INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Upon starvation for one or more nutrients, cells of the gram-positive bacterium Bacillus subtilis differentiate into metabolically dormant spores which are adapted to resist environmental damage during dormancy (6, 27). The spore's dormancy and resistance properties ensure its survival through conditions that are not conducive to cell growth. When nutrient-rich conditions return, spore dormancy is broken and the spore is converted back to a vegetative cell through spore germination and outgrowth (9, 14). During that process, the spore loses its dormancy and resistance properties and consequently becomes vulnerable to its environment. Thus, before a spore initiates germination, it must ascertain that the environment is conducive to cell growth.

Many studies have shown that dormant spores use small molecules and ions as indicators of conditions that permit cell growth (35). These indicator molecules, called germinants, are by themselves sufficient to initiate spore germination, and their identity differs significantly between spores of different species. In B. subtilis, L-alanine or a combination of L-asparagine, D-fructose, D-glucose, and K⁺ ions (AFGK) acts as a germinant to initiate spore germination (32-34). Because many germinants are metabolites, they were originally proposed to reactivate spore metabolism by supplying substrates for spore enzymes (7, 19). However, that hypothesis was challenged by subsequent work which showed that radiolabeled germinants are not significantly metabolized early in germination (25, 26) and that nonmetabolizable analogs of germinants also trigger germination (21, 28). Moreover, investigation of the germination-initiating properties of derivatives and isomers of the known germinants suggested that these molecules probably initiate germination by binding to and activating receptors that are present in the spore (35, 37).

Candidates for the hypothesized spore germinant receptor(s) were identified in genetic screens for ger mutations that blocked spore germination (8, 15, 30). Of the ger mutations that were identified in those screens, mutations in gerA, gerB, and gerK conferred a germinant-specific defect in germination. For example, gerA mutants failed to germinate only in L-alanine, whereas gerB and gerK mutants exhibited a defect only in AFGK-induced germination (8, 15). These mutant phenotypes were best explained by a model in which the gerA product(s) were required for L-alanine recognition, while the gerB and gerK products were required for AFGK recognition (16). Subsequent work showed that gerA, gerB, and gerK are homologous tricistronic operons, indicating that these three loci might encode proteins with similar functions (3, 14, 38). In addition, the first two proteins in each operon are predicted to be integral membrane proteins (3, 38), which is consistent with them being receptors for environmental signals. Thus, it was proposed that the gerA, gerB, and gerK operons encode homologous components of distinct germinant receptors (16).

Although attractive, the idea that gerA, gerB, and gerK encode germinant receptors has not been substantiated, and it is not clear whether all three proteins encoded by each of these loci are required for recognition and binding of the germinant. In this work, we tried to address these issues by designing a genetic screen to specifically isolate mutations that affect the germinant receptor(s). We identified three mutations, two of which affected the GerBA protein and one of which affected the GerBB protein. Thus, our studies strongly support a model in which the gerB operon (and probably also the gerA and the gerK operons) encodes components of a spore germinant receptor.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Strains, plasmids, and media used. B. subtilis strains used in this study are listed in Table 1. B. subtilis strains were constructed by transformation with either chromosomal DNA or plasmid DNA as previously described (1). Escherichia coli TG1 and DH5F' were used for production of plasmids as described elsewhere (23). The rich media LB and 2×YT were used for growth of E. coli and for vegetative growth of B. subtilis (23). 2×SG medium was used for B. subtilis sporulation at 37°C, and spores were harvested, cleaned, and stored as described elsewhere (18). B. subtilis spores that were used in the germination assays were prepared by the resuspension method at 30°C (29). When necessary, growth media were supplemented with (per liter) 50 or 100 mg of ampicillin; 100 mg of spectinomycin; 1 mg of erythromycin and 25 mg of lincomycin (MLS); or 5 mg of chloramphenicol.

View larger version (13K): [in this window] [in a new window]   FIG. 1.   Restriction map of the 5.3-kb genomic region which contains the gerB operon and the strategy used to clone that DNA fragment. The large bar denotes the 5.3-kb genomic region which includes the gerB operon demarcated by the solid region within the bar; the solid line represents flanking genomic DNA. The plasmid vector denoted by the thick solid line and the spc cassette represented by the hatched bar are not drawn to scale. Restriction enzyme sites: B, BamHI; Cl, ClaI; H, HindIII; R, EcoRV; Ss, SstI; St, StuI. ApR denotes the ampicillin resistance marker carried on the plasmid.

Mutagenesis. Mutagenized cultures of B. subtilis PS832 were generated by ethyl methanesulfonate mutagenesis of exponentially growing cells as described previously (5). The mutagenized cultures were sporulated by nutrient exhaustion, and the spores were harvested, cleaned, and stored as described elsewhere (18).

Separation of germinated and ungerminated spores. Germinated and ungerminated B. subtilis spores were separated on a metrizoic acid gradient on the basis of buoyant density (18). The gradient was prepared in a 2.5-ml ultraclear ultracentrifuge tube (Beckman, Fullerton, Calif.) by sequential layering of 0.1 ml of 70%, 0.5 ml of 60%, 0.2 ml of 50%, 0.2 ml of 40% and 0.2 ml of 30% metrizoic acid solutions. The spore suspension which was to be separated (in 0.2 ml of 20% metrizoic acid) was layered on top of the gradient, which was centrifuged at 13,000 rpm in a TLS-55 rotor (TL100 ultracentrifuge) for 45 min at 4°C. The deceleration was set at 8 to avoid disturbing the gradient at the end of the run. The dormant spores concentrated in the 70% layer at the bottom of the gradient, whereas germinated spores formed a band in the 50% layer. For purification of dormant spores, the 70% layer (0.1 to 0.2 ml) was recovered with a Pasteur pipette, diluted 10-fold in water, and centrifuged for 20 s to pellet the spores. The dormant spores were washed 10 times with 1 ml of water before use. For enrichment of germinated spores, the 50% layer (0.2 ml) was recovered with a Pasteur pipette and inoculated into 5 ml of 2×YT broth. After the culture had grown to saturation at 37°C, it was divided into 1-ml aliquots which were either frozen for storage, plated out for screening individual colonies, or subcultured into 200 ml of 2×SG medium for sporulation.

Assays of spore germination. A modification of a previously described filter assay (12, 15) was used to identify B. subtilis colonies whose spores germinated in D-alanine. Briefly, B. subtilis colonies were patched onto 2×SG agar plates (wrapped in a plastic bag to reduce drying) and sporulated by incubation at 37°C for 5 days. The sporulated colonies were lifted onto nitrocellulose filters, which were then baked at 65°C for 3 h to kill vegetative cells and heat activate dormant spores. After cooling to room temperature, the filters were placed on a Whatman 3MM paper disc soaked in germination solution (10 mM Tris-HCl [pH 8.4]), 1 mg of 2,3,5-triphenyltetrazolium chloride per ml, 2.5 mM glucose, 10 mM test germinant) and incubated at 37°C for 4 to 8 h. Colonies that contained germinating spores developed a red color because germinated but not dormant spores can reduce the tetrazolium dye (12, 15). Glucose was included in the germination solution because it enhanced red color development in the control studies which were used to standardize the protocol.

Genetic mapping. The B. subtilis mapping strains, 1A627 to 1A645, were obtained from the Bacillus Genetic Stock Center, Ohio State University, and phage PBS1 stock was obtained from Wayne Nicholson, University of Arizona. Standard procedures were used for phage PBS1 manipulation (4), except that 2× nutrient broth was used in place of brain heart infusion broth to culture B. subtilis cells for infection.

Recovery of the gerB operon from wild-type and mutant B. subtilis strains. To recover the gerB operon from B. subtilis strains, the 5' region of the gerB locus was PCR amplified from strain PS832 chromosomal DNA with primers gerB06 (5' GGTGATTGCGTCATGATCC) and gerB18 (5' GAAATGGCCATTCTAGTCGG), which hybridize starting at (underlined positions) nt 279 and +950 relative to the gerBA translation start site (+1). A 643-bp HindIII-EcoRV fragment contained within the PCR fragment was subcloned between the same sites in plasmid pJL74 (13) to create plasmid pFE16. Plasmid pFE16 was used to transform the B. subtilis strain whose gerB operon was to be recovered to spectinomycin resistance. Transformants in which plasmid pFE16 had inserted at the gerB locus by Campbell integration (Fig. 1) were identified by Southern blot analysis and are designated gerB:spc:gerB because they contain a gerB operon, which is truncated at the first EcoRV site in Fig. 1, followed by the spc cassette and then a full-length gerB operon with an intact promoter (2) (Fig. 1). Chromosomal DNA from those transformants was linearized with SstI and ligated, and the ligation mix was used to transform E. coli TG1 to Ap resistance (Fig. 1). Plasmids carrying the 5.3-kb gerB fragment from the different strains are designated as follows: pFE24, wild-type strain PS832; pFE23, mut4 strain FB8; pFE25, mut8 strain FB9; pFE26, mutb1 strain FB10; pFE28, mutb2 strain FB11; and pFE29, muta2 strain FB12.

Site-directed mutagenesis. The 1.5-kb BamHI fragment from the wild-type gerB operon was cloned at the BamHI site in pUC19 to generate plasmid pFE45, which was mutagenized by using a Transformer site-directed mutagenesis kit (Clontech, Palo Alto, Calif.). The selection primer pUC19-RI/RV (5' CGGCCAGTGATATCGAGCTCGG) was used in combination with one of three mutagenic primers, gerBMut4 (5' CATTTATTTGCCCAGTCTGTATATTTCTC), gerBMut8 (5' GCAGGCTTAACGTATCATTCCCGCC), or gerBMutb1 (5' TATTGAACGAATTGATTTGTTCTTACAG), to introduce the gerBA1*, gerBA2*, or gerBB1* mutation, respectively. Each mutagenized 1.5-kb BamHI region was sequenced completely to ensure that it carried only the site-directed mutation. The BamHI fragment from each mutant plasmid pFE67 (gerBA1*), pFE68 (gerBA2*), and pFE69 (gerBB1*), was then used to replace the BamHI fragment from the gerB operon in plasmid pFE24 to construct the single-mutant gerB operon plasmids pFE70 (gerBA1*), pFE71 (gerBA2*), and pFE72 (gerBB1*). The gerBA1* gerBB1* double-mutant plasmid (pFE76) was constructed by replacing a 1.6-kb ClaI-StuI fragment (Fig. 1) in pFE72 with the same fragment from pFE70. The gerBA2* gerBB1* double-mutant plasmid (pFE77) was similarly constructed from pFE72 and pFE71.

Integration of wild-type and mutant gerB operons at the amyE locus. The wild-type and mutant gerB operons were cloned into plasmid pDG364 (5) in two steps. Initially, we constructed pFE96, which is a pDG364 derivative containing a wild-type gerB operon (including its own promoter [2]) lacking the internal 1.5-kb BamHI fragment. In the second step, the 1.5-kb BamHI fragments from the wild-type (pFE24) and mutant (pFE70, pFE71, pFE72, pFE76, pFE77) gerB plasmids were cloned in the correct orientation into pFE96 to generate plasmids pFE97 through pFE102, respectively. Each plasmid was linearized with BglII and used to transform a gerB::spc strain, FB41, to chloramphenicol resistance. Transformants in which the plasmid-borne gerB operon had integrated at the amyE locus were identified by their amy phenotype and Southern blot analysis.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Isolation of D-alanine responsive mutants. To identify spore germinant receptor(s), we decided to isolate B. subtilis mutants whose spores germinated in the novel germinant D-alanine because we expected such mutants to arise as the result of mutations in a gene encoding a preexisting germinant receptor. As it is difficult to identify rare mutant spores that germinate in D-alanine within a population of wild-type spores, we initially enriched a spore population for mutants that could germinate in D-alanine. The enrichment was achieved by separating germinated and dormant spores on the basis of their differential migration in a buoyant density gradient (18). The separation protocol was standardized for spores of our wild-type strain PS832 by centrifuging a mixture of germinated (in 10 mM L-alanine) and ungerminated spores in a 20 to 70% metrizoic acid gradient. After centrifugation, the spores were concentrated in two major bands (data not shown); the dormant spores migrated to the 70% metrizoic acid layer, while the germinated spores concentrated in the 50% metrizoic acid layer. The resolution of the two bands was further improved by increasing the height of the intervening 60% metrizoic acid layer (Materials and Methods).

Response of the mutants to different germinants. While we hoped that the mutant spores were germinating specifically in D-alanine, it was possible that they were simply unstable and had a tendency to germinate nonspecifically. To address this possibility, wild-type and mutant spores were purified, heat activated, and incubated at 37°C in a germination mix (10 mM Tris-HCl [pH 8.4], 1 mM D-glucose) with or without added D-alanine. Germination of the spore suspensions was followed by measurement of the OD₆₀₀, which decreases as the phase-bright dormant spores germinate and become phase dark. In the germination reaction lacking D-alanine, neither wild-type nor mutant spore suspensions showed a significant change in OD₆₀₀ (<2%) (Fig. 2A and data not shown), indicating that none of those spores germinated in the absence of D-alanine. When 10 mM D-alanine was added to the germination reaction, spores from all five mutants but not wild-type spores germinated (Fig. 2A and data not shown). The requirement for D-alanine seemed to be saturable since germination of the mutant spores in 10 mM D-alanine was comparable to that in 20 mM D-alanine but faster than that in 1 mM D-alanine (data not shown). These observations suggested that germination of the mutant spores in D-alanine was not due to spore instability and was dependent on the presence of D-alanine in the germination reaction. Nevertheless, germination of the mutant spores in D-alanine was slower than in L-alanine (Fig. 3A; see below), suggesting that D-alanine was not an optimal germinant.

View larger version (16K): [in this window] [in a new window]   FIG. 2.   Germination of mutant spores in the novel germinant, D-alanine, in the presence (A) or absence (B) of D-glucose. Spores from wild-type strain PS832 () or mutant strain FB8 (mut4) (, ), FB9 (mut8) (), or FB10 (mutb1) (, ) were heat activated and subsequently incubated in 10 mM Tris-HCl (pH 8.4) buffer (open symbols) or buffer supplemented with 10 mM D-alanine (closed symbols) with (A) or without (B) 1 mM D-glucose at 37°C. The OD600 (shown here as A600) of each sample was measured periodically and plotted as a fraction of the initial OD600 [A600(t)/A600(init)] versus time. Spores from all strains produced overlapping, reasonably flat curves when incubated in buffer alone, and only one representative curve is shown ( in panel A and  in panel B). Wild-type PS832 spores produced identical curves in D-alanine in the presence or absence of D-glucose, and only one representative curve ( in panel B) is shown.

View larger version (17K): [in this window] [in a new window]   FIG. 3.   Germination of mutant spores in the germinants L-alanine (A) and L-asparagine (B). Spores from wild-type strain PS832 () or mutant strain FB8 (mut4) (, ), FB9 (mut8) (), or FB10 (mutb1) () were assayed for germination in buffer alone (open symbols), 10 mM L-alanine (closed symbols in panel A), or 10 mM L-asparagine (closed symbols in panel B) as described in the legend to Fig. 2. Note that D-glucose was not present in these germination reactions. Spores from all strains produced overlapping, reasonably flat curves when incubated in buffer alone, and only one representative curve is shown ( in both panels).

Genetic mapping of the mutant loci. Because preliminary characterization of the mutants suggested that they might contain mutations in a germinant receptor, we genetically mapped the mutations. Initially, we used PBS1-mediated generalized transduction to determine the linkage between the mut4 mutation in strain FB8 and the MLS resistance marker in 19 B. subtilis mapping kit strains, each of which carries the MLS resistance marker at a unique chromosomal location (31). PBS1 transducing lysates made in each mapping strain (1A627 to 1A645) were used to transduce strain FB8 to MLS resistance, and spores from at least 50 MLS-resistant transductants were tested for germination in D-alanine by the plate assay. We found that 85% of the MLS-resistant transductants obtained from lysates made in strain 1A644 had lost the mutant phenotype. Thus, the MLS resistance marker and the wild-type allele of the mut4 mutation from strain 1A644 cotransduced 85% of the time, suggesting that the two loci were linked. Consistent with this finding, none of the MLS resistance markers in the 18 other mapping strains showed any linkage to the mut4 locus. To determine if the remaining four mutations (mut8, mutb1, mutb2, and muta2) mapped within the same region, we measured the frequency at which they cotransduced with the MLS resistance markers from strains 1A644 and 1A645. Again, all four mutations cotransduced 80 to 90% of the time with the MLS resistance marker in strain 1A644 but showed no significant cotransduction with the MLS resistance marker in strain 1A645. Thus, all five mutations were linked to the MLS resistance marker located at 316 degrees on the chromosome in strain 1A644.

mut4 gerB

Effect of a gerB mutation on the mutant germination phenotype. The tight linkage of the mut4 mutation to the gerB locus suggested that the mutation might affect a gerB cistron. This idea was also consistent with the response of the mutant spores to AFGK and L-asparagine. We reasoned that if a mutant GerB protein was indeed responsible for the mut4 phenotype, then disruption of the gerB operon would eliminate the mutant phenotype. To test this prediction, the gerB operon was disrupted in the mut4 mutant, and spores from the mut4 strain FB8 and its gerB derivative strain FB34 were tested for germination in various germinants. Unlike the mut4 spores, the mut4 gerB double-mutant spores failed to germinate in D-alanine (Fig. 4A). The germination defect of the double-mutant spores was specific to D-alanine and was not apparent in other germinants such as L-alanine (Fig. 4B) or a rich medium (data not shown). Thus, the mut4 spores required an intact gerB operon for germination in D-alanine. Moreover, that requirement was specific to the gerB operon because disruption of gerA, which is highly homologous to gerB (3), did not affect germination of mut4 spores in D-alanine (Fig. 4A).

View larger version (14K): [in this window] [in a new window]   FIG. 4.   Effect of gerA or gerB disruption on mut4 spore germination in D-alanine (A) and L-alanine (B). Spores from the mut4 strain FB8 (, ), mut4 gerA::spc strain FB22 (), or mut4 gerB::spc strain FB34 () were assayed for germination in buffer (open symbols) with either 10 mM D-alanine and 1 mM D-glucose (closed symbols in panel A) or 10 mM L-alanine alone (closed symbols in panel B) as described in the legend to Fig. 2. Spores from all strains produced overlapping, flat curves when incubated in buffer alone, and only one representative curve is shown ( in panel B).

Recovery of the gerB operon from the mutant strains. Because a variety of criteria suggested that the mutations which allowed spore germination in D-alanine affected the gerB operon, we decided to localize the mutations within the gerB operon. For this purpose, a 5.3-kb genomic DNA fragment, which contained the gerB operon and 1.6 kb of downstream DNA, was recovered from wild-type and mutant strains by a two-step integration-recovery method (Fig. 1). The recovered DNA, which was not linked to a selectable marker, was then introduced into the wild-type strain PS832 by congression (Table 2) with the unlinked, chromosomal MLS resistance marker from strain 1A640. Spores from at least 49 MLS-resistant transformants were tested for germination in D-alanine by the plate assay. As expected, all transformants obtained by introduction of the wild-type 5.3-kb DNA fragment produced only wild-type spores (Table 2). However, when the genomic fragment derived from the mut4 mutant was used, 21% of the MLS-resistant transformants produced spores that germinated in D-alanine (Table 2). Thus, the genomic fragment containing the gerB operon and some downstream DNA from the mut4 strain conferred the mutant phenotype on an otherwise wild-type strain. Similar experiments showed that the same 5.3-kb genomic fragment from each mutant was sufficient to confer the mutant phenotype in a wild-type strain (Table 2), suggesting that all five mutations lay within the same 5.3-kb region of the chromosome.

Sequences of the gerB operons from mutant and wild-type strains. Because the mutations mapped within the 1.5-kb BamHI region in the gerB operon, we identified the mutations at the DNA level by sequencing that region of the gerB operon from wild-type and mutant strains. Compared to the wild-type sequence, the gerB operon from the mut4 mutant showed a single GA transition which resulted in a Gly₂₉₇ (GGU)Ser (AGU) substitution in the gerBA open reading frame (Fig. 5). The sequence of the gerB operon from the mut8 mutant differed from the wild-type sequence by a single CT transition in the gerBA open reading frame (Fig. 5). This transition produces a Pro₃₂₆ (CCA)Ser (UCA) alteration in the predicted GerBA protein (Fig. 5). The gerB operon from the mutb1 mutant showed no alteration in the gerBA cistron but contained a single TA transversion which produced a Phe₂₆₉ (TTT)Ile (ATT) substitution in the gerBB open reading frame (Fig. 5). The gerB operons from the mutb2 and muta2 mutants contained the same TA transversion, indicating that these three mutants probably arose as the result of a single mutagenic event. Thus, the screen yielded three independent mutations in the gerB operon, two in the gerBA cistron and one in the gerBB cistron, that allowed spores to germinate in D-alanine. The mutant alleles from the mut4, mut8, and mutb1 strains will be referred to as gerBA1*, gerBA2*, and gerBB1*, respectively, in the remainder of the text.

View larger version (40K): [in this window] [in a new window]   FIG. 5.   Locations of the mutations within the gerB operon. The DNA sequence of the 1.5-kb BamHI fragment from strain PS832 is shown together with the predicted protein sequences of the gerBA and gerBB open reading frames. The locations of the gerBA1* (Gly297 [GGU]Ser [AGU]), gerBA2* (Pro326 [CCA]Ser [UCA]), and gerBB1* (Phe269 [UUU]Ile [AUU]) mutations are represented by boldface underlined letters. Deviations of the gerB sequence from that of our wild-type strain PS832 and the published sequence (11, 17) and the resulting amino acid changes (if any) are underlined.

Introduction of each gerB* mutation into a wild-type strain. Although the data presented above strongly indicated that the mutations which we had identified by DNA sequence analysis were solely responsible for the mutant phenotype, we felt it important to prove this point conclusively. To this end, each mutation was first engineered by site-directed mutagenesis into a plasmid containing the wild-type 1.5-kb BamHI fragment. The entire mutagenized DNA fragment was sequenced to ensure that it contained only the appropriate mutation and then used to replace the BamHI fragment in the wild-type gerB plasmid, pFE24. The resulting plasmid was introduced into the wild-type strain PS832 by congression with the chromosomal MLS resistance marker from strain 1A640, and spores from at least 50 MLS-resistant transformants were scored for germination in D-alanine. When a plasmid carrying any one of the three mutagenized BamHI fragments was used, 15 to 50% of the colonies yielded spores that germinated in D-alanine (Table 3). By comparison, none of the MLS-resistant transformants obtained by introduction of the wild-type gerB plasmid, pFE24, exhibited the mutant phenotype (Table 3). Thus, each of the three mutations allowed otherwise wild-type spores to germinate in D-alanine.

View larger version (13K): [in this window] [in a new window]   FIG. 6.   Dominant/recessive nature of the gerB mutations. (A) Germination of the gerB amyE::gerBA1* haploid spores (FB44) (), gerB amyE::gerBA1* merodiploid spores (FB50) (), or gerB amyE::gerB haploid spores (FB43) () in 10 mM D-alanine-1 mM D-glucose was assayed as described in the legend to Fig. 2. (B) Germination of gerB amyE::gerBA2* haploid spores (FB45) () and gerB amyE::gerBA2* merodiploid spores (FB51) () in 10 mM D-alanine-1 mM D-glucose. (C) Germination of gerB amyE::gerBB1* haploid spores (FB46) () and gerB amyE::gerBB1* merodiploid spores (FB52) () in 10 mM D-alanine-1 mM D-glucose. Germination curves of gerB amyE::gerB merodiploid spores (FB49) in 10 mM D-alanine-1 mM D-glucose and of all spores in buffer alone were identical to that of the gerB amyE::gerB haploid ( in panel A) and are not shown.

Dominant/recessive nature of the gerBA and gerBB mutations. To determine if the phenotype conferred by the gerB* mutations could be attributed to a new function gained by the mutant GerBA* and GerBB* proteins, we examined if the gerB* mutations were dominant over the wild-type gerB allele. Haploid strains (which contained a single-mutant gerB* operon) and merodiploid strains (which contained a wild-type gerB and a mutant gerB* allele) were constructed by inserting mutant gerB operons at the amyE locus in either a gerB::spc and a wild-type strain, respectively. The strains were sporulated by resuspension, and the spores were assayed for germination in D-alanine. Spores from all three merodiploids germinated in D-alanine (Fig. 6), indicating that all three mutations were dominant over the wild-type gerB allele. Thus, the mutant phenotype probably results from a function gained by the mutant GerBA* and GerBB* products.

Combination of mutations in gerBA and gerBB. To determine the interaction between the gerBA* and gerBB* mutations, we examined the germination characteristics of spores containing mutations in both genes. Double-mutant gerBA1* gerBB1* or gerBA2* gerBB1* operons were derived from the single-mutant gerB* plasmids and inserted at the amyE locus in the gerB::spc strain FB41. While preparing spores from the double-mutant strains, we observed that 20 to 30% of the spores germinated in the distilled water used to wash the spores. This anomalous germination of the double-mutant spores was independent of the sporulation conditions and was not apparent in any of the single-mutant spores. Thus, the mutations in gerBA and gerBB seemed to enhance one another. Consistent with this idea, the double-mutant spores turned red much faster (in less than one-fifth the time) than the single mutants (data not shown) in the plate assay for D-alanine-induced germination.

View larger version (17K): [in this window] [in a new window]   FIG. 7.   Combinations of gerBA* and gerBB* mutations. (A) Germination of gerB amyE::gerBA1* (FB50) (, ), gerB amyE::gerBB1* (FB52) (), and gerB amyE::gerBA1* gerBB1* (FB56) (, ) spores in 10 mM Tris-Cl (pH 8.4) in the absence (open symbols) or presence (solid symbols) of 10 mM D-alanine and 1 mM D-glucose was assayed as described in the legend to Fig. 2. (B) Germination of gerB amyE::gerBA2* (FB51) (, ), gerB amyE::gerBB1* (FB52) (), and gerB amyE::gerBA2* gerBB1* (FB57) (, ) spores was assayed as described above.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Accurate recognition of germinants is critical to ensure that dormant spores germinate only under favorable environmental conditions. In B. subtilis spores, recognition of the germinant L-alanine or AFGK is thought to be mediated by specific receptors (14). In this report we have described a new strategy to genetically identify putative germinant receptor(s) in B. subtilis. Our findings suggest that two proteins encoded by the gerB operon are components of a germinant receptor, and thus our work supports previous studies (16, 24) which had proposed a role for gerB in germinant recognition. In addition, our studies suggest that the germinant receptor is a complex of at least two proteins, both of which are most likely integral membrane proteins.

The gerB locus was originally implicated in AFGK recognition because inactivating mutations at that locus specifically blocked AFGK-induced germination (24). In this study, we identified three dominant mutations in the gerB operon which allowed spores to germinate in the novel germinant D-alanine. Whereas loss of gerB function blocked germination in AFGK (3, 15, 16), gain-of-function gerB mutations allowed spores to germinate in D-alanine. These findings are best explained by a model in which gerB encodes one or more components of a receptor required for AFGK-induced germination. In this model, a dysfunctional AFGK receptor could account for the germination defect of gerB mutant spores, whereas a subtle structural alteration of the receptor could explain why our dominant gerB* mutations allow spores to germinate in D-alanine (see below). But why would alterations in the AFGK receptor allow it to recognize D-alanine? In addition to AFGK, gerB was shown to mediate germination in a mixture of L-alanine, D-fructose, D-glucose, and K⁺ ions (AlaFGK) (24). Moreover, in both mixtures, AFGK and AlaFGK, gerB was implicated in recognizing the amino acid (3). This ability of the gerB receptor to recognize a range of amino acids could account for its repeated isolation in our screen for mutations that produce a D-alanine-responsive receptor.

The gerB operon encodes three putative proteins, GerBA, GerBB, and GerBC, all of which are required for AFGK-induced germination (3). However, it is not clear which, if any, of these proteins are part of the germinant receptor. In this study, we identified mutations in gerBA and gerBB that allowed spores to germinate in D-alanine. All of these mutations were dominant, indicating that both mutant GerBA* and mutant GerBB* proteins could affect germinant recognition. Thus, both GerBA and GerBB seem to be components of the germinant receptor, suggesting that the receptor is actually a complex of several proteins. Such a model would account for the genetic interaction between the gerB and gerB* alleles, as the ability of the wild-type gerB allele to partially mask the phenotype of gerBA* and gerBB* could result from competition between wild-type and mutant proteins for incorporation into the receptor complex. For example, if the receptor was a GerBA-GerBB dimer, then all of the GerBA* and GerBB* molecules would be incorporated into GerBA*-GerBB* double-mutant receptors in gerBA* gerBB* haploid spores. However, only one-half of the mutant products would form double-mutant receptors in merodiploids because the remaining molecules would be incorporated into GerBA*-GerBB or GerBA-GerBB* receptors, and thus the merodiploids would have fewer double-mutant receptors. On this basis, we propose that the germinant receptor is a complex of GerBA and GerBB proteins, both of which play a role in recognition of the germinant. It is possible that the receptor complex also contains products of genes which were not identified in the screen because of a low frequency of gain-of-function mutations, and further studies will be needed to elaborate the constitution of the receptor complex.

The predicted GerBA and GerBB proteins contain 5 and 10 putative membrane-spanning domains, respectively (3), suggesting that they are probably integral membrane proteins. Thus, it is tempting to speculate that the germinant receptor complex is associated with and transduces a germinant signal across a spore membrane. The spore is surrounded by an inner membrane that is derived from the forespore and an outer membrane which originates from the mother cell (16). Because the integrity of the outer membrane is questionable (16), it is not clear which of the two membranes forms the outermost barrier across which the germinant signal must be transduced (16). Thus, it is not currently possible to predict the location of the germinant receptor. Moreover, recent studies attempting to localize the GerA proteins, which are also proposed to constitute a germinant receptor (38), gave contradictory results about the membrane in which those proteins are located (14, 22, 36). Thus, identification of the membrane that harbors the germinant receptor, and presumably marks the site where the germination reaction is initiated, remains an important issue to be addressed.

In addition to gerB, previous genetic studies identified two other operons, gerA and gerK, that were implicated in germinant recognition (8, 24). Both of these operons encode proteins that are homologous to the gerB products and therefore could perform a similar function (14). The gerA operon is required for germination in L-alanine and might encode a germinant receptor that is dedicated to L-alanine recognition. Consistent with gerA and gerB encoding two distinct receptors, we found that a gerA disruption did not affect the gerB* mutant phenotype. The gerK operon probably encodes a distinct glucose receptor, as gerK was proposed to mediate the effects of D-glucose in AFGK- and AlaFGK-induced germination (8). In addition, the Bacillus genome sequence (11, 17) has revealed two more operons, yndDEF and yfkQRT, that share sequence homology with the gerB operon. Thus, it is likely that B. subtilis spores contain a family of germinant receptors that mediate responses to diverse germinants.

In conclusion, we propose that the gerB operon and its homologues encode a family of multicomponent receptors that recognize environmental germinants and trigger germination. Further biochemical studies of the proteins encoded by gerB should allow us to test various predictions of the model presented here and refine our understanding of the germinant receptor. In addition, the dominant gerB mutations identified here can be used in genetic epistasis tests to define the ger loci that act downstream of the receptor. The identification of those loci should provide us with insights into how the receptor ultimately triggers germination.