Identifying how bacterial cells find their middle: a new perspective



Bacterial cell division begins with the polymerization of the FtsZ protein to form a Z ring at the division site. This ring subsequently recruits the division machinery to allow cytokinesis. How the Z ring is positioned correctly remains a challenging question in biology and our knowledge in this area has been restricted to a few model species. Spatial regulation of division in these bacteria has been considered to be negatively controlled, with Z rings assembling in the area of least inhibition: the cell centre. An article in this issue of Molecular Microbiology reports the discovery of a new protein in Myxococcus xanthus, called PomZ (Positioning at midcell of FtsZ), that is required for the efficient recruitment of the Z ring to the division site. PomZ is a member of the Mrp/Min family of P loop ATPases that includes a diverse range of proteins involved in spatial regulation in bacteria. PomZ is the first positive regulator of Z ring positioning to be identified in vegetatively growing bacterial cells. Positive spatial regulation of division has previously been observed during sporulation in Streptomyces coelicolor and has been suggested to occur in Bacillus subtilis. Perhaps this will emerge as a common theme in the future.

All organisms must control their proliferation to ensure that progenitor cells have the correct genetic complement. This is crucial for survival of the species; and relies on exquisite regulation of the cell cycle in both time and space. Cell division, or cytokinesis, must occur at the right time and the right place to ensure equal partitioning of the DNA into newborn cells. The regulation of cell division in such seemingly simple organisms as bacteria has been relatively difficult to decipher. One of the most important questions that remain is how do bacterial cells identify the division site? While many proteins have been localized to the division site, how they get there is still one of the most intriguing questions in this field.

The earliest event in bacterial cell division is the polymerization of the highly conserved tubulin-like protein, FtsZ, to form a ring, called the Z ring, at midcell. This ring recruits about 20 proteins of the division machinery, collectively known as the divisome (Adams and Errington, 2009; de Boer, 2010). The Z ring then contracts and is accompanied by ingrowth of the cell envelope to generate the two newborn cells. Thus the Z ring is the divisome component that marks the position of the division site; and all proteins known to influence division site positioning in bacteria, do so by controlling the position of the Z ring. FtsZ is a GTPase and in the presence of GTP polymerizes to form long filaments in vitro (Mukherjee and Lutkenhaus, 1994; Erickson et al., 2010; Mingorance et al., 2010).

While the FtsZ protein is almost universally conserved throughout bacteria, the proteins that regulate it are not. Research in this field has mainly involved three well-characterized rod-shaped species, Escherichia coli, Bacillus subtilis and Caulobacter crescentus. In E. coli and B. subtilis, division site positioning has been generally believed to be controlled by the combined action of two negative regulatory systems: the Min system and nucleoid occlusion (Rothfield et al., 2005; Harry et al., 2006; Barak and Wilkinson, 2007; Wu and Errington, 2012). Together these systems prevent Z ring assembly all along the cell except at the centre, restricting Z ring assembly to this region. The Min system prevents division close to the cell poles by inhibiting the polymerization of FtsZ (Barak and Wilkinson, 2007; Lutkenhaus, 2007; Bramkamp and van Baarle, 2009). MinC prevents assembly of the Z ring through direct interaction and associates with membrane-bound MinD, a member of the Mrp/MinD family of P loop ATPases. Topological determinants, DivIVA in B. subtilis and MinE in E. coli, direct MinCD to the cell poles. Polarly localized DivIVA pilots MinCD to the poles in B. subtilis whereas in E. coli, MinE oscillates from pole to pole, redistributing MinCD towards the poles. Thus in both systems, the concentration of MinCD is highest at the poles and lowest at the centre, preventing Z ring assembly at the poles, while allowing it at the cell centre ( 1).

Figure 1.

Spatial regulation of Z ring assembly: a diversity of mechanisms. In B. subtilis and E. coli, the Min system and nucleoid occlusion (NO) influence Z ring positioning by inhibiting FtsZ assembly at inappropriate locations. The concentration of the Min proteins (blue) is highest at the cell poles, while the nucleoid (red) inhibits Z ring assembly in the central region of the cell until a nucleoid-free gap is formed through chromosome segregation. Evidence from B. subtilis suggests that a positive signal (yellow), whose activation is linked to the initiation of DNA replication, also plays a key role in directing the Z ring (green) to midcell. In C. crescentus, the FtsZ inhibitor MipZ (purple) associates non-specifically with DNA at the cell pole via its binding partner ParB. In newborn cells, MipZ is concentrated at a single pole, restricting FtsZ (green) to the opposite pole. Replication and segregation of the chromosomal origin regions sets up a bipolar MipZ gradient that dislodges FtsZ from the pole and restricts Z ring formation to midcell. In sporulating hyphae of S. coelicolor, the SsgB protein (dark blue) localizes to division sites (via SsgA) then recruits FtsZ to form the Z ring (green). SsgB remains associated with the divisome. In M. xanthus, PomZ (magenta) plays a key role in Z ring positioning. In newborn cells, PomZ exhibits a patchy localization pattern over the nucleoid. Following chromosome segregation, PomZ localizes at the cell centre and recruits FtsZ to form the Z ring (green).

Nucleoid occlusion prevents Z rings assembling over the nucleoid, or chromosome. This occlusion is relieved when the two newly replicated chromosomes segregate, allowing a Z ring to form between two chromosomes (1). Nucleoid occlusion proteins localize over the nucleoids by binding to DNA, and include the non-homologous Noc in B. subtilis (Wu and Errington, 2004; Wu et al., 2009) and SlmA in E. coli (Bernhardt and de Boer, 2005; Cho et al., 2011; Tonthat et al., 2011). It is not entirely clear how these proteins function, but they appear to work differently in their respective organisms (Wu and Errington, 2012). The existence of other DNA-binding nucleoid occlusion factors has been proposed (Wu and Errington, 2004; Bernhardt and de Boer, 2005; Bernard et al., 2010).

Interestingly, neither the Min system nor the Noc/SlmA proteins are essential in E. coli or B. subtilis, although division is much less efficient in cells without both of them (Wu and Errington, 2004; Bernhardt and de Boer, 2005). In B. subtilis Z rings are still positioned correctly at the midcell division site even in the complete absence of both Min and nucleoid occlusion, suggesting that Min and Noc do not identify the division site. Rather, they are responsible for its efficient utilization (Rodrigues and Harry, 2012). It has been proposed that a positive signal links progress of the initiation phase of DNA replication with identification of the division site at midcell in this organism (1; Moriya et al., 2010; Rodrigues and Harry, 2012).

Many bacteria do not have Min or nucleoid occlusion proteins. How do these organisms position the division site? In C. crescentus which lacks both these systems, division site positioning is also negatively regulated, but by different proteins. Z ring positioning occurs by a bipolar gradient of the FtsZ polymerization inhibitor, MipZ, also a member of the Mpr/MinD P loop family of ATPases (Thanbichler and Shapiro, 2006). However, unlike MinD which acts through MinC to inhibit Z ring assembly, MipZ acts directly on FtsZ (Thanbichler and Shapiro, 2006). The MipZ bipolar gradient relies on interaction and regulation of its ATPase activity with the chromosome partitioning protein, ParB. It is proposed that polar ParB complexes promote formation of ATP-bound MipZ dimers, which subsequently bind non-specifically to DNA at the cell poles (1; Du and Lutkenhaus, 2012; Kiekebusch et al., 2012). MipZ is conserved in all α proteobacteria lacking MinCD orthologues (Thanbichler and Shapiro, 2006). In sporulating cells of Streptomyces coelicolor, which also lacks the Min and Noc/SlmA control systems, Z ring positioning occurs via positive regulation using a completely different set of proteins, SsgA and SsgB (Willemse et al., 2011). Membrane-associated SsgB recruits FtsZ to the division site through direct interaction, presumably by promoting polymerization of FtsZ to form the ring (1). The localization of SsgB is mediated through the orthologous SsgA (Traag and van Wezel, 2008; Willemse et al., 2011). These two proteins are only present in certain Actinobacteria (Traag and van Wezel, 2008).

None of the above proteins and systems that spatially regulate Z ring assembly is universally conserved and there are bacteria that lack homologues of all of them. One such organism is Myxococcus xanthus, a rod-shaped δ-proteobacterium. In this issue, Treuner-Lange et al. (2012) report the identification of a novel protein, called PomZ (Positioning at midcell of FtsZ), that plays a role in division site positioning in M. xanthus. Cells deleted for the pomZ gene grow normally and appear able to replicate and segregate DNA, but are filamentous (elongated without forming division septa) and form minicells (DNA-less cells) resulting from misplaced septa.

PomZ shows an interesting localization pattern that changes over the course of the cell cycle (1). It localizes over the nucleoid in newborn cells, forms a focus off-centre following partial segregation of the nucleoids, then forms a transverse band at midcell at the completion of chromosome segregation, where it remains during constriction or cytokinesis. As with other bacteria, FtsZ in M. xanthus localizes as a ring at the division site at midcell between segregated chromosomes.

The authors then make the intriguing discovery that PomZ is actually required for efficient Z ring formation and for recruiting FtsZ to midcell. PomZ localizes to the division site both prior to and independently of FtsZ. The localization of PomZ to midcell and its FtsZ-recruiting role is similar to what is observed for the unrelated SsgA/B proteins in Streptomyces that localize to the division site, but is in stark contrast to the Min and MipZ/ParB systems of E. coli, B. subtilis and C. crescentus that localize to cell poles and prevent Z ring formation there by inhibiting FtsZ polymerization (Thanbichler and Shapiro, 2006; Barak and Wilkinson, 2007; Lutkenhaus, 2007; Bramkamp and van Baarle, 2009). PomZ is therefore one of very few proteins identified to date that is both able to localize to the division site in bacteria independently of FtsZ and to be required for its recruitment. The authors propose that PomZ is involved in the identification of the division site, recruiting FtsZ to this site, and in stabilizing the Z ring. Consistent with its function as a positive regulator of Z ring positioning, and unlike the Min system, M. xanthus cells that lack PomZ do not have more than one Z ring.

The simplest hypothesis for how PomZ might function to recruit FtsZ to midcell is through direct interaction. Indeed, this is supported by the ability of His-tagged PomZ to pull down FtsZ from M. xanthus cell extracts. However, purified FtsZ and PomZ did not interact with each other in vitro, suggesting involvement of another interaction partner. A most unexpected result was the lack of detectable polymerization of M. xanthus FtsZ into filaments in vitro under conditions that would normally promote filament formation of FtsZ from other bacteria. This was despite the fact that FtsZ from this organism hydrolysed GTP in a concentration-dependent and cooperative manner. The authors convincingly showed that C. crescentus FtsZ was indeed able to polymerize under the conditions used. As suggested by the authors, the M. xanthus FtsZ filaments may have been too small to be detected and this might be expected for an FtsZ whose ring formation is positively controlled.

Like MinD and MipZ, PomZ is a member of the Mrp/MinD family of P loop ATPases that self-associate and can bind ATP to form filaments (Leipe et al., 2002; Leonard et al., 2005; Gerdes et al., 2010). His-tagged PomZ purifies as an oligomer and bound ATP and GTP in vitro. However nucleotide hydrolysis was not observed. This was despite the demonstration that a mutant of PomZ, predicted from other studies with homologues to be deficient in ATP hydrolysis, did not support division. Nor did PomZ induce polymerization of FtsZ into filaments or affect its GTPase activity in vitro.

These results have led the authors to speculate that other proteins, yet to be identified, link these FtsZ and PomZ proteins directly, or are responsible for nucleotide hydrolysis, to enable FtsZ polymerization, and stabilization of the Z ring in M. xanthus. That PomZ does not act alone is consistent with the finding that overproduction of FtsZ did not rescue pomZ-deleted cells. This is in contrast to the Min system in E. coli in which overproduction of FtsZ rescues the min filamentous phenotype; and MinC directly interacts with FtsZ to inhibit its polymerization (Bi and Lutkenhaus, 1990; Huang et al., 1996; Bernhardt and de Boer, 2005). The authors propose that the nucleotide bound state of PomZ, and therefore its localization, is cell-cycle regulated. According to this model, PomZ only at the midcell location can stimulate Z-ring formation. Future challenges will be to determine how PomZ localization and activity is regulated in M. xanthus.

PomZ is the first positive regulator of Z ring positioning to be identified in vegetatively growing bacterial cells. Up until now only one example of positive regulation of division site placement had been revealed, and was proposed to be actinomycete-specific (Willemse et al., 2011). A positive regulation mechanism has also been suggested for growing B. subtilis cells. Perhaps positive spatial regulation of division in bacteria will become an emerging theme in the future.

The cell division protein FtsZ is highly conserved in bacteria. The proteins involved in the spatial control of Z ring assembly are not highly conserved but, with the exception of SsgA and SsgB which are actinomycete-specific, they share in common a member of the Mrp/MinD family of P loop ATPases (Leipe et al., 2002; Kiekebusch et al., 2012). We can now add PomZ to this list of proteins. MinD and MipZ have binding partners (MinE and ParB respectively) that regulate their ATPase activity but do so in different ways. It will be interesting to identify the putative PomZ ATPase regulator and determine the mechanism.

Several members of the MrpMinD family of P loop ATPases are involved in spatial regulation of processes in bacteria that include cell division, chromosome and plasmid segregation and segregation of cytoplasmic protein clusters (Lutkenhaus, 2007; Gerdes et al., 2010). As proposed by Treuner-Lange et al. (2012), the variation on a theme for positioning the Z ring could have evolved from a precursor ATPase acting as a spatial regulator of macromolecular localization that was subsequently adapted to regulate Z ring positioning to suit the diversity of lifestyles and shapes of different bacteria.