Читать книгу Structure and Function of the Bacterial Genome - Charles J. Dorman - Страница 20

The daughter chromosomes have to be moved to locations in the mother cell that correspond to the emerging daughter cells. At the end of the movement period, it should be possible to close the cell division septum between the nascent daughter cells without damaging the chromosome copies by guillotining them. The chromosome segregation process follows the Origin‐to‐Terminus axis, just as the replication process does (Bouet et al. 2014) (Figure 1.9).

Figure 1.9 The choreography of chromosome movement in E. coli during the cell cycle. (a) In slow‐growing E. coli cells, chromosome replication is initiated at the origin (Ori, green/yellow lozenge) located at mid‐cell, and proceeds bidirectionally, copying the left (LR, light blue) and right (RR, dark blue) replichores and ending in the terminus (red), where the products of replication are decatenated prior to segregation into the daughter cells. Arrows drawn alongside the chromosome are used to indicate the direction of replication fork movement. (b) Rapidly growing E. coli cells have multiple rounds of chromosome replication underway simultaneously. Instead of the mid‐cell position seen in slow‐growing E. coli cells, the Ori is at the cell pole in the rapidly growing bacteria. The newly born cell has its Ter region displaced towards one pole of the cell and this undergoes a transition to the mid‐cell. A second round of chromosome replication starts before the previous one is complete and multiple replication forks can be observed. The final separation of the daughter chromosomes is thought to exert a force at the terminus that moves this part of the chromosome to an eccentric position that is maintained in the daughter cell immediately after its birth (Youngren et al. 2014).

The immediate products of replication behind the moving fork and replisome are catenated, positively supercoiled, interwound, double‐stranded DNA molecules. Topo IV will attempt to decatenate these (Deibler et al. 2001; Espéli et al. 2003; Khodursky et al. 2000; Lopez et al. 2012). However, in a 300‐ to 400‐kb sliding window immediately in the wake of fork passage, the DNA is only hemimethylated. Any hemimethylated SeqA sites can bind the SeqA protein, which has the potential both to bridge DNA molecules and to exclude Topo IV, preventing decatenation (Joshi et al. 2013). The result is cohesion of the chromosome copies (Joshi et al. 2011; Nielsen et al. 2006a).

Two factors are thought to contribute to cohesion: exclusion of Topo IV by SeqA and bridging of daughter chromosomes by SeqA bound to hemimethylated copies of its binding sites. In support of the model, a positive correlation has been described between the time delay in ending cohesion and the density of SeqA‐binding sites at oriC and at two so‐called ‘snap’ loci (in the Right Replichore in E. coli) at which particularly strong physical tethers seem to fail suddenly as the replication‐and‐segregation process proceeds (Joshi et al. 2011). Chromosome cohesion must be overcome if segregation is to proceed to completion and this involves efficient decatenation of interwound chromosome copies and the breaking of any inter‐chromosome bridges. The resulting segregation process appears to include a series of successive jerks as each tether breaks in turn (Fisher et al. 2013). Dam‐mediated methylation of the newly synthesised DNA strands seems to be a key step in preventing SeqA binding and using its capacity to bridge chromosome copies and to interfere with Topo IV access to catenated DNA substrates behind the fork (Joshi et al. 2011). Cohesion does not require the Structural Maintenance of Chromosome (SMC) proteins, such as MukBEF in E. coli, despite their potential to encircle DNA duplexes (Adachi et al. 2008; Danilova et al. 2007; Joshi et al. 2013) (Figure 1.10).

Figure 1.10 MukBEF structure and function. (a) The Dimeric MukB protein (intermediate grey) consists of a head domain with ATPase activity that is attached to a hinge domain by a coiled‐coil. A ‘cap’ region in the head domain interacts with carboxyl terminus of MukF (light grey) while the central segment of MukF interacts with a MukE dimer (black). When ATP is absent, the complex has two copies of MukB, MukF and four of MukE. One MukF protein is displaced following ATP binding by MukB. This form of the MukBEF complex may in turn dimerise (not shown) (Nolivos and Sherratt 2014). (b) Clusters of MukBEF complexes, here represented by just two, have the potential to organise and segregate the ori regions of sister chromosomes following replication. The catenated ori regions are decatenated by Topo IV (dark grey), working in combination with the DNA segregation activity of the MukBEF motor.

The presence or absence of a ParAB‐parS system is a major determinant of the pattern of chromosome segregation seen in a bacterium. E. coli does not possess such a system and forces of mutual repulsion acting on the chromosome copies as they emerge in the confined space of the rod‐shaped cell may drive them to segregate, perhaps aided by the imprinted structural and super‐structural features of the chromosomes (Jun and Mulder 2006; Jun and Wright 2010; Junier et al. 2014; Pelletier et al. 2012; Wiggins et al. 2010).

ParAB‐parS systems may be useful rather than essential in bacteria that have just one chromosome, unless the bacterium is sporulating (Ireton et al. 1994) or going through a growth phase transition (Godfrin‐Estevenon et al. 2002). If the microbe has more than one chromosome, then the partitioning system is essential if the segregation of its different chromosomes is to be properly coordinated, as, for example, in the case of V. cholerae (Yamaichi et al. 2007) or members of the Burkholderias (Passot et al. 2012). The roles of chromosome‐encoded ParAB‐parS systems as functioning partitioning machines was confirmed in early work where it was demonstrated that they could replace the native plasmid par systems on single copy episomes (Godfrin‐Estevenon et al. 2002; Lin and Grossman 1998; Yamaichi and Niki 2000).

The cis‐acting parS centromere‐like sequences, and the genes that encode the ParA and ParB proteins, typically are found close to the oriC region of those bacterial chromosomes that harbour these systems (Livny et al. 2007; Reyes‐Lamothe et al. 2012). This centromere positioning ensures that the first part of the chromosome to be duplicated is going to be delivered to the appropriate cellular site for the orientation of the segregation of the rest of the chromosome. Positioning varies and can be at mid‐cell or quarter‐cell (Fogel and Waldor 2005; Webb et al. 1997); in the case of chromosomes with oriC tethering at the poles of rod‐shaped cells, it will be at the cell pole (Bowman et al. 2008; Fogel and Waldor 2006; Harms et al. 2013) (Figure 1.11). Dimeric ParB is in excess compared with its multiple 16‐bp parS binding sites and it seems that ParB can spread beyond parS, perhaps by a bridging mechanism that has the effect of folding the centromere region into a tightly organised complex (Graham et al. 2014; Sanchez et al. 2013).

Figure 1.11 Spatiotemporal dispositions of chromosomes in model bacteria other than E. coli. (a) In vegetative cells of Bacillus subtilis, the origins of chromosome replication are at mid cell in newborn cells before moving to the poles. The origins are at quarter cell positions at the onset of cell division. (b) In sporulating cells, the origins are tethered at the poles as one chromosome copy (left) is segregated into the developing fore‐spore. (c) Streptomyces species have linear chromosomes with the origin of replication at the midsection. (d) Caulobacter crescentus has two cell types: planktonic cells with a polar flagellum and sessile cells that adhere to the substratum via a polar holdfast. The origin of chromosome replication is tethered to the pole where the flagellum/holdfast is located. (e) Vibrio cholerae has two chromosomes. ChrI is the primary chromosome and its origin of replication is tethered to the cell pole where the single flagellum is located. The second chromosome, or chromid (ChrII), is plasmid‐like and synchronises its replication with that of ChrI.

The ParA protein drives bidirectional segregation of the parS‐ParB complexes in an ATP‐dependent manner. It can form filaments, and these have been proposed to be a factor in chromosome segregation (Bouet et al. 2007; Hui et al. 2010; Ptacin et al. 2010). However, it is also possible that ParA‐driven chromosome segregation works by a diffusion‐ratchet‐type mechanism that has been described for its plasmid‐encoded counterparts (Hwang et al. 2013; Vecchiarelli et al. 2014) or a trans‐nucleoid relay mechanism (Lim et al. 2014).

Ter and associated genetic loci replicate at mid‐cell and cohere for an extended period as a result of the matS‐MatP nucleoprotein complex (Dupaigne et al. 2012; Reyes‐Lamothe et al. 2008; Mercier et al. 2008; Stouf et al. 2013; Wu et al. 2019). The C‐terminal domain of MatP interacts with the ZapB protein in the cell division apparatus and this interaction contributes to the extended cohesion of the Ter domain copies (Dupaigne et al. 2012; Espéli et al. 2012). MatP is displaced from matS by the action of the FtsK motor (or SpoIIIE in B. subtilis) as it drives the Ter domain copies into the nascent daughter cells (Deghorain et al. 2011; Graham et al. 2010; Marquis et al. 2008; Massey et al. 2006; Sherratt et al. 2010). FtsK uses its dif‐oriented KOPS repeats to bind and to guide this process; the XerCD recombination dif site is the final component of the chromosome to be segregated (Stouf et al. 2013). The formation of FtsK hexamers, triggered by the onset of cell division, is a critical step in FtsK's own activation (Bisicchia et al. 2013) and its activation of the XerCD recombination apparatus (Zawadzki et al. 2013), illustrating the exquisite integration of the Ter segregation and the chromosome dimer resolution systems.