The spoIIE Locus Is Involved in the Spo0A-Dependent Switch in the Location of FtsZ Rings in Bacillus subtilis

Abstract

A switch in the location of FtsZ ring structures from medial to polar is one of the earliest morphological indicators of sporulation in Bacillus subtilis. This switch can be artificially caused during vegetative growth by induction of an active form, Sad67, of the transcription regulator, Spo0A (P. A. Levin and R. Losick, Genes Dev. 10:478–488, 1996). We have used immunofluorescence microscopy to show that the switch in FtsZ ring location during vegetative growth caused by Sad67 induction is blocked by a spoIIE deletion mutation. The spoIIE mutation also impaired polar FtsZ ring formation during sporulation. These results suggest that SpoIIE mediates the Spo0A-directed formation of polar FtsZ rings.

A central issue in developmental biology is how the progeny of a single cell division assume dissimilar fates. An attractive experimental system in which this problem can be addressed is sporulation in Bacillus subtilis, in which the formation of an asymmetrically positioned septum partitions the developing cell into a prespore and a mother cell (12, 25, 26, 30). The smaller compartment, the prespore, becomes the spore, whereas the larger compartment, the mother cell, participates in the maturation of the spore but eventually lyses to release the spore when development is complete. We explore here the mechanism responsible for the switch from medial division during vegetative growth to asymmetric division at the onset of sporulation.

One of the earliest morphological indicators of the onset of sporulation is the switch in the location of cell division protein FtsZ ring structures from mid-cell to near the cell pole (21). Although no direct regulator of septum position in B. subtilis has been identified, activation of the Spo0A transcription factor (16) is a prerequisite for the formation of polar FtsZ rings (21). Mutations in spo0A arrest sporulation at stage 0, prior to the asymmetric division (26). Spo0A is activated by phosphorylation at the onset of sporulation (stage 0) through the action of a phosphorelay (16, 17). Deletions in the amino-terminal region of the protein cause it to be active even in the absence of phosphorylation (18). Levin and Losick (21) showed that the expression of one such spo0A mutant allele, sad-67, during vegetative growth was sufficient to cause the switch in FtsZ ring location from medial to polar.

Spo0A is the earliest-acting and most pleiotropic sporulation transcription factor and is directly or indirectly responsible for activation or repression of a number of B. subtilis genes. It seemed plausible that its effect on septum position was mediated by some locus expressed early in sporulation. Phosphorylated Spo0A is known to activate directly transcription of three spo loci that are transcribed before septation (16, 30). Two of them (spoIIA and spoIIG) contain the structural genes for the first prespore-specific and mother-cell-specific transcription sigma factors, ς^F and ς^E, respectively (30). However, they are not involved in septation. The third locus directly activated by Spo0A is spoIIE, and the role of this locus in septation is explored here.

SpoIIE is a bifunctional protein, which contains multiple membrane-spanning domains in its N-terminal portion and a cytoplasmic tail in its C terminus (5). The SpoIIE C-terminal domain is a serine protein phosphatase, which plays a key role in the activation of ς^F (1, 2, 7, 8). The transcription of ς^F-controlled genes commences shortly after the formation of the polar septum (14, 32) and is strictly confined to the prespore by a mechanism operating at the level of the activity of the ς^F protein (reviewed in reference 30). SpoIIE is responsible for dephosphorylation (and thereby activation) of SpoIIAA-P. SpoIIAA is an anti-anti-sigma factor that counteracts the inhibitory effect of SpoIIAB by binding to the SpoIIAB-ς^F complex and by causing the release of free and active ς^F. In the predivisional sporangium, SpoIIE localizes to both sites of potential polar division. A septum is formed at one of the poles, and SpoIIE disappears from the distal pole while persisting at the septum. Thus, at the time of septation, SpoIIE localizes at the boundary between the mother cell and the prespore (2, 5, 29). It is suspected that polar localization of SpoIIE in the sporulation septum might be the cause of prespore-specific activation of ς^F.

SpoIIE is also involved in spore septum formation (26). Barák and Youngman (6) and Feucht et al. (13) characterized a group of null spoIIE mutants in which as much as 80% of the cells had no detectable septum formation after the initiation of sporulation, while about 20% had aberrant thick septa; formation of these septa was delayed (13). There is evidence of interaction between SpoIIE and FtsZ (29), and it seemed possible that SpoIIE mediates the Spo0A-dependent switch in the FtsZ ring location. Evidence supporting such a role is presented in this paper.

MATERIALS AND METHODS

Media.

B. subtilis was grown in modified Schaeffer’s sporulation medium (MSSM), in Luria broth with glucose, and on Schaeffer’s sporulation agar (27). When required, chloramphenicol at 3 μg/ml, neomycin at 3 μg/ml, and erythromycin at 1 μg/ml were added.

Strains.

The B. subtilis 168 strain BR151 (trpC2 metB10 lys-3) was used as the parent strain. The B. subtilis strains used are listed in Table 1. Escherichia coli DH5α (GIBCO-BRL) was used to maintain plasmids.

TABLE 1.

B. subtilis strains used

Strain	Relevant characteristics	Origin or reference
BR151	trpC2 metB10 lys-3	Lab stock
SL6422	trpC2 metB10 lys-3 spoIIR::neo	Margaret Karow
SL7240	trpC2 metB10 lys-3 spoIIE::neo	This paper
SL7243	trpC2 metB10 lys-3 Pspac-spoIIE	This paper
SL7260	trpC2 metB10 lys-3 Pspac-sad-67	Petra Levin (18)
SL7261	trpC2 metB10 lys-3 Pspac-sad-67, spoIIE::neo	SL7240→SL7260a
SL7275	trpC2 metB10 lys-3 Pspac-sad-67, spoIIR::neo	SL6422→SL7260a

^aStrain constructed by transformation (arrow indicates donor to recipient strain). 

The spoIIE gene was cloned on a 2.8-kb fragment by PCR into pBluescript SK⁺ with the following primers: TGTAGCATGCAAGCGGGTCTTCCCC and CAAGCGGGTCTTCCCCATGG. A spoIIE disruption was constructed by replacing the PstI fragment extending from codons 142 to 628 within the spoIIE open reading frame (5) with a neo cassette in the opposite orientation and by isolating a Neo^r transformant of BR151 in which spoIIE had been disrupted by double-crossover recombination (strain SL7240).

A 0.67-kb EcoRI-PstI fragment containing the spoIIE promoter region and extending into the 5′ end of the gene was cloned into the SmaI site of pDH88, a plasmid designed for placing genes under the control of the P_spac promoter (15). Integration of the resulting plasmid at spoIIE by single crossover into BR151 produced a strain (SL7243) in which spoIIE was under the control of P_spac.

Immunofluorescence microscopy.

Cells were prepared and fixed for immunofluorescence microscopy essentially as described elsewhere (21, 32). Briefly, cells were fixed in 30 mM NaPO₄ buffer (pH 7.5) with a final concentration of 2.5% (vol/vol) paraformaldehyde for 15 min at room temperature and 45 min on ice prior to being washed in phosphate-buffered saline (PBS). Localization of FtsZ utilized affinity-purified polyclonal antibodies. Rabbit antibodies generated against the B. subtilis FtsZ protein (kindly provided by J. Lutkenhaus) were used in a 1:300 dilution in PBS with 2% bovine serum albumin. Secondary antibodies coupled to the Cy3 fluorophore were purchased from Jackson ImmunoResearch (Bar Harbor, Maine). FtsZ structures were visualized as bands in micrographs of longitudinal cells. These bands were inferred to represent ring-like structures that circle the rod-shaped organism (21, 23). Cells were visualized by phase-contrast microscopy with a yellow conversion filter for daylight color film. Photography and quantitation of cell types were performed as described previously (32).

DNA manipulation.

The procedure for transformation of B. subtilis was described previously (28). Other DNA manipulations were based on the procedures described by Ausubel et al. (3). Other methods have been described previously (27).

RESULTS

Disruption of spoIIE impairs the switch in FtsZ ring position caused by expression of sad-67.

A constitutively active form of Spo0A is known to induce the formation of polar FtsZ bands during vegetative growth (21). In agreement with this observation, we found that inducing the expression of the gene, sad-67, for one such mutant Spo0A protein switched the pattern of FtsZ assembly from medial to polar within 1 h of induction from an isopropyl-β-d-thiogalactopyranoside (IPTG)-inducible P_spac promoter during vegetative growth in MSSM. At this time, the FtsZ bands for 90% of cells were located at cell poles in the induced culture; the corresponding figure was 1% for the uninduced culture sampled at the same time (Table 2; Fig. 1A and B; typically, polar bands were not as sharp as medial bands). As noted by Levin and Losick (21), many cells in the induced culture had a bipolar rather than a unipolar band pattern (data not shown). Under the same conditions, FtsZ bands remained located predominantly at the middle of the cell in the strain containing a spoIIE deletion-insertion mutation in addition to the P_spac-sad-67 construction (Table 2; Fig. 1C and D), with very few cells (0 to 5% in different experiments) exhibiting a polar FtsZ distribution. Samples taken 1.5 h after induction gave a very similar result, with spoIIE largely preventing formation of polar FtsZ bands (Table 2). In contrast, the spoIIE mutation did not prevent formation of polar FtsZ bands in the 3-h sample (Table 2), which was taken approximately 2 h after the estimated start of sporulation.

TABLE 2.

Influence of spoIIE deletion on the change in pattern of FtsZ distribution caused by P_spac-sad-67 induction with IPTG^a

spoIIE genotypec	Time of sample (h after IPTG addition)	IPTG	No. of organisms showing the following FtsZ band locations:			% of cells with FtsZ bands in which the bands are polar
			Medial	Polarb	None
+	1	−	80	1	6	1
+	1	+	14	130	32	90
Δ	1	−	47	0	8	0
Δ	1	+	94	1	5	1
+	1.5	−	40	6	23	13
+	1.5	+	1	45	25	98
Δ	1.5	−	56	1	15	2
Δ	1.5	+	24	1	14	4
+	3	−	4	18	52	82
+	3	+	1	25	47	96
Δ	3	−	45	15	11	25
Δ	3	+	10	10	10	50

^aIPTG (to 1 mM) was added 1 h before the end of exponential growth in MSSM. 

^bIncludes cells showing bipolar and unipolar FtsZ bands. As noted by Levin and Losick (21), the pattern was predominantly bipolar in the early samples and predominantly unipolar at 3 h. 

^c+, strain SL7260; Δ, strain SL7261. 

FIG. 1.

Immunolocalization of FtsZ. Scale bar, 1 μm. The photographs are of cells immunostained with affinity-purified antibodies against the B. subtilis FtsZ protein (A, C, E, and F) and viewed by phase-contrast microscopy with a yellow filter (B and D). A secondary antibody conjugated to the red fluorophore Cy3 was used to visualize FtsZ. Localization of FtsZ in the strains containing P_spac-sad-67 1 h after IPTG addition during vegetative growth in spoIIE⁺ (SL7260) (A and B [arrows indicate polar FtsZ bands]) and spoIIE::neo (SL7261) (C and D [arrow indicates a medial FtsZ band]) backgrounds. FtsZ localization in the strain containing P_spac-spoIIE without (E [arrow indicates a medial FtsZ band]) and 30 min after (F [arrows indicate medial and polar FtsZ bands within a single cell]) IPTG addition.

The medium used in the studies discussed above, MSSM, supports efficient sporulation, and we also tested the effect of P_spac-sad-67 induction on bacteria grown in Luria broth, which does not support sporulation. Induction in Luria broth also caused a switch in the location of the FtsZ band (Table 3), although a longer induction period was required for the switch than in MSSM (data not shown). This switch in location was again dependent on spoIIE. The switch was not affected by disruption of another sporulation locus, spoIIR (Table 3).

TABLE 3.

Comparison of the effect of spoIIE and spoIIR mutations on the P_spac-sad-67-induced change in FtsZ band position^a

Fluorescence patternb	No. of cells showing pattern
	spo+ (SL7260)		spoIIE (SL7261)		spoIIR (SL7275)
	−	+	−	+	−	+
Midcell	35	45	44	120	20	8
Polarc	2	65	4	16	2	111
No localization	4	33	4	10	5	18

^aStrains were grown in Luria broth, and IPTG was added 3 h before the predicted end of exponential growth. 

^bValues were determined 3 h after IPTG addition and at the same time in parallel cultures with no IPTG. − and +, absence and presence of 1 mM IPTG. 

^cIncludes cells showing bipolar and unipolar FtsZ bands. 

The role of spoIIE in sporulation conditions was examined further with a spoIIE mutant, SL7240, that did not contain the P_spac-sad-67 construction. This strain also formed polar FtsZ bands in sporulation conditions. However, with this strain, as compared to the isogenic spoIIE⁺ strain (Fig. 2), their formation was delayed, and the proportion of bacteria with FtsZ bands was substantially reduced. Thus, in sporulation conditions, SpoIIE contributes to but is not essential for the switch to a polar site of FtsZ ring assembly. However, the spoIIE mutation did not appear to impede the loss during sporulation of the centrally located assembly site (Fig. 2), in agreement with a Spo0A-dependent, SpoIIE-independent mechanism for blocking medial septation.

FIG. 2.

The position of FtsZ bands under sporulation conditions of a spo⁺ strain and a spoIIE mutant. Bacteria were grown in MSSM. The start of sporulation was defined as the end of exponential growth. The pattern of FtsZ localization was visualized with immunofluorescence microscopy. Results are expressed as percentages of all cells. Open symbols, cells with medial FtsZ bands; closed symbols, cells with polar bands (bipolar or unipolar); squares, spo⁺ strain (BR151); circles, the isogenic spoIIE::neo strain (SL7240).

Expression of SpoIIE during vegetative growth disturbs the pattern of FtsZ distribution.

Because SpoIIE clearly mediates the effect of Spo0A on the FtsZ ring position, we suspected that SpoIIE overexpression might also disturb the pattern of FtsZ distribution. To this end, we constructed a strain, SL7243, containing spoIIE under the control of the IPTG-inducible P_spac promoter. Thirty minutes after induction of spoIIE expression during exponential growth, multiple FtsZ bands were detected in approximately 15% of the cells (Fig. 1F); induction for 60 min did not significantly increase this percentage. No cells with multiple FtsZ bands were detected when IPTG was not added (Fig. 1E). This result fits well with the report of Barák et al. that the induction of spoIIE during vegetative growth led to the formation of multiple septa (4, 5). Interestingly, in all of the cells with multiple FtsZ bands, the major band was still in the middle of the cell; the other bands were less intense, but were clearly visible and usually located at the poles. Thus, induction of SpoIIE during vegetative growth altered the profile of FtsZ distribution in cells but did not cause a complete switch of the pattern of FtsZ distribution from medial to polar. The observed phenotype differs significantly from the one resulting from induction of an active form of Spo0A, in which the medial FtsZ band was lost, and some 90% of FtsZ bands were located at the cell pole (Table 2). It is thought likely that some Spo0A-induced gene other than spoIIE is required in order to deactivate the mechanism for the assembly of an FtsZ band at the middle of the cell. However, the amount of SpoIIE was not determined, so that it is also possible that quantitative (or qualitative) differences in SpoIIE could account for the loss of the medial FtsZ band in the P_spac-sad-67 strain and not in the P_spac-spoIIE strain.

DISCUSSION

The switch of the site of assembly of FtsZ rings from the midcell to the cell pole is one of the first detectable morphological events after the initiation of sporulation (21, 23). Expression of a constitutively active form of the early sporulation-specific transcriptional factor Spo0A (encoded by sad-67) is sufficient to induce this switch even during vegetative growth (21). We have demonstrated here that disruption of spoIIE under these conditions largely prevented the change in position of FtsZ band localization during vegetative growth. The SpoIIE protein is known to have two roles, i.e., one as a phosphatase for SpoIIAA-P in the process of ς^F activation (8) and the other as a determinant of the location and structure of the sporulation septum (6, 13, 26). Levin and Losick (21) have shown that mutation in spo0H has no effect in the sad-67-induced switch in FtsZ ring position during vegetative growth, and Wu et al. (31) showed that Spo0H is required for transcription of the operon encoding ς^F. Thus, we think that the effect of SpoIIE on FtsZ ring location is a direct manifestation of its role in determining the location and structure of the sporulation septum and not an indirect consequence of ς^F activation.

It had previously been shown that under sporulation conditions, deletion of spoIIE substantially reduced the frequency of polar septum formation but did not prevent it. We observed that spoIIE deletion had a similar effect on polar FtsZ band formation (Fig. 2). Moreover, the spoIIE mutant showed a delay in formation of the polar FtsZ band, which matches well the delay previously reported for septum formation (13). Our results indicate that SpoIIE acts at or before the assembly of the FtsZ ring (FtsZ structures are visualized as bands which are inferred to represent ring-like structures [21, 23]), that is to say, in the initial stages of polar septum formation. While SpoIIE is not essential for the change in FtsZ position in sporulation conditions, it nevertheless contributes to polar FtsZ ring formation.

We speculate that during the normal sporulation process, there may be two alternative pathways leading to formation of the polar FtsZ ring, both of which depend on activation of Spo0A (Fig. 3). One of them requires the Spo0A-dependent recruitment of SpoIIE, while the second requires some other factor(s) (X) which functions only under sporulation conditions and is independent of SpoIIE. The second pathway is, by itself, inefficient, and the septa formed by this pathway resemble vegetative septa rather than sporulation septa in structure. Under the artificial conditions of sad-67 expression during vegetative growth, X is absent, leaving the SpoIIE-dependent pathway as the only option for altering the location of FtsZ band assembly by Spo0A activation. The localization of FtsZ to the middle of vegetative cells is ordinarily determined by the minCD genes (24). It is not at present clear how the Spo0A-SpoIIE system interacts with the MinCD system.

FIG. 3.

Schematic representation of the paths leading to polar FtsZ ring formation in B. subtilis. Spo0A-P is the active, phosphorylated form of Spo0A that is made at the start of sporulation. The Sad67 mutant form of Spo0A is active without phosphorylation and can substitute for Spo0A-P. The major path from active Spo0A is via SpoIIE. A minor SpoIIE-independent pathway can also function under sporulation conditions.

Since SpoIIE mediates the effect of Spo0A on septum location, one might expect that SpoIIE overexpression would also affect septum location, and this has been observed by Barák et al. (4, 5). In agreement with the observation of Barák et al., we have found that SpoIIE induction during vegetative growth induced the formation of FtsZ bands located near the cell poles in approximately 15% of the cells. We also noted that induction of spoIIE resulted in approximately 0.3% of the cells being anucleate minicells (data not shown). The persistence of the medial band as the major FtsZ band in these cells contrasts with its disappearance when sad-67 was induced during vegetative growth. This difference is consistent with the previously postulated role of Spo0A in blocking medial septation (9).

In contrast to our results and those of Barák et al. (4, 5), those of Levin et al. (22) demonstrated that expression of spoIIE during vegetative growth from a xylose-inducible promoter did not cause additional FtsZ bands to appear. The difference in results from the two sets of experiments most likely can be explained by the differences in the level of expression of SpoIIE protein under the weaker xylose-inducible and stronger P_spac promoters (19). A high level of SpoIIE expression might be required to induce the additional FtsZ ring formation because of, for example, competition with a MinCD-dependent mechanism restricting FtsZ rings to the middle of the cell.

Prior to spore septum formation, SpoIIE localizes in ring-like structures near cell poles (2, 5, 29). These SpoIIE structures coincide with the FtsZ rings. Levin et al. (22) showed that the localization of SpoIIE in such structures is dependent on FtsZ, because no ring-like SpoIIE structures were observed in sporulating cells from which FtsZ protein had been depleted, and instead SpoIIE seemed to be dispersed throughout the plasma membrane at the cell poles. We think it possible, nevertheless, that SpoIIE itself determines the polar location of the ring assembly during sporulation. SpoIIE and FtsZ may work together in a feedback mechanism, with SpoIIE determining the position of FtsZ ring assembly (for example, by providing the sites of initiation of FtsZ polymerization) and FtsZ determining the macrostructure of the rings with which SpoIIE is associated. In the absence of FtsZ structures, SpoIIE is unable to maintain its association with these nucleation sites.

Thus, we consider that SpoIIE either is required for (during vegetative growth) or facilitates (during sporulation) the Spo0A-mediated formation of polar FtsZ rings. The FtsZ protein is the procaryotic homolog of tubulin, and FtsZ rings are the analogs of eucaryotic microtubules (10, 11). Most microtubules undergo rapid assembly and disassembly. Eucaryotic cells contain many randomly oriented microtubules, which are stabilized by binding to structures such as kinetochores, whereas other microtubules not attached to such structures fall apart (20). It is possible that SpoIIE acts in a similar way to stabilize FtsZ structures at cell poles. SpoIIE involvement in localization and formation of the sporulation septum may in turn be crucial for coupling asymmetric septum formation with compartment-specific transcription activation.