¡vIVA la DivIVA!

Reproduction in the bacterial kingdom predominantly occurs through binary fission—a process in which one parental cell is divided into two similarly sized daughter cells. How cell division, in conjunction with cell elongation and chromosome segregation, is orchestrated by a multitude of proteins has been an active area of research spanning the past few decades.

ABSTRACT

Reproduction in the bacterial kingdom predominantly occurs through binary fission—a process in which one parental cell is divided into two similarly sized daughter cells. How cell division, in conjunction with cell elongation and chromosome segregation, is orchestrated by a multitude of proteins has been an active area of research spanning the past few decades. Together, the monumental endeavors of multiple laboratories have identified several cell division and cell shape regulators as well as their underlying regulatory mechanisms in rod-shaped Escherichia coli and Bacillus subtilis, which serve as model organisms for Gram-negative and Gram-positive bacteria, respectively. Yet our understanding of bacterial cell division and morphology regulation is far from complete, especially in noncanonical and non-rod-shaped organisms. In this review, we focus on two proteins that are highly conserved in Gram-positive organisms, DivIVA and its homolog GpsB, and attempt to summarize the recent advances in this area of research and discuss their various roles in cell division, cell growth, and chromosome segregation in addition to their interactome and posttranslational regulation.

INTRODUCTION

Cell division in bacteria usually gives rise to two similarly sized progenies, although exceptions to this rule exist (1–3). On the basis of the presence of nutrients and/or stress signals, a multiprotein complex, called the divisome, orchestrates the division process (4). Division in bacterial cells is precisely timed by regulators tied to cell size and the status of DNA replication and segregation (5). The identity and nature of many of the factors involved in regulating cell size, shape, and division have been discovered in the Gram-negative organisms Escherichia coli and Caulobacter crescentus and in the Gram-positive organism Bacillus subtilis (4, 5). However, there are still gaps in our knowledge of these model organisms (6), and a remarkable increase in studies of nontraditional model organisms of various cell shapes and division modes has recently highlighted the presence of diverse regulators and a multitude of underpinning regulatory mechanisms (3, 4). In this review, we focus on two highly conserved proteins in Gram-positive organisms—DivIVA and a DivIVA-like-domain-containing protein, GpsB. DivIVA is a coiled-coil protein broadly conserved in Gram-positive Firmicutes and Actinobacteria and also, to a limited extent, in the phyla Cyanobacteria (7), Tenericutes (8), Deltaproteobacteria (9, 10), and Deinococcus (11, 12). Unlike DivIVA, GpsB is present only among Firmicutes and was likely acquired via divIVA gene duplication (13–15). We aim to shine light on the diverse roles performed by DivIVA and GpsB in B. subtilis, in which they were first characterized, and summarize their roles, known interaction partners, and posttranslational regulation in other organisms (Fig. 1). A list of interaction partners (including negative results) and of the methods used to test the interactions is provided in Table S1 in the supplemental material.

FIG 1.

Localization of DivIVA and GpsB and the overlapping network of their interaction partners in Gram-positive bacteria. Venn diagrams show the interaction partners of DivIVA (A) and GpsB (B) discussed in this article. (A) (Top panel) Clockwise from the top right, localization patterns of DivIVA in actively dividing cells of B. subtilis, L. monocytogenes, S. aureus, and S. pneumoniae are shown. (Bottom panel) Clockwise from the top, localization patterns of DivIVA in actively growing cells of Streptomyces, Mycobacterium, and Corynebacterium species are depicted. Dashed circles indicate alternate sites of localization. The Streptomyces cell is depicted in an open-ended manner to indicate the mycelial growth pattern. (B) Clockwise from the top right, localization patterns of GpsB in actively dividing cells of B. subtilis, L. monocytogenes, S. aureus, and S. pneumoniae are presented. The subcellular locations of DivIVA and GpsB are illustrated in green. A comprehensive list of DivIVA and GpsB partners and methods used to test interactions, including negative results, is provided in Table S1. *, the specific names of penicillin-binding proteins are shown in Table S1.

Discovery of DivIVA.

Studies of cell division mutants in B. subtilis isolated in the early 1970s (div IV-A1 and div IV-B1 [16–18]) led to the eventual mapping and discovery of the divIVA gene more than 2 decades later (19, 20). The name “DivIVA” was chosen on the basis of the recommendation that division mutants that exhibited abnormal off-center septum placement be named Div IV mutants (21). The mutation in another Div IV mutant, div IV-B1, which mapped to the minCD cluster (adjoining mreBCD genes encode cell shape-determining proteins), specifically disrupted the minD gene (22–24). MinCD, together with the help of the spatial determinant DivIVA (and of MinJ), regulate the assembly of the FtsZ ring (Z-ring), a central component of the cell division machinery required for cell division (25–27).

DIVERSE ROLES OF DivIVA IN FIRMICUTES

Bacillus subtilis.

DivIVA is capable of self-association and forms large multimeric structures (28, 29). Subsequent structural analysis showed that DivIVA is a tetramer (30). Genetic and crystallographic studies have pinpointed a key phenylalanine residue (F17) as critical for DivIVA-membrane association (30). Several other mutations that disrupt DivIVA localization and/or function have been described in many published studies; we have provided the summary of reported DivIVA (and GpsB) mutations in B. subtilis and other organisms in Table S2 in the supplemental material. In B. subtilis, DivIVA is capable of localization by sensing negative (concave) membrane curvature within the cell, presumably independently of the action of FtsZ at cell poles (31, 32), and at sites where the division septum intersects the lateral cell surface (33, 34). In fact, an earlier report indicated that DivIVA targets the sites of cytokinesis even when artificially expressed in unrelated organisms—E. coli and fission yeast Schizosaccharomyces pombe (35). It is possible that DivIVA is able to localize to nascent division sites by recognizing negative curvature elicited by the constriction of Z-rings (27, 31). Superresolution microscopy revealed that DivIVA localizes and forms two rings at midcell, likely sandwiching a constricting Z-ring (27). DivIVA localization appears to be dynamic in actively dividing cells, and transfer of molecules between the old and the nascent division sites has been observed (36). DivIVA is localized at division sites during vegetative growth and prevents aberrant assembly of Z-rings immediately adjacent to newly formed septa (27, 37–39). Furthermore, localization of DivIVA, a peripheral membrane protein, appears to depend on the presence of a translocase subunit, SecA, of the protein secretion apparatus (40).

DivIVA is a versatile protein which performs multiple roles by interacting with several partners at various points during the cell cycle and stress response (41) (Table S1A). At the onset of endospore formation (sporulation), DivIVA prelocalizes to the cell poles and interacts with RacA and the Soj/Spo0J (ParA/ParB) system to anchor the origins of the segregating chromosomes to the cell poles (42–49). DivIVA facilitates the asymmetric positioning of the septum at the onset of sporulation, where it localizes and sequesters SpoIIE, a regulatory phosphatase (50), to allow the subsequent compartment-specific differential gene expression (51). When B. subtilis cells enter a competent state, DivIVA sequesters the posttranscriptional regulator ComN (52, 53) and interacts with Maf to stall cell division (54). Additional interactions between ComN-MinJ and ComN-MinD (45) highlight the influence of the competence pathway on cell division in B. subtilis. Cells lacking divIVA exhibit impaired swarming motility, most likely due to inefficient cell division (37).

Other than the interactions described above, DivIVA was previously shown in bacterial two-hybrid screens to interact with PrkC, a division septum-localized S/T kinase (55–57), EzrA, a cell division protein (58), and GpsB (56). Although S109 of DivIVA is capable of being phosphorylated (59), based on in vitro experiments, PrkC may not be the kinase responsible (56). An arginine kinase, McsB, may phosphorylate R102 of DivIVA (60). Investigations involving phosphomimetic (R102E) and phosphoablative (R102K) variants of DivIVA suggested that the arginine in this position, irrespective of the type of mutation, is important for the role played by DivIVA in sporulation (possibly for RacA binding) but not for vegetative cell division (61). Thus, the physiological significance of DivIVA phosphorylation by an S/T kinase or McsB remains to be elucidated. A tyrosine kinase, PtkA, is able to interact with MinD, localize to cell poles in a MinD-dependent manner, and phosphorylate DivIVA (62). Another tyrosine kinase involved in the sporulation process, YabT, is able to phosphorylate RacA, while SpoIIE is able to dephosphorylate it (62)—the last two, as described earlier, are known interaction partners of DivIVA. Additionally, the DnaK chaperone, another substrate of PtkA (63), appears to protect B. subtilis DivIVA from degradation based on protein turnover analysis (64).

Streptococcus spp.

On the basis of phylogenetic analysis, coccoid cell morphology is the result of mutations in rod-shaped bacteria (65). For instance, non-rod-shaped Firmicutes typically lack the Min system; however, they have retained the B. subtilis Min system regulatory protein DivIVA (66), as is the case in the ovococcoid Streptococcus pneumoniae. Note that Streptococcus spp. lack the cell elongation factor MreB (66–68). The deletion phenotypes of genes contained within the division and cell wall (dcw) cluster region suggest that DivIVA is important for cell division, morphology, and nucleoid segregation in S. pneumoniae (69).

DivIVA localizes to the midcell and cell poles in S. pneumoniae, resembling the pattern seen in B. subtilis (70, 71). As in other organisms, pneumococcal DivIVA also directly interacts with a cohort of cell division proteins (70, 72) (Fig. 1; see also Table S1A), including GpsB and EzrA, to help regulate cell elongation and cell separation (70, 73). In Streptococcus suis, cells lacking divIVA divide along multiple planes, implying a role for DivIVA in division plane selection in this organism (74). Regarding DivIVA turnover, clues from a study in Streptococcus mutans strongly suggest the involvement of ClpXP protease machinery (75); however, how DivIVA is recognized as a substrate remains to be determined.

Chromosome segregation also plays a role in the determination of the cell division site in S. pneumoniae (76), and a prior report indicated that S. pneumoniae DivIVA interacts with Spo0J/ParB (70). In addition to exhibiting a cell separation defect (70, 73), the divIVA null strain also produces anucleate cells (70). Given that ParB and SMC (structural maintenance of chromosomes) condensin facilitate nucleoid segregation (77), pneumococcal DivIVA may play an additional role in nucleoid segregation similar to that played by B. subtilis DivIVA (44, 45); however, this possibility remains to be further evaluated. Of note, unlike certain model organisms, S. pneumoniae lacks the ParA ATPase component of the chromosome segregation machinery (78).

Streptococcal StkP, an S/T kinase, regulates cell elongation and division in multiple species, and DivIVA is one of its prominent substrates (71, 74, 79–81). Although phosphorylation of residue T201 of pneumococcal DivIVA was shown to be important for the maintenance of cell morphology (82), it appears to be dependent on the strain type (83, 84). Another substrate of StkP, MapZ/LocZ, is key for streptococcal division site selection (85–87), as originally proposed (88, 89). Finally, in another ovococcoid organism, Enterococcus faecalis, which also lacks the Min system, DivIVA is essential and plays a role in division site selection, cell separation, and chromosome segregation (90).

Other Firmicutes.

In the rod-shaped bacterium Listeria monocytogenes, the absence of DivIVA leads to a cell separation defect (91) which mimics the chaining phenotype seen in a S. pneumoniae divIVA null strain (70, 73). Defective cell separation was attributed to the dependency of autolysins p60 and MurA on DivIVA for subsequent secretion via the SecA2 pathway (91) (Table S1A). Furthermore, an L. monocytogenes strain lacking divIVA is unable to swarm (a process which is independent of MinJ, unlike the case in B. subtilis [37]) or build biofilms and exhibits attenuated virulence (91, 92). DivIVA also appears to play a role in cell division in L. monocytogenes by directly interacting with MinD, thereby circumventing the need for MinJ (93); the direct MinD-DivIVA interaction has also been documented in Clostridium spp. (94).

In the spherical bacterium Staphylococcus aureus (which also lacks min genes, similarly to other non-rod-shaped bacteria), the absence of DivIVA does not lead to significant changes in cell morphology or produce cell division defects (95), even though S. aureus DivIVA localizes to division sites (64, 95). A role for DivIVA in chromosome segregation, possibly through a direct interaction with SMC condensin, was observed previously (64). Bacterial two-hybrid and protein abundance analyses revealed that DivIVA interacts with and is stabilized by DnaK (64). Additional DivIVA-interacting proteins include the cell division proteins FtsZ, FtsA, EzrA, DivIC, penicillin-binding protein 1 (PBP1), and GpsB (64) (Fig. 1; see also Table S1A).

MULTIPLE FUNCTIONS OF DivIVA IN ACTINOBACTERIA

Streptomyces coelicolor.

In Streptomyces coelicolor, a filamentous bacterium that undergoes polar growth, DivIVA is an essential protein and localizes to the tips of the hyphae and sites of branching and plays a role in promoting peptidoglycan synthesis at cell poles (96, 97). Time-lapse microscopy showed that hyphal-tip-localized DivIVA could detach to mark future branching sites (98). The site of the germ tube (vegetative cell) emerging from germinating spores is also marked by DivIVA (96). As observed in some Firmicutes, S. coelicolor DivIVA is also phosphorylated by at least one of the 34 S/T kinases encoded by this organism, AfsK, and the colocalization of DivIVA and AfsK has been reported previously (99). DivIVA of S. coelicolor interacts with other polar-localized intermediate filament-like proteins, Scy and FilP (100, 101) (see Table S1B in the supplemental material). In Streptomyces venezuelae, FilP displays enhanced localization at regions proximal to the tips of growing hyphae (which is similar to S. coelicolor [101]), and lack of filP results in an irregular growth pattern that is likely a consequence of the presence of differing levels of DivIVA accumulation at poles (102), suggesting a crucial role for FilP and DivIVA in hyphal growth. Consistent with the chromosome tethering/segregation functions described in Firmicutes, DivIVA appears to be involved in anchoring the origin of the linear S. coelicolor chromosomes via ParA indirectly (103, 104). An interaction between S. coelicolor DivIVA and ParB proteins was observed in E. coli and by the use of multiple in vitro experiments (105); however, this interaction was not confirmed in its natural host (103).

Corynebacterium glutamicum.

DivIVA is essential in Corynebacterium glutamicum (previously referred to as Brevibacterium lactofermentum), and its overproduction leads to polar bulging and other morphological defects (106). DivIVA localizes to the cell poles and division sites subsequent to septation in this bacterium (106, 107). To support robust polar growth, DivIVA recruits RodA, a lipid II flippase, to the cell poles (108, 109) (Table S1B). Thus, in Corynebacterium, which does not encode actin-like cytoskeletal proteins (110, 111), DivIVA appears to play a major role in cell shape regulation. As in other organisms, chromosome segregation and cell division are intertwined in C. glutamicum (112, 113). In this organism, the origins of the replicated chromosomes are tethered to cell poles via the ParAB machinery (114). While absence of ParA, PldP (a ParA-like protein), or ParB results in the production of minicells lacking nucleoids, overexpression of ParA or PldP leads to increased cell length (114). In fact, a direct interaction between FtsZ and ParB was noted through bacterial two-hybrid assay (114). At least when artificially expressed in E. coli, DivIVA and ParB of C. glutamicum interact with each other (105). However, the physiological function of DivIVA in chromosome segregation in this organism remains to be elucidated.

Although there is no evidence for the phosphorylation of DivIVA in this organism (109, 111, 115), there are reports that indicate a role for S/T kinases in cell division and growth (116–118). Interestingly, a coiled-coil domain containing cell shape morphogenetic protein RsmP (which is overexpressed upon DivIVA depletion [115]) is phosphorylated by at least one of the 4 S/T kinases encoded by C. glutamicum (117), namely, PknA, and the phosphorylation leads to its accumulation at cell poles (115). How RsmP influences peptidoglycan synthesis at poles and whether RsmP works together with DivIVA for this purpose need to be evaluated.

Mycobacterium spp.

DivIVA, also known as Wag31, was initially characterized as the highly immunogenic antigen Ag84 in Mycobacterium tuberculosis and Mycobacterium leprae strains (119, 120). In Mycobacterium species (which, unlike Firmicutes, undergo apical growth [121]), DivIVA is essential (122–124) and localizes to the septum, cell poles, and branching sites (124–127). Overproduction of DivIVA leads to polar bulging and branching in Mycobacterium smegmatis (125), a fast-growing cousin of M. tuberculosis, where DivIVA dictates the site of peptidoglycan insertion (128, 129). An indirect link between DivIVA and Z-ring assembly was revealed by the elucidation of direct interactions between DivIVA and a cell wall synthesis protein (CwsA) and FtsZ-interacting FtsI/PBP3 (130–133) (Table S1B). Localization of DivIVA depends on Ami1 amidase, as there was a higher rate of unusual lateral budding in an Ami1 amidase mutant due to DivIVA mislocalization to the sites of lateral branching (134), which resembles the localization pattern of DivIVA in Streptomyces. In addition to assisting in cell wall synthesis, DivIVA also interacts with proteins involved in the biosynthesis of mycolic acids (128, 135, 136). Mycolic acids are an integral part of the mycomembrane and play a crucial role in pathogenesis (137).

A role for S/T kinases and cognate phosphatases in regulating cell shape and division in M. tuberculosis has been observed previously (138, 139). Of the 11 S/T kinases encoded by this organism (140), at least one of them, PknA, phosphorylates DivIVA at T73 (138, 141). It was also elucidated that a phosphomimetic variant of DivIVA displays an increased rate of peptidoglycan synthesis compared to the phosphoablative mutant (126). Furthermore, the phosphorylation status of DivIVA is also regulated by the essential MtrAB two-component system and FtsI (133). Similarly to DivIVA in other organisms, M. smegmatis DivIVA also interacts with the chromosome segregation machinery—in this case, with the ParA ATPase component (142) and possibly with ParB as well (105). In M. smegmatis, which lacks the Min system, a paralog of DivIVA, SepIVA, is involved in cell division regulation (143). Apart from the functions described above, there are additional roles for DivIVA in oxidative stress response (130) and bacteriophage infection (144).

PROCESSES INFLUENCED BY GpsB

Bacillus subtilis.

In 2008, two independent groups explored the possible roles of B. subtilis GpsB (formerly YpsB), a DivIVA-like protein conserved in Firmicutes (14, 15). A synthetic-lethal screen to identify genes that become essential in cells that lack a nonessential cell division/elongation factor, EzrA, led to the discovery of GpsB, and its role in guiding PBP1 (a bifunctional penicillin-binding protein) shuttling from cell poles to division sites to regulate cell wall synthesis was uncovered (15). Bacterial two-hybrid analysis revealed the interactions between GpsB and MreC, a known cell shape regulator (145), EzrA, and PBP1 (15) (see Table S1C in the supplemental material). B. subtilis GpsB was also independently investigated by another group, and their data showed that GpsB localizes to division sites and becomes synthetically essential in the absence of FtsA, an FtsZ-anchoring protein (14). Although GpsB becomes important in the absence of EzrA or FtsA, both of which directly interact with FtsZ, it is believed that the predominant role of GpsB is in regulating peptidoglycan synthesis (58).

B. subtilis GpsB is a hexamer (146), and further structural analysis of GpsB and its binding partner PBP1 uncovered the molecular basis of their interaction (147). On the basis of a signature sequence motif, GpsB was discovered to interact with two lesser-known nonessential membrane proteins, YpbE and YrrS, with links to peptidoglycan binding and/or synthesis (147). Similarly to what was observed for DivIVA (51), GpsB also localizes to asymmetric division sites during sporulation, at least when produced under the control of an inducible system (14). The precise function of GpsB during sporulation, which may be mediated by YrrS activity, remains to be investigated (147). However, sporulation progresses in an unaffected manner in the absence of GpsB (14).

Similarly to its homolog, DivIVA, links have been established between the S/T kinase PrkC and GpsB. Phosphorylation of GpsB by PrkC apparently occurs at residue T75, and the unphosphorylated form of GpsB enhances the autophosphorylation activity of PrkC and its kinase activity toward its other substrates (56). Additionally, the same study showed that GpsB, regardless of the status of phosphorylation, interacts with PrkC, EzrA, and DivIVA (Table S1C). Thus, it appears that GpsB, via PrkC, is the main driver connecting cell division and other processes (57) and that PrpC, the cognate phosphatase of PrkC, could reset this process through dephosphorylating GpsB during different stages of the cell cycle (56).

Listeria monocytogenes.

The Halbedel group was the first to report that GpsB is dispensable in L. monocytogenes under conditions of growth at 30°C and 37°C (146). However, at 37°C, cell lysis and increases in cell width and length were observed in cells lacking gpsB. At 42°C, GpsB was deemed to be essential for growth (146). Roles for GpsB in virulence (146) and in lysozyme resistance for immune evasion (148) have also been noted. Deletion of both gpsB and divIVA leads to cell elongation and decreased cell width (146). In fact, similarly to the GpsB counterparts in B. subtilis and S. pneumoniae, listerial GpsB also binds a class A bifunctional PBP (PBP A1) and, based on genetic analysis, GpsB appears to negatively regulate the activity of PBP A1 (146, 147). Isolation of suppressor mutants that were able to overcome the growth defect at 42°C appears to have confirmed the GpsB-mediated regulation of PBP A1 (149). Full-length GpsB of L. monocytogenes functions as a hexamer (146), and this hexameric structure could allow GpsB to bind multiple PBP1-like molecules (150). Apart from the well-studied interaction with PBP A1, listerial GpsB is also capable of binding multiple partners, including the known FtsZ-tethering proteins ZapA, SepF, and EzrA; cell shape regulators MreC and MreBH; late-arriving divisomal proteins DivIB and DivIC; and several penicillin-binding proteins—PBP A2, PBP B1, PBP B2, and PBP B3 (147) (Table S1C). As seen with the other organisms described here, the S/T kinase PrkA is important for maintaining cell shape and virulence in L. monocytogenes (151). Whether GpsB interacts with or is phosphorylated by PrkA remains to be determined.

Streptococcus pneumoniae.

GpsB is essential in S. pneumoniae, but accumulation of suppressor mutations that allow GpsB to become nonessential in some strain backgrounds has been clearly delineated (73, 84). Depletion of GpsB results in cell elongation and formation of constriction-deficient Z-rings (152). In a different strain background, deletion of gpsB resulted in cell elongation because of improper helical localization of division proteins and concomitant peptidoglycan insertion (73). This phenotype appears to be corrected in cells in which divIVA is also absent, highlighting that coordination of cell wall synthesis in pneumococcus requires both DivIVA and GpsB (73). However, the corrective effect was not reproducible in other strain types (84).

The interaction between GpsB and PBP2a has been established by genetic, biochemical, and structural methods, and the residues involved in the interaction between GpsB and PBP2a are similar to those of B. subtilis GpsB and PBP1 (84, 147). In addition, GpsB was found to be in a complex with EzrA, MreC, StkP, and multiple penicillin-binding proteins: PBP1a, PBP2b, and PBP2x (84, 147) (Table S1C). The interaction between GpsB and PBP2b observed via coimmunoprecipitation analysis may not be direct, as the interaction was not seen in bacterial two-hybrid or fluorescence polarization assays (147). Together, multiple lines of evidence indicate that GpsB plays a vital role in linking peptidoglycan synthesis at division sites for septation and cell periphery for cell elongation (84).

The S/T kinase StkP plays a critical role in cell wall synthesis and virulence in multiple Streptococcus species (153–155). Known StkP substrates include GpsB, DivIVA, FtsA, EzrA, MapZ, and FtsZ (153, 155). Bacterial two-hybrid and surface plasmon resonance experiments revealed that pneumococcal GpsB interacts with both EzrA and DivIVA and that GpsB is required for proper localization and autophosphorylation of StkP (73). In a different pneumococcal strain, however, the interaction between GpsB and DivIVA (based on coimmunoprecipitation assay) and GpsB-mediated localization of StkP were not evident (84). StkP autophosphorylation and phosphorylation of its substrates, on the other hand, are dependent on GpsB in multiple strain backgrounds (84).

Staphylococcus aureus.

A genome-wide transposon mutagenesis experiment indicated that transposon insertion was not tolerated at certain parts of gpsB (corresponding to the region coding for DivIVA-like N-terminal domain); therefore, it was determined that GpsB is a domain-essential protein in S. aureus (156). Bacterial two-hybrid analysis led to the finding that S. aureus GpsB self-interacts and also interacts with EzrA (157), similarly to the GpsB of other organisms discussed here (Table S1C). Our group showed that overproduction of GpsB results in cell enlargement in S. aureus (158), indicative of cell division inhibition and improper incorporation of peptidoglycan along the cell periphery (159). Using superresolution microscopy, we observed that GpsB preferentially localized to the leading edge of the invaginating membrane during septation (158). It appears that the localization of FtsZ and GpsB are codependent in S. aureus. Furthermore, using multiple in vitro analyses, we noted that GpsB transiently bundles FtsZ polymers and subsequently enhances the GTPase activity of FtsZ to promote its depolymerization. We speculate that this function of GpsB may be important for kick-starting FtsZ treadmilling—a process in which the concerted action of Z-ring constriction and peptidoglycan synthesis allows the creation of division septum (160–162). This role of GpsB may be broadly conserved in other bacteria, albeit with the help of FtsZ-tethering proteins such as EzrA (163). Given that teichoic acid (TA) synthesis impairment significantly affects division site positioning in staphylococcal cell division (164) and that the EzrA interacts with many of the key TA synthesis machinery components (165), it is likely that EzrA, together with its interaction partners, acts as a transducer of signals between multiple processes such as cell wall/TA synthesis and cell division (157, 166).

In S. aureus, the S/T kinase, Stk (also known as PknB and Stk1), is targeted to the division septum (167) and, together with its cognate phosphatase, Stp, plays a role in the regulation of peptidoglycan and capsule biosynthesis (168–170). Although a direct interaction between Stk and GpsB is absent, a prominent interaction between EzrA and Stk (and also between EzrA and other binding partners of Stk) has been noted (169). Nevertheless, GpsB of S. aureus is phosphorylated at multiple S and T residues as well as at least one R residue, including the following: T21, T75, S76, R77, S85, T93, and S105 (171). It is therefore conceivable that the phosphorylation status of GpsB in S. aureus may be used to regulate and hook Z-ring constriction and the peptidoglycan synthesis machinery either directly or indirectly via EzrA-like protein partners.

CONCLUDING THOUGHTS

Studies over the past 2 decades have uncovered multiple critical roles played by DivIVA and GpsB in Gram-positive bacteria. Although their function in peptidoglycan synthesis regulation is well documented, much remains to be elucidated in regard to other processes.

(i) Chromosome segregation.

Although the link between DivIVA and DNA segregation is relatively well understood in B. subtilis (45), how DivIVA affects DNA segregation in other organisms needs further clarification. Apart from the organisms that belong to the Firmicutes and Actinobacteria phyla, interactions between DivIVA and chromosome segregation machinery components have also been in noted in Deinococcus radiodurans (172). Identification of new factors that link cell division and chromosome segregation may shed more light on the interplay between these two processes (173).

(ii) Protein secretion.

It would be interesting to see if the link between DivIVA and SecA-mediated protein secretion that is seen in B. subtilis and L. monocytogenes (40, 91) is also conserved in other species.

(iii) Regulation of FtsZ.

Although the role of GpsB in regulating peptidoglycan incorporation became increasingly clear in recent years, whether the function of GpsB in FtsZ regulation that was identified previously in S. aureus (158) is also conserved in other organisms remains to be investigated. Given the conserved nature of the binding partners of GpsB, it appears that this function may operate in an indirect manner via EzrA (FtsA, SepF, and/or ZapA). It is also possible that GpsB (together with other proteins) acts as an anchor to link the Z-ring and septal cell wall biosynthesis machinery for FtsZ treadmilling during septation.

(iv) Posttranslational regulation.

While the emerging trend suggests that DivIVA and GpsB utilize S/T kinase-mediated phosphorylation, or vice versa, to direct cell wall synthesis at appropriate sites in multiple organisms, more biochemical and structural studies are needed to fully appreciate their precise mechanisms of action and modes of regulation. In B. subtilis, several lines of evidence point to a role for PrkC phosphorylation-mediated regulation of lipoteichoic acid synthesis, carbon metabolism, stress response, and cell wall synthesis (174–178). Given that GpsB regulates the phosphorylation status of PrkC in B. subtilis and StkP in S. pneumoniae (56, 57, 73), it is possible that GpsB performs a larger role in linking multiple processes by modulating the S/T kinase activity. Indeed, this may be the case in S. pneumoniae, as a mutation that disrupts the phpP phosphatase was shown previously to render gpsB nonessential (84). The significance of arginine phosphorylation in DivIVA/GpsB (60, 171) and the role of tyrosine kinases in chromosome segregation and cell division (62, 179) warrant further investigation.

(v) Future directions.

Further research to reveal and validate the complete set of interaction partners of DivIVA and GpsB during various stages of cell cycle and under different types of stress conditions would provide us a better understanding of the versatility of these two proteins. Solving the multicomplex structure(s) of GpsB, DivIVA, EzrA, FtsA, FtsZ, S/T kinase, PBP(s), and other protein partners involved in division/growth would give us the critically needed molecular details of the protein-protein interactions. Similar structures were recently reported for lipopolysaccharide export in E. coli by the use of advanced structural biology techniques (180, 181).

Given the ominous rise in antibiotic resistance, particularly among Gram-positive pathogens (182, 183), more in-depth understanding of the regulation of cell division and cell wall synthesis is urgently needed to reveal novel drug targets. Thus, in addition to enhancing our fundamental comprehension of basic bacterial processes, knowledge of the key players involved in cell growth and division would be immensely beneficial for the targeted development of new lines of antibiotics and antibiotic adjuvants.

ADDENDUM

While the manuscript was in preparation, a complementary review article discussing the structural features of DivIVA and GpsB and the nature of their interaction with multiple partners was published (184).