Special Reviews

Bacterial Chromosome Dynamics

See allHide authors and affiliations

Science  08 Aug 2003:
Vol. 301, Issue 5634, pp. 780-785
DOI: 10.1126/science.1084780


Bacterial chromosomes are highly compacted structures and share many properties with their eukaryote counterparts, despite not being organized into chromatin or being contained within a cell nucleus. Proteins conserved across all branches of life act in chromosome organization, and common mechanisms maintain genome integrity and ensure faithful replication. The principles that underlie chromosome segregation in bacteria and eukaryotes share similarities, although bacteria segregate DNA as it replicates and lack a eukaryote-like mitotic apparatus for segregating chromosomes. This may be because the distances that newly replicated bacterial chromosomes move apart before cell division are small as compared to those in eukaryotes. Bacteria specify positional information, which determines where cell division will occur and which places the replication machinery and chromosomal loci at defined locations that change during cell cycle progression.

Central to the life process is the organization of DNA into chromosomes that allow the DNA to be managed in concert with other cellular activities. Negative supercoiling of DNA and SMC (structural maintenance of chromosomes) proteins play ubiquitous roles in chromosome organization, whereas the use of histones in chromosome compaction and organization is restricted to eukaryotes and archaea. The integrity of DNA is maintained using conserved repair and recombinational mechanisms, and highly accurate DNA replication is a prelude to the faithful transmission of newly replicated chromosomes to daughter cells before cell division. The mechanism of DNA replication and its control through initiation is conserved through all branches of life, although bacterial chromosomes control replication from a single origin per chromosome, whereas eukaryote chromosomes contain multiple origins. Ubiquitous AAA proteins [adenosine triphosphatases (ATPases) associated with various cellular activities] (1) play key roles in eukaryote and prokaryote DNA replication initiation, replication fork progression, recombination and chromosome segregation, by using the energy of ATP hydrolysis to translocate DNA, separate DNA strands, and remodel nucleoprotein complexes.

The segregation of bacterial chromosomes, commonly containing 1000 to 5000 genes within a single DNA molecule, may seem to be less of a challenge than the segregation of the multiple chromosomes of eukaryote cells. However, bacteria have no mitotic apparatus and segregate chromosomes to daughter cells as they replicate. Furthermore, some bacteria have multiple chromosomes, and many contain plasmids that encode proteins crucial to that organism's life-style. The replication and segregation to daughter bacterial cells of multiple replicating units (replicons) must be precisely controlled and coordinated with the cell cycle if the descendants of a cell are to maintain the genetic identity of their parent (2, 3). Remarkably, some plasmids have evolved to be transferred to and maintained within diverse bacterial species, indicating autonomy of replication control, though they retain an ability to interact with diverse host environments.

Most studies of bacterial chromosome organization and processing have been directed at Escherichia coli, Bacillus subtilis, and Caulobacter crescentus. The developmental program that leads to cell type–specific chromosome behavior in the latter two species has been invaluable for probing chromosome biology (4). The combination of experimental data with the analysis of the genome sequences derived from more than 130 bacterial species indicates that all bacteria use the same principles for chromosome organization and segregation, replication, repair, and recombination, although details of how these processes are regulated and integrated into the life-style of a particular organism differ. Archaea share features of their chromosomal biology with bacteria and eukaryotes (5). Together, bacteria and archaea constitute more than 90% of Earth's cellular nitrogen-phosphorus, up to 50% of the cellular carbon, and they are responsible for most bioconversions. Furthermore, their huge population size (>1030), along with their diversity (∼106 species), has provided enormous opportunities for evolutionary change (6). Therefore, it is instructive to find such conserved mechanisms for DNA organization and processing.

Although knowledge about the mechanisms that transcribe, replicate, recombine, and repair DNA is extensive, molecular understanding of the processes that organize, position, and segregate bacterial DNA is only now beginning to emerge. In part this is because genetic approaches have not revealed genes whose function is dedicated to any one of these latter processes. Nevertheless, the application of tools that allow the visualization of genes and proteins in living cells and the availability of extensive genomic data along with the use of genetics and molecular biology, are rapidly furthering understanding of how chromosome organization and processing are integrated into the overall biology of the bacterial cell.

Chromosome Organization

Bacterial chromosomal DNA is compacted into a cytologically visible nucleoid, which occupies a substantial proportion of the cell volume; for example, the E. coli nucleoid of ∼0.2 μm3 resides in a cell whose volume is ∼1 μm3 (7) The linear compaction of 103- to 104-fold is comparable to that in eukaryote chromosomes, despite an absence of histones; consequently ∼50% of DNA supercoiling is unconstrained (7, 8). A range of abundant, yet nonessential, DNA binding proteins, which include HU, IHF, HNS, and SMC, may facilitate compaction and organization. Typically, the nucleoid in growing E. coli occupies the central part of a cell as a diffuse, often bilobed, mass. Inhibition of transcription or translation leads to a further compaction into a precisely positioned ovoid nucleoid (7). Within the nucleoid ∼5% of the volume is occupied by DNA, which is organized into separate 20- to 100-kilobase pairs (kbp) supercoiled loops. These are dynamic and are unlikely to have sequence-specific domain boundaries (8). Eukaryote chromosomes have similar supercoiled domains. Negative superhelicity is maintained by the action of topoisomerases, in particular by the ability of gyrase to remove the positive supercoils generated during replication and transcription (911).

SMC proteins (Muk in E. coli) contribute to chromosome organization and processing in all branches of life (1214). SMCs are large, V-shaped, homodimeric or heterodimeric proteins, which have a conserved architecture and are likely to share their biochemical action (Fig. 1). Each arm of the V consists of two antiparallel coiled-coil chains derived from a monomer, with a constant length of ∼50 nm; these could be the template for the wrapping and organization of DNA. Monomers dimerize through interaction at the V base. The heads at the open end of the V contain the ATP binding and hydrolysis domains and interact with ubiquitous accessory proteins. These may help provide specificity with regard to particular biological function, whether it is in the organization of DNA (condensins), in cohesion of newly replicated sisters (cohesins), or in processing broken DNA ends (13, 14). SMCs are unlikely to use ATP hydrolysis to move or segregate chromosomal DNA directionally. Despite the role of SMCs in bacterial chromosome organization, they are often dispensable during growth at low temperatures.

Fig. 1.

SMC architecture. SMCs, which act in chromosome organization, sister cohesion, and processing of DNA ends, may be homodimers (bacteria) or heterodimers (eukaryotes). The “hinge” is the dimerization region. Interactions of the N- and C-terminal regions form a functional ATPase. Two or more accessory proteins interact with the C-N-terminal regions. These regions may also interact to form a closed ring (14).

Control of DNA Replication Through Initiation

The strategies for controlling prokaryote and eukaryote DNA replication through initiation are conserved even though bacterial chromosomes have a single defined replication origin, whereas eukaryote chromosomes have many origins, which are not necessarily specified by defined DNA sequence. Bacteria can have several overlapping replication cycles under way in the same cell, because the time it takes to replicate a complete chromosome can be longer than the generation time.

Replication initiation requires the opening of duplex DNA strands to allow access and assembly of a functional replication complex, the replisome. The bacterial initiation protein (DnaA), when complexed with ATP, binds to specific origin sequences, where in a complex set of interactions it opens them to allow access to the helicase loader (DnaC), which facilitates loading of the replicative helicase (DnaB) (15, 16). The sequential loading of these three AAA proteins plays an important role in controlling replication initiation to ensure that it only occurs once per cell generation. Subsequently, other replisome components assemble onto this scaffold. In eukaryotes and archaea, the Orc complex acts as the DnaA homolog and Cdc6 performs the role of DnaC, in facilitating the loading of the replicative helicase, Mcm (1517). Understanding of how coordinated initiation frequency responds to bacterial physiological state is still murky, although the availability of initiator protein and its accessibility to origins both play a role. In E. coli and its relatives, SeqA sequesters GATC-rich origins and prevents premature reinitiation by DnaA-ATP, when the As within duplex GATC are hemimethylated (15).

Under some conditions, highly transcribed DNA regions and recombination intermediates can additionally act as E. coli DNA replication initiation sites. In the latter case, it is particularly crucial that replication control permits this essential fork reassembly and replication reinitiation. Therefore, it is not surprising that these reinitiations require the alternative initiation-loading complex, DnaC-PriA, rather than DnaC-DnaA (18, 19). E. coli, like other bacteria and budding yeast, preferentially uses defined replication origins, although other “openable” sequences that are highly transcribed or are the sites for homologous recombination events can be used, a situation reminiscent of that in those eukaryotes that initiate replication from many different sequences. Whatever the organism, under normal circumstances the replication initiation “licensing” system has to ensure that each replication origin of a single copy chromosome fires no more than once per generation (15, 17).

Replication Fork Progression

DNA replication is highly processive, one pair of replication forks being able to replicate the whole bacterial chromosome bidirectionally from a defined origin at a rate of up to 1000 nucleotides/s. The replication mechanism is essentially identical in all branches of life, with more than 15 proteins participating in the elongation reaction, in which leading and lagging strand synthesis must be coordinated (16). Whereas E. coli uses a single polymerase to copy both leading and lagging strands, B. subtilis dedicates different polymerases for leading and lagging strand synthesis, much like eukaryotes (20). Both replication forks are associated with the same replication factory, which assembles on origins at specific cellular positions at initiation and disassembles when replication is complete (2123).

Progression of the replication fork requires the removal of positive supercoils that accumulate ahead of it (9) (Fig. 2A). Ultimately the linkage between the two strands of the DNA double helix must be reduced to zero in order for the two daughter chromosomes to be separated at the completion of replication; this requires the removal of up to 100 links/s during replication. Three of the four E. coli topoisomerases may function in this unlinking process (911, 24). DNA gyrase acts to remove positive supercoils as they accumulate ahead of the progressing replication fork, whereas topoisomerase IV can act both in separation of topologically linked chromosomes (decatenation) and in the removal of positive supercoils. If positive supercoils ahead of the replication fork diffuse behind the replication fork in vivo, they will form precatenanes in the newly replicated DNA; these may be the major substrate for decatenation. Formation of precatenanes requires rotation of the replication fork and presumably the whole replisome and may be largely restricted to the replication termination region (9, 10). Topoisomerase III can also remove precatenanes by acting at the single-stranded DNA at the replication fork (24).

Fig. 2.

Replication fork (RF) progression and restart by interconversion of RFs and HJ recombination intermediates. (A) As a replication fork progresses, the unreplicated DNA undergoes an increase in twist, which is accommodated as positive (+ve) supercoiling ahead of the fork, or as precatenanes in the replicated DNA. Precatenane formation by the diffusion backwards of (+ve) supercoils requires rotation of the RF and presumably the associated replisome. Gyrase removes (+ve) supercoils (SC) ahead of the fork, and topoisomerases IV and III are implicated in removing precatenanes. If precatenanes are not removed, they becomes catenanes in the completely replicated sister chromosomes. Accumulation of positive supercoils ahead of a fork can facilitate conversion to a HJ, in which the two newly replicated stands have annealed with each other (top). This interconversion is only likely to occur if a functional replisome has disassembled. The action of RecA, helicases and DNA topology can promote the interconversion of a RF and a HJ. (B). RF and HJ interconversion and RF reassembly. The major pathways for processing these intermediates during replication fork rebuilding and during homologous recombination are outlined. A replication fork can encounter a strand discontinuity (right side) or a block to progression (purple triangle; Embedded Image left side). A broken end can invade the unbroken duplex in a reaction promoted by RecA-like proteins; this invasion requires the prior nuclease processing of the end to produce a single-strand tail. The invasion forms a HJ and a new RF. The HJ is resolved by junction-specific nuclease action (green arrowhead;Embedded Image). A stalled RF can be converted to a HJ by several related pathways. The ensuing HJ can be resolved to produce a broken chromosome end that can be processed as above; alternatively, the free end may be degraded (Pac Man symbol; Embedded Image) (RecBCD), reforming a RF, or act as a substrate for reinvasion into the homologous sister region by homologous recombination, thereby producing a double HJ structure that can be resolved back to a RF. Lastly, the HJ may be converted back to a RF by reannealing promoted by RecA, helicases or supercoiling; if the blocking lesion had been on the leading strand and the lagging strand polymerase had replicated beyond the lesion, the HJ formation provides the opportunity for copying beyond the lesion by using the newly replicated strand as a template. At least some reassembled replication forks will use PriA-DnaC, rather than DnaA-DnaC, to reload the replicative helicase, DnaB. For further details see (19, 26).

Termination of Replication and Sister Chromosome Separation

As converging replication forks approach each other, positive supercoiling accumulates in the unreplicated DNA between them as a consequence of reduced topoisomerase access, and replication may stall. It was originally thought that this topological barrier to the completion of replication and to the separation of newly replicated chromosomes is relieved by introducing nicks into the unreplicated DNA. However, it is now believed that the positive supercoils diffuse behind the replication fork to form precatenanes, which contain the final links initially present in the two strands of the unreplicated DNA. Replication can now be completed, but the newly replicated sisters will be interlinked by catenation. Topoisomerases remove these final links by acting on the precatenanes and/or the final catenated replication products. If positive supercoiling is allowed to accumulate ahead of replication forks, they may be converted to Holliday junction (HJ) recombination intermediates, which have to be converted back to forks if replication is to be completed (25) (Fig. 2).

Although the separation of newly replicated chromosomes requires decatenation irrespective of whether they are circular or linear, the separation of circular chromosomes can also be compromised by crossing over during homologous recombination, which occurs about once every six generations in E. coli. This generates dimeric chromosomes, which need to be converted to monomers before chromosome segregation by a conserved dimer resolution machinery. This uses two related tyrosine recombinases, XerC and XerD, which act at the recombination site, dif, located in the replication termination region of circular chromosomes (26). The removal of a crossover by XerCD action at dif converts dimers to monomers in a reaction that is spatially regulated by FtsK, a large multidomain protein that coordinates cell division with chromosome segregation. The FtsK N-terminal domain, an integral membrane protein, is essential for cell division and localizes FtsK to the septum through interaction with FtsZ, the tubulin that forms the division ring at mid-cell (2). The FtsK C-terminal domain, which contains the AAA motif, acts in DNA processing events associated with the completion of chromosome segregation. In addition to dimer resolution, which requires specific interaction with XerCD-dif, these include the translocation of DNA away from the septal space and facilitation of decatenation. All of these processes require ATP hydrolysis and can be reconstituted in vitro (27).

Intriguingly, Borrelia burgdorferi, the causative agent of Lyme disease, has linear chromosomes with hairpin ends. Replication of a linear molecule generates head-to-head circular dimers, which are resolved in a site-specific breaking-joining reaction to hairpin monomers by enzymes related to XerCD (28).

Interplay Between Recombination and Replication

Although it has been known for more than 30 years that homologous recombination is intimately involved in DNA replication of bacterial viruses and, reciprocally, that DNA replication plays key roles in homologous recombination, it has only recently been recognized that recombination proteins are used universally to reassemble broken or stalled replication forks, which form during many replication cycles (18).

Chromosome breaks occur if a replication fork encounters a discontinuity in one of the template strands (Fig. 2B). Repair by homologous recombination can integrate the broken end and reform a functional replication fork. A replication fork will stall when it encounters a lesion in one or both template strands, or some other block to fork progression. Under these conditions, the processes outlined may convert the three-way fork to a four-way branched HJ recombination intermediate. The action of recombination proteins can convert such intermediates back to replication forks, with or without a complete recombination reaction or crossing over. In these situations, it is essential to be able to rebuild a functional replisome in a way that is independent of the normal replication initiation controls. It is also important that no upstream replication fork collides with the fork that is undergoing reassembly. Those bacteria that are able to continuously replicate their DNA may be restricted to homologous recombination processes to repair breaks in DNA, whereas others, which have extensive periods without DNA replication, may also use nonhomologous end-joining (NHEJ) to repair breaks (29). This situation is reminiscent of that in eukaryotes, where homologous recombination between sisters may be used preferentially to repair breaks in S-G2, whereas NHEJ is used in G1, when sister chromosomes are not available to direct homologous events.

As different types of recombination evolved, their first functions were probably DNA duplication and its vertical or horizontal transmission. Many retain such functions. Genetic transposition by both DNA and RNA retro-elements is inevitably duplicative, either as a direct consequence of the transposition process or indirectly as a consequence of gene conversion by homologous recombination at an excised locus. Site-specific recombination systems frequently function in the stable maintenance and propagation of the genetic systems that encode them. Homologous recombination during meiosis is often essential for normal chromosome segregation, whereas many eukaryote homologous recombination processes are essential for cell viability because of their role in rebuilding stalled or broken replication forks.

Chromosome Positioning and Its Relation to Segregation and Cell Division

A key element of bacterial chromosome segregation is that it occurs concomitantly with replication, with origins moving away from the division plane soon after their replication (Fig. 3A). The driving force that moves unreplicated DNA into the replication factory and newly replicated DNA away from mid-cell and toward the poles during segregation is likely to be at least partly DNA replication, although transcription and cotranscriptional translation and protein translocation into the membrane may also contribute (7, 21, 30). This active segregation of chromosomal DNA is in contrast to the passive segregation proposed in the original replicon model (3). SMCs, ParAB-like partition proteins, and negative supercoiling of newly replicated DNA may all facilitate segregation by condensing and organising the newly replicated DNA, once its new cellular locations have been taken up. Correct intracellular positioning is crucial to replication and segregation and their coordination with cell division. The molecular nature of this positioning information remains unclear, although actin- and tubulin-like proteins, and a family of ATP-hydrolyzing proteins that oscillate in the cell longitudinal axis, may be involved (3134).

Fig. 3.

Bacterial chromosome positioning and segregation. (A) Schematic of the E. coli cell cycle at moderate (40 min generation time), fast (20 min generation time), and slow-growth rates (120 min generation time), indicating the positions of origins (green circles), replication factories (blue ovoids), replication termini [red circles; based on (23, 39, 43)]. For clarity, the DNA (black lines) is not shown for the 20 min generation time. The positions of future septations are shown by the grey vertical lines (thick, thin, thinner, respectively). It is proposed that a single positioning determinant marks the positions that are sequentially occupied by newly separated origins, replication factory and terminus. Representative E. coli cells with a generation time of ∼40 min, with labeled origins (green), termini (red), and replication factory (green), and which correspond to the various cell cycle stages are also shown. Under these conditions, a FtsZ ring forms at the 1/4 positions at stages 6 and 7, as the central ring disassembles (23). FtsK is shown in orange. An S period of 40 min and G2-M of 20 min is assumed. The stages indicate points at which changes in fluorescent marker positioning occur, rather than fixed time increments. (B) Condensed nucleoids, revealed by DAPI (4′,6′-diamidino-2-phenylindole) staining, in normal cells (1), in cells defective in chromosome segregation (FtsKC-) (2), and in cells defective in cell division (FtsZts) (3).

The oscillating membrane-associated E. coli Min proteins ensure that cells divide at mid-cell rather than near the poles, although in the absence of Min cell divisions occur at mid-cell or at the poles, and not at intermediate positions (32). Therefore, additional mechanisms must determine the mid-cell position, one of which may be the ability of the nucleoid to occlude septum ring formation by FtsZ, although it is difficult to see how this would provide the observed precision of ring placement (32, 34). An actin-like cytoskeleton is responsible for maintaining cell shape in E. coli and B. subtilis (31, 33) and could function indirectly in chromosome organization and positioning, as well as in defining mid-cell; however, in E. coli, at least, MreBactin is nonessential.

Knowledge of dedicated bacterial replicon segregation processes is largely derived from studies of plasmid-encoded segregation determinants (35). These fall into two families, one of which is typified by ParM-ParR-parC of plasmid R1. ParM forms actin-like filaments whose polymerization may move sister plasmids apart when it interacts with ParR bound to its centromere-like site, parC (36). This is the nearest thing to a bacterial mitotic spindle interacting with a centromere-like sequence through an accessory protein, yet members of this family are not known to participate in bacterial chromosome segregation.

The other family, typified by ParA-ParB-parS of plasmid P1 and its homolog in plasmid F, consists of a specific DNA site, its DNA-binding protein and an additional protein that oscillates within the cell (3). Members of this family are present in many bacterial chromosomes, the best characterized of which are the Soj-Spo0J-parS determinants of B. subtilis and ParAB-parS of C. crescentus (4). Soj has been shown to oscillate between nucleoids and between the cell poles, reactions that require the partner protein, Spo0J, and its interaction with parS sequences, located proximal to the replication origin (37). These interactions lead to an apparent condensation of DNA in the region of the origins, but are not directly involved in origin positioning (38). Interactions of Soj with membrane-associated MinD, a ParA/Soj relative, are required for the polar oscillation (37). Although a functional Soj-Spo0J-parS system is nonessential, its absence leads to some chromosome segregation defects. ParAB homologs are absent in E. coli and its relatives, indicating that other mechanisms for ensuring efficient chromosome segregation exist.

Although the original replicon model proposed a specific membrane attachment of bacterial chromosomes (3), the roles of membrane in chromosome positioning and segregation remain elusive. The specific positioning of chromosomes is readily visualized when the chromosomes are condensed, and chromosome-chromosome positioning is maintained in filaments in which the septum is unable to invaginate as a consequence of a FtsZts mutation or in which septa are mispositioned with respect to nucleoids as a consequence of a mutation in the C-terminal domain of FtsK (Fig. 3B). It is likely that the specific positioning of replication origins and termini directs overall nucleoid localization. In newborn E. coli, growing at moderate growth rates, genes close to the replication origin lie close to the cell quarter positions, whereas genes in the replication terminus are close to mid-cell (23, 39, 40) (Fig. 3A). A similar situation holds in B. subtilis, although as sporulation initiates origins move to the cell poles and, consequently, the nucleoid becomes elongated in the longitudinal axis and extends between the two poles (41). Interactions between the pole-located protein DivIVA and RacA, a chromosomal binding protein, are required for this polar attachment of origins (42). In C. crescentus, the replication of a polar origin is followed by the rapid transfer of one of the sister origins to the distal pole (4).

In C. crescentus, and probably in many other bacteria, there is a mandatory separation of the DNA synthesis (S) phase from the G1- and G2-M phases (2, 4, 43), as in eukaryotes. In contrast, E. coli and other bacteria capable of fast growth rates have these phases separate only when growing slowly (43). In E. coli growing at 37°C, S phase is typically ∼40 min and G2-M is ∼20 min. The minimum generation time, which cannot be less than the G2-M, is close to 20 min. When the generation time is less than S + G2-M, two or more overlapping replication cycles are under way within the same chromosome, and partially replicated chromosomes are segregated to daughter cells (Fig. 3A). In contrast to the situation in eukaryotes, it follows that the chromosome segregation and cell division machineries in these bacteria act while chromosomes undergo replication.

The factory model for DNA replication (22), as refined in Fig. 3A, provides a mechanistic model for the positioning of origins and termini and how these influence bacterial chromosome segregation and subsequent cell division. The replication machinery assembles at precisely positioned origins, and the two newly replicated origins move away in opposite directions from the replication factory after replication initiation, with other regions following after their replication (2224, 41). In newborn E. coli cells with a single replication factory, it is invariably located close to mid-cell. If replication has already initiated, which will happen when generation time is less than S + G2-M, the origins will already have segregated to the 1/4 positions, whereas in slow-growing cells, the origin and replication factory will colocalize at mid-cell. At very fast growth rates, newborn cells will have three replication factories and four origins (Fig. 3A). Although most studies suggest origin separation soon after replication, under some conditions, newly replicated sister regions may remain associated for much of the E. coli replication cycle, consistent with Muk being able to have a cohesin-like role (44).

I propose that a key element in bacterial chromosome segregation in rod-shaped bacteria like E. coli is the process that localizes origins to the cell 1/4 positions after their replication at mid-cell (or at the 1/8, 3/8, 5/8, and 7/8 positions in fast growing cells, after their replication at the 1/4 positions) (Fig. 3A). Because the 1/4 position is where a future septum will form, the same positional information may sequentially localize newly replicated origins, a subsequent replication factory, the future septum, and the terminus. Under our experimental conditions, a new FtsZ ring forms at the 1/4 position when DNA replication is under way there, yet the termini remain close to mid-cell (23) (Fig. 3A). This model provides a mechanism for coordinating replication with chromosome segregation and cell division by ensuring that septum invagination is dependent on origins having replicated and separated away from mid-cell, on replication being completed and the replication factory disassembled, thereby restricting invagination to nucleoid-free areas.


The overall strategies for organizing DNA into chromosomes and ensuring the faithful duplication and segregation of DNA use highly conserved mechanisms, which are tuned to the biological needs of a particular organism and which extend through all branches of life. AAA proteins play key roles in the initiation of DNA replication, in DNA fork elongation, and in chromosome segregation, and SMC proteins function globally in chromosome organization. Unknown mechanisms, which may involve cytoskeletal elements and/or oscillating proteins, specify intracellular positional information that is central to the segregation process by positioning newly replicated origins away from the division plane. These are likely to help coordinate segregation with cell division.

Even though bacteria can replicate and segregate multiple replicons to daughter cells, each autonomous DNA molecule contains a unique replication origin. The replication terminus region also functions in chromosome segregation, by providing the substrate for dimer resolution, for decatenation, and for other processes mediated by FtsK, which clear the septal region of DNA before completion of cytokinesis. Regulatory processes, which signal when replication is complete and chromosomes are completely separated, may control the final steps of cell division and the next new round of DNA replication.

The presence of a single origin and terminus in bacterial replicons means that either of these regions could be functionally equivalent to eukaryotes centromeres. The repositioning of origin regions away from the division plane after their replication suggests that the origin region is the centromere equivalent. It is difficult to know whether centromeres and mitosis evolved once individual chromosomes had acquired multiple origins, or whether the evolution of mitosis allowed chromosomes to develop multiple origins. Once centromeres and the mitotic apparatus had evolved, cells could move sister chromosomes large distances apart at mitosis and thereby grow in size. Intriguingly, spirochaete Borrelia are ∼20 μm long; it will be interesting to discover whether these have eukaryote-like mechanisms for segregating their chromosomes to distant regions of their cytoplasm.

References and Notes

View Abstract

Stay Connected to Science

Navigate This Article