Partition of the linear plasmid N15: interactions of N15 partition functions with the sop locus of the F plasmid

Abstract

A locus close to one end of the linear N15 prophage closely resembles the sop operon which governs partition of the F plasmid; the promoter region contains similar operator sites, and the two putative gene products have extensive amino acid identity with the SopA and -B proteins of F. Our aim was to ascertain whether the N15 sop homologue functions in partition, to identify the centromere site, and to examine possible interchangeability of function with the F Sop system. When expressed at a moderate level, N15 SopA and -B proteins partly stabilize mini-F which lacks its own sop operon but retains the sopC centromere. The stabilization does not depend on increased copy number. Likewise, an N15 mutant with most of its sop operon deleted is partly stabilized by F Sop proteins and fully stabilized by its own. Four inverted repeat sequences similar to those of sopC were located in N15. They are distant from the sop operon and from each other. Two of these were shown to stabilize a mini-F sop deletion mutant when N15 Sop proteins were provided. Provision of the SopA homologue to plasmids with a sopA deletion resulted in further destabilization of the plasmid. The N15 Sop proteins exert effective, but incomplete, repression at the F sop promoter. We conclude that the N15 sop locus determines stable inheritance of the prophage by using dispersed centromere sites. The SopB-centromere and SopA-operator interactions show partial functional overlap between N15 and F. SopA of each plasmid appears to interact with SopB of the other, but in a way that is detrimental to plasmid maintenance.

FIG. 1

N15 prophage and the locus of homology with the F plasmid sop operon. The 46,375-bp bacteriophage sequence (13) was converted to the prophage form by in silico ligation of the ends followed by cutting at bp 24802 and -3. The prophage sketch is oriented following Lobocka et al. (19), with the enlargement of the sop homologue locus inverted to correspond to the usual representation of F sop. The rounded ends represent telomeric structures (telL and telR). Other loci are assigned on the basis of sequence data (13) and are drawn only approximately to scale.

FIG. 2

sop locus variants of mini-F. The pDAG114 plasmid is represented as a linear molecule; only the sop regions and restriction sites used in construction are shown below. ccdB is inactive, as indicated by square brackets. Light shading indicates genes and sites of mini-F, and dark shading indicates those of N15. Gaps show the extent of deleted DNA. The thin double line after the N15 sop operon is a natural IR. The block labelled IR2 is an inserted IR (see Materials and Methods). The sopB deletions at the bottom were introduced into sop operons inserted in the pZS*21 vector.

FIG. 3

Alignment of Sop homologues of N15 and F. In sopOP, the sop promoter-operator top-strand alignment is based on pairing the 5′-CTTTG SopA binding sites (dark shading) (24), whose relative orientation is indicated by arrows; the light shading denotes promoter and Shine-Dalgarno sequences and the sopA start codon. In SopA, amino acid sequences were aligned with the Multalin program. Shading denotes identical residues. The nucleotide and Mg binding sites are labelled ATP I (Walker A box) and II, respectively. Motifs 2 and 3 are conserved in members of the SopA/ParA partition protein family (25). The underlined sequence and that in SopB below are predicted helix-turn-helix structures (6). In SopB, aligned as for SopA, minor changes in parameters alter gap positions, most notably aligning the N15 C-terminal KNKEKK with the F C-terminal KELEKP. The large box (dotted outline) shows an F SopB peptide which protects sopC DNA; the small box (solid outline) indicates a C-terminal deletion which eliminates the protection (10).

FIG. 4

Ability of N15 sopAB to substitute for F sopAB. (a) Segregation of wild-type mini-F (pDAG114 [open circles]), mini-F ΔsopPABC (pDAG202 [solid circles]), and mini-F carrying the sop operon locus of N15 (pDAG219 [solid triangles]). (b) Segregation of mini-F ΔsopAB (pDAG205) from cells also carrying pDAG221 (sopAB_F [solid triangles]), pDAG222 (sopAB_N15 [open triangles]), or the vector (pZS*21 [solid circles]). (c) Relative pDAG205 copy number. In an experiment duplicating that shown in panel b, samples were withdrawn from cultures after three to four generations of growth for estimation of the fraction of cells still carrying pDAG205 (fr. cells mF+) and for measurement of copy number by hybridization of radioactively labelled mini-F and chromosomal probes to Southern blots of extracted DNA (see Materials and Methods). The PflMI site adjacent to repE is poorly cleaved (PflM), resulting in two mini-F fragments being detected; the radioactivity from both bands was summed for the calculation of miniF/chromosome ratios (mF/ldc). Correction for the fraction of each population that had lost the mini-F plasmid allowed calculation of the number of mini-F plasmids per cell relative to that in the strain with no sopAB (−) [mF/cell (rel.)]. This in turn allowed calculation of the contribution of the ∼10% copy number increase to stability: assuming random distribution of the partition-defective plasmid pDAG205 to daughter cells, the observed loss rate of 5.5% corresponds to a copy number of 4.2 at division (loss rate = 0.5^copy-number), leading to an expected loss rate for the 1.1-fold-higher copy number of 4.1% in the absence of active partition. (d) Segregation of mini-F plasmids carrying sopC (pDAG209 [diamonds]) or sopP (pDAG210 [squares]) in the presence (open symbols) or absence (solid symbols) of a plasmid (pDAG222) carrying the N15 sop operon.

FIG. 5

Western blot analysis of SopA production. Exponential cultures of strains harboring wild-type mini-F (pDAG114), mini-F ΔsopA_F (pDAG206), pZS*21 carrying the sop operon of F (pDAG221) or N15 (pDAG222), or no plasmid (−) were grown in LB broth with appropriate antibiotics and sampled for preparation of cell extracts. A series of twofold concentration increments of each sample was subjected to SDS-polyacrylamide gel electrophoresis and Western blot analysis with anti-SopA antibody. Lanes showing approximately the same level of cross-reacting host protein(s) were chosen. The band intensity was proportional to amounts applied to the gel for the film exposure times used.

FIG. 6

Identification of sopC-like centromere sequences in N15 DNA. (a) The midi-N15 plasmid, pG54. The dashed box shows the 25.5-kb BglII deletion which has removed sopB and the 3′ end of sopA. IR1 to -4 denote sopC-like IRs detailed in panel c. (b) Segregation of pG54 from cells also carrying the mini-Fs pDAG202 [Δsop(PABC)_F (solid circles)] and pDAG216 (sopPAB_N15 [open circles]) or the pZS*21-based plasmid pDAG241 (sopAB_F⁺ [open squares]). (c) Comparison of N15 IR sequences with the sopC IR. The underlined bases show differences from the sopC sequence. (d) Segregation of mini-F plasmids carrying the N15 sop operon without (pDAG219 [solid circles]) or with (pDAG215 [open circles]) IR2.

FIG. 7

Effect of SopA on the stability of ΔsopA miniplasmids. For clarity, the rates of loss of pDAG206 (mini-F ΔsopA_F) measured in separate experiments are shown in panels a and b, with the results for certain complementing plasmids shown in both. Note the different ranges on the ordinate of each panel. (a) Segregation of pDAG206 from cells which also carry pZS*21 (no sopAB [solid triangles]), pDAG221 (sopAB_F⁺ [open circles]), pDAG222 (sopAB_N15⁺ [open triangles]), pDAG225 (sopA_F⁺ [open inverted triangles]), or no other plasmid (solid circles). (b) Same as panel a plus pDAG226 (sopA_N15⁺ [solid inverted triangles]). (c) Segregation of pDAG217 (mini-F ΔsopA_N15) from cells which also carry the plasmids shown in panels a and b.