Postdivisional Synthesis of the Sporosarcina ureae DNA Translocase SpoIIIE either in the Mother Cell or in the Prespore Enables Bacillus subtilis To Translocate DNA from the Mother Cell to the Prespore

Abstract

The differentiation of vegetative cells of Bacillus subtilis into spores involves asymmetric cell division, which precedes complete chromosome partitioning. The DNA translocase SpoIIIE is required to translocate the origin distal 70% of the chromosome from the larger mother cell into the smaller prespore, the two cells that result from the division. We have tested the effect of altering the time and location of SpoIIIE synthesis on spore formation. We have expressed the spoIIIE homologue from Sporosarcina ureae in B. subtilis under the control of different promoters. Expression from either a weak mother cell-specific (σ^E) promoter or a weak prespore-specific (σ^F) promoter partly complemented the sporulation defect of a spoIIIE36 mutant; however, expression from a strong prespore-specific (σ^F) promoter did not. DNA translocation from the mother cell to the prespore was assayed using spoIIQ-lacZ inserted at thrC; transcription of spoIIQ occurs only in the prespore. Translocation of thrC::spoIIQ-lacZ into the prespore occurred efficiently when spoIIIE_Su was expressed from the weak σ^E- or σ^F-controlled promoters but not when it was expressed from the strong σ^F-controlled promoter. It is speculated that the mechanism directing SpoIIIE insertion into the septum in the correct orientation may accommodate slow postseptational, prespore-specific SpoIIIE synthesis but may be swamped by strong prespore-specific synthesis.

The formation of spores from vegetative cells of Bacillus subtilis has become a paradigm for the study of cell differentiation in prokaryotes. In common with many such systems, an early stage is an asymmetric division that yields two cells with distinct developmental fates. These cells are the smaller prespore, which develops into the mature spore, and the larger mother cell, which is necessary for spore formation but ultimately lyses (14, 17). One of the most surprising features of the sporulation division is that septum formation precedes complete partitioning of a chromosome into the prespore. As a result, only the origin-proximal ∼30% of a chromosome is present in the prespore (also called the forespore) at the time of septation (24, 26). A DNA translocase, SpoIIIE, is required for the postseptational translocation of the origin-distal 70% of that chromosome from the mother cell to the prespore (2). The spoIIIE gene is transcribed during vegetative growth (6). However, spoIIIE mutants generally display defects only during spore formation, and they are blocked at stage III of sporulation (engulfment of the prespore by the mother cell) (14, 20). Consequently, it is thought that SpoIIIE functions primarily during spore formation.

The B. subtilis SpoIIIE protein is a 787-residue membrane protein (6). During spore formation, it is localized at the middle of the sporulation division septum, where it is required for translocation of a complete chromosome into the prespore (2, 25). It remains associated with the septal membrane and later moves to the cell pole, where it facilitates the completion of engulfment of the prespore by the mother cell (20). The mutant SpoIIIE36 protein is also localized at the middle of the sporulation division septum but does not migrate to the cell pole (20, 25). Chromosome translocation and completion of engulfment are blocked in spoIIIE36 mutants (20, 24, 25), but compartmentalization of gene expression, directed by σ^F in the prespore and by σ^E in the mother cell, is not affected (18, 24, 27). In contrast to the situation in a spoIIIE36 missense mutant, in spoIIIE null mutants compartmentalization of the activities of σ^F and σ^E is disrupted, as are DNA translocation and engulfment (18, 24, 27). Thus, the SpoIIIE protein is associated with three important processes during sporulation: DNA translocation, engulfment, and compartmentalization of gene expression. Here, we test the effect of altering the time and location of SpoIIIE synthesis on spore formation and, in particular, on DNA translocation.

Because DNA is translocated from the mother cell to the prespore, it seemed plausible that the location of SpoIIIE in the sporulation septum (20, 25) would have a polarity that favors such unidirectional chromosomal transfer. We have tested this possibility by manipulating the transcription of the spoIIIE gene from Sporosarcina ureae in B. subtilis. S. ureae is unusual among spore formers in that the volumes of the prespore and the mother cell are very similar when they are first formed (28). Consequently, gross asymmetry in volume between the two compartments cannot ordinarily determine the orientation of SpoIIIE insertion in S. ureae. It has been shown that spoIIIE from S. ureae, subsequently referred to as spoIIIE_Su, can complement spoIIIE mutants of B. subtilis and, importantly, does not undergo homologous recombination or homogenotization with B. subtilis spoIIIE (reference 5 and unpublished observations); homogenotization between alleles differing by a point mutation can occur in 25 to 50% of the progeny of a transformation (4). These properties have facilitated experiments, reported here, in which we test the effect of restricting expression of spoIIIE_Su to specific cell types during the sporulation of B. subtilis spoIIIE mutants. Compartmentalized activation of RNA polymerase sigma factors is a hallmark of spore formation. Two are activated after septum formation, σ^F exclusively in the prespore and σ^E exclusively in the mother cell (reviewed in reference 17). In order to test the effect of expressing spoIIIE exclusively in each of the cell types, the spoIIIE_Su gene was placed under the control of σ^E- and σ^F-directed promoters. We find that expression of spoIIIE_Su in either the mother cell or the prespore facilitates DNA translocation from the mother cell into the prespore and is able to substantially complement the sporulation defect of a spoIIIE36 mutant of B. subtilis. However, when spoIIIE_Su is overexpressed in the prespore from a strong σ^F-directed promoter, it does not correct the spoIIIE36 defect.

MATERIALS AND METHODS

Strains and plasmids.

The B. subtilis 168 strain BR151 (trpC2 metB10 lys3) was used as the parent strain for the study. Other B. subtilis strains used are listed in Table 1. All plasmids were maintained in Escherichia coli DH5α (GIBCO-BRL) unless otherwise stated. Plasmid pVK141 is a derivative of pDG793 (a gift from P. Stragier, Institut Biologie Physico Chimique, Paris, France) designed to insert a spoIIQ-lacZ transcriptional fusion at thrC by double crossover; it was constructed by inserting the spoIIQ promoter region, −200 to +47 (1), as a BamHI-EcoRI fragment into BamHI-EcoRI-digested pDG793. SL7653 is a derivative of BR151 in which thrC::spoIIQ-lacZ has been introduced by double crossover using linearized pVK141.

TABLE 1.

B. subtilis strains used

Strain	Relevant genotype	Source and/or referencea
SL7375	spoIIIE::spc	PL412 (Petra Levin) → BR151
SL7543	spoIIIE36	5
SL7653	thrC::PspoIIQ-lacZ	See text
SL7760	amyE::Pspac-spoIIIESuspoIIIE36	5
SL7870	amyE::Pspac-spoIIIESuthrC::PspoIIQ-lacZ spoIIIE36	See text
SL7980	amyE::PspoIIQ-spoIIIESuthrC::PspoIIQ-lacZ spoIIIE36	pVK149b → SL7870
SL7983	amyE::PspoIIQ-spoIIIESuthrC::PspoIIQ-lacZ spoIIIE::spc	SL7375 → SL7980
SL8309	amyE::PspoIIQ-spoIIIESu-gfp thrC::PspoIIQ-lacZ spoIIIE36	pVK161 → SL7980
SL8311	amyE::PspoIIQ-spoIIIESu-gfp thrC::PspoIIQ-lacZ spoIIIE::spc	SL7375 → SL8309
SL8509	thrC::PspoIIQ-lacZ	SL7653 → BR151
SL8776	thrC::PspoIIQ-lacZ spoIIIE36	SL8509 → SL7543
SL8808	thrC::PspoIIQ-lacZ ppsD::PspoIID-spoIIIESuspoIIIE36	pVK179 → SL8776
SL8816	amyE::PspoIIQ-spoIIIESuthrC::PspoIIQ-lacZ ppsD::PspoIID-spoIIIESuspoIIIE36	SL7980 → SL8808
SL8824	thrC::PspoIIQ-lacZ ppsD::PspoIID-spoIIIESu-gfp spoIIIE36	pVK183b → SL8808
SL9319	amyE::PspoIIQ-lacZ spoIIIE36	pEIA99b → SL7543
SL9559	ppsD::PspoIID-lacZ spoIIIE36	pVK205b → SL7543
SL10406	amyE::PspoIIR-spoIIIESuthrC::PspoIIQ-lacZ spoIIIE36	pVK217b → SL7980
SL10463	amyE::PspoIIR-spoIIIESu-gfp thrC::PspoIIQ-lacZ spoIIIE36	pVK161 → SL10406
SL10616	amyE::PspoIIR-lacZ spoIIIE36	pMLK215b → SL7543
SL11134	amyE::PspoIIQ-lacZ spoIIIE::spc	SL7375 → SL9319

^ax → y indicates the donor DNA (x) and the recipient strains (y) used in the strain construction.

^bPlasmids linearized by restriction enzyme digestion prior to transformation of the recipient strain to favor integration by double crossover.

Construction of B. subtilis spoIIIE strains expressing spoIIIE_Su from σ^F- and σ^E-directed promoters.

A B. subtilis strain expressing spoIIIE_Su from the σ^F-directed spoIIQ promoter was constructed as follows. A derivative of strain SL7870 expressing spoIIIE_Su under the control of the spoIIQ promoter was constructed by transformation with linearized pVK149 (Fig. 1). Strain SL7870 contains P_spac-spoIIIE_Su linked to cat and inserted by double crossover at the amyE locus; it is derived from SL7760 (5) and also contains a spoIIQ-lacZ transcriptional fusion inserted at thrC by transformation with DNA from SL7653. Plasmid pVK149 contains the 5′ end of spoIIIE_Su under the control of P_spoIIQ. To construct pVK149, a 5′ portion of spoIIIE_Su that includes its putative ribosome binding site, but no sequences upstream of the ribosome binding site, was initially cloned as a 750-bp PCR fragment in pVK50 to give pVK142; pVK50 was derived from pVK48 (5) by deletion of the NaeI fragment containing the 3′ end of amyE. A 1.6-kb SmaI fragment containing the 3′ end of amyE, a neo cassette, and the P_spoIIQ promoter was isolated from pEIA99 (containing the spoIIQ promoter region from −200 to +9 cloned into pEIA96 [1]) and inserted upstream of spoIIIE_Su in pVK142 to give pVK149. Double-crossover recombination between pVK149 and SL7870 yielded strain SL7980 (Fig. 1). Screening for a change of the antibiotic-resistance marker from cat to neo ensured that replacement of the promoter for spoIIIE_Su was by double crossover.

FIG. 1.

Schematic representation of the replacement of the P_spac promoter driving expression of spoIIIE_Su transcription in B. subtilis SL7870 by double-crossover recombination with linearized pVK149 to yield SL7980.

A similar method was used to replace the spoIIQ promoter in SL7980 with the spoIIR promoter by transformation with linearized pVK217. Plasmid pVK217 is a derivative of pMLK215 (kindly provided by M. Karow) that has the same 5′ region of spoIIIE_Su used in pVK142 cloned downstream of the spoIIR promoter (−561 to +366) and linked to cat. Double-crossover recombination of SL7980 with linearized pVK217 yielded SL10406. Further details of this and other constructs are available on request.

A different approach was used to place spoIIIE_Su under the control of the σ^E-directed spoIID promoter. Plasmid pVK179 was designed to insert P_spoIID-spoIIIE_Su by single crossover at ppsD, which is located at 169°, in the terminus region of the B. subtilis chromosome. It contains a promoterless copy of the complete spoIIIE_Su open reading frame as a 3-kb PCR fragment cloned downstream of the spoIID promoter (−291 to +22 in pMLK5169, kindly provided by M. Karow); it also contains part of ppsD as a 450-bp fragment and a selectable cat marker. Strain BR151 was transformed with pVK179, selecting for Cm^r and screening for single-crossover (Campbell-like) integration of the plasmid at ppsD. The site of integration of pVK179 was confirmed by linkage analysis. DNA was prepared from one such transformant and used to transform strain SL8776 (Table 1), producing strain SL8808.

Construction of B. subtilis strains carrying spoIIIE_Su-gfp transcriptional fusions.

The 3′ region of spoIIIE_Su was cloned as a 700-bp PCR fragment in pVK47 (5) to produce pVK160. The gfp cassette from pGreenTIR (12) was then cloned downstream of this spoIIIE_Su fragment to produce pVK161. Plasmid pVK161 was used to transform strains SL7980, SL7983, and SL10406 to introduce gfp immediately downstream of spoIIIE_Su (inserted at amyE) by a single crossover (Campbell-like recombination), producing strains SL8309, SL8311, and SL10463, respectively. Plasmid pVK183, a derivative of pVK161, was designed so that the linearized plasmid could be used to introduce gfp by double crossover immediately downstream of P_spoIID-spoIIIE_Su (which had been introduced by Campbell-like recombination of pVK179 into the ppsD locus) in strain SL8808, producing SL8824.

Fluorescence microscopy.

Nucleoid morphology was studied by fluorescence microscopy of samples fixed in 0.37% formalin and stained with DAPI (4′-6′-diamidino-2-phenylindole) (3). Cultures for gfp expression were grown at 30°C. At appropriate times, samples were examined by fluorescence microscopy without sample fixation.

Media.

E. coli was maintained on Luria-Bertani (LB) agar supplemented with ampicillin (100 μg/ml), spectinomycin (100 μg/ml), or neomycin (75 μg/ml) when required. B. subtilis was maintained on Schaeffer's sporulation agar (SSA) or on LB agar. When appropriate, B. subtilis was grown in the presence of antibiotics at the following concentrations: chloramphenicol, 5 μg/ml; neomycin, 3.5 μg/ml on SSA and 12 μg/ml on LB agar; erythromycin, 1.5 μg/ml; spectinomycin, 100 μg/ml; and tetracyline, 10 μg/ml.

Sporulation.

The induction of sporulation was carried out using modified Schaeffer's sporulation medium (MSSM) essentially as described previously (15). Growth was followed by measuring the optical density at 600 nm, which was converted to milligrams (dry weight) of bacteria using a standard calibration curve. Time was indicated in hours after the end of exponential growth, which was considered the start of spore formation (T₁, 1 h; T₂, 2 h; etc.). Cultures were analyzed by phase-contrast microscopy and by heat kill assay in order to determine the extent of spore formation 20 h after the end of exponential growth.

β-Galactosidase activity.

Samples were assayed with o-nitrophenyl-β-d-galactopyranoside as a substrate as described previously (8). Specific β-galactosidase activity was expressed as nanomoles of o-nitrophenyl-β-d-galactopyranoside hydrolyzed per minute per milligram (dry weight) of bacteria.

Other methods.

The methods used for transformation and for chromosomal and plasmid DNA isolation have been described previously (15, 16).

RESULTS

It had been shown that induction of P_spac-spoIIIE_Su during exponential growth by addition of IPTG (isopropyl-β-d-thiogalactopyranoside) complemented the sporulation defect caused by spoIIIE mutations in B. subtilis (5), and the strains produced about 2 × 10⁸ spores per ml. Addition of IPTG as late as 2 h after the end of exponential growth, that is, after the start of sporulation, still produced substantial spore formation (although only ∼20% of the sporulation of a spo⁺ strain as determined by phase-contrast microscopy). We therefore initiated a study of the effect of expressing spoIIIE_Su only after spore septum formation in either the mother cell or the prespore.

Mother cell-specific expression of spoIIIE_Su after spore septum formation from the spoIID promoter partly complemented the sporulation defect caused by the spoIIIE36 mutation in B. subtilis.

In order to test the effect of expressing spoIIIE only after sporulation septum formation, and exclusively in the mother cell, the σ^E-directed spoIID promoter (19) was used. Introduction of the P_spoIID-spoIIIE_Su fusion substantially restored the ability to sporulate to the spoIIIE36 mutant (strain SL8808) (Table 2). Spore formation increased to 5 × 10⁶ to 9 × 10⁶ per ml compared to <10 per ml for the isogenic spoIIIE36 mutant (SL8776). Compartmentalization of the green fluorescent protein (GFP) signal in a derivative strain with a transcriptional P_spoIID-spoIIIE_Su-gfp fusion confirmed that expression was localized exclusively in the mother cell (Table 3, strain SL8824), consistent with previous studies of σ^E activity in spoIIIE36 mutant strains (18). The spore titer of strains SL8808 was between 1 and 4% of that of the isogenic spo⁺ strain in different experiments (SL8509 [Table 2]). Consistent with this titer, we did not observe any substantial correction of the spoIIIE nucleoid phenotype (24), as visualized by DAPI staining (not shown). Expression of spoIIIE_Su from the spoIID promoter in both the prespore and the mother cell following septation was achieved by replacing spoIIIE36 with spoIIIE::spc. (In a spoIIIE null background, σ^E activity is detected in both the prespore and the mother cell [18, 24, 27].) It did not substantially increase spore formation over the frequency obtained with expression only in the mother cell (data not shown).

TABLE 2.

Effect of expression of spoIIIE_Su after septation on sporulation of B. subtilis spoIIIE mutants

Relevant genotypea (strain)	No. of heat-resistant spores per mlb			Frequency (%) relative to isogenic spo+ strain
	Series 1	Series 2	Series 3
spo+ (SL8509)	2.4 × 108	1.3 × 108	4.8 × 108	100
spoIIIE36 (SL8776)	<10	<10	ND	<10−5
spoIIIE36 PspoIID-spoIIIESu (SL8808)	8.9 × 106	4.9 × 106	5.3 × 106	1-4
spoIIIE36 PspoIIR-spoIIIESu (SL10406)	2.4 × 106	3.0 × 106	ND	1-3
spoIIIE36 PspoIIQ-spoIIIESu (SL7980)	<10	1.2 × 103	3.9 × 102	<0.001
spoIIIE36 PspoIIQ-spoIIIESu; PspoIID-spoIIIESu (SL8816)	4.9 × 106	4.2 × 106	ND	2-3
spoIIIE::spc PspoIIQ-spoIIIESu (SL7983)	5.9 × 106	1.5 × 105	ND	0.1-2
spoIIIE::spc (SL7375)	<10	<10	ND	<10−5

^aAll strains except SL7375 contain thrC::P_spoIIQ-lacZ.

^bSporulation was assessed 20 h after the end of exponential growth. Heat treatment was for 20 min at 80°C. The results of three separate series of experiments are shown. Relative frequencies are estimated within a series. ND, not determined.

TABLE 3.

Localization of spoIIIE_Su transcription as indicated by expression of a transcriptional gfp fusion

Relevant genotype (strain)	Observed GFP signala
	PS	MC	WC
spoIIIE36 ppsD::PspoIID-spoIIIESu-gfp (SL8824)	1	67	3
spoIIIE36 amyE::PspoIIR-spoIIIESu-gfp (SL10463)	300	0	3
spoIIIE36 amyE::PspoIIQ-spoIIIESu-gfp (SL8309)	104	0	0
spoIIIE::spc amyE::PspoIIQ-spoIIIESu-gfp (SL8311)	5b	0	132

^aSamples were analyzed by fluorescence microscopy and scored as showing prespore-specific (PS), mother cell-specific (MC), or whole-cell (WC) fluorescence. The results combine data from several experiments; between 35 and 65% of bacteria were fluorescing in the different experiments.

^bWeak signal also seen in the mother cell.

Prespore-specific expression of spoIIIE_Su after spore septum formation from the spoIIR promoter, but not the spoIIQ promoter, partly complemented the sporulation defect caused by the spoIIIE36 mutation in B. subtilis.

In order to test the effect of expressing spoIIIE exclusively in the prespore, spoIIIE_Su was placed under the control of the σ^F-directed spoIIR (strain SL10406), and spoIIQ (SL7980) promoters. The fusions were located at the origin-proximal amyE locus. Sporulation of strain SL10406, about 3 × 10⁶ spores per ml, was similar to that of strain SL8808, in which spoIIIE_Su was under the control of the σ^E-directed spoIID promoter (Table 2). However, sporulation of strain SL7980 was very poor, at most 1.2 × 10³ spores per ml in three different experiments, >1,000-fold lower than for SL10406 (Table 2). Analysis of derivative strains with gfp placed immediately downstream from spoIIIE_Su confirmed that expression from both the spoIIR and the spoIIQ promoters was confined to the prespore (Table 3) and that both promoters were active.

DNA translocation from the mother cell to the prespore is facilitated by postseptation expression of spoIIIE_Su in either the prespore or the mother cell.

The SpoIIIE protein has several roles during spore formation. Its DNA translocase function has received the most attention (2, 24, 26). However, it also has a role in the engulfment of the prespore by the mother cell (20), and it is involved in the compartmentalization of gene expression during sporulation (18, 24, 25, 27). Spore formation presumably requires that it function effectively in each of these roles. We have analyzed the DNA translocase function of SpoIIIE_Su in B. subtilis separately from its other roles by monitoring the expression of a spoIIQ-lacZ transcriptional fusion located at the thrC locus. The spoIIQ-lacZ fusion is transcribed exclusively by E-σ^F (10). In spoIIIE36 mutants, the 30% of the chromosome that is proximal to the origin of replication (oriC) is located in the prespore, but the remaining 70% is trapped in the mother cell (26). As thrC is located at 283°, it does not gain access into the prespore and remains trapped in the mother cell. Consequently, in spoIIIE36 mutants, E-σ^F-dependent genes located at thrC are normally not expressed because E-σ^F activity is confined to the prespore (23, 26). Thus, monitoring the expression of the spoIIQ-lacZ fusion located at thrC provides a good indicator of its translocation across the septum from the mother cell to the prespore. Consistent with this conclusion, expression of β-galactosidase from thrC::spoIIQ-lacZ in a spo⁺ strain (SL8509) (Fig. 2) occurs about 2.5 h after the start of sporulation in MSSM and is not observed in a spoIIIE36 mutant (SL8776).

FIG. 2.

Translocation of thrC::P_spoIIQ-lacZ from the mother cell into the prespore. Bacteria were incubated in MSSM and assayed for β-galactosidase activity, which is induced upon entry of thrC::P_spoIIQ-lacZ into the prespore. All strains contained the thrC::P_spoIIQ-lacZ fusion. Solid circles, spo⁺ (SL8509); open triangles, spoIIIE36 (SL8776); solid squares, spoIIIE36 P_spoIID-spoIIIE_Su (SL8808); open squares, spoIIIE36 P_spoIIR-spoIIIE_Su (SL10406); open circles, spoIIIE36 P_spoIIQ-spoIIIE_Su (SL7980).

The thrC::spoIIQ-lacZ region was translocated efficiently into the prespore in the spoIIIE36 derivative SL8808, as indicated by β-galactosidase activity (Fig. 2). In this strain, the spoIIIE_Su gene was expressed in the mother cell (Table 3) from the σ^E-directed spoIID promoter. The translocation may be slightly delayed compared to that observed for the spo⁺ strain (SL8509). Such a delay is consistent with the late transcription of spoIIIE_Su in SL8808 compared to transcription of spoIIIE from its natural, vegetatively expressed (6) promoter (in SL8509). However, it is difficult to assess any such delay, because the thrC::spoIIQ-lacZ fusion became hyperexpressed in SL8808 compared to the spo⁺ strain (Fig. 2). This hyperexpression is typical of origin-proximal σ^F-directed genes in a spoIIIE36 background. It suggests that although thrC was translocated efficiently into the prespore, the rest of the chromosome was not; this suggestion was consistent with DAPI staining and with the inefficient sporulation of SL8808 (about 1 to 4% of that of the spo⁺ strain [Table 2]). The thrC::spoIIQ-lacZ region was translocated into the prespore at efficiency comparable to that of SL8808 when spoIIIE_Su was expressed in the prespore from the σ^F-directed spoIIR promoter (SL10406) (Fig. 2). In sharp contrast to the situation with P_spoIIR and P_spoIID, there was little translocation into the prespore in the strain with spoIIIE_Su under the control of the prespore-specific spoIIQ promoter (SL7980) (Fig. 2). Two experiments confirmed that spoIIIE_Su was indeed expressed from the spoIIQ promoter and that a functional SpoIIIE_Su was formed in SL7980. (i) A derivative strain containing a transcriptional gfp fusion immediately downstream from P_spoIIQ-spoIIIE_Su, strain SL8309, resulted in GFP fluorescence localized exclusively in the prespore (Table 3). (ii) The strain formed by replacing the spoIIIE36 mutation in strain SL7980 with the spoIIIE::spc null mutation, SL7983, displayed SpoIIIE_Su activity, as indicated by spore formation (Table 2).

The spoIIQ promoter is much more strongly expressed than the spoIID and spoIIR promoters.

The relative strengths of the promoters used are critical to our interpretation of the results described above. The native B. subtilis spoIIIE promoter is weak (6), and the spoIID and spoIIR promoters are also weak (7, 11, 19). In contrast, spoIIQ has the strongest of the σ^F-directed promoters in a spo⁺ background (1, 10). Further, expression of σ^F-directed promoters is greatly enhanced in spoIIIE36 mutants (7, 11, 24), about 20-fold for spoIIR (7, 11) and for spoIIQ (our studies). It seemed possible that the greater activity of the spoIIQ promoter compared to the other promoters might explain the failure of expression of spoIIIE_Su from P_spoIIQ to restore spore formation to the spoIIIE36 mutant.

To compare the promoter strengths, spoIIIE36 strains were constructed in which lacZ replaced spoIIIE_Su under the control of P_spoIIQ (inserted at amyE [SL9319]), P_spoIIR (inserted at amyE [SL10616]), or P_spoIID (inserted at ppsD [SL9559]). Expression of β-galactosidase started at about the same time during sporulation from the three promoters. Expression from P_spoIIQ was substantially greater (about fivefold 4 h after the start of spore formation) than from P_spoIID or P_spoIIR (Fig. 3). Expression from P_spoIID was very similar to that from P_spoIIR (Fig. 3). Thus, overexpression of SpoIIIE_Su from P_spoIIQ in the spoIIIE36 mutant could be the cause of the failure to complement the sporulation defect.

FIG. 3.

Relative strengths of promoters assessed with transcriptional lacZ fusions. Bacteria were incubated in MSSM and assayed for β-galactosidase activity. Solid circles, spoIIIE36 (SL8776); filled triangles, spoIIIE36 ppsD::P_spoIID-lacZ (SL9559); open circles, spoIIIE36 amyE::P_spoIIR-lacZ (SL10616); open squares, spoIIIE36 amyE::P_spoIIQ-lacZ (SL9319); solid squares, spoIIIE::spc amyE::P_spoIIQ-lacZ (SL11134).

Strong expression of spoIIIE_Su in the prespore did not block the complementation of spoIIIE36 resulting from expression of spoIIIE_Su in the mother cell.

Strain SL8816 was constructed containing the spoIIIE36 mutation and both P_spoIID (at ppsD) and P_spoIIQ (at amyE) driving spoIIIE_Su expression. In strain SL8816, thrC::spoIIQ-lacZ was strongly expressed, indicating efficient translocation into the prespore. Further, sporulation was at a frequency similar to that of strain SL8808, in which spoIIIE_Su was expressed only from P_spoIID (Table 2). Thus, high expression of spoIIIE_Su from the spoIIQ promoter did not counteract the effect of expression from the spoIID promoter. Consistent with this finding, inclusion of P_spoIIQ-spoIIIE_Su in a strain that was genetically spo⁺ did not impair spore formation (unpublished observations).

Replacing spoIIIE36 with spoIIIE::spc resulted in uncompartmentalized GFP expression with either σ^F- or σ^E-directed promoters driving gfp expression (Table 3 and our unpublished observations). These observations support earlier reports that spoIIIE null mutations disrupt compartmentalization (18, 24, 25, 27). The substitution did not impair the activity of the spoIIQ promoter (in fact, strain SL11134 showed somewhat greater activity than SL9319 [Fig. 3]). However, expression of spoIIIE_Su from the spoIIQ promoter (SL7983) was able partly to complement the sporulation defect caused by spoIIIE::spc, although it had not complemented spoIIIE36 (SL7980) (Table 2). We interpret this complementation to be a consequence of uncompartmentalized expression of spoIIIE_Su in the spoIIIE::spc background. The behavior of strain SL7983 is consistent with the conclusion that high expression of spoIIIE_Su from the spoIIQ promoter is not detrimental, provided spoIIIE_Su is expressed in the mother cell. We were unable to obtain a mother cell-specific promoter expressed at a strength comparable to that of P_spoIIQ in order to test the effect of overexpression of spoIIIE_Su exclusively in the mother cell.

DISCUSSION

The experiments reported here demonstrate that expression of the spoIIIE gene of S. ureae, spoIIIE_Su, under various conditions in B. subtilis can complement the sporulation defect caused by the missense mutation spoIIIE36 and by the knockout mutation spoIIIE::spc. The SpoIIIE_Su protein complemented B. subtilis spoIIIE mutants efficiently when it was expressed during vegetative growth or soon after the start of spore formation (reference 5 and unpublished data). Even when expression of spoIIIE_Su was delayed until after formation of the spore septum, sporulation was increased to >10⁶ spores per ml compared to <10 per ml for the parental spoIIIE mutant, reaching ∼1% of the wild-type level (Table 2). This complementation occurred whether spoIIIE_Su was expressed exclusively in the mother cell from the σ^E-directed spoIID promoter or exclusively in the prespore from the σ^F-directed spoIIR promoter. In contrast, complementation was not achieved when expression was exclusively in the prespore from the much stronger (Fig. 3) σ^F-directed spoIIQ promoter (Table 2). The results were corroborated by analysis of the translocation of thrC::spoIIQ-lacZ into the prespore. We think it likely that the failure to complement is a consequence of the strength of the spoIIQ promoter. An alternative explanation, which we think less likely but cannot exclude, is that the spoIIQ promoter is expressed later than the spoIIR promoter. We did not detect any such difference in the times of their expression (unpublished results), but a delay of a few minutes would be below the level of our detection and might be critical to SpoIIIE_Su function.

The SpoIIIE protein is required for chromosome translocation into the prespore and subsequently for engulfment of the prespore by the mother cell (20, 24). The distinction between the two σ^F-directed promoters was apparent at the translocation stage (Fig. 2). DNA translocation was assessed by measuring β-galactosidase resulting from expression of a spoIIQ-lacZ transcriptional fusion inserted at thrC. In a spoIIIE36 mutant, this fusion is ordinarily not expressed because it is located outside the origin-proximal portion of the chromosome that is in the prespore, and so it is inaccessible to E-σ^F, which is active only in the prespore (26). Efficient translocation of thrC::spoIIQ-lacZ into the prespore was obtained in the spoIIIE36 mutant whether spoIIIE_Su was expressed in the mother cell from the spoIID promoter or in the prespore from the spoIIR promoter. In contrast, little translocation was observed when expression was from the spoIIQ promoter.

The DNA translocation assay indicated efficient translocation of thrC into the prespore when spoIIIE_Su was expressed from either the spoIID or spoIIR promoter. However, DAPI staining revealed little difference in nucleoid morphology between the spoIIIE36 mutant and its P_spoIID-spoIIIE_Su and P_spoIIR-spoIIIE_Su derivatives (unpublished observations), suggesting that chromosome translocation was not completed efficiently. We were not able to test the completeness of translocation with spoIIQ-lacZ or spoIIR-lacZ located near the chromosome terminus, as such fusions were poorly expressed even in a spo⁺ background (reference 9 and our unpublished observations). Inefficient completion of translocation may explain why sporulation reached only ∼1% of the wild-type level when spoIIIE_Su was expressed from either the spoIID or spoIIR promoter.

The activities of σ^F and σ^E are largely uncompartmentalized in spoIIIE null mutants, in contrast to the situation with a spoIIIE36 mutant (18, 24, 27) (Table 3). The change from spoIIIE36 to spoIIIE::spc did not substantially affect the sporulation of P_spoIID-spoIIIE_Su or P_spoIIR-spoIIIE_Su strains. Thus, their reduced sporulation (compared to that of a spo⁺ strain) was not a consequence of compartmentalization of their action. Rather, we think, sporulation reached only ∼1% of the wild-type level because of the late synthesis and heterologous nature of the SpoIIIE_Su protein. There are several examples where delaying the expression of some sporulation genes relative to others reduces the efficiency of spore formation. For example, delaying the time of expression of the σ^F-directed spoIIR locus as little as 20 min (by delaying its translocation into the prespore) has been shown to reduce sporulation to ∼2% of that of the parent strain (9, 29).

It seems likely that transcription of spoIIIE_Su in the mother cell results in insertion of SpoIIIE_Su into the sporulation septum from the mother cell side and, conversely, that transcription in the prespore results in insertion from the prespore side. If so, then the side of the septum on which SpoIIIE_Su insertion occurs is not important when spoIIIE_Su is transcribed after septum formation from the spoIIR or spoIID promoter. It is possible that the orientation of SpoIIIE_Su in the septum is not important for its function. Indeed, the similar E. coli DNA translocase, FtsK, is apparently able to pump the DNA into both dividing cells when helping to resolve chromosome dimers during division (22). It is also possible that SpoIIIE_Su, which comes from a species that divides symmetrically during sporulation (28), can accommodate different orientations, whereas SpoIIIE from B. subtilis cannot. Nevertheless, it is possible that the final orientation of SpoIIIE_Su in the septum is important for correct function in translocating DNA from the mother cell to the prespore. Two mechanisms are suggested that might give a final directed orientation of SpoIIIE_Su without directed insertion. First, the SpoIIIE36 protein is inactive as a translocase but is inserted into the spore septum in the correct location for DNA translocation (25). This preexisting SpoIIIE36 structure may provide a template for correct assembly of newly synthesized SpoIIIE_Su. Second, it is also possible that asymmetry across the sporulation septum (13) may direct the final orientation of SpoIIIE_Su insertion. Either mechanism might accommodate slow synthesis in the prespore from the spoIIR promoter and direct the insertion of SpoIIIE_Su in the correct orientation. However, either mechanism might be swamped by the much stronger synthesis in the prespore from the spoIIQ promoter. Such a suggestion is clearly speculative, but it does provide an explanation for the otherwise paradoxical result that weak expression of spoIIIE_Su in the prespore (from the spoIIR promoter) works whereas strong expression (from the spoIIQ promoter) does not. Pertinent to this suggestion are the two other results with P_spoIIQ-spoIIIE_Su. First, when P_spoIID-spoIIIE_Su is present in a spoIIIE36 mutant along with P_spoIIQ-spoIIIE_Su, P_spoIID-spoIIIE_Su is epistatic to P_spoIIQ-spoIIIE_Su; that is, the P_spoIID-spoIIIE_Su phenotype is expressed with respect to both thrC translocation and spore formation—the weak is epistatic to the strong. Second, in a spoIIIE::spc background, P_spoIIQ-spoIIIE_Su partly restores the ability to form spores; that is, when σ^F activity is present in both mother cell and prespore, high-level expression from the spoIIQ promoter is no longer a problem. The latter result, obtained with no “chaperoning” SpoIIIE36, suggests that inherent asymmetry across the spore septum (13) may direct the correct orientation of SpoIIIE insertion.

Sharp and Pogliano (21) have recently reported experiments somewhat similar to those reported here. They expressed the B. subtilis spoIIIE gene, rather than the S. ureae gene, from mother cell- and prespore-specific promoters. Expression from the mother cell-specific spoIID promoter was reported to restore wild-type sporulation to a spoIIIE null mutant and to restore normal nucleoid translocation as indicated by DAPI staining. This is more efficient than the complementation that we observed for postseptation expression of spoIIIE_Su from the spoIID promoter and may result from expression of a homologous rather than a heterologous spoIIIE. Expression from the prespore-specific spoIIR promoter in a spoIIIE null mutant gave ∼1% sporulation (21), similar to the frequencies we have found. The authors argue that this difference in sporulation frequency is a consequence of the location of spoIIIE expression. This conclusion is tentative, because spoIIIE null mutants display σ^E and σ^F activities that are substantially uncompartmentalized (18, 24, 27) (Table 3). The authors also reported that expression of spoIIIE from a σ^E-directed promoter but not from σ^F-directed promoters in a spoIIIE36 background largely corrected the nucleoid partitioning defect. They did not give data for spore formation and assessed relative promoter strengths only in a spoIIIE⁺ background. As discussed above, expression from σ^F-directed, but not σ^E-directed, promoters is greatly enhanced in spoIIIE mutant backgrounds (7, 11, 24); consequently, overexpression from the σ^F-directed promoter in the spoIIIE36 background might have impaired complementation. However, their data clearly indicate that expression of spoIIIE only in the mother cell almost completely complements the spoIIIE36 translocation defect. We did not obtain comparable complementation, and we think it probable that this reflects our use of the heterologous spoIIIE_Su gene. While the latter functioned effectively when expressed in vegetative growth or early in sporulation, its heterologous nature may have made a significant difference when trying to rescue the sporulation defect after the spore septum has formed. Nevertheless, we have shown that expression of spoIIIE_Su in the prespore could complement the spoIIIE36 defect as effectively as expression in the mother cell.