MreB proteins are bacterial actin homologues thought to have a role in cell shape determination by positioning the cell wall synthetic machinery. Many bacteria, particularly Gram-positives, have more than one MreB isoform. Bacillus subtilis has three, MreB, Mbl and MreBH, which colocalize in a single helical structure. We now show that the helical pattern of peptidoglycan (PG) synthesis in the cylindrical part of the rod-shaped cell is governed by the redundant action of the three MreB isoforms. Single mutants for any one of mreB isoforms can still incorporate PG in a helical pattern and generate a rod shape. However, after depletion of MreB in an mbl mutant (or depletion of all three isoforms) lateral wall PG synthesis was impaired and the cells became spherical and lytic. Overexpression of any one of the MreB isoforms overcame the lethality as well as the defects in lateral PG synthesis and cell shape. Furthermore, MreB and Mbl can associate with the peptidoglycan biosynthetic machinery independently. However, no single MreB isoform was able to support normal growth under various stress conditions, suggesting that the multiple isoforms are used to allow cells to maintain proper growth and morphogenesis under changing and sometimes adverse conditions.
Bacteria have a wide range of characteristic shapes, and the peptidoglycan (PG) layer of the cell wall is a major determinant of cell shape. It is built from long glycan strands cross-linked by peptide cross bridges (Goffin and Ghuysen, 1998; Höltje, 1998; Vollmer et al., 2008). The precursor for PG synthesis is synthesized in the cytosol and then transferred to the outside of the cytoplasmic membrane. The newly synthesized PG is incorporated into the existing PG meshwork by a combination of transglycosylation and transpeptidation reactions of the penicillin-binding proteins (PBPs). Recent studies using fluorescent derivatives of the antibiotics vancomycin (Van-FL) and ramoplanin, which label newly externalized or incorporated PG precursors, have revealed an underlying helical pattern of PG synthesis along the lateral cell wall of rod-shaped bacterium Bacillus subtilis (Daniel and Errington, 2003; Formstone and Errington, 2005; Tiyanont et al., 2006). Similar results were obtained by use of d-cysteine labelling in Escherichia coli (Varma et al., 2007).
Effects of loss of MreB isoforms on synthesis of PG in the lateral cell wall
Previous experiments with Van-FL and ramoplanin have established that PG synthesis probably occurs in a helical pattern along the lateral cell wall of B. subtilis (Daniel and Errington, 2003; Formstone and Errington, 2005; Tiyanont et al., 2006). It seemed likely that this pattern was governed by the helical MreB cytoskeleton but reports have differed on the effects of mbl mutations on the pattern: Daniel and Errington (2003) reported that PG precursor accumulation was largely redirected to cell division sites in an mbl mutant background, whereas no significant change in staining pattern was seen by Tiyanont et al. (2006). Meanwhile, several other papers have shown that the Mg2+ concentration in the medium has an important effect on the phenotypic penetrance of strains mutated for mreB and its isologues (Formstone and Errington, 2005; Carballido-López et al., 2006; Schirner and Errington, 2009). The findings that mbl mutants are lethal under typical lab conditions but can readily pick up suppressor mutations that restore viability (Schirner and Errington, 2009; Schirner et al., 2009) also prompted a systematic reappraisal of the effects of mreB family members on cell wall synthesis. To this end, various mutants were generated under high Mg2+ conditions (to avoid spontaneous suppressor mutations) and then grown in the presence or absence of 10 mM Mg2+ and stained with Van-FL. Typical cells are shown in Fig. 1. As expected, wild-type cells showed Van-FL fluorescent bands at ongoing or recent division sites and helical staining at the lateral cell wall in the presence or absence of extra Mg2+ (Fig. 1A and B). The mreB mutant cells showed similar patterns of staining to the wild-type cells in the presence of Mg2+ (Fig. 1C). However, in the absence of added Mg2+, the mreB mutant showed abnormally bright fluorescence at the division sites and bulging poles (Fig. 1D). Consistent with previous findings that PBP1 accumulates at the cell poles in mreB mutants under low Mg2+ conditions (Kawai et al., 2009), the unusual bulging and polar staining disappeared in an mreB ponA double mutant (Fig. 1F). In addition, less prominent but detectable staining was still present in the cylindrical parts of mreB mutant cells (Fig. 1D and F).
In the presence of Mg2+, the staining pattern of mbl mutant cells was similar to that of wild-type cells (Fig. 1G). In the absence of added Mg2+, staining appeared enhanced at division sites and concomitantly reduced but still readily detectable in the cylindrical parts of the cells (Fig. 1H). The mreBH mutant was indistinguishable from the wild type in the presence or absence of added Mg2+ (Fig. 1I and J). Staining of an mbl mreBH double mutant was similar to that of the mbl single mutant (Fig. 1K and L). Other double mutant combinations could not be constructed, even in the presence of high Mg2+ (see below). These results indicated that MreB and Mbl are required for normal organization of PG synthesis in the lateral cell wall but no single deletion (or mbl mreBH double) completely eliminates lateral wall staining.
Inhibition of rod-shaped growth and lateral cell wall synthesis in cells lacking MreB and Mbl
As mentioned above, an mbl mreBH double mutant is viable but other double mutant combinations are apparently not (Defeu Soufo and Graumann, 2006; A. Formstone and J. Errington, unpublished). To test the effects of loss of both MreB and MreBH, and MreB and Mbl on lateral wall PG synthesis, we constructed strains lacking these gene combinations in a background containing an extra copy of mreB fused to a histidine tag and controlled by a xylose-inducible promoter, Pxyl, at the amyE locus. As shown in Fig. 2A, these strains grew on plates containing both 10 mM Mg2+ and 0.5% xylose. However, in the absence of xylose (i.e. with synthesis of MreB repressed) they showed a strong growth defect even in the presence of high Mg2+, although a control strain with only mreB deleted grew well. The severe growth defect was consistent with the mreB homologues having partially redundant roles in cell morphogenesis. The morphology of these strains was examined by phase-contrast microscopy of cultures grown in PAB medium. In the presence of added Mg2+/xylose, these strains maintained their rod shape reasonably well (Fig. 2C and D). In the absence of added Mg2+/xylose, cells of the ΔmreBH background still maintained rod shape, but frequently showed polar bulges and lysis, similar to that of the isogenic mreBH+ strain (compare Fig. 2E and F). However, cells of the Δmbl background lost the ability to grow as rods and underwent substantial bulging, becoming almost round, with much of the cell population undergoing lysis (Fig. 2G). We do not understand why the cells show the ballooning but it could be due in part to continued autolytic activity in the pre-existing wall.
We examined whether impaired PG synthesis in the lateral wall might be responsible for the loss of rod shape and reduced cell elongation in the ΔmblΔmreB background. As shown in Fig. 2H, in the presence of added Mg2+/xylose cells displayed a normal Van-FL staining pattern (compare with wild type, Fig. 1A). At 60 min after removal of added Mg2+/xylose, abnormal bright fluorescence was evident at division sites, and reduced and slightly disorganized fluorescence was seen in the cylindrical parts of the cells. Further cultivation (> 90 min) of cells resulted in loss of the rod shape and the staining along the lateral cell wall had virtually disappeared (Fig. 2H and Fig. S1). These results provide strong support for the idea that the overlapping activities of MreB and Mbl are crucial for the helical organization of lateral PG synthesis and maintenance of the rod-shaped form of B. subtilis. Because of the functional redundancy of mreB and mbl, both proteins need to be eliminated before the lethal cell shape and elongation defects are revealed.
Elimination of PBP1 suppresses the lethality of an mbl mutant, but not of an mbl mreB double mutant
MreB has a critical role in the control of PG synthesis by directly recruiting PBP1 to the lateral cell wall (Kawai et al., 2009). The delocalization of PBP1 in the absence of MreB is associated with incorrectly localized PG synthesis, bulging and eventually cell death. Therefore, we suggest that inactivation of PBP1 restores the viability of mreB mutants grown under low Mg2+ conditions by preventing bulge formation (Kawai et al., 2009). To test whether this was also case for lethality of mbl mutant under low Mg2+ conditions, we examined the viability of an mbl ponA double mutant on PAB plates with no added Mg2+. As shown in Fig. 2I, a null mutant of mbl showed a strong defect in cell viability on PAB plates, as reported previously (Schirner and Errington, 2009). In contrast, the viability of an mbl ponA double mutant was similar to that of wild-type strain, although the colonies were smaller. Thus, inactivation of PBP1 can rescue mbl mutants as well as mreB mutants, suggesting that Mbl might also contribute to the regulation of PBP1 localization.
Cells of mreB ponA and mbl ponA double mutant grew as rods (Kawai et al., 2009; data not shown). Because of the functional redundancy of MreB and Mbl, lateral wall PG synthesis in the mutants in each case could be supported by the remaining MreB isoforms. If so, simultaneous disruption of both mreB and mbl might be lethal even in the ponA mutant background. To test this, the only copy of mbl was replaced under the control of the IPTG-inducible Pspac promoter and introduced into an mreB ponA double mutant strain. As shown in Fig. 2J, the resultant strain grew on PAB plates only in the presence of IPTG (i.e. in the presence of Mbl). These results show that either MreB or Mbl can support cell elongation and growth under low Mg2+ conditions in ponA mutant background. However, although a ponA deletion can suppress the lethality of mreB or mbl single mutants, it does not restore viability when both isologues are lost.
Overexpression of MreB or MreBH is lethal
As mentioned above, an mbl mreBH double mutant was the only viable double mutant combination. Therefore, MreB alone can support growth and viability but neither of the other isologues can do so. We wondered whether Mbl and MreBH could support viability and cell elongation when expressed at higher levels. We therefore constructed fusions of these isologues to the very strong (although poorly repressed) PspacHY promoter (Quisel et al., 2001). For comparison, we also made a PspacHY–mreB fusion. We first tested the effects of overexpression in otherwise wild-type cells. Figure 3B shows that MreB and Mbl, at least, were overproduced in the presence of inducer (0.5 mM IPTG); we do not have antibodies against MreBH but assume from results described below that it was also overproduced. As shown in Fig. 3A, all three strains grew well on plates in the absence of inducer. However, in the presence of inducer (0.5 mM IPTG) only the mbl-overexpressing strain grew. Therefore, overexpression of mreB or mreBH is toxic in B. subtilis. (Fig. S2 shows some quantitative data on the levels of protein overproduction in various constructs and defects on cell viability.) To examine the phenotypic effects of overexpression on cells, all three strains were grown in the presence and absence of inducer. Figure 3C–E shows that addition of inducer to wild-type cells with no PspacHY fusion had no effect on morphology. Figure 3F, I and L shows that cell morphology was normal for all three strains in the absence of inducer. In the presence of inducer the mreB- and mreBH-overexpressing strains showed progressive bulging and lysis (Fig. 3G, H, M and N), suggesting impairment of cell width control and explaining the lethal effects on growth. There was also an apparent impairment in cell division, as many of the cells were unusually long. Only a slight increase in cell length and little or no change of cell width were seen following mbl overexpression (Fig. 3J and K).
All three single mreB isologues can support cell viability and rod-shaped growth
We then examined whether cells having these overexpression constructs could tolerate sequential deletion of the three endogenous mreB family genes. This was found to be readily achievable, and the transformation frequencies were high enough to exclude the requirement for acquisition of additional suppressor mutations (data not shown). Figure 4A shows that the three resultant strains could grow on PAB plates with no added Mg2+ provided inducer was present within a concentration range that varied according to the strain. Thus, the mreB overexpression strain (henceforth, ‘B-only’) grew optimally at 0.005 mM IPTG; mbl overexpression (L-only) at 0.02 mM; and mreBH (H-only) at 0.05 mM or more. In the presence of those IPTG concentrations, MreB and Mbl, at least, were overproduced (Fig. S2). Remarkably, when the cells were grown in liquid medium at their optimum IPTG concentration, in all three cases the cells appeared as relatively normal rods (Fig. 4B, D and F). Van-FL staining in the three strains gave a similar pattern of staining to the wild-type cells (compare Fig. 5A withFig. 5B, E and H). Removal of inducer provided a means of assessing the effects of near complete absence of MreB proteins. As shown in Fig. 4C, E and G, after the removal of IPTG, triple mutant cells containing any of the overexpression constructs became almost round and lytic, and PG synthesis in the lateral cell wall was largely inhibited (Fig. 5D, G and J). Therefore, all three of the mreB isologues are capable, independently, of supporting cell elongation and of specifying a rod shape by organizing PG synthesis in the lateral cell wall. This highlights and confirms that there is a substantial degree of functional overlap between the three proteins.
It has been reported that depletion of MreB affects chromosomal segregation in B. subtilis (Soufo and Graumann, 2003). In the course of carrying out these experiments we also stained the cells with DAPI and examined their DNA (Fig. 5, right). However, no significant evidence of anucleate cell formation was detected, consistent with a recent report that mreB mbl mreBH triple mutants (suppressed by rsgI deletion) also do not have a significant DNA segregation defect (Schirner and Errington, 2009).
MreB, Mbl and MreBH form a single complex
To test for the existence of complexes containing three MreB isoforms, we isolated protein complexes from cells expressing, in turn, His-tagged MreB, Mbl and MreBH. MreB–His was expressed from a Pxyl promoter at the amyE locus, to avoid the complication of having the important mreC and mreD genes downstream from the mreB gene (which was deleted in frame). Mbl–His and MreBH–His fusions were expressed from their natural promoters at their native loci. The growth rates and cell shapes of these strains were indistinguishable from that of the wild type (cells expressing MreB–His were cultivated in the presence of 0.25% xylose), indicating that all three fusion proteins are functional (data not shown). The theoretical molecular mass of MreB, Mbl and MreBH is 35.8, 35.7 and 35.5 kDa respectively. Purified proteins from cell cultures without cross-linking treatment were detected as a major single band of just under 40 kDa for MreB and Mbl, and a weak band of just over 40 kDa for MreBH, corresponding to the approximate molecular masses of His-tagged MreB, Mbl and MreBH fusions (Fig. 6A, lanes 1–3). After cross-linking, several co-purified proteins were also detected in these complexes (Fig. 6A, lanes 4–6). The most prominent additional bands were, in each case, slightly higher in mobility than the His-tagged species, likely corresponding to the untagged MreB-family proteins. For the MreB–His and Mbl–His complexes we anticipated that the second bands might correspond to Mbl and MreB respectively. In accordance with this prediction, these bands disappeared when complexes were purified from isogenic strains bearing deletions of the corresponding genes (i.e. mbl and mreB respectively; Fig. 6B, a and b). In both cases liquid chromatography tandem mass spectrometry analysis (LC-MS/MS) confirmed the identities of these band (data not shown). Since MreBH was less abundant in the cells (see Fig. 6A, lane 3), it was hard to detect as a clear visible band in the MreB or Mbl complexes. However, LC-MS/MS analysis detected MreBH in both the MreB–His and Mbl–His complexes, and also demonstrated the presence of MreB and Mbl in the MreBH complex (Fig. 6A, lane 6 and arrows). Furthermore, the intensity of the major band in the Mbl complex was increased when MreBH was overexpressed (Fig. 6B, c, arrow). Therefore, all three MreB isoforms can readily be detected in complex with each other. As a control, cells expressing His-tagged DnaC (a component of the replisome) were analysed in parallel, as shown previously (Kawai et al., 2009). However, no MreB isoforms were detected in equivalent samples from the MreB–His complexes.
Mbl associates with the PG synthetic machinery
We have shown that MreB associates with several PBPs in a complex probably via direct interaction (Kawai et al., 2009). The result that Mbl is also involved in the control of lateral PG synthesis prompted us to test for the existence of complexes containing Mbl and PBPs. Cells expressing His-tag fusions were treated with the fluorescent penicillin analogue Bocillin FL (Zhao et al., 1999) and formaldehyde to cross-link proteins, then disrupted by sonication, as described previously (Kawai et al., 2009). To detect specific interactions between either MreB or Mbl and the PBPs, MreB or Mbl complexes were purified from isogenic strains bearing deletions of the other isologues (i.e. mbl and mreBH, or mreB and mreBH respectively). Therefore, these strains contain only a single mreB isologue present in both wild-type and his-tagged states, which as shown above, allows cell growth. As a control, we first purified MreB complexes in the presence of Mbl and MreBH. Consistent with previous results (Kawai et al., 2009), fluorescently labelled bands corresponding approximately to the molecular masses of PBP1, PBP2a and PBP4, at least, were clearly detected in the MreB complexes (Fig. 7, lane 5). A band possibly corresponding to PBP5 was detected in the complexes, although the signal was weaker than other bands. These bands were also detected in both MreB and Mbl complexes purified from cells lacking other mreB homologues (Fig. 7, lanes 6 and 7). These results indicate that MreB and Mbl can associate with the PG synthetic machinery independently.
Inhibition of cell viability and rod-shaped growth in cells with single mreB isologues under stress conditions
The results described above indicated that all three single mreB isologues can support cell viability and rod-shaped growth under normal growth conditions (Fig. 4). As one approach to test whether the mreB isologues have any specialization of roles, we examined the viability of the B-only, L-only and H-only strains described above, under various stress conditions. As shown in Fig. 8A, in all stress conditions tested the wild-type strain grew well, similar to growth under normal conditions. In contrast, the H-only strain showed a defect in cell viability on PAB plates during heat stress. In alkaline stress, the H-only and B-only strains showed a strong defect in cell viability. In the case of salt stress, all mutant strains showed a complete inhibition of cell growth. The severe defects in cell viability suggest that the mreB isologues are functionally differentiated and that their different properties are required to cope with stress conditions. Phase-contrast microscopy showed that when the H-only strain was incubated at 50°C for 60 min it lost the ability to grow as a rod, becoming rounded and lytic (Fig. 8K), whereas the other strains showed relatively normal cell morphology (Fig. 8B, E and H). Under alkaline stress conditions, the B-only and H-only strains showed bulging and lysis (Fig. 8F and L), whereas the wild-type and L-only cells were relatively unperturbed (Fig. 8C and I). Under salt stress, all of the mutant strains showed severe bulging and lysis (Fig. 8G, J and M). Thus, no single mreB isologue is sufficient for normal growth and morphology of B. subtilis under the stress conditions tested.
Helical insertion of nascent PG synthesis is governed by the redundant action of MreB isoforms
Previous work and our new results showed that mbl mreBH double mutants are viable but that other double mutant combinations are not, even at high Mg2+ conditions (Defeu Soufo and Graumann, 2006; A. Formstone and J. Errington, unpublished). We previously reported that PBP1 shows dynamic localization during the cell cycle and that its abnormal accumulation at the cell poles generates bulging poles and eventually cell lysis (Claessen et al., 2008; Kawai et al., 2009). The unusual cell wall synthesis at cell poles associated with PBP1 was observed in mreB mutant cells, but not in single or double mutants of mbl and mreBH. We also showed that MreB is required for normal localization of PBP1 at higher Mg2+ concentrations (< 1 mM), while Mbl and MreBH are only required at low Mg2+ concentrations (< 0.1 mM) (Kawai et al., 2009). These observations suggest that MreB alone can support the dynamic localization of PBP1 at high Mg2+ conditions, and could explain why mbl mreBH was the only viable double mutant combination among the mreB homologues.
The discovery that PG is inserted in a helical manner provided strong support for the idea that the MreB filaments direct the synthesis of new PG in a spatially controlled manner (Daniel and Errington, 2003). However, helical incorporation of PG along the lateral cell wall was still evident in mreB mutant cells. Continued lateral wall PG synthesis was also observed in single and double mutants of the mreB isologues, mbl and mreBH. On the other hand, formation of spherical cells after depletion of MreB in an mbl mutant strongly supported the idea of a functional overlap between mreB and mbl in the control of cell elongation. Strikingly, the helical pattern of PG staining was virtually absent in these mutant cells. Thus, either MreB or Mbl can support lateral wall PG synthesis and maintenance of the rod-shaped form of B. subtilis cells. In contrast, depletion of MreB in an mreBH mutant still maintained rod-shaped growth under normal conditions (at 30°C or 37°C in PAB medium), although polar bulges were frequently observed. Interestingly, during heat stress (at 48°C) the mutant cells showed a much stronger defect in cell shape with progressive bulging and lysis (data not shown). In addition, a single mreBH mutant also showed severe bulging under heat stress (data not shown). A recent report showed that the transcription of mreBH is activated by heat stress in a SigI-dependent manner (Tseng and Shaw, 2008). Therefore, the contribution of MreBH to cell elongation could be especially important during heat stress.
All three single MreB isoforms can control cylindrical cell wall synthesis by organizing the cell elongation machinery
We found here that overexpression of any one of the MreB isoforms enabled us to delete all three endogenous mreB isologues. In each case the cells maintained their rod shape, with apparently helical incorporation of PG along the lateral cell wall. Absence of all three MreB isoforms led to the loss of lateral wall PG synthesis and acquisition of a spherical shape, similar to that reported for an mreB mbl mreBH triple mutant observed in the presence of a lethality suppressing rsgI mutation (Schirner and Errington, 2009). Therefore, all three of the MreB isoforms can independently support cell elongation and specify a rod shape.
All three MreB isoforms colocalize, forming a helical structure spanning the length of the cell (Carballido-López et al., 2006). Interactions between the three proteins have been also detected by BiFC and FRET experiments (Defeu Soufo and Graumann, 2006). We have now found that the three MreB isoforms can be detected in close association by cross-linking in vivo. These findings suggest that the MreB isoforms physically interact and might form a single mixed heteropolymeric cable in the cell. The fact that the proteins occupy similar space in the cell would again suggest the idea of functional redundancy. It has been shown that the discrete localizations of several PBPs are independent of either MreB or Mbl individually in B. subtilis (Scheffers et al., 2004). However, based on our new findings, these PBPs could be positioned by any of the three MreB isoforms, acting redundantly. Indeed, direct interactions between MreB and several PBPs have been detected (Kawai et al., 2009). Although it is not clear whether interactions between Mbl or MreBH and PBPs are direct or indirect, several PBPs were detected in complexes with Mbl even in the absence of MreB and MreBH. Therefore, Mbl can associate with the PG synthetic machinery by itself. Since MreBH was less abundant, it was hard to detect clear PBP bands in the MreBH complexes by our pull-down experiments (data not shown). Nevertheless, LC-MS/MS analysis identified the presence of PBP1 and PBP2a, at least, in the MreBH complexes (data not shown). Thus, all three MreB isoforms cooperate and play partially redundant roles in orchestrating the helical elongation of the lateral cell wall of B. subtilis, probably by organizing localization of the cell wall synthetic machinery.
Multiple MreB isoforms are required for rod-shaped growth of B. subtilis under various stresses
All single MreB isoforms have the ability to support rod-shaped cell growth in normal growth conditions. However under several general stress conditions, including heat, alkaline or high salt stress, growth of mutant cells expressing single mreB isologues was largely inhibited and the cells showed strong defects in cell morphology. The results demonstrate that the multiplicity of MreB isoforms becomes important under stress conditions. Although it seems premature to assign specific functions to the isoforms on the basis of the present data, the results support the general idea that the three isoforms have partially differentiated roles in adaptation to stresses. In general, adaptation to environmental stresses is crucial for the survival of bacteria in nature. Therefore, bacteria cope with stress by changing their spatial organization, including their morphology. It has been reported that changes in the expression levels and localization of MreB in Vibrio parahaemolyticus, under various stresses, are related to changes in subcellular architecture and cell morphology (Chiu et al., 2008). In E. coli, a general stress response gene, bolA, inhibits cell elongation by repressing expression of mreB (Freire et al., 2009). In B. subtilis, mreBH gene is a member of the SigI regulon and induced by heat stress (Tseng and Shaw, 2008). In addition, B. subtilis mreB is upregulated by overexpression of several extracytoplasmic function (ECF) sigma factors (Asai et al., 2003), which are required in response to changes in stress conditions (Lonetto et al., 1994). MreB and Mbl are also upregulated in the K-state, which is a growth-arrested state distinct from sporulation and a unique adaptation to stress (Berka et al., 2002). These findings suggest that B. subtilis cells respond to stresses by reconfiguring or reconstituting their MreB cytoskeleton, to adjust cell wall synthesis to the changing needs of the cell.
Bacterial strains, growth conditions and media
Bacillus subtilis strains used in this study are listed in Table 1. Bacillus strains were grown at 30°C and 37°C in LB medium or Difco antibiotic medium 3 (PAB) supplemented, where required, with IPTG, xylose and/or 10 mM MgSO4. If necessary to ameliorate the lytic phenotype of mreB mutants, 0.3 M sucrose was added to PAB medium. Transformation of B. subtilis cells was performed according to the protocol previously reported (Moriya et al., 1998). Bacillus transformants were selected on PAB agar supplemented with 50 μg ml−1 spectinomycin, 5 μg ml−1 chloramphenicol, 5 μg ml−1 kanamycin, 1 μg ml−1 erythromycin and/or 25 μg ml−1 lincomycin. DNA manipulation and E. coli transformation were performed by standard methods (Sambrook et al., 1989).
The mbl or mreBH gene containing the Shine–Dalgano sequence was amplified by PCR from genomic DNA of wild-type strain 168, then cloned between the SmaI and BglII, or HindIII and SphI sites of plasmid pPL82 (Quisel et al., 2001), creating pPL82–mbl and pPL82–mreBH. The resulting plasmid was used to transform 168, with selection for chloramphenicol resistance, to generate ASK4601 or ASK4602 in which the full-length mbl or mreBH are expressed from the IPTG-inducible promoter PspacHY at the amyE locus on the B. subtilis chromosome.
Construction of a triple mutant of mreB isologues
To obtain strain YK754 (ΔmreBΔmblΔmreBH amyE::PspacHY–mreB), an mbl mutation (mbl::spc) was first introduced into strain YK735 (ΔmreB amyE::PspacHY–mreB), and the transformants were selected on LB plates with kanamycin, spectinomycin and 10 mM MgSO4. Then, an mreBH mutation (mreBH::erm) was introduced into the resultant strain YK745 (ΔmreBΔmbl amyE::PspacHY–mreB), and the transformants were selected on LB plates with kanamycin, spectinomycin, erythromycin and 10 mM Mg2+.
To obtain strain YK755 (ΔmreBΔmblΔmreBH amyE::PspacHY–mbl), an mreBH null mutation (mreBH::erm) was first introduced into strain YK721 (Δmbl amyE::PspacHY–mbl), and the transformants were selected on LB plates with spectinomycin, erythromycin and 10 mM Mg2+. Then, an mreB mutation (3725) was introduced into the resultant strain YK752 (ΔmblΔmreBH amyE::PspacHY–mbl), and the transformants were selected on LB plates with kanamycin, spectinomycin, erythromycin, 0.1 mM IPTG and 10 mM Mg2+.
To obtain strain YK756 (ΔmreBΔmblΔmreBH amyE::PspacHY–mreBH), an mbl mutation (mbl::spc) was first introduced into strain YK722 (ΔmreBH amyE::PspacHY–mreBH), and the transformants were selected on LB plates with erythromycin, spectinomycin and 10 mM Mg2+. Then, an mreB mutation (3725) was introduced into the resultant strain YK753 (ΔmblΔmreBH amyE::PspacHY–mreBH), and the transformants were selected on LB plates with kanamycin, spectinomycin, erythromycin, 0.1 mM IPTG and 10 mM Mg2+. Based on efficient transformation frequencies, and backcross experiments, we do not think that the formation of these multiple mutants required a suppressor mutation.
Construction of histidine-tagged fusions of mbl and mreBH
The pMUTinHis plasmid allows a translational fusion of 12 × histidine-tag to the 3′-terminus of target gene (Ishikawa et al., 2006). The C-terminal regions of the mbl or mreBH gene were amplified by PCR from the wild-type strain 168 genomic DNA using primers mbl–his-F (which contains a EcoRI site, 5′- GAAGAATTCAGTACAAGCTGCTGATCG-3′) and mbl–his-R (which contains a XhoI site, 5′-CTCCTCGAGGCTTAGTTTGCGTTTAGG-3′), or mreBH–his-F (which contains a EcoRI site, 5′-GAAGAATTCTGGAAATTGGCCATGCTC-3′) and mreBH–his-R (which contains a XhoI site, 5′-CTCCTCGAGTTTAATTGCCTTTTGCAGC-3′), then cloned between the EcoRI and XhoI sites of plasmid pMUTinHis, creating pMUTinHis–mbl and pMUTinHis–mreBH respectively. The resulting plasmid was used to transform 168, with selection for erythromycin resistance, to generate YK403 or YK404.
Depletion of Mg2+, IPTG and/or xylose
To remove extra Mg2+, IPTG and xylose in PAB liquid medium, cells were grown in PAB medium with added Mg2+, IPTG and/or xylose to mid-exponential phase, and collected by centrifugation. The cells were then washed once in PAB medium without added Mg2+, IPTG and/or xylose, and then diluted back to an OD600 of 0.02–0.05 into fresh PAB medium.
Fluorescent vancomycin staining
Fluorescent vancomycin staining was performed, essentially as described previously (Daniel and Errington, 2003). To label sites of nascent PG synthesis, cells were grown at 30°C in PAB medium with 0.3 M sucrose, treated with fluorescently labelled vancomycin (BODIPY FL Conjugate, Invitrogen), and then fixed in 3.6% formaldehyde.
For fluorescence microscopy, cells from overnight culture in PAB liquid medium or on PAB plates were diluted into fresh PAB medium and grown to mid-exponential phase at 30°C or 37°C. For live cell imaging, cells were mounted on microscope slides covered with a thin film of 1.2% agarose in water, essentially as described previously (Glaser et al., 1997). Images were acquired with a Sony Cool-Snap HQ cooled CCD camera (Roper Scientific) attached to a Zeiss Axiovert 200 M microscope. The images were acquired and analysed with METAMORPH version 6 software.
Purification of protein complexes
An overnight liquid culture of B. subtilis cells expressing histidine-tagged protein in LB medium containing 0.5 μg ml−1 erythromycin, 0.25% xylose, 0.1 mM IPTG and/or 10 mM MgSO4, if necessary, at 37°C was inoculated into 400 ml of the same medium. When the cells, grown at 37°C, reached an OD600 of 0.5–0.6, the culture was treated with formaldehyde (1% final concentration) for 10 min. The culture was incubated for an additional 10 min after addition of 150 mM glycine. Purification of protein complex with His-tagged protein and label of PBPs by Bocillin FL was performed as described previously (Ishikawa et al., 2006; Kawai et al., 2009). In-gel enzymatic digestion, peptide analysis and protein identification by LC-MS/MS were carried out as described by Kuwana et al. (2002).
We thank the group members for helpful discussions, and particularly K. Schirner and R.A. Daniel for critical reading of the manuscript. We also thank K. Schirner, M. Leaver and A. Formstone for the gift of strains, and S. Ishikawa and M. Kuwano for technical assistance. This work was supported by a grant from the UK Biotechnology and Biological Sciences Research Council.