The DNA binding domains of P1 ParB and the architecture of the P1 plasmid partition complex

Stable maintenance of P1 plasmids in Escherichia coli is mediated by a high affinity nucleoprotein complex called the partition complex, which consists of ParB and the E. coli integration host factor (IHF) bound specifically to the P1 parS site. IHF strongly stimulates ParB binding to parS, and the minimal partition complex contains a single dimer of ParB. To examine the architecture of the partition complex, we have investigated the DNA binding activity of various ParB fragments. Gel mobility shift and DNase I protection assays showed that the first 141 residues of ParB are dispensable for the formation of the minimal, high affinity partition complex. A fragment missing only the last 16 amino acids of ParB bound specifically to parS, but binding was weak and was no longer stimulated by IHF. The ability of IHF to stimulate ParB binding to parS correlated with the ability of ParB to dimerize via its C terminus. Using full and partial parS sites, we show that two regions of ParB, one in the center and the other near the C terminus of the protein, interact with distinct sequences within parS. Based on these data, we have proposed a model of how the ParB dimer binds parS to form the minimal partition complex.     

Plasmid partitioning and the spreading of P1 partition protein ParB

Bacterial plasmids of low copy number, P1 prophage among them, are actively partitioned to nascent daughter cells. The process is typically mediated by a pair of plasmid-encoded proteins and a cis-acting DNA site or cluster of sites, referred to as the plasmid centromere. P1 ParB protein, which binds to the P1 centromere (parS), can spread for several kilobases along flanking DNA. We argue that studies of mutant ParB that demonstrated a strong correlation between spreading capacity and the ability to engage in partitioning may be misleading, and describe here a critical test of the dependence of partitioning on the spreading of the wild-type protein. Physical constraints imposed on the spreading of P1 ParB were found to have only a minor, but reproducible, effect on partitioning. We conclude that, whereas extensive ParB spreading is not required for partitioning, spreading may have an auxiliary role in the process.     

ATP-regulated interactions between P1 ParA, ParB and non-specific DNA that are stabilized by the plasmid partition site, parS

Localization of the P1 plasmid requires two proteins, ParA and ParB, which act on the plasmid partition site, parS. ParB is a site-specific DNA-binding protein and ParA is a Walker-type ATPase with non-specific DNA-binding activity. In vivo ParA binds the bacterial nucleoid and forms dynamic patterns that are governed by the ParB-parS partition complex on the plasmid. How these interactions drive plasmid movement and localization is not well understood. Here we have identified a large protein-DNA complex in vitro that requires ParA, ParB and ATP, and have characterized its assembly by sucrose gradient sedimentation and light scattering assays. ATP binding and hydrolysis mediated the assembly and disassembly of this complex, while ADP antagonized complex formation. The complex was not dependent on, but was stabilized by, parS. The properties indicate that ParA and ParB are binding and bridging multiple DNA molecules to create a large meshwork of protein-DNA molecules that involves both specific and non-specific DNA. We propose that this complex represents a dynamic adaptor complex between the plasmid and nucleoid, and further, that this interaction drives the redistribution of partition proteins and the plasmid over the nucleoid during partition.     

Stoichiometry of P1 plasmid partition complexes

The P1 plasmid prophage is faithfully partitioned by a high affinity nucleoprotein complex assembled at the centromere-like parS site. This partition complex is composed of P1 ParB and Escherichia coli integration host factor (IHF), bound specifically to parS. We have investigated the assembly of ParB at parS and its stoichiometry of binding. Measured by gel mobility shift assays, ParB and IHF bind tightly to parS and form a specific complex, called I + B1. We observed that as ParB concentration was increased, a second, larger complex (I + B2) formed, followed by the formation of larger complexes, indicating that additional ParB molecules joined the initial complex. Shift Western blotting experiments indicated that the I + B2 complex contained twice as much ParB as the I + B1 complex. Using mixtures of ParB and a larger polyhistidine-tagged version of ParB (His-ParB) in DNA binding assays, we determined that the initial I + B1 complex contains one dimer of ParB. Therefore, one dimer of ParB binds to its recognition sequences that span an IHF-directed bend in parS. Once this complex forms, a second dimer can join the complex, but this assembly requires much higher ParB concentrations.     

P1 and P7 plasmid partition: ParB protein bound to its partition site makes a separate discriminator contact with the DNA that determines species specificity.

The cis-acting P1 and P7 parS sites are responsible for active partition of P1 and P7 plasmids to daughter cells. The two sites are similar but function only with ParB proteins from the correct species. Using hybrid ParB proteins and hybrid parS sites, we show that specificity is determined by contacts between bases that lie within two parS hexamer boxes and a region in the ParB C-terminus. We refer to these contacts as discriminator contacts. The P7 discriminator contacts were mapped to 3 and 2 bp respectively within the two parS hexamer boxes, and a 10 amino acid region of P7 ParB. Similarly placed residues of different sequence are responsible for the P1 discriminator contact. The discriminator contacts are distinct from previously identified DNA binding contacts which involve different ParB and parS regions. Deletion of the ParB C-terminus that makes the discriminator contact does not diminish in vitro binding but renders it species independent. The discriminator contact is therefore a negative function, interfering with binding of the wrong ParB, but not providing energy for the binding of the correct one. Similar discriminator contacts might be responsible for specificities seen among families of eukaryotic DNA binding proteins that share common binding motifs.     

The homologous operons for P1 and P7 plasmid partition are autoregulated from dissimilar operator sites

The plasmid-partition regions of the P1 and P7 plasmid prophages in Escherichia coli are homologues which each encode two partition proteins, ParA and ParB. The equivalent PI and P7 proteins are closely related. In each case, the proteins are encoded by an operon that is autoregulated by the ParA and ParB proteins in concert. This regulation is species-specific, as the P1 proteins are unable to repress the P7 par operon and vice versa. The homologous ParA proteins are primarily responsible for repression and bind to regions that overlap the operon promoter in both cases. The DNA-binding domain of the P7 auto-repressor lies in the amino-terminal end of the P7 ParA protein. This region includes a helix-turn-helix motif that has a clear counterpart in the P1 ParA sequence. However, despite the common regulatory mechanism and the similarity of the proteins involved in repression, the promoter-operator sequences of these two operons are very different in sequence and organization. The operator is located downstream of the promoter in P1 and upstream of it in P7, and the two regions show little, if any, homology. How these differences may have arisen from a common ancestral form is discussed.     

Altered ParA partition proteins of plasmid P1 act via the partition site to block plasmid propagation

The partition system of the P1 plasmid, P1 par consists of the ParA and ParB proteins and a cis -acting site, parS . It is responsible for the orderly segregation of plasmid copies to daughter cells. Plasmids with null mutations in parA or parB replicate normally, but missegregate. ParB binds specifically to the parS site, but the role of ParA and its ATPase activity in partition is unclear. We describe a novel class of parA mutants that cannot be established or maintained as plasmids unless complemented by the wild-type gene. One, parAM314I , is conditional: it can be maintained in cells in minimal medium but cannot be established in cells growing in L broth. The lack of plasmid propagation in L broth-grown cells was shown to be caused by a ParB-dependent activity of the mutant ParA protein that blocks plasmid propagation by an interaction at the parS site. Thus, ParA acts to modify the ParB– parS complex, probably by binding to it. Partition is thought to involve selection of pairs of plasmids before segregation, either by physical pairing of copies or by binding of copies to paired host sites. We suggest that ParA is involved in this reaction and that the mutant ParA protein forms paired complexes that cannot unpair.     

Effects of the P1 plasmid centromere on expression of P1 partition genes

The partition operon of P1 plasmid encodes two proteins, ParA and ParB, required for the faithful segregation of plasmid copies to daughter cells. The operon is followed by a centromere analog, parS, at which ParB binds. ParA, a weak ATPase, represses the par promoter most effectively in its ADP-bound form. ParB can recruit ParA to parS, stimulate its ATPase, and significantly stimulate the repression. We report here that parS also participates in the regulation of expression of the par genes. A single chromosomal parS was shown to augment repression of several copies of the par promoter by severalfold. The repression increase was sensitive to the levels of ParA and ParB and to their ratio. The increase may be attributable to a conformational change in ParA mediated by the parS-ParB complex, possibly acting catalytically. We also observed an in cis effect of parS which enhanced expression of parB, presumably due to a selective modulation of the mRNA level. Although ParB had been earlier found to spread into and silence genes flanking parS, silencing of the par operon by ParB spreading was not significant. Based upon analogies between partitioning and septum placement, we speculate that the regulatory switch controlled by the parS-ParB complex might be essential for partitioning itself.     

Fine-structure analysis of the P1 plasmid partition site.

P1 plasmid partition requires two plasmid-encoded Par proteins and a cis-acting site. The site, parS, lies in a region consisting of a 13-bp palindrome and an adjacent AT-rich sequence. A series of point mutations were analyzed for their effects on partition site activity. The results indicated that only the left arm of the palindrome and some adjacent bases were needed. The limits of the functional site were further refined to a maximum of 22 bp, which includes binding sites for the P1 ParB protein. Mutations in the 22-bp site cause concomitant defects in partition and the ability to exert partition-mediated incompatibility. Like the region immediately to the left of the 22-bp region, the right arm of the palindrome is not essential for partition but does contain information that affects the specificity of incompatibility.     

Biochemical activities of the ParA partition protein of the P1 plasmid

The unit-copy P1 plasmid depends for stability on a plasmid-encoded partition region called par, consisting of the parA and parB genes and the parS site. ParA is absolutely required for partition, but its partition-critical role is not known. Purified ParA protein is shown to possess an ATPase activity in vitro which is specifically stimulated by purified ParB protein and by DNA. ParA is responsible for regulation of expression of parA and parB, and purified ParA has an ATP-dependent, site-specific DNA binding activity which recognizes a sequence that overlaps the parA promoter. The role of the ATP-dependence of the binding activity, as well as other possible functions of the ATPase activity in partition, is discussed.     

The P1 plasmid in action: time-lapse photomicroscopy reveals some unexpected aspects of plasmid partition

The prophage of bacteriophage P1 is a low copy number plasmid in Escherichia coli and is segregated to daughter cells by an active partition system. The dynamics of the partition process have now been successfully followed by time-lapse photomicroscopy. The process appears to be fundamentally different from that previously inferred from statistical analysis of fixed cells. A focus containing several plasmid copies is captured at the cell center. Immediately before cell division, the copies eject bi-directionally along the long axis of the cell. Cell division traps one or more plasmid copies in each daughter cell. These copies are free to move, associate, and disassociate. Later, they are captured to the new cell center to re-start the cycle. Studies with mutants suggest that the ability to segregate accurately at a very late stage in the cell cycle is dependent on a novel ability of the plasmid to control cell division. Should segregation be delayed, cell division is also delayed until segregation is successfully completed.     

Species and incompatibility determination within the P1par family of plasmid partition elements

The P1par family of active plasmid partition systems consists of at least six members, broadly distributed in a variety of plasmid types and bacterial genera. Each encodes two Par proteins and contains a cis-acting parS site. Individual par systems can show distinct species specificities; the proteins from one type cannot function with the parS site of another. P1par-versus-P7par specificity resides within two hexamer BoxB repeats encoded by parS that contact the ParB protein near the carboxy terminus. Here, we examine the species specificity differences between Yersinia pestis pMT1parS and Escherichia coli P1 and P7parS. pMT1parS site specificity could be altered to that of either P1 or P7 by point mutation changes in the BoxB repeats. Just one base change in a single BoxB repeat sometimes sufficed. The BoxB sequence appears to be able to adopt a number of forms that define exclusive interactions with different ParB species. The looped parS structure may facilitate this repertoire of interaction specificities. Different P1par family members have different partition-mediated incompatibility specificities. This property defines whether two related plasmids can coexist in the same cell and is important in promoting the evolution of new plasmid species. BoxB sequence changes that switch species specificity between P1, P7, and pMT1 species switched partition-mediated plasmid incompatibility in concert. Thus, there is a direct mechanistic link between species specificity and partition-mediated incompatibility, and the BoxB-ParB interaction can be regarded as a special mechanism for facilitating plasmid evolution.     

A plasmid partition system of the P1-P7par family from the pMT1 virulence plasmid of Yersinia pestis

The complete sequence of the virulence plasmid pMT1 of Yersinia pestis KIM5 revealed a region homologous to the plasmid partition (par) region of the P7 plasmid prophage of Escherichia coli. The essential genes parA and parB and the downstream partition site gene, parS, are highly conserved in sequence and organization. The pMT1parS site and the parA-parB operon were separately inserted into vectors that could be maintained in E. coli. A mini-P1 vector containing pMT1parS was stably maintained when the pMT1 ParA and ParB proteins were supplied in trans, showing that the pMT1par system is fully functional for plasmid partition in E. coli. The pMT1par system exerted a plasmid silencing activity similar to, but weaker than those of P7par and P1par. In spite of the high degree of similarity, especially to P7par, it showed unique specificities with respect to the interactions of key components. Neither the P7 nor P1 Par proteins could support partition via the pMT1parS site, and the pMT1 Par proteins failed to support partition with P1parS or P7parS. Typical of other partition sites, supernumerary copies of pMT1parS exerted incompatibility toward plasmids supported by pMT1par. However, no interspecies incompatibility effect was observed between pMT1par, P7par, and P1par.     

Recognition of the P1 plasmid centromere analog involves binding of the ParB protein and is modified by a specific host factor.

The P1 plasmid partition system is responsible for segregation of daughter plasmids during division of the Escherichia coli host cell. The P1-encoded elements consist of two essential proteins, ParA and ParB, and the cis-acting incB region. The incB region determines partition-mediated incompatibility and contains the centromere-like site parS. We have isolated and purified the two proteins. ParB binds specifically to the incB region in vitro. DNase I footprinting assays place a strong binding site over the 35-bp parS sequence previously shown to be sufficient for partition when the Par proteins are supplied in trans. A weaker site lies within the incB region in sequences that are important for specifying incompatibility, but are not essential for partition. Gel band retardation assays show that a host factor binds specifically to the incB sequence. The factor strongly stimulates binding of ParB. Cutting the region at a site between the two ParB binding sites yields two fragments that can bind ParB but not host factor. Thus, information for host-factor binding lies in the region determining the specificity of plasmid incompatibility. The roles of parB and the host factor in partition and the specificity of plasmid incompatibility are discussed.     

Fine-structure analysis of the P7 plasmid partition site.

The par region of bacteriophage P7 is responsible for active partition of the P7 plasmid prophage into daughter cells. The cis-acting partition site was defined precisely as a 75-bp sequence that was necessary and sufficient to promote correct segregation of an unstable vector plasmid when the two P7 partition proteins, ParA and ParB, were supplied in trans. Roughly the same region was necessary to exert partition-mediated incompatibility. The minimal site contains an integration host factor (IHF) protein binding site bracketed by regions containing heptamer repeat sequences that individually bind ParB. An additional sequence forms the left boundary of the site. Site-directed mutations in the latter sequence, as well as the IHF motif and the rightmost ParB box, blocked site function. Although the P7 site shares 55% sequence identity with its counterpart in bacteriophage P1, functional interactions between the partition sites and the Par proteins of the two plasmids were entirely species specific in vivo. The P1 sequence has similar IHF and ParB binding motifs, but the left boundary sequence differs radically and may define a point of species-specific contact with the Par proteins. No evidence was found for the existence of a functional P7 analog of the P1 parS core, a small subregion of the P1 site that, in isolation, acts as an enfeebled partition site with modified incompatibility properties.     

P1 ParB Domain Structure Includes Two Independent Multimerization Domains

ParB is one of two P1-encoded proteins that are required for active partition of the P1 prophage in Escherichia coli. To probe the native domain structure of ParB, we performed limited proteolytic digestions of full-length ParB, as well as of several N-terminal and C-terminal deletion fragments of ParB. The C-terminal 140 amino acids of ParB form a very trypsin-resistant domain. In contrast, the N terminus is more susceptible to proteolysis, suggesting that it forms a less stably folded domain or domains. Because native ParB is a dimer in solution, we analyzed the ability of ParB fragments to dimerize, using both the yeast two-hybrid system and in vitro chemical cross-linking of purified proteins. These studies revealed that the C-terminal 59 amino acids of ParB, a region within the protease-resistant domain, are sufficient for dimerization. Cross-linking and yeast two-hybrid experiments also revealed the presence of a second self-association domain within the N-terminal half of ParB. The cross-linking data also suggest that the C terminus is inhibitory to multimerization through the N-terminal domain in vitro. We propose that the two multimerization domains play distinct roles in partition complex formation.     

Probing the structure of complex macromolecular interactions by homolog specificity scanning: the P1 and P7 plasmid partition systems.

The P1 plasmid partition locus, P1 par, actively distributes plasmid copies to Escherichia coli daughter cells. It encodes two DNA sites and two proteins, ParA and ParB. Plasmid P7 uses a similar system, but the key macromolecular interactions are species specific. Homolog specificity scanning (HSS) exploits such specificities to map critical contact points between component macromolecules. The ParA protein contacts the par operon operator for operon autoregulation, and the ParB contacts the parS partition site during partition. Here, we refine the mapping of these contacts and extend the use of HSS to map protein-protein contacts. We found that ParB participates in autoregulation at the operator site by making a specific contact with ParA. Similarly, ParA acts in partition by making a specific contact with ParB bound at parS. Both these interactions involve contacts between a C-terminal region of ParA and the extreme N-terminus of ParB. As a single type of ParA-ParB complex appears to be involved in recognizing both DNA sites, the operator and the parS sites may both be occupied by a single protein complex during partition. The general HSS strategy may aid in solving the three-dimensional structures of large complexes of macromolecules.     

P1 plasmid replication: initiator sequestration is inadequate to explain control by initiator-binding sites.

The unit-copy plasmid replicon mini-P1 consists of an origin, a gene for an initiator protein, RepA, and a control locus, incA. Both the origin and the incA locus contain repeat sequences that bind RepA. It has been proposed that the incA repeats control replication by sequestering the rate-limiting RepA initiator protein. Here we show that when the concentration of RepA was increased about fourfold beyond its normal physiological level from an inducible source in trans, the copy number of a plasmid carrying the P1 origin increased about eightfold. However, when the origin and a single copy of incA were present in the same plasmid, the copy number did not even double. The failure of an increased supply of RepA to overcome the inhibitory activity of incA is inconsistent with the hypothesis that incA inhibits replications solely by sequestering RepA. We propose that incA, in addition to sequestration, can also restrain replication by causing steric hindrance to the origin function. Our proposal is based on the observation that incA can bind to a RepA-origin complex in vitro.