* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download SepF, a novel FtsZ-interacting protein required for a late step in cell
Survey
Document related concepts
Cell nucleus wikipedia , lookup
Cell membrane wikipedia , lookup
Extracellular matrix wikipedia , lookup
Cell encapsulation wikipedia , lookup
Protein moonlighting wikipedia , lookup
Cell culture wikipedia , lookup
Organ-on-a-chip wikipedia , lookup
Magnesium transporter wikipedia , lookup
Cellular differentiation wikipedia , lookup
Cell growth wikipedia , lookup
Endomembrane system wikipedia , lookup
Signal transduction wikipedia , lookup
Cytokinesis wikipedia , lookup
Transcript
Blackwell Science, LtdOxford, UKMMIMolecular Microbiology0950-382X© 2005 The Authors; Journal compilation © 2005 Blackwell Publishing Ltd? 2005593989999Original ArticleSepF, a novel FtsZ-interacting proteinL. W. Hamoen et al. Molecular Microbiology (2006) 59(3), 989–999 doi:10.1111/j.1365-2958.2005.04987.x First published online 30 November 2005 SepF, a novel FtsZ-interacting protein required for a late step in cell division Leendert W. Hamoen,1* Jean-Christophe Meile,1,2 Wouter de Jong,1† Philippe Noirot2 and Jeff Errington1 1 Sir William Dunn School of Pathology, University of Oxford, South Parks Road, Oxford, OX1 3RE, UK. 2 Laboratoire de Génétique Microbienne, INRA, domaine de Vilvert, 78352 Jouy en Josas Cedex, France. Summary Cell division in nearly all bacteria is initiated by polymerization of the conserved tubulin-like protein FtsZ into a ring-like structure at midcell. This Z-ring functions as a scaffold for a group of conserved proteins that execute the synthesis of the division septum (the divisome). Here we describe the identification of a new cell division protein in Bacillus subtilis. This protein is conserved in Gram positive bacteria, and because it has a role in septum development, we termed it SepF. sepF mutants are viable but have a cell division defect, in which septa are formed slowly and with a severely abnormal morphology. Yeast two-hybrid analysis showed that SepF can interact with itself and with FtsZ. Accordingly, fluorescence microscopy showed that SepF accumulates at the site of cell division, and this localization depends on the presence of FtsZ. Combination of mutations in sepF and ezrA, encoding another Z-ring interacting protein, had a synthetic lethal division effect. We conclude that SepF is a new member of the Gram positive divisome, required for proper execution of septum synthesis. Introduction Cell division in nearly all bacteria is initiated by the polymerization of the conserved tubulin-like protein FtsZ into a ring-like structure at midcell. This so-called Z-ring functions as a scaffold for the proteins that execute the synthesis of the division septum. The assembly of this complex and dynamic protein structure, also referred to as the divisome, has been thoroughly investigated in the model organisms Escherichia coli and Bacillus subtilis Accepted 1 November, 2005. *For correspondence. E-mail [email protected]; Tel. (+44) 1865 275 562; Fax (+44) 1865 275 556. †Present address: Groningen Biomolecular Sciences and Biotechnology Institute, University of Groningen, Kerklaan 30, 9751 NN Haren, the Netherlands. © 2005 The Authors Journal compilation © 2005 Blackwell Publishing Ltd (reviewed in Errington et al., 2003). Most of the components of the divisome are conserved; however, there are distinct differences between the mechanisms of divisome assembly in B. subtilis and E. coli. It is likely that this reflects structural differences between the cell-envelopes of Gram positive and Gram negative bacteria. We will focus here on cell division in B. subtilis. The divisome complex spans the cytoplasmic membrane. At the core of the cytoplasmic site of the divisome lies the Z-ring, assembly of which is supported by at least two other cytosolic proteins. Most important is the conserved actin-like division protein FtsA. This protein is essential and interacts with the C-terminus of FtsZ (Erickson, 2001). It has recently been shown that the conserved C-terminus of FtsA contains an amphipathic helix that is essential for targeting of the protein to the membrane. This led to the suggestion that FtsA serves as the principal membrane anchor for the Z-ring (Pichoff and Lutkenhaus, 2005). In addition, there is a small 85-amino-acid cytoplasmic protein, ZapA, which promotes polymerization of FtsZ. However, a deletion of zapA showed no apparent phenotype in B. subtilis or E. coli (Gueiros-Filho and Losick, 2002; Johnson et al., 2004). EzrA is another conserved protein that binds to the Z-ring (Levin et al., 1999). This large protein has a single N-terminal transmembrane domain, but this domain is not required for interaction with FtsZ (Haeusser et al., 2004). The role of EzrA in the activity of the divisome is not entirely clear. An ezrA mutant is viable, yet it tends to assemble multiple Z-rings both at polar and midcell positions. It is assumed that this protein is a negative regulator of Z-ring formation (Levin et al., 1999). At least two of the conserved B. subtilis divisome proteins are involved in peptidoglycan synthesis: the penicillin-binding protein Pbp2B and the integral membrane protein FtsW (R.A. Daniel and J. Errington, unpublished; Daniel et al., 2000; Aarsman et al., 2005). The 10 membrane spanning domains of FtsW suggest a possible role in transport, and it has been speculated that FtsW translocates the lipid-linked precursor for the septal peptidoglycan matrix (Holtje, 1998). The periplasmic part of the divisome comprises, aside from Pbp2B, three other conserved cell division proteins: FtsL, DivIB and DivIC. Like Pbp2B, these proteins are attached to the membrane by a single transmembrane domain. Not much is known about the function of these three proteins. They seem to 990 L. W. Hamoen et al. be intrinsically unstable and biochemical data suggest that they form heteromultimers that ensure their stability (Sievers and Errington, 2000; Robson et al., 2002). Therefore, it has been proposed that these proteins may fulfil a regulatory role in divisome assembly and/or disassembly. The divisomes of B. subtilis and E. coli appear to be comparable in composition (Errington et al., 2003). However, the E. coli divisome comprises several more components, such as FtsK, FtsE, FtsQ, FtsN and AmiC (Aarsman et al., 2005). This complexity could be a consequence of the fact that in E. coli cytokinesis requires constriction of outer membrane, as well as inner membrane, and peptidoglycan layers. An intriguing difference between B. subtilis and E. coli is the way their divisomes assemble. In E. coli, the recruitment of the different components seems to occur in an ordered manner, whereby the depletion of one cell division protein leads to the absence of the downstream proteins in the interaction pathway [the order being: FtsZ (FtsA, ZipA, ZapA), FtsK, FtsQ (FtsL, FtsB), FtsW, FtsI, FtsN and AmiC] (Aarsman et al., 2005; Goehring et al., 2005). This linearity of recruitment is not observed in B. subtilis. The assembly of the B. subtilis divisome seems to depend on the presence of all the essential division proteins, and depletion of FtsA, DivIC, FtsL or Pbp2B, abolishes the positioning of the other cell division proteins at midcell. Only the polymerization of the Z-ring does not require the other division proteins (Errington et al., 2003). Although much is known about the assembly of the divisome, the actual mechanism of constriction remains unclear. The homology of FtsZ and FtsA with tubulin and actin, respectively, makes it plausible that these proteins lead the constriction of the divisome. However, whether they are sufficient to drive this process, or whether the synthesis of the peptidoglycan provides the ultimate force for constriction, remains to be established (Bramhill, 1997; Errington et al., 2003). In this paper we describe the identification of a new FtsZ-interacting cell division protein, which we designated SepF. Combination of a sepF and ezrA mutation results in a synthetic-lethal division defect, suggesting that one or the other of these proteins is necessary for proper functioning of the Z-ring. However, sepF single mutants affect only the later stages of septum constriction indicating that SepF is important throughout the process of division. While this work was in progress, we learned that Ogasawara and coworkers had obtained similar results implicating SepF in cell division of B. subtilis. Results Deletion of the ylmB-H region has a mild cell division defect In rod-shaped bacteria the conserved protein couple MinC/MinD prevents aberrant Z-ring assembly close to the poles of the cell. Mutations in these proteins result in polar division and anucleate minicells. MinC is the actual inhibitor of FtsZ assembly but requires the membrane-bound ATPase MinD for its activity (Hu et al., 1999; Marston and Errington, 1999). In B. subtilis the activity of this protein couple is confined to the cell poles due to interactions with the polar localized protein DivIVA (Edwards and Errington, 1997). How DivIVA accumulates at the poles is unknown. In many Gram positive bacteria the conserved ylm genes are found upstream of divIVA (Fig. S1). Several of the ylm genes have been disrupted previously (Kobayashi et al., 2003) (see also, http://genome.jouy.inra.fr/cgi-bin/micado/ index.cgi). However, none of the mutants were reported to have a phenotype. The genes ylmE and ylmH were not tested, and to assess a potential role of the ylm genes in DivIVA localization, we replaced the complete ylmB-H region with an antibiotic resistance marker (hereafter referred to as ylm mutant, Fig. 1). Light microscopic observations of the resulting strain (B. subtilis 3317) seem to indicate a normal wild-type phenotype, although the cells appeared longer. To analyse this in more detail, we measured cell length of an ylm mutant and wild-type culture. As shown in Fig. 1C and D, the difference was modest during exponential growth, with an average length of mutant cells approximately 20% longer than wild-type cells (4.6 and 5.5 µm for wild-type and ylm mutant respectively). In the stationary phase this difference increased to about 40% (2.8 and 4.0 µm for wild-type and ylm mutant respectively). It was clear that the ylm genes were not required for DivIVA localization because a normal polar and septal fluorescent pattern was observed when the ylm deletion was introduced into a strain carrying a divIVA-gfp fusion (Fig. 1F). However, there was a clear difference in spacing of green fluorescent protein (GFP) bands, which reflects the increase in cell length of the mutant. Disruption of the ylm locus has a synthetic lethal effect when combined with ezrA It seemed possible that the mild mutant phenotype produced by the ylm deletion might be exacerbated when combined with other mutations that impair cell division. To test whether the ylm mutant had increased sensitivity to FtsZ levels, the ylm deletion was introduced into cells in which ftsZ expression is under control of the IPTG-inducible Pspac promoter. Although low IPTG concentrations resulted in elongated cells, the presence of an ylm deletion did not substantially aggravated this phenotype. Various other mutations in genes that regulate FtsZ polymerization, such as minC and zapA, were combined with an ylm deletion, but the effects on cell length were again marginal (data not shown). In contrast, the number of transformants was much lower than expected when the ylm mutant was transformed with chromosomal DNA from © 2005 The Authors Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 989–999 SepF, a novel FtsZ-interacting protein 991 2 kb A ylmA -G -B -C -D -E -F -H sigE sigG divIVA B ylmA ileS kmR ylm mutant sigE sigG divIVA Fig. 1. Phenotype of an ylm mutant. The organization of the ylm locus and adjacent genes is shown in the upper panel (A). In the ylm mutant (strain 3317) the ylmB-H region has been replaced by a kanamycin-resistance marker (kmR) (B). The positions of putative terminator sequences are indicated. Graphs C and D show cell length measurements of a wild-type culture and an ylm mutant culture. Samples were taken during logarithmic growth (C), and after 2 h in stationary phase (D). The numbers of cells measured (n) are indicated, and cell lengths are given in µm. The fluorescence microscopy pictures show DivIVA-GFP expression in wild-type (E) and ylm mutant cells (F). Scale bar indicates 5 µm. ileS Deletion of the ylm locus leads to impaired septa 25 ylm (n = 273) C frequency (%) 20 wild type (n = 260) 15 10 5 < 1-1 1 1.5 .5 2.0 2.0 -2. 2.5 5 3.0 3.0 -3. 5 3.5 4-4 4 . 4.5 5 -5 5-5 .5 5.5 6-6 6 .5 6.5 -7 7-7 . 7.5 5 -8 8-8 . 8.5 5 -9 9-9 . 9.5 5 > 0 cell length 35 ylm (n = 265) frequency (%) 30 wild type (n = 282) D 25 20 In view of the possible deficiency in cell division resulting from deletion of the ylm locus, the ultrastructure of division septa was analysed using electron microscopy (EM). It appeared that in ylm mutant cells septa showed strong deformations (Fig. 2). This was most apparent in the early stages of septum synthesis, as shown in Fig. 2D and E. Division seemed to proceed normally when septation has been completed, although distortions in matured septa were still observed (Fig. 2F). These distortions did not seem to lead to deformed cell poles (data not shown). The ylmD-H genes comprise an operon Inspection of the DNA sequence of the conserved ylm region revealed only a single large space likely to accom- 15 wt 10 5 <1 1-1 1.5 .5 -2 2.0 .0 -2. 2.5 5 -3 3.0 .0 -3. 5 3.5 -4 4-4 .5 4.5 -5 5-5 .5 5.5 -6 6-6 .5 6.5 -7 7-7 .5 7.5 -8 8-8 .5 8.5 -9 9-9 .5 9.5 > 0 cell length E wt F Dylm A B C E F H I D D ylmF G an ezrA mutant. The few colonies that emerged appeared to have lost the kanamycin-resistance marker, and therefore had been co-transformed with the wild-type ylm locus from the donor strain. This finding supported the view that one or more of the ylm genes plays a role in cell division. Fig. 2. Transmission electron microscopy (TEM) images of division septa in wild-type cells (A–C), in the ylm mutant (B. subtilis 3317) (D– F) and in the ylmF:pMUTIN4 integration mutant (B. subtilis BFA2863) grown in the presence of 1 mM IPTG (G–I). © 2005 The Authors Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 989–999 992 L. W. Hamoen et al. ylmD wt mut ylmG wt mut ylmH wt mut divIVA wt mut Fig. 3. Northern analysis of the ylmD-divIVA region. Total RNA from wild-type (wt), and a B. subtilis mutant (mut) that contains the Pspac promoter immediately upstream of the ylmD open reading frame (B. subtilis 4055), was hybridized with radioactively labelled probes directed against ylmD, ylmG, ylmH and divIVA respectively. B. subtilis 4055 was grown in the absence of IPTG. modate a promoter: a 165 bp non-coding region between ylmC and ylmD. To test whether this region contains a promoter, and simultaneously, to examine whether ylmB and/or ylmC contributed to the cell division defect, two deletion mutants were constructed: strains 4054 and 4055. In strain 4055, ylmB, and ylmC including the 165 bp non-coding region, were replaced with the IPTG-inducible Pspac promoter. Strain 4054 was similar except that the 165 bp potential promoter region remained in position upstream of ylmD. When an ezrA mutation was introduced into these strains, the resulting transformants grew normally in the presence of IPTG. However, transformants from strain 4055 became elongated and eventually lysed when IPTG was omitted, whereas those of 4054 retained normal cell length and viability (data not shown). These results showed that loss of ylmB and ylmC does not give a division phenotype, and strongly suggested that the gene or genes responsible for the division effect lay downstream of the 165 bp region and required transcription from that promoter (ylmD promoter). To test which of the genes downstream of the 165 bp region were transcribed from the putative promoter, RNA was extracted from wild-type B. subtilis, and strain 4055 (grown in the absence of IPTG). As shown in Fig. 3, probes directed against ylmD, ylmG and ylmH gave a clear signal when RNA from wild-type cells was used, but almost no signal when RNA from the mutant was used. This suggests that there is a single polycistronic mRNA, and that the promoter for this mRNA is absent from strain 4055. In contrast, a probe against the downstreamlocated gene divIVA hybridized only to a faster migrating band that was present in both strains. The latter supported previous findings that divIVA has its own promoter (Edwards and Errington, 1997). ylmF (sepF) is responsible for the division defect To determine which of the ylm genes was responsible for the effects on cell division, single gene disruptions were constructed. To reduce polar effects, the integration vector pMUTIN4 (Vagner et al., 1998), which places downstream-located genes under control of the Pspac promoter, was inserted by single (Campbell type) crossover. ylmG proved to be too small for efficient homologous recombination and was deleted by double crossover. Chromosomal DNA from the various ylm mutants (listed in Table 1) was used to transform an ezrA mutant, and transformants were selected on plates containing 1 mM IPTG. Normal transformation percentages were found with DNA from all ylm integration mutants except for the ylmF integration. Construction of an ylmF ezrA double mutant proved not to be possible. Electron microscopic inspection of the ylmF:pMUTIN4 integration mutant (Fig. 2G–I) indicated that deletion of this gene was responsible for the aberrant structure of septa observed in the ylm mutant. As it appeared that ylmF is involved in septum synthesis hereafter we refer to it as ‘sepF ’. SepF interacts with itself and FtsZ In most Gram positive bacteria sepF lies in a conserved gene cluster together with ylmE, ylmG and ylmH homologues (see Fig. S1). This suggested that their gene products might be functionally related. To assess whether SepF, YlmE, -G and -H physically interacted, the corresponding open reading frames were expressed in yeast as fusions to the GAL4 activation domain (AD), and to the GAL4 binding domain (BD). Interactions were tested by a yeast two-hybrid mating assay, as described in Experimental procedures. The result of a typical mating experiment is shown in Fig. 4A. The self-activation of the interaction reporter genes by the BD-YlmH fusion prevented any interpretation for this protein. In contrast, BDSepF interacted specifically with AD-SepF, suggesting that the protein can dimerize or oligomerize. Various division proteins were then tested for interaction with SepF, YlmE and YlmG, including: DivIB, DivIC, DivIVA, EzrA, FtsA, FtsL, FtsZ, Pbp2B and ZapA. Only a single interaction emerged from this extensive screening experiment: a clear interaction between SepF and FtsZ (Fig. 4B). This result provided strong support for the notion that SepF is part of the divisome complex. SepF localizes to division sites dependent on FtsZ but not later components of the divisome To further test the interaction of SepF with FtsZ, we examined the localization of the protein in vivo by fusing it to GFP. As shown in Fig. 5A, cells expressing a sepF-gfp fusion showed a pattern of regular transverse bands (arrowheads) and dots (arrows) corresponding to septa in the early and late stages of division. Unfortunately, the © 2005 The Authors Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 989–999 SepF, a novel FtsZ-interacting protein 993 Table 1. Bacillus subtilis strains used in this study. B. subtilis Relevant genotypea,b Construction or reference 3317 1803 3318 1801 3362 4054 4055 4076 4077 BFA2813 4214 BFA2863 4094 4215 4181 4216 BFA2817 4217 2020 4218 3122 4219 ylmB-H::km divIVA::(PdivIVA-gfp divIVA+ cat) ylmB-H::km divIVA::(PdivIVA-gfp divIVA+ cat) chr::(Pspac-ftsZ ble) ezrA::tet ylmBC::(pMut erm) ylmBC::(pMut Pspac-ylmD erm) ylmBC::erm ezrA::tet ylmBC::(erm Pspac-ylmD) ezrA::tet ylmD:(pMUTIN Pspac-ylmE erm) ylmE:(pMut Pspac-sepF erm) sepF(ylmF):(pMUTIN Pspac-ylmG erm) ylmG::(pMut Pspac-ylmH erm) ylmH:(pMut erm) amy::(Pxyl-sepF-gfp spc) amy::(Pxyl-sepF-gfp spc) chr::(Pspac-ftsZ ble) yllB:(pMUTIN Pspac-ylxA-ftsL-pbp2B erm) amy::(Pxyl-sepF-gfp spc) yllB:(pMUTIN Pspac-ylxA-ftsL-pbp2B erm) amyE::(Pxyl-gfp-ftsZ spc) ylmBC::(erm Pspac-ylmD) ezrA::tet amyE::(Pxyl-gfp-ftsZ spc) pbp2B::(Pxyl-gfp-pbp2B cat) ylmBC::(erm Pspac-ylmD) ezrA::tet pbp2B::(Pxyl-gfp-pbp2B cat) This study Thomaides et al. (2001) 3317 transformed with 1803 Marston et al. (1998) This study This study This study 4054 transformed with 3362 4055 transformed with 3362 BFA project This study BFA project This study This study This study 4181 transformed with 1801 BFA project 4181 transformed with BFA2817 J. Sievers (unpublished) 4077 transformed with 2020 Scheffers et al. (2004) 4077 transformed with 3122 a. All strains are based on 168 and carry the trpC2 marker. b. Antibiotics used: km, kanamycin; cat, chloramphenicol; ble, phleomycin; erm, erythromycin; spc, spectinomycin; and tet, tetracycline. SepF-GFP fusion did not complement a sepF disruption in an ezrA background. Nevertheless, because the fusion localized at division sites, independent of the presence of wild-type SepF (data not shown), we assume that the localization pattern, at least partially, reflects the normal localization of the protein. To test whether the localization was dependent on FtsZ, the GFP fusion construct was moved into a strain in which ftsZ could be repressed. Depletion of FtsZ resulted in the formation of long filaments devoid of septal SepF-GFP (Fig. 5C). To examine whether SepF localization also required a later step in divisome assembly, the sepF-gfp construct was introduced into a strain in which the division proteins FtsL and Pbp2B could be simultaneously depleted. This leads to a rapid block in assembly of the divisome, but Zrings remain present at intervals in the resulting elongated cell filaments (Daniel et al., 1998; 2000). The membrane staining in Fig. 5F shows that depletion of FtsL and Pbp2B results in elongated cells devoid of septa. However, the GFP signal in Fig. 5E indicates that in these cells SepF bands remained visible. Thus, SepF does not require the later assembling components of the divisome to be recruited to the Z-ring. SepF acts late in the division process To determine at what step SepF intervenes in division, we first tested whether Z-rings are still formed in the absence of SepF (and EzrA). As shown in Fig. 6B, cells became elongated and no longer divided when the expression of SepF was repressed in the absence of EzrA. However, Z- rings still assembled, as visualized with an ftsZ-gfp fusion (Fig. 6A). Some of the Z-rings appeared abnormal, like double structures or short helices (arrowheads). Such abnormal structures were also observed in ezrA single mutants, but they appeared more often when SepF was depleted, as well. Nevertheless, the conclusion must be that SepF is not essential for the formation of Z-rings in an ezrA mutant. To test whether SepF affects a later step in division, the experiment was repeated with a gfp-pbp2B fusion. As shown in Fig. 6E and F, bands of GFP-Pbp2B (arrows) were still visible after depletion of SepF, albeit with a reduced intensity. The persistence of the GFPPbp2B banding pattern suggests that SepF is not required for assembly of the divisome under these conditions. The GFP experiments suggested that the sepF ezrA arrest in division occurs at a very late step. EM was used to examine the ultrastructure of the arrested cells. In the presence of IPTG division septa had a fairly normal appearance, indicating that deletion of ezrA does not significantly affect the formation of the division septum (Fig. 6I–K). In the absence of IPTG, normal division septa were essentially absent. In some cells the early stages of constriction were visible (Fig. 6L–N). It is likely that these are the remains of septa that had been initiated just before SepF became limiting. Discussion The ylm locus in B. subtilis and related bacteria In many bacteria the genes of the main players in cell division and cell wall synthesis are located together in a © 2005 The Authors Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 989–999 994 L. W. Hamoen et al. A YlmE SepF YlmG YlmH ad YlmE SepF YlmG YlmH bd B FtsZ FtsA EzrA ZapA DivIVA ad YlmE SepF YlmG Most Gram negative bacteria contain homologues of ylmE and ylmG, but they lack genes that show similarities to sepF or ylmH. Interestingly, cyanobacteria harbour cell division proteins that are typical of both Gram positive and Gram negative bacteria. In the genomes of several sequenced cyanobacteria, homologues of both divIVA and minE are present, and sepF, ylmE, ylmG and ylmH are all conserved (Miyagishima et al., 2005). However, these genes are not clustered at a single locus in the genome. Despite their outer membrane, molecular phylogenetic studies now group the cyanobacteria separately from the other Gram negatives (Woese et al., 1990; Miyagishima et al., 2005). It was recently shown that in the cyanobacterium Synechococcus elongatus, an interruption of sepF resulted in a doubling in cell length (Miyagishima et al., 2005). The relatively mild effect of sepF disruption on cell length in B. subtilis may explain why this gene had not been picked up in previous mutant searches. Deletion of sepF resulted in aberrant septum synthesis according to the EM images of Fig. 2. These structural deformations could delay maturation of the division septum, which may explain why sepF mutations result in longer cells. A late role for SepF in cell division bd Fig. 4. Yeast two-hybrid interactions. Diploid yeast cells expressing the indicated combinations of proteins fused to the GAL4 BD, and to the GAL4 AD, were subjected to selection for expression of the ADE2 interaction reporter. Binary interactions appeared as pairs of growing colonies on selective SC-LUA medium (Synthetic Complete medium, lacking leucine, uracil and adenine). The upper matrix (A) presents an example of an interaction study between SepF and the conserved Ylm proteins. The lower matrix (B) shows a screen in which possible interactions with other division proteins are tested. On top of the matrix the activator fusions are indicated, and at the left site the BD fusions are indicated. Every BD/AD combination was tested in duplicate. Control matings with the GAL4 BD (bd) and GAL4 AD (ad) were performed to detect self-activation. cluster that is well conserved, the DCW cluster (Massidda et al., 1998). In many Gram positive bacteria the ylm locus is located next to the DCW cluster (Fig. S1). An indication that the ylm genes might have a role in cell division came from a deletion study in Streptococcus pneumoniae (Fadda et al. 2003). In this organism ftsA, ftsZ, the ylm genes, and divIVA form an operon. Deletion of divIVA resulted in a strong morphological effect with incomplete septa, strongly deformed cells and occasional anucleate cells. Insertions in the individual ylm genes resulted in thin septa, and a slightly altered cell shape. Unfortunately, interpretation of these data was difficult due to potential polar effects on the expression of ylm genes and divIVA. In this study we have shown that SepF is an important cell division protein. The conserved genomic localization of sepF upstream of divIVA might suggest a functional relationship. However, deleting sepF does not result in a minicell phenotype, and the typical targeting of DivIVA to division sites and cell poles was unaffected in such mutant. In addition, deletions of divIVA and minC/D did not interfere with the localization of SepF (data not shown). This suggests that SepF functions in a different pathway from DivIVA. The information that can be extracted from the primary sequence of the ylm operon is rather limited. Only ylmE shows sequence homology that could point towards a specific activity in the cell division process, because this protein has homology with pyridoxal phosphate-dependent enzymes. The cofactor pyridoxal phosphate is involved in a wide variety of catalytic reactions, including the conversion of L-alanine to Dalanine by alanine racemases, which is one of the first steps in the synthesis of peptidoglycan (Shaw et al., 1997). Nevertheless, the yeast two-hybrid study did not reveal an interaction between SepF and YlmE, YlmG or YlmH. In addition, none of the ylmE, ylmG or ylmH mutants showed a filamentous phenotype in an ezrA background that would resemble a sepF ezrA double knockout. So far, our data do not provide support for the assumption that SepF and the other Ylm proteins function in the same pathway, as the conserved operon organization would suggest. © 2005 The Authors Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 989–999 SepF, a novel FtsZ-interacting protein 995 A wt Fig. 5. In vivo localization of SepF. Upper panel (A) shows SepF-GFP expressed in wild-type cells (B. subtilis 4181). The arrowheads and arrows indicate septa in early and late stages of division respectively. Membranes were stained with the fluorescent membrane dye FM5-95 (B). The fluorescence images C (SepFGFP) and D (membranes) show cells that were depleted for FtsZ (B. subtilis 4216). The fluorescence images E (SepF-GFP) and F (membranes) show cells that were depleted for FtsL and Pbp2B (B. subtilis 4217). Depletion was achieved by growing cells for 2 h in the absence of IPTG at 30°C. Scale bars indicate 5 µm. B C E - ftsZ - ftsL/pbp2B D F To facilitate the discussion on the function of SepF, we dissect the process of septum formation into three sequential steps: (i) Z-ring formation, (ii) divisome assembly and (iii) septum synthesis (Fig. 7). The first step, the formation of the Z-ring, begins with the polymerization of FtsZ monomers into protofilaments that subsequently assemble into a Z-ring. FRAP studies (fluorescence recovery after photobleaching) have shown that the Z-ring is a dynamic structure with a turnover of subunits within seconds (Stricker et al., 2002; Anderson et al., 2004). This remodelling of the Z-ring appears to depend on the hydrolysis of GTP by FtsZ polymers (Stricker et al., 2002; Anderson et al., 2004). Several proteins control the assembly of the Z-ring in B. subtilis such as MinC and ZapA (Fig. 7). In vitro experiments with MinC of E. coli have shown that this protein interacts with FtsZ and inhibits polymerization (Hu et al., 1999). EzrA is also a negative regulator of Zring formation, and recent biochemical experiments showed that purified EzrA inhibits the polymerization of FtsZ into protofilaments (Haeusser et al., 2004). ZapA has a positive effect on Z-ring formation and in vitro studies showed that this protein strongly stimulates the bundling of protofilaments (Gueiros-Filho and Losick, 2002). Interestingly, an ezrA zapA double knockout is highly filamen- tous (Gueiros-Filho and Losick, 2002). The reason for this is not clear, but this phenotype could be used to argue that SepF might have a comparable function with that of ZapA. However, we do not think that this is the case. A zapA mutant is sensitive to reduced FtsZ concentrations, and this was not observed for the ylm mutant. Also the introduction of minC or zapA deletions into the ylm mutant did not affect cell length more than the mutations did individually. Finally, in the absence of both SepF and EzrA, Z-rings are still formed. Therefore, we assume that SepF is not involved in Z-ring formation, but is active in a later step of septum synthesis. Interestingly, SepF requires only the Z-ring for its localization, which is so far unique for a late cell division protein in B. subtilis. After the Z-ring has formed the divisome assembles. In E. coli there is a considerable time delay between these two steps: 15–20 min, depending on the growth rate (Aarsman et al., 2005). Whether this is also the case for B. subtilis is not known. As described in Introduction, assembly of the divisome in this organism is a cooperative process and requires the presence of a number of essential division proteins (Fig. 7). SepF is not essential for this step of the process, because the ylm mutant grew normally. Even when both SepF and EzrA are absent, and © 2005 The Authors Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 989–999 996 L. W. Hamoen et al. A B + IPTG C - IPTG D E F - IPTG G + IPTG H + IPTG I J K - IPTG septation is blocked, the divisome still assembles, as the fluorescent GFP-Pbp2B bands in Fig. 6E indicated. Thus, it is unlikely that the main function of SepF is to stimulate the assembly of the divisome complex. However, the EM images of Fig. 2 suggest that the synthesis of the division septum occurs in an irregular manner when SepF is absent. Such aberrant septa have not been described for B. subtilis cell division mutants. Although mutants that affect Z-ring formation can result in misplaced septa, they do not show such strong deformations and thickening of division septa. For example, Fig. 6I–K illustrates that septa have a wild-type appearance in an ezrA background, whereas it is clear from Fig. 2 that maturation is strongly affected in a sepF mutant septum. Therefore, we favour a model in which SepF participates in the third step of the cell division process: the constriction of the divisome and/or synthesis of the septum wall. The reasoning behind the model of Fig. 7 presents us with a problem, namely, how to explain the synergistic effect of an ezrA sepF double knockout. The precise role that EzrA fulfils in the division process is not entirely clear. Fluorescence microscopy studies have shown that EzrA primarily localizes at the site of cell division, and that this localization is FtsZ-dependent (Levin et al., 1999). This suggests that EzrA is active during synthesis of the septum, which is surprising because EzrA is an inhibitor of FtsZ polymerization. However, biochemical experiments provided an explanation for this paradox because EzrA is unable to disassemble preformed FtsZ-polymers in vitro (Haeusser et al., 2004). The polymerization of FtsZ is very sensitive to differences in cellular FtsZ concentrations, and even a twofold increase can result in doublet and polar Z-rings (Weart and Levin, 2003). A local increase in the FtsZ concentration in the vicinity of a dynamic and contracting Z-ring could therefore lead to the assembly of a doublet, which could eventually result in the formation of a minicell. Possibly, EzrA represses formation of adja- Z-ring L M Divisome N Fig. 6. Depletion of SepF in an ezrA background. The effect on Zring assembly is presented In the upper panel. B. subtilis strain 4218 (ezrA, Pspac-ylmD-H, Pxyl-gfp-ftsZ) was grown in the absence (A, B) or presence of 1 mM IPTG (C, D), for 2 h at 30°C. Arrowheads indicate aberrant Z-rings. The fluorescence images in the middle panel were taken from cells expressing a GFP-Pbp2B fusion (B. subtilis 4219: ezrA, Pspac-ylmD-H, Pxyl-gfp-pbp2B). The strain was grown in the absence (E, F) or presence of 1 mM IPTG (G, H), for 2 h at 30°C. Some of the GFP-Pbp2B bands are pointed out by arrows. Membranes were stained with fluorescent membrane dye FM5-95 (B, D, F, H). Scale bars indicate 5 µm. The TEM images in the lower panel show some division septa of B. subtilis 4077 (ezrA, Pspac-ylmD-H) grown in the presence of 1 mM IPTG (I J, K) or absence of IPTG (L, M, N). MinC ( EzrA ) ZapA FtsZ FtsA FtsL DivIB DivIC FtsW Pbp2B SepF Fig. 7. A model of the different stages leading to septum synthesis. The process is divided in three subsequent steps: Z-ring formation, divisome assembly, and septum synthesis accompanied by constriction of the Z-ring. Grey and black lines indicate cell wall and cytoplasmic membrane respectively. Open circles indicate FtsZ molecules, and the different proteins that constitute the divisome are depicted as open rectangles. © 2005 The Authors Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 989–999 SepF, a novel FtsZ-interacting protein 997 cent Z-rings when septum synthesis is in progress. Thus, EzrA could be most important during the period when SepF is also active. Blocking both processes may therefore be too much to bear for the division machinery. In conclusion, SepF is a new member of the divisome of Gram positive bacteria. We postulate that it acts as a late division protein, even though it is an early recruit to the Z-ring. We now need to know whether the protein is required for divisome constriction, or plays a more direct role, perhaps in septal peptidoglycan synthesis. Experimental procedures General methods and materials Molecular cloning, polymerase chain reactions (PCR) and E. coli transformations were carried out using standard techniques. B. subtilis strains were grown in Difco antibiotic medium 3 (PAB). E. coli was used as cloning intermediate. B. subtilis chromosomal DNA for transformation and PCR reactions was purified as described by Venema et al. (1965). Transformation of competent B. subtilis cells was performed using an optimized two-step starvation procedure based on the method of Anagnostopoulos and Spizizen (Anagnostopoulos and Spizizen, 1961; Hamoen et al., 2002). Selection for E. coli and B. subtilis transformants was carried out on nutrient agar supplemented with: 100 µg ml−1 ampicillin, 5 µg ml−1 kanamycin, 4 µg ml−1 erythromycin, 5 µg ml−1 chloramphenicol, 50 µg ml−1 spectinomycin, 0.2 µg ml−1 phleomycin or 5 µg ml−1 tetracyclin. ylmC1 and ylmF2. The amplified products were digested, ligated and transformed to B. subtilis 168. In case of B. subtilis 4054 only the ylmB-C region was replaced by double crossover by pMut. For this, the upstream region was amplified by primers sigE1 and ylmA2, and the downstream region was amplified by primers ylmD3 and ylmF2. Insertion deletion of ylmE (B. subtilis 4214) and ylmH (B. subtilis 4215) were obtained by single crossover of pMut. An internal fragment of ylmE was amplified using primers ylmE8 and ylmE9. pMut was again amplified from pMUTIN4 using primers pmut1 and pmut2. After digestion and ligation the ligation mixture was first transformed to E. coli, and plasmid DNA from the correct clone was transformed to B. subtilis 168. The same procedure was followed for the construction of the ylmH integration except that pMut was amplified using primers pmut4 and pmut5, and an internal ylmH fragment was obtained using primers ylmH11 and ylmH14. In case of the ylmG deletion (B. subtilis 4094), a double crossover was required. The upstream region was amplified by primers ylmD1 and ylmF8, and the downstream region was amplified by primers ylmH15 and ileS10. pMut was again amplified from pMUTIN4 using primers pmut1 and pmut2. The amplified products were digested, ligated and transformed directly to B. subtilis 168. All chromosomal integrations were verified by PCR, restriction digestion, and sequencing. For the construction of a sepF-gfp fusion (B. subtilis 4181), we made use of the B. subtilis GFP-cloning vector pSG1154 (Lewis and Marston, 1999). The sepF open reading frame was amplified using primers ylmF1 and ylmF4. Cloning this fragment into pSG1154 resulted in a C-terminal GFP fusion under control of the xylose-inducible Pxyl promoter. After transformation to B. subtilis 168, the gene fusion was integrated at the amy locus by double crossover. Strain construction The relevant B. subtilis strains are listed in Table 1. Primers used for PCR reactions are listed in Table S1. Deletion of the ylmB-H region (B. subtilis 3317) was accomplished by double crossover of a kanamycin marker. The upstream region was amplified by PCR using primers IIGA and ylmB2, the downstream region was amplified using primers ylmH1 and ileS2, and the kanamycin marker was amplified using primers Km3 and Km4. As template for PCR reactions, chromosomal DNA from B. subtilis 168 was used. The kanamycin marker was amplified from an aph3 containing plasmid. The PCR fragments were digested, ligated, and the ligation product was directly transformed to competent 168 cells. An ezrA deletion (B. subtilis 3362) was constructed by double crossover of a tetracycline marker. The upstream region was amplified by PCR using primers nifZ2 and ezrA1, and the downstream region was amplified using primers ezrA2 and ytrP1. The tetracycline marker was cut out of pBEST309, ligated with the upstream and downstream fragments, and the ligation product was directly transformed to competent B. subtilis 168 cells. B. subtilis 4055 was constructed as follows. The ylmBC region plus the ylmD promoter region was replaced with the pMUTIN4 derivative pMut by double crossover. pMut is pMUTIN4 without the lacZ gene, and was derived by PCR reaction on pMUTIN4 with primers pmut1 and pmut2. The ylmB-C upstream region was amplified by primers sigE1 and ylmA2, and the downstream region was amplified by primers Fluorescence light microscopy Cells from overnight cultures grown in PAB were inoculated into fresh PAB. The following B. subtilis strains required 1 mM IPTG for growth: 4077, 4216, 4217, 4218 and 4219. B. subtilis 4219 also required 0.5% xylose. To induce the sepF-gfp, gfp-ftsZ and gfp-pbp2B fusions, 0.1%, 0.2% and 0.5% xylose was used respectively. Depletion of FtsZ, FtsL, Pbp2B or SepF was achieved by growing exponentially growing cultures for 2 h in the absence of IPTG. For fluorescence microscopy, cultures were grown in PAB at 30°C, samples were taken at exponentially growth, and mounted onto microscope slides coated with a thin layer of 1.5% agarose. Images were acquired with a Zeiss Axiovert 200M coupled to a CoolsnapHQ CCD camera, and using Metamorph imaging software (Universal Imaging). When required, cells were grown in the presence of the membrane dye FM5-95 (90 µg ml−1, Molecular Probes). For cell length measurements, cells were mixed with the membrane dye Nile Red (2.5 µg ml−1, Molecular Probes), prior to microscopic examinations. Electron microscopy For transmission electron microscopy (TEM), cultures were grown in PAB at 37°C to mid-exponential phase and fixed by © 2005 The Authors Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 989–999 998 L. W. Hamoen et al. the addition of gluteraldehyde to a final concentration of 2.5%. Cells were then pelleted and fixation was continued overnight at 4°C. Cell pellets were washed with 200 mM phosphate buffer, post fixed with 1% osmium tetroxide in 100 mM phosphate buffer, and incubated for 1 h at 4°C. Pellets were then washed in distilled water and stained with 2% aqueous uranyl acetate for 1 h at 4°C in the dark. The fixed cells were washed twice in water, dehydrated in acetone, and washed in propylene dioxide. Samples were imbedded in epon-araldite resin that was allowed to polymerize overnight at 65°C. Sections of 80 nm were cut using an ultracut microtome (Reichert and Jung), and examined with a Zeiss EM 912 OMEGA Electron Microscope. Northern blotting Total RNA was isolated from exponentially growing cells using the FastRNA Pro-kit (Qbiogene). For Northern blotting, 8 µg of RNA was denaturated using 10 µl denaturation-mix (10× FA buffer, 37% formaldehyde and formamide; 20:34:100) at 55°C for 15 min. 10× FA buffer contains 200 mM MOPS (pH 7), 50 mM Na-acetate and 10 mM EDTA. RNA was separated on formaldehyde agarose gels (0.5 g agarose, 4 ml 10× FA buffer, 720 µl 37% formaldehyde, in 40 ml), run in 1× FA buffer (100 ml 10× FA buffer and 20 ml 37% formaldehyde, in 1 l), and blotted onto a nylon membrane (Amersham) using 20× SSC blot buffer (3 M NaCl, 0.3 M Na-citrate pH 7). RNA was immobilized at 55°C for 3 h, and hybridized with radioactively labelled probes directed against ylmD, ylmG, ylmH and divIVA. Probes were synthesized using the Prime-a-gene labelling kit (Promega). Blots were hybridized at 67°C and subsequently exposed to a phosphor screen. Yeast two-hybrid assay The different proteins that have been tested for interaction in the yeast two-hybrid assay are listed in the Results section. The related open reading frames were PCR amplified from chromosomal DNA, and PCR products were cloned into pGBDU (bait) and pGAD (prey) vectors (James et al., 1996) using the EcoRI and BamHI restriction sites. The resulting vectors were checked by DNA sequencing, and the yeast strains PJ69–4a (bait) and PJ69–4α (prey) (James et al., 1996) were transformed by the bait and prey vectors respectively. Bait-containing cells were selected on SC-U medium (Synthetic Complete lacking uracil), and prey-containing cells were selected on SC-L medium (Synthetic Complete lacking leucine). For each construct, two independent transformants were stored for further mating experiments. Prey- and baitcontaining strains were grown on fresh selective plates at 30°C for 48 h, and cells were resuspended in 5 ml YEPD liquid medium. Mating was carried out by mixing 50 µl of prey- and bait-containing cells in 96 well plates, and depositing, with a replicating tool, approximately 3 µl of the cell mixtures (prey + bait) onto YEPD plates, which were incubated at 30°C for 48 h. Cells were transferred onto SC-LU medium (Synthetic Complete lacking leucine and uracil) for diploid selection, and the plates were incubated for 2–3 days at 30°C. Diploid colonies were then transferred onto media selecting for the expression of the HIS3 and ADE2 interaction reporters (SC-LUH and SC-LUA; Synthetic Complete lacking leucine, uracil and histidine or adenine respectively). Interaction phenotypes were scored after 5 days of growth at 30°C. Acknowledgements We would like to thank the members of the group for helpful advice and discussions, and Michael Shaw for expert help with the EM studies, and Naotake Ogasawara for communicating results prior to publication. This research was supported by an EMBO Long-Term Fellowship (LH), a grant from the UK Biotechnology and Biological Science Research Council (JE), and a fellowship from the EPA Cephalosporin Fund and INRA (JM). References Aarsman, M.E., Piette, A., Fraipont, C., Vinkenvleugel, T.M., Nguyen-Disteche, M., and den Blaauwen, T. (2005) Maturation of the Escherichia coli divisome occurs in two steps. Mol Microbiol 55: 1631–1645. Anagnostopoulos, C., and Spizizen, J. (1961) Requirements for transformation in Bacillus subtilis. J Bacteriol 81: 741– 746. Anderson, D.E., Gueiros-Filho, F.J., and Erickson, H.P. (2004) Assembly dynamics of FtsZ rings in Bacillus subtilis and Escherichia coli and effects of FtsZ-regulating proteins. J Bacteriol 186: 5775–5781. Bramhill, D. (1997) Bacterial cell division. Annu Rev Cell Dev Biol 13: 395–424. Daniel, R.A., Harry, E.J., Katis, V.L., Wake, R.G., and Errington, J. (1998) Characterization of the essential cell division gene ftsL (yIID) of Bacillus subtilis and its role in the assembly of the division apparatus. Mol Microbiol 29: 593– 604. Daniel, R.A., Harry, E.J., and Errington, J. (2000) Role of penicillin-binding protein PBP 2B in assembly and functioning of the division machinery of Bacillus subtilis. Mol Microbiol 35: 299–311. Edwards, D.H., and Errington, J. (1997) The Bacillus subtilis DivIVA protein targets to the division septum and controls the site specificity of cell division. Mol Microbiol 24: 905– 915. Erickson, H.P. (2001) The FtsZ protofilament and attachment of ZipA-structural constraints on the FtsZ power stroke. Curr Opin Cell Biol 13: 55–60. Errington, J., Daniel, R.A., and Scheffers, D.J. (2003) Cytokinesis in bacteria. Microbiol Mol Biol Rev 67: 52–65. Fadda, D., Pischedda, C., Caldara, F., Whalen, M.B., Anderluzzi, D., Domenici, E., and Massidda, O. (2003) Characterization of divIVA and other genes located in the chromosomal region downstream of the dcw cluster in Streptococcus pneumoniae. J Bacteriol 185: 6209– 6214. Goehring, N.W., Gueiros-Filho, F., and Beckwith, J. (2005) Premature targeting of a cell division protein to midcell allows dissection of divisome assembly in Escherichia coli. Genes Dev 19: 127–137. Gueiros-Filho, F.J., and Losick, R. (2002) A widely con- © 2005 The Authors Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 989–999 SepF, a novel FtsZ-interacting protein 999 served bacterial cell division protein that promotes assembly of the tubulin-like protein FtsZ. Genes Dev 16: 2544– 2556. Haeusser, D.P., Schwartz, R.L., Smith, A.M., Oates, M.E., and Levin, P.A. (2004) EzrA prevents aberrant cell division by modulating assembly of the cytoskeletal protein FtsZ. Mol Microbiol 52: 801–814. Hamoen, L.W., Smits, W.K., de Jong, A., Holsappel, S., and Kuipers, O.P. (2002) Improving the predictive value of the competence transcription factor (ComK) binding site in Bacillus subtilis using a genomic approach. Nucleic Acids Res 30: 5517–5528. Holtje, J.V. (1998) Growth of the stress-bearing and shapemaintaining murein sacculus of Escherichia coli. Microbiol Mol Biol Rev 62: 181–203. Hu, Z., Mukherjee, A., Pichoff, S., and Lutkenhaus, J. (1999) The MinC component of the division site selection system in Escherichia coli interacts with FtsZ to prevent polymerization. Proc Natl Acad Sci USA 96: 14819–14824. James, P., Halladay, J., and Craig, E.A. (1996) Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast. Genetics 144: 1425– 1436. Johnson, J.E., Lackner, L.L., Hale, C.A., and de Boer, P.A. (2004) ZipA is required for targeting of DMinC/DicB, but not DMinC/MinD, complexes to septal ring assemblies in Escherichia coli. J Bacteriol 186: 2418–2429. Kobayashi, K., Ehrlich, S.D., Albertini, A., Amati, G., Andersen, K.K., Arnaud, M., et al. (2003) Essential Bacillus subtilis genes. Proc Natl Acad Sci USA 100: 4678–4683. Levin, P.A., Kurtser, I.G., and Grossman, A.D. (1999) Identification and characterization of a negative regulator of FtsZ ring formation in Bacillus subtilis. Proc Natl Acad Sci USA 96: 9642–9647. Lewis, P.J., and Marston, A.L. (1999) GFP vectors for controlled expression and dual labelling of protein fusions in Bacillus subtilis. Gene 227: 101–110. Marston, A.L., and Errington, J. (1999) Selection of the midcell division site in Bacillus subtilis through MinDdependent polar localization and activation of MinC. Mol Microbiol 33: 84–96. Marston, A.L., Thomaides, H.B., Edwards, D.H., Sharpe, M.E., and Errington, J. (1998) Polar localization of the MinD protein of Bacillus subtilis and its role in selection of the mid-cell division site. Genes Dev 12: 3419–3430. Massidda, O., Anderluzzi, D., Friedli, L., and Feger, G. (1998) Unconventional organization of the division and cell wall gene cluster of Streptococcus pneumoniae. Microbiology 144 (Pt 11): 3069–3078. Miyagishima, S.Y., Wolk, C.P., and Osteryoung, K.W. (2005) Identification of cyanobacterial cell division genes by comparative and mutational analyses. Mol Microbiol 56: 126– 143. Pichoff, S., and Lutkenhaus, J. (2005) Tethering the Z ring to the membrane through a conserved membrane targeting sequence in FtsA. Mol Microbiol 55: 1722–1734. Robson, S.A., Michie, K.A., Mackay, J.P., Harry, E., and King, G.F. (2002) The Bacillus subtilis cell division proteins FtsL and DivIC are intrinsically unstable and do not interact with one another in the absence of other septasomal components. Mol Microbiol 44: 663–674. Scheffers, D.J., Jones, L.J., and Errington, J. (2004) Several distinct localization patterns for penicillin-binding proteins in Bacillus subtilis. Mol Microbiol 51: 749–764. Shaw, J.P., Petsko, G.A., and Ringe, D. (1997) Determination of the structure of alanine racemase from Bacillus stearothermophilus at 1.9-A resolution. Biochemistry 36: 1329– 1342. Sievers, J., and Errington, J. (2000) Analysis of the essential cell division gene ftsL of Bacillus subtilis by mutagenesis and heterologous complementation. J Bacteriol 182: 5572–5579. Stricker, J., Maddox, P., Salmon, E.D., and Erickson, H.P. (2002) Rapid assembly dynamics of the Escherichia coli FtsZ-ring demonstrated by fluorescence recovery after photobleaching. Proc Natl Acad Sci USA 99: 3171–3175. Thomaides, H.B., Freeman, M., El Karoui, M., and Errington, J. (2001) Division site selection protein DivIVA of Bacillus subtilis has a second distinct function in chromosome segregation during sporulation. Genes Dev 15: 1662–1673. Vagner, V., Dervyn, E., and Ehrlich, S.D. (1998) A vector for systematic gene inactivation in Bacillus subtilis. Microbiology 144 (Pt 11): 3097–3104. Venema, G., Pritchard, R.H., and Venema-Schroeder, T. (1965) Fate of transforming deoxyribonucleic acid in Bacillus subtilis. J Bacteriol 89: 1250–1255. Weart, R.B., and Levin, P.A. (2003) Growth rate-dependent regulation of medial FtsZ ring formation. J Bacteriol 185: 2826–2834. Woese, C.R., Kandler, O., and Wheelis, M.L. (1990) Towards a natural system of organisms: proposal for the domains Archaea, Bacteria, and Eucarya. Proc Natl Acad Sci USA 87: 4576–4579. Supplementary material The following supplementary material is available for this article online: Fig. S1. Comparison of the ylm locus in different Gram positive bacteria. The conserved ftsA, ftsZ and ileS genes are also indicated. The B. subtilis ylm locus is located approximately 8 kb downstream of ftsZ. divIVA is indicated in black, and the conserved ylm genes (ylmE, -F, -G and -H) are marked with different shades of grey. Table S1. Primers used for PCR reactions. This material is available as part of the online article from http://www.blackwell-synergy.com © 2005 The Authors Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 989–999