relaxation complex formation, suggesting that the Mob proteins are selective for their cognate plasmids in vivo.

FIG. 1.

Organization of plasmid pC223 (4,608 bp). The map is oriented with the nick site of the double-stranded replication origin (dso) at the +1 position. Novel sequence data obtained in the present study are indicated by the solid bar on the map; amendments to previously available data are concentrated in the shaded region. The principal reading frames are indicated by the open arrows. sso indicates the site and orientation of the single-stranded (−) replication origin. The nick site within the origin of transfer determined in the present study is given as nic. The layout of pC221 is similar, although the two plasmids have considerably reduced sequence identity in the region shown by the dashed line.

FIG. 2.

Effects of plasmid mutations on relaxosome formation. (A) Agarose gel electrophoresis of plasmid variants. Control lanes show negatively supercoiled pC221 (lane 1), pC221cop903 (lane 8), and pC223 (lane 14). Other lanes show samples derived from whole-cell lysates of S. aureus RN4220 carrying different plasmids: lane 2, pC221; lane 3, pC221mobA1; lane 4, pC221mobA4; lane 5, pC221mobAB2; lane 6, pC221mobC1; lane 9, pC221cop903; lane 10, pC221mobA2; lane 11, pC221mobB1; lane 12, pC221mobC3; lane 15, pC223; lane 16, pC223mobA3; lane 17, pC223mobAB1; lane 18, pC223mobC4; and lane 19, pC223mobC5. Lysates from plasmid-free cells are shown in lanes 7, 13, and 20. CHR, chromosomal DNA; OC, SC, open circular and negatively supercoiled plasmid DNA. (B) Restriction sites used to generate mob plasmid mutations, and the relative locations of the mob genes involved. pC221 and pC221cop903 are identical in this region. Conserved motifs I, II, and III are indicated by the shaded regions within mobA. Note that pC221mobA1 results from deletion of 75 bp between the two MwoI sites encoding motif I, pC223mobA3 has lost 461 bp between the BsgI-BbvI sites deleting motifs II and III, and in pC223mobAB1 the mutation causes a frameshift fusing mobA to mobB at the position indicated.

FIG. 3.

Location of the nic site within pC221. (A) Denaturing polyacrylamide gel electrophoresis of 5′-end-labeled restriction products. Negatively supercoiled pC221 previously digested with HpaII (lanes 1 to 5) or pC221 obtained from whole-cell lysates (lanes 6 to 10) was further digested with BstBI (lanes 1 and 6), BstEII (lanes 2 and 7), BstXI (lanes 3 and 8), ClaI (lanes 4 and 9), or XbaI (lanes 5 and 10). 5′-End-labeled fragments were detected by autoradiography. Fragment sizes are indicated in nucleotides; the doublet seen in lane 3 results from the uneven upper and lower strand lengths after digestion. (B) Location of restriction sites within the 692-bp AluI fragment previously shown to contain oriT (49). Larger fragments were not resolved on the gel shown. (C) Mapping of the labeled (★) fragments BE and BX locates the nic site 3′ to the HpaII site and in the sense strand with respect to cat and the mob genes; failure to generate similar fragments in lanes 6, 9, and 10 indicates the 5′ end at nic is blocked (•).

FIG. 4.

Identification of the nic sites of pC221cop903 and pC223. (A) Primer extension analysis against the positive-strand of pC221cop903. The same negative-strand primer was used for all tracks. HpaII, plasmid template cut with HpaII; ATCG, sequencing ladder corresponding to coordinates 3100 to 3170 of the pC221 sequence; pC221(+), pC221cop903 obtained by whole-cell lysate procedure as a template. Doublets resulting from primer extension are correlated with nucleotide sequence by the arrows. (B) Same as for panel A, but with a positive-strand primer. (C and D) Same as for panels A and B, respectively, but with pC223 as a template; this plasmid does not possess a corresponding HpaII site. (E) Interpretation of the primer extension doublets: HpaII cuts the positive-strand as indicated; primer extension products of the expected length (n) and one nucleotide longer (n + 1) are generated by terminal transferase activity of the polymerase. The same interpretation is compatible with a unique nic site at the position indicated.

FIG. 5.

Specificity for oriT in relaxosome formation. Whole-cell lysates were analyzed by gel electrophoresis. Plasmids are shown as follows: lane 1, pC221; lane 2, pC221mobA4; lane 3, pC223; lane 4, pC221cop903; lane 5, pC223oriT1; lane 6, pC223oriT2; lane 7, pC221oriT3; and lane 8, pC221oriT4. CHR, chromosomal DNA; OC and SC, open circular and negatively supercoiled plasmid DNA, respectively. Loadings have been normalized to compensate for the higher copy number of the pC221cop903 replicon.

FIG. 6.

Alignment of oriT sequences. (A) Orientation of features between coordinates 3000 and 3250 of pC221. The locations of nic, restriction sites HpaII and BstBI, putative −35 and −10 signals, and the start of the mobC reading frames are shown, as are five prominent inverted repeats (labeled 1 to 5). The shaded region indicates the sequences present in pT181, shown in panel B. (B) Alignment of pC221-like, pSK639-like, and pSN2-like plasmids with pT181. Selected sequences identified by BLAST searches (1) were aligned by using CLUSTAL X (61). The coordinate (and strand) of the first nucleotide in each row are derived from the present study (pC221 and pC223) or GenBank accession numbers X06627 (pS194), J01764 (pT181), AE015931 (pSE12228-02), AF045240 (pIP1629), U40259 (pSK639), X55798 (pOX2000), and V01282 (pSN2). The numbers at the end of the line represent the distance to nic (or RS_A in the case of pT181). (C) Alignment of sequences immediately upstream of nic (contiguous with panel B above). The consensus nic sequence obtained from these gram-positive plasmids is compared to the consensus for gram-negative plasmids of the IncP family (66). (D) Potential promoter region for mobC-encoding plasmids. The separation of the initiator codon from the start of the transcript is indicated.