The Products of the spoVA Operon Are Involved in Dipicolinic Acid Uptake into Developing Spores of Bacillus subtilis

Abstract

Bacillus subtilis cells with mutations in the spoVA operon do not complete sporulation. However, a spoVA strain with mutations that remove all three of the spore’s functional nutrient germinant receptors (termed the ger3 mutations) or the cortex lytic enzyme SleB (but not CwlJ) did complete sporulation. ger3 spoVA and sleB spoVA spores lack dipicolinic acid (DPA) and have lower core wet densities and levels of wet heat resistance than wild-type or ger3 spores. These properties of ger3 spoVA and sleB spoVA spores are identical to those of ger3 spoVF and sleB spoVF spores that lack DPA due to deletion of the spoVF operon coding for DPA synthetase. Sporulation in the presence of exogenous DPA restored DPA levels in ger3 spoVF spores to 53% of the wild-type spore levels, but there was no incorporation of exogenous DPA into ger3 spoVA spores. These data indicate that one or more products of the spoVA operon are involved in DPA transport into the developing forespore during sporulation.

A characteristic feature of the endospores of various Bacillus and Clostridium species is the presence of high levels of pyridine-2,6-dicarboxylic acid (dipicolinic acid [DPA]) (2). DPA generally comprises ≥10% of the dry weight of these spores, and most of the DPA is likely in a 1:1 chelate with divalent cations, predominantly Ca²⁺ (2). DPA is synthesized in the mother cell compartment of a sporulating cell from an intermediate in the lysine biosynthetic pathway, and the final synthetic step is catalyzed by DPA synthetase, the product of the two cistrons of the spoVF operon (1, 2, 3). DPA is located in the spore protoplast or core and is excreted in the first minute of spore germination (2, 21).

In strains lacking DPA (for example, spoVF strains of Bacillus subtilis) sporulation is not completed as the developing spores lyse during sporulation (2, 3, 4, 18). However, spoVF spores can be stabilized by additional mutations that remove either the three functional nutrient germinant receptors (termed the ger3 mutations) or a major spore cortex lytic enzyme, SleB (17, 18, 19). ger3 spoVF and sleB spoVF spores have higher levels of core water and consequently are less wet heat resistant than wild-type dormant spores (7, 17, 18; B. Setlow and P. Setlow, unpublished data). This suggests that in addition to stabilizing the spore’s dormant state, DPA is also important in reducing the spore’s core water content and thereby increasing its heat resistance (7).

Although the location and pathway of DPA synthesis are known and there has been some general indication of the function of DPA in the developing and dormant spore, it is unclear how DPA gets into the developing forespore from its site of synthesis in the mother cell. It is also unclear how the dormant spore’s DPA is excreted in the first minute of spore germination. The latter process is of interest, since DPA movement out of the dormant spore is regulated. DPA is normally retained in the dormant spore for long periods, but it is excreted within minutes either when nutrients bind to germinant receptors on the spore’s inner membrane (9, 20, 21) or when the spore’s peptidoglycan cortex is hydrolyzed (17). However, the proteins involved in DPA movement during sporulation and germination are not known. To identify genes coding for proteins involved in DPA transport during forespore development and spore germination, we identified five criteria that such genes would likely meet. (i) Mutation of the genes should result in DPA-less spores and thus a spo phenotype analogous to that of an spoVF strain. (ii) Mutation of the genes should not eliminate DPA synthesis during sporulation. (iii) The genes should encode one or more membrane proteins, as during sporulation DPA must traverse both the inner and outer forespore membranes; during spore germination at least the inner spore membrane must be traversed, although it is not clear if the outer spore membrane is a permeability barrier in the dormant spore (2). (iv) At least one gene should be expressed in the forespore compartment of the sporulating cell to facilitate insertion of the encoded protein into the inner forespore membrane. And (v) the genes should be expressed at or before the time of DPA synthesis.

There is one set of genes, the cistrons of the spoVA operon, which fulfills all of the criteria given above. The spoVA operon encodes six proteins, several of which appear to be membrane proteins (6, 24). Mutation of any of the first five cistrons of the spoVA operon, but probably not spoVAF, results in an spo phenotype in which immature forespores appear but then lyse, although spoVA mutants do synthesize DPA during sporulation (3, 4, 6, 13). The spoVA operon is also transcribed exclusively in the forespore at about the time of DPA synthesis in the mother cell (3, 5, 14, 25; Setlow and Setlow, unpublished). While the functions of the proteins encoded by the spoVA operon are unknown, it was suggested some time ago that SpoVA proteins are involved in DPA transport during sporulation (3). In support of this suggestion, recent genome sequence data have shown that at least SpoVAC, SpoVAD, and SpoVAE are well conserved in Bacillus and Clostridium species.

Since the spoVA operon seemed to be a likely candidate for an operon that encodes components of a DPA transport system, we constructed an spoVA mutation and combined it with deletion of all three nutrient germinant receptors (the ger3 mutations), reasoning that ger3 spoVA spores should be stable, as ger3 spoVF spores are (18). Strains lacking both spoVA and either CwlJ or SleB, the two enzymes involved in cortex degradation during spore germination (either CwlJ or SleB is sufficient for this process) (10, 15, 21), were also constructed. The B. subtilis strains used in this work are derivatives of strain 168, and the wild-type strain used was PS832, a trp⁺ revertant of strain 168. Strains FB72 (ΔgerA::spc ΔgerB::cat ΔgerK::ermC), FB108 (ΔgerA::spc ΔgerB::cat ΔgerK::ermC ΔspoVF::tet), and FB112 (ΔsleB::spc) have been described previously (17, 18, 19). Strain PS3349 (ΔspoVF::tet) was prepared by transforming strain PS832 to tetracycline resistance with plasmid pFE229 (18) and by using Southern blot analysis to identify a tetracycline-resistant transformant that arose by a double-crossover event. The plasmid used to construct a deletion of the cwlJ gene was derived from plasmid pFE243 (17), which contains upstream and downstream regions of cwlJ with most of the coding region replaced by a tet marker. The tet marker in pFE243 was removed by digestion with BamHI and EcoRI and was replaced with a BamHI-EcoRI fragment containing the spc marker from plasmid pJL74 (11). The new plasmid, designated pPS3400, was isolated in Escherichia coli TG1. The ΔspoVA::tet plasmid pPS3306 was derived from plasmid pDG1515 (8), which contains a tet marker flanked by BamHI and PstI sites on one side and EcoRI and HindIII sites on the other side. The 5prime; end of the spoVA operon from bp 26 to 428 relative to the spoVA translation start site (defined as bp 1) was amplified by PCR performed with Vent DNA polymerase by using PS832 chromosomal DNA and primers A-Bam (5prime;-GGATCCGGCTTCGCCACCGAGTGC-3prime;) and A-Pst (5prime;-CTGCAGGGATAGTCATTCCTCTCTCCCG-3prime;), which contained extra 5prime; residues (underlined) with BamHI or PstI sites. The PCR fragment was cloned into SmaI-cut plasmid pUC19 in E. coli JM83, giving plasmid pMC1, and the correct DNA sequence of the insert in this plasmid was verified. The insert was excised from plasmid pMC1 by digestion with BamHI and PstI and cloned between BamHI and PstI sites in plasmid pDG1515 in E. coli, giving plasmid pMC2. The 3prime; end of the spoVA operon spanning bp 1099 to 1453 relative to the spoVAF translation start site (defined as bp 1) was PCR amplified as described above but with primers A-Eco (5prime;-GAATTCGCCGCTGTTCTGATCGGC-3prime;) and A-Hind (AAGCTTGGGGTGAACGATGCTGGG-3prime;), which contained extra residues at their 5prime; ends (underlined) with either EcoRI or HindIII sites. The PCR product was cloned in SmaI-cut plasmid pUC19 in E. coli, giving plasmid pMC3, and the correct DNA sequence of the insert was verified. The insert was removed by digestion with EcoRI and HindIII and cloned between EcoRI and HindIII sites in pMC2 in E. coli, giving the ΔspoVA::tet plasmid pPS3306. This plasmid was used to transform B. subtilis strains FB72 (ger3) and PS832 (wild type) to tetracycline resistance. Southern blot analysis was used to identify transformants that arose by double crossover into the spoVA locus with replacement of the majority of the spoVA operon with the tet marker. One such ger3 spoVA transformant was designated PS3317, and one spoVA transformant was designated PS3318. Strain PS3318 was further transformed to spectinomycin resistance with plasmid pPS3400, and Southern blot analysis of one resulting transformant confirmed that the chromosomal structure was the structure expected. This transformant was designated PS3405 (cwlJ spoVA). Chromosomal DNA from strain FB112 (sleB) was used to transform strain PS3318 to resistance to both tetracycline and spectinomycin, and Southern blot analysis confirmed the expected chromosomal structure of one transformant, which was designated PS3406 (sleB spoVA).

B. subtilis strains were sporulated at 37°C on 2× SG plates (16, 18) without antibiotics. Neither strain PS3318 (spoVA) nor strain PS3349 (spoVF) formed spores that could be isolated, as expected (3, 4, 18). However, spores of strain PS3317 (ger3 spoVA) could be isolated, and sporulation of this strain was identical to sporulation of strain FB72 (ger3 spoVF). The spores produced by strains FB72, PS832, and PS3317 were harvested, purified, and stored as described previously (16, 18), and all of the spore preparations used were free (≥96%) of sporulating cells, germinated spores, and cell debris. The DPA contents of the purified spores were determined as described previously (16, 18, 22), and as found previously (18), ger3 spoVF spores lacked DPA, as did ger3 spoVA spores (Table 1). The core wet densities of the various types of spores, which reflected the core water contents of the spores (the higher the spore core wet density, the lower the spore core water content) (7), were determined as described previously (12). As found previously (18), ger3 spoVF spores had a lower core wet density than wild-type spores, presumably due to the absence of DPA (Table 1); the core wet density of ger3 spoVA spores was almost identical to that of ger3 spoVF spores (Table 1). While the core wet densities of the ger3 spoVA and ger3 spoVF spores were significantly lower that than that of wild-type spores, the value for fully germinated wild-type spores (1.194 g/ml [26]) is well below the value obtained for the DPA-less spores. While fully germinated wild-type spores lose their DPA, they also degrade their peptidoglycan cortex; the latter event results in an increase in spore core volume that is accompanied by an increase in core water content and thus a further decrease in the core wet density of the spores.

TABLE 1.

Core wet densities and DPA contents of spores of various B. subtilis strains^a

Strain	Sporulated without DPA		Sporulated with DPA (100 μg/ml)
	Core wet density (g/ml)b	DPA content (μg/OD600)c	Core wet density (g/ml)b	DPA content (μg/OD600)c
PS832 (wild type)	1.332	11.4	NDd	ND
FB72 (ger3)	1.330	11.5	ND	ND
FB108 (ger3 spoVF)	1.275	<0.5	1.309	6.1
PS3317 (ger3 spoVA)e	1.263	<0.5	1.263	<0.5

^aSpores were prepared, cleaned, stored, and analyzed as described in the text. There was a significant level of Ca²⁺ in the sporulation medium used. Thus, it is not clear whether during sporulation DPA was taken up independent of this cation or as the Ca²⁺ chelate.

^bThe values are the values from one determination. In other experiments, the values varied by less than 1% from those shown.

^cThe values are from one determination. In other experiments, the values varied by less than 6% from those shown. OD₆₀₀, unit of optical density at 600 nm.

^dND, not determined. Previous work (18) has shown that DPA in the sporulation medium has no effect on the core wet density or DPA content of spores of strain PS832 or FB72.

^espoVA strains produce DPA, but the amount of DPA produced is significantly less than 100 μg/ml (4).

Spores of ger3 strains do not germinate in response to nutrients (19). Consequently, spore viability was assessed by germinating spores in a 1:1 mixture (60 mM each) of Ca²⁺ and DPA, as described previously (19), prior to plating on Luria-Bertani medium agar plates (23) and incubation overnight at 37°C. This analysis showed that wild-type, ger3 spoVF, and ger3 spoVA spores had similar viabilities (≥4 × 10⁷ CFU/unit of optical density at 600 nm) (data not shown). Previous work has shown that wild-type spores have greater wet heat resistance than ger3 spoVF spores, presumably because of the higher core water content of the latter spores (7, 18). When spore wet heat resistance was determined as described previously (18), we obtained similar results and also found that ger3 spoVA spores had wet heat resistance that was essentially identical to that of ger3 spoVF spores (Fig. 1). While the wet heat resistance of the ger3 spoVA or ger3 spoVF spores was less than that of ger3 or wild-type spores (18) (Fig. 1), it was greater than that of fully germinated spores, which are killed (>99%) in 10 min at 60°C (26). The ger3 spoVA and ger3 spoVF spores also had identical UV resistance characteristics (data not shown).

FIG. 1.

Wet heat resistance of spores of various B. subtilis strains. Spores of strain FB72 (ger3), FB108 (ger3 spoVF), and PS3317 (ger3 spoVA) were prepared without exogenous DPA and cleaned, and wet heat resistance was determined by incubation at 72°C as described in the text. Previous work (17) has shown that spores of the wild-type and FB72 strains have identical heat resistance characteristics. Symbols: □, FB72; •, FB108; ○, PS3317.

The results described above are consistent with the hypothesis that the spoVA operon encodes one or more proteins involved in forespore DPA uptake. If this is indeed the case, then exogenous DPA should not be taken up by sporulating ger3 spoVA cells. To test this prediction, the various strains were sporulated with 100 μg of DPA per ml in the medium as described previously (18). As found previously (18), ger3 spoVF spores accumulated exogenous DPA, and the spore DPA level reached 53% of the level in wild-type spores (Table 1). In contrast, the ger3 spoVA spores accumulated no exogenous DPA (Table 1), which is consistent with the previous observation that exogenous DPA does not restore production of heat-resistant spores by an spoVA mutant strain (4).

Previous work has shown that spoVF spores can also be stabilized by inactivation of the sleB gene, which encodes one of the two enzymes involved in cortex hydrolysis during spore germination (15, 17). The second enzyme involved in cortex hydrolysis is CwlJ, and either CwlJ or SleB is sufficient for cortex hydrolysis during nutrient germination of wild-type spores (10, 15, 21). However, a cwlJ mutation does not stabilize spoVF spores, as expected, since CwlJ requires DPA for activity while SleB does not (17; Setlow and Setlow, unpublished). Consequently, if SpoVA proteins are indeed involved in DPA transport during sporulation, then sleB spoVA spores should be stable and cwlJ spoVA spores should not be stable. This was indeed the case, as developing cwlJ spoVA spores lysed too rapidly to allow their isolation, while stable sleB spoVA spores were readily isolated. In addition, the latter spores lacked DPA and had viability and a core wet density identical to those of ger3 spoVA spores (data not shown). Recent work (I. Bagyan and P. Setlow, unpublished data) has shown that CwlJ is located in one of the spore coat layers and thus should be accessible to DPA synthesized in the mother cell during sporulation. However, since CwlJ is normally not activated during sporulation, presumably either DPA levels in the mother cell never become high enough to activate CwlJ during sporulation or there is some requirement for CwlJ activation in addition to DPA.

The results of this study strongly suggest that one or more of the proteins encoded by the spoVA operon are involved in some aspect of uptake of DPA by sporulating cells. It is of course possible that SpoVA proteins are not involved in DPA transport and that the lack of DPA in spoVA spores is due to a pleiotropic effect of a spoVA mutation on the level or activity of the actual DPA transport proteins. However, we feel that this is extremely unlikely because of the lack of an effect of a spoVA mutation on gene expression during sporulation (3, 4), as well as the absence of other forespore-specific genes that might encode proteins involved in DPA transport. We presume that the SpoVA proteins are involved in DPA movement out of the spore during spore germination as well, but we have no direct evidence for this yet. Since the spoVA operon is expressed only in the forespore (3, 5, 27), the SpoVA proteins might be involved only in DPA transport across the inner spore membrane. This may be all that is needed for DPA exit from the spore during germination, since the outer spore membrane may not be a permeability barrier in the dormant spore (2). However, there must be some mechanism for moving DPA across the outer forespore membrane during sporulation, as this membrane is almost certainly a permeability barrier in sporulating cells (2). It is of course possible that SpoVA proteins are in both the outer and inner forespore membranes, but the location of the SpoVA proteins in spores is not yet known. While our data strongly indicate that at least one SpoVA protein is involved in DPA transport, it is not clear whether all six proteins of the operon are involved in this process. Indeed, it appears that the SpoVAF protein is not even essential for sporulation (3, 13). Similarly, it is not clear whether the SpoVA proteins are sufficient for DPA transport or function in conjunction with other proteins to effect this process. Unfortunately, analysis of the amino acid sequences of the various SpoVA proteins is not particularly helpful for suggesting how one or more of these proteins could play a role in DPA transport. However, with identification of one or more of the SpoVA proteins as proteins that play a role in DPA transport, it should be possible to obtain answers to these questions, and this work is in progress.