The “Pro” Sequence of the Sporulation-Specific ς Transcription Factor ς^E Directs It to the Mother Cell Side of the Sporulation Septum

Abstract

ς^E, a mother cell-specific transcription factor of sporulating Bacillus subtilis, is derived from an inactive precursor protein (pro-ς^E). Activation of ς^E occurs when a sporulation-specific protease (SpoIIGA) cleaves 27 amino acids from the pro-ς^E amino terminus. This reaction is believed to take place at the mother cell-forespore septum. Using a chimera of pro-ς^E and green fluorescent protein (GFP) to visualize the intracellular location of pro-ς^E by fluorescence microscopy, and lysozyme treatment to separate the mother cell and forespore compartments, we determined that the pro-ς^E::GFP signal, localized to the forespore septum prior to lysozyme treatment, is restricted to the mother cell compartment after treatment. Thus, pro-ς^E::GFP had been sequestered to the mother cell side of the septum. This segregation of pro-ς^E::GFP, and presumably pro-ς^E, to the mother cell is likely to be the reason why ς^E activity is restricted to that compartment.

At an early stage in endospore formation, Bacillus subtilis portions itself into two compartments of unequal size contained within a common cell wall. Each of these compartments has a unique developmental fate. The smaller, prespore compartment is engulfed by the larger, mother cell compartment, which nurtures it during subsequent stages of development. When the sporulation process is complete, the mother cell lyses and the mature spore is freed into the environment. The individualized program of gene expression for each of these compartments is controlled by the sequential appearance of compartment-specific transcription factors: ς^E and then ς^K in the mother cell and ς^F and then ς^G in the forespore (reviewed in reference 34). The first of the mother cell-specific ς factors (ς^E), as well as its counterpart in the prespore (ς^F), is synthesized at the onset of sporulation, but neither ς^E nor ς^F becomes active until after the septation event establishes the two compartments (6, 10, 19, 21, 24, 25, 32, 35, 36, 37). ς^E and ς^F are each silenced by distinct mechanisms. ς^F is held inactive in a complex with an inhibitor (SpoIIAB), while ς^E is synthesized as an inactive proprotein (1, 5, 7, 8, 21, 22, 25, 31, 33).

ς^F is freed from SpoIIAB by the action of a second protein (SpoIIAA), which binds to SpoIIAB in lieu of ς^F (1, 5, 8). In the preseptal cell, SpoIIAA is phosphorylated and inactive (8, 9, 25). SpoIIAA remains inactive in the mother cell but is activated in the prespore by SpoIIE, a membrane-bound phosphatase (2–4, 7). In contrast, pro-ς^E is activated when 27 amino acids are removed from its amino terminus (21, 26). Pro-ς^E processing is catalyzed by the sporulation-specific protease SpoIIGA (15, 29). SpoIIGA, like pro-ς^E, is present in the preseptal cell but is inactive until the septum forms (28). Pro-ς^E and SpoIIGA localize to the septum (11, 13, 16, 17). Processing is initiated when SpoIIR, a ς^F-dependent gene product, traverses the septum and triggers SpoIIGA to cleave the “pro” sequence from ς^E (14, 18, 23, 39).

The mechanisms by which ς^E and ς^F become activated in only one of the two compartments have been the subject of much speculation (1–3, 13, 14, 17, 24, 31, 39). Fusions of the phosphatase (SpoIIE) that is responsible for ς^F activation and green fluorescent protein (GFP) localize to the forespore septum (3, 4, 7). Recently, Wu et al. reported that septum-bound SpoIIE-GFP is preferentially released into the forespore compartment following lysozyme treatment (38). This was interpreted as evidence for sequestration of SpoIIE to the forespore face of the septum and provided a plausible explanation for how the activation of ς^F could be limited to the forespore. The conclusion that SpoIIE is sequestered to the forespore has been questioned by others, who argue that the larger volume of the mother cell could reduce the intensity of the GFP signal, leading to the impression of forespore localization even if similar amounts of SpoIIE-GFP are released from both sides of the septum (20). Although the interpretation of the SpoIIE-GFP result is controversial, the technique of separating forespore from mother cell by lysozyme treatment remains useful.

We previously used a fusion of the ς^E pro sequence and GFP to show that the pro sequence can tether proteins to the forespore septum (16, 17). In the present work, we employed the same fusion to determine whether protoplasting of these cells would reveal a preferential localization of the SigE-GFP fusion to either mother cell or forespore. The B. subtilis strain that we examined carries a sigE::gfp fusion in which the coding element for 55 amino acids from the amino terminus of SigE is joined to GFP and integrated into the B. subtilis chromosome at sigE (i.e., P_spoIIG::sigE55::gfp) (Table 1). Due to the integration event, the strain expresses the fusion from the sigE promoter in lieu of sigE. The strain also carries a null mutation in the processing-essential spoIIAC gene (spoIIAC::erm) so as to insure that the pro sequence is not cleaved from GFP due to normal sporulation processing (16). Without SpoIIAC or a source of intact ς^E, the strain is Spo⁻ and does not proceed past stage II of sporulation (i.e., the sporulation septum forms but the forespore is not engulfed by the mother cell). Instead of prespore engulfment, a second septum is laid down at the pole of the cell opposite that at which the first septum appeared. This disporic morphology, with two forespore compartments bracketing the former mother cell, is characteristic of cells that lack ς^E. Figure 1A consists of micrographs of this strain at the disporic stage. As we had previously observed (16, 17), the pro-ς^E55::GFP protein preferentially localizes to the two sporulation septa (Fig. 1A₂). To determine whether this septum-bound material is present on both sides of the septa, we treated samples of the culture with lysozyme to protoplast the cells and partially separate the compartments. When this was done, the GFP signal was seen to localize exclusively to the membrane of the mother cell compartment (Table 2). This is evident in Fig. 1B, where the GFP signal (Fig. 1B₂) forms a ring around the central mother cell compartments (Fig. 1B₁), while the DAPI (4′,6-diamidino-2-phenylindole)-stained chromosomes are partitioned to the forespore compartments that bracket them (Fig. 1B₃).

TABLE 1.

B. subtilis strains and plasmids

Strain or plasmid	Relevant genotype or features	Source, construction,a or reference
Strains
SMY	trpC	Laboratory strain
SEK84	kan/sigEΔ84	Laboratory strain
SFG1	PdacF::sigE55::gfp	16
SFG7	kan/sigEΔ84 PdacF::sigE55::gfp	SEK84→SFG1
SPF3	PspoIIG::sigE55::gfp spoIIAC::erm	15
SPF5	PspoIIG::sigE335::gfp	pJM335GM→SMY
Plasmid
pJM335GM	bla cat sigE335-gfp	This study

^aAn arrow indicates construction of the strain by transformation of DNA from the source on the left of the arrow into the strain on the right of the arrow. 

FIG. 1.

Localization of pro-ς^E55::GFP in sporulating B. subtilis. Stationary-phase B. subtilis cells were diluted 1/200 in DS medium and incubated for 12 h at 30°C, an interval in which wild-type B. subtilis reaches stage III to stage IV (15) and the SigE⁻ P_spoIIG::sigE55::gfp strain displays a disporic morphology. Samples were prepared to maximize the GFP signal, as previously described (16); stained with DAPI; and either directly examined by microscopy (A) or treated on the slide with lysozyme (34) (4 mg/ml for 30 to 60 s) and then examined (B). The cells were visualized by phase-contrast microscopy (A₁ and B₁) and by fluorescence microscopy to visualize either the pro-ς^E::GFP fusion (A₂ and B₂) or DAPI-stained DNA (A₃ and B₃). The images in each series depict the same cells under each of the viewing conditions.

TABLE 2.

GFP localization

Fusion	No. of cellsa
	Untreated		Lysozyme treated
	Cytosol	Septum	MC	FS	Both
PspoIIG::sigE55::gfp	5	62	42	0	1
PspoIIG::sigE335::gfp	77	0	2	0	84
PdacF::sigE55::gfp	56	0	0	34	41

^aFor cells not treated with lysozyme (untreated), the numbers of cells observed with GFP either throughout the cell (cytosol) or at the septum are indicated. For lysozyme-treated cells, the numbers of cells with localization of the fluorescent signal to the mother cell (MC), forespore (FS), or both compartments are given. Cells were grown at 30°C in DS medium for a time equivalent to that in which a wild-type culture would reach stage III or IV. Untreated samples were prepared and examined as previously described (16). Lysozyme-treated samples were incubated on the slide in observation buffer with lysozyme (4 mg/ml) for 30 s to 1 min before the image was captured. At least 90% of cells in each culture displayed a GFP signal. 

As a test of whether the apparent sequestration of the GFP signal to the mother cell is due to a directed localization of pro-ς^E55::GFP to the mother cell side of the septum or is an artifact arising from differences in the sizes of the two compartments and the amount of GFP that could be trapped in each, we repeated the experiment using a GFP fusion protein in which gfp was joined to the amino terminus of a sigE allele (sigE335) which lacks the first 15 amino acids of the ς^E pro sequence (27). Unlike the fusion protein which carried the intact pro sequence, the GFP in B. subtilis expressing P_spoIIG::sigE335::gfp did not localize to the septum (Fig. 2A₂) and was easily discernible in both the mother cell and forespore compartments following protoplasting (Fig. 2B₂) (Table 2). Thus, membrane tethering is required for mother cell localization.

FIG. 2.

Localization of the P_spoIIG::sigE335::gfp product in sporulating B. subtilis. B. subtilis SPF5 (P_spoIIG::sigE335::gfp) was grown and examined as described in the legend to Fig. 1. Cells were untreated (A) or lysozyme treated (B) and viewed with phase-contrast microscopy (A₁ and B₁) and with GFP-enhanced (A₂ and B₂) or DAPI-enhanced (A₃ and B₃) fluorescence microscopy.

As a further test of our ability to visualize GFP in the forespore, we placed the sigE55::gfp fusion under control of the ς^F-dependent dacF promoter (36, 37). This construction does not make the strain Spo⁻. To give the cells the same terminal phenotype as the strains we had used in the experiments illustrated by Fig. 1 and 2, we introduced a null allele of sigE (sigEΔ84) into the strain (27). When the sigEΔ84 P_dacF::sigE55::gfp strain was examined, the GFP signal could be seen to preferentially accumulate in the forespore compartments (Fig. 3). There was, however, a weaker but discernible GFP signal in the mother cell (Table 2; Fig. 3A₂ and B₂). This is likely due to active ς^F in the mother cell compartment of this mutant strain. The SpoIIE phosphatase, which activates ς^F and which normally disappears from the mother cell following septation, has been shown to persist in the mother cell compartment if sporulation is blocked due to the absence of ς^E (30). Thus, partial activation of ς^F and limited expression of dacF in this mutant’s mother cell compartment would not be unexpected. An additional feature of the GFP pattern is that one of the forespore compartments typically gave a stronger signal than the other (Fig. 3B₂). Presumably, the stronger signal represents the compartment that formed as a result of the first asymmetric division, with the second appearing later and accumulating less GFP.

FIG. 3.

Localization of the P_dacF::sigE55::gfp product in sporulating B. subtilis. B. subtilis SFG7 (sigEΔ84 P_dacF::sigE55::gfp) was grown and examined as described in the legend to Fig. 1. Cells were untreated (A) or lysozyme treated (B) and viewed with phase-contrast microscopy (A₁ and B₁) and GFP-enhanced (A₂ and B₂) or DAPI-enhanced (A₃ and B₃) fluorescence microscopy.

Translocation of pro-ς^E to the mother cell side of the septum was proposed by Hofmeister as a plausible mechanism for its mother cell specificity (13). Our present finding that the fluorescent signal of pro-ς^E::GFP, previously localized at the forespore septum, becomes restricted to the mother cell compartment following lysozyme treatment supports this model. The conclusion that the GFP signal is localized to this compartment is not subject to the criticism leveled at the apparent localization of SpoIIE-GFP to the forespore (20). The mother cell is the larger of the two compartments and, as such, is the compartment in which the intensity of a GFP signal would be more likely to dissipate when released from the septum (20). We conclude that the ς^E pro sequence not only targets ς^E to the sporulation septum but also allows ς^E to be preferentially sequestered to the mother cell side of the forespore septum. In earlier studies (17), we found that SigE, but not SigE-GFP, is preferentially degraded in the forespore compartment. Based on that result, we suggested that ς^E degradation could be responsible for the absence of ς^E activity in the forespore (17). The present data argue that pro-ς^E mobilization to the mother cell side of the septum is likely to be the primary vehicle that places ς^E activity in the mother cell and that the degradation of ς^E in the forespore probably represents a secondary device to destroy any ς^E which might inadvertently form in the forespore.

The ability of the SigE pro sequence to target pro-ς^E to the mother cell side of the septum is intriguing. The primary sequence of the pro region suggests that it has an alpha-helical structure with hydrophobic and positively charged faces (27). In this regard, it resembles antimicrobial peptides that are believed to associate with membrane phospholipids via their positively charged faces (12). Presumably, such a structure would allow the pro sequence to be targeted to cell membranes. Once membrane bound, more specific amino acid residues might make contact with yet-to-be-defined membrane proteins for sequestration to the developing septum. It is interesting to note that treatment with lysozyme caused a dispersal of pro-ς^E55::GFP over the mother cell membrane (Fig. 1B₂) from its previous location at the septal poles (Fig. 1A₂). Apparently, an organizing element is lost when the integrity of the peptidoglycan is degraded.