Dynamique des réplicons bactériens

 

 

Follows, a brief summary of the Lane group's research programme

Follows, a brief summary of the Lane group's research programme and the specific projects that compose it; the post-doc should feel free to consider any of the areas listed (biochemical and biophysical approaches particularly encouraged), with the understanding that a project will be defined in consultation with Dr Lane.

Bacteria, no less than eukaryotic cells, must reliably transmit replicas of their genomes to daughter cells arising at each division if they are to survive and thrive. The last ten years have seen notable advances in our knowledge of bacterial mitosis, due largely to synergistic application of fluorescence microscopy and biochemical techniques. Nevertheless, despite significant recent progress, it is still a process that fundamentally we do not understand.

Two classes of function govern bacterial replicon segregation, or partition. The first is characteristic of low copy-number plasmids which, because they are dispensable for cell survival, were the main tools of early partition studies. These functions are specified by three determinants constituting a specific locus (par) which, although generally situated adjacent to the plasmid replication region, behaves as an independent unit capable of functioning when transferred to another replicon. The typical par locus comprises a two-gene operon, parAB, whose transcription is autoregulated, and a cis-acting site, parS, that functions as a centromere. The ParA protein is an ATP/ADP-binding protein of either the Walker box (e.g. plasmids F, P1, RK2) or actin (R1) type. ParB binds specifically to the parS centromere to form a partition complex. (The plasmid P1 nomenclature is used here for general description.) Partition is thought to be based on an interaction between partition complexes which pairs copies of the plasmid, followed by intervention of ParA to unpair the complexes and to impel the plasmid molecules into each daughter cell. Transitions between the ATP- and ADP-bound states of ParA appear to be fundamental to the protein's ability to regulate partition.

Genome sequencing has revealed that most bacterial chromosomes carry parAB homologues, usually linked to their replication origins. These differ from plasmid systems mainly in being associated with multiple parS sites dispersed over the ori-par region and in lacking the autoregulated expression seen in plasmid versions. They also differ in that whereas plasmid parABS functions appear to be solely responsible for partition of their replicons, chromosomal partition involves other actors. These constitute the second, rather diverse, class of function, and include homologues of the cytoskeletal proteins of eukaryotes: MukB, a dynamin-like SMC protein which (with MukE and F) compacts DNA, and MreB, an actin-like protein which forms a membrane-associated spiral structure apparently involved in a poleward movement of chromosomes in which RNA polymerase also participates. They also include non-parS centromere-like sequences or zones and the proteins which bind to them, polarized sequence elements needed for chromosome organization, and dimer-resolution functions that ensure the mid-cell is cleared for passage of the division septum. The relative importance to partition of these factors and how they are related to ParABS function is not yet clear.

The overall objective of our research is to understand how these functions interact to generate the mechanism that segregates bacterial replicons. We work on two experimental models: the E.coli plasmid F, which has a classic Walker-box ATPase partition system (termed SopABC), and the ParABS systems specified by each replicon of the three-chromosome genome of Burkholderia cenocepacia. We are seeking the answers to certain fundamental problems:

– What are the essential properties of the partition complex formed by the centromere (parS) and the protein that binds to it (ParB) that allow it to function in partition and to activate the ParA ATPase? There is no evidence for the general assumption that the pairing of partition complexes observed in vitro and in vivo is needed for partition, despite its tempting parallel to eukaryotic metaphase, nor does a straightforward in vivo assay exist. The following projects address these issues.

3D-structure of the F SopB protein. To date, partial structures of three proteins of the ParB family have been reported. In collaboration with the Dyda group at NIH we have obtained crystals of the SopB-sopC complex and are thus well positioned to determine the first entire structure. Our aim is to understand SopB function through analysis of mutant proteins alone and in co-crystals with SopA and DNA.

Requirement for partition complex pairing. To determine whether, as generally assumed but not shown, partition depends on prior pairing of plasmid replicas, we have devised a test based on single-infection by phage that carry the sopC centromere and a tag allowing localisation.

Assays of partition complex pairing. Assuming that even if pairing is not essential it plays a facilitating role in partition, we intend to examine the formation and activity of the paired structure by setting up a single-molecule assay (in collaboration with Mikhail Grigoriev of this institute). In addition, we shall refine an in vivo assay based on enhancement of IS911 transposase-mediated dimer formation recently developed in this lab.

– What do the dynamic properties of ParA proteins signify at the molecular level and how are they related to partition? Walker-box ParA proteins in the presence of their cognate partition complex exhibit ATP-dependent oscillation over the nucleoid region. The mechanism is unknown.

Polymerisation of the F SopA protein. Our in vitro characterisation of SopA has revealed the relationship between SopA filament formation and DNA binding and the influence of SopB protein on both. The model of SopA dynamics suggested by our results extends existing proposals by defining the role of DNA binding by SopB, and we shall test it using a combination of electron microscopy, FRET and biochemistry.

Role of the nucleoid. Although it has long been clear that Walker-box systems oscillate over the nucleoid area, this aspect of partition has not been explored. By targeted disruption of nucleoid structure we shall test the dependence of plasmid positioning and SopA movement on the presence of non-specific DNA.

– Contrary to classic experimental models, an increasing number of bacterial species (many of them pathogens) carry a multipartite genome. This raises the question of how the cell organizes the orientation and active displacement into daughter cells of two or more bulky DNA molecules which share the same potential attachment surfaces (membrane, nucleoid, …) and, given the uniformity of parS sequences in most sequenced species, a common centromere. Our group is the first to have investigated the activity of ParABS systems from a multi-chromosome bacterium. We have shown that each of the three chromosomes of B.cenocepacia carries a slightly different centromere, on which the cognate ParB proteins act to confer segregational specificity. Other groups have observed independent localisation of replication origins.

ParABS action in chromosome partition. The capacity of plasmid-like partition systems (ParABS) to direct chromosome segregation remains unclear. We considered that the demands of segregating multiple chromosomes could require the specificity and precision that these systems manifest. The study of partition in multichromosome bacteria may thus clarify the role of ParABS loci as well as that of other contributors to bacterial mitosis. We used a heterologous (E.coli) system to demonstrate the partition competence of the B.cenocepacia ParABS systems. We shall examine partition in B.cenocepacia itself, by mutating the par loci, localizing various chromosomal regions via FISH and fluorescent-protein-binding tags, and following displacement of the ParA proteins.

Regulation of parAB expression. Despite the known disruptive effects of Par protein excess on partition, almost no studies of parAB regulation have been reported. A recent report described developmental control in Streptomyces. We observed that mutating parS-like sites in or near parAB altered partition behaviour and that two ParB proteins silence gene expression. We shall analyze the role of these elements in governing Par protein levels and partition behaviour.

Specificity and cross-talk among ParABS systems. Although the ParB-parS pairs are functionally exclusive in E.coli, the question of whether they cross-react in B.cenocepacia, to co-regulate partition for example, remains open. We shall mutate the parB genes and parS sites to map the interaction surfaces. This study will profit from knowledge of the structure of the SopB-sopC complex (above). Interactions between non-cognate ParA-ParB pairs will also be investigated, using silencing, partition and co-immunoprecipitation assays.