Role of the gerA operon in L-alanine germination of Bacillus licheniformis spores

Background The genome of Bacillus licheniformis DSM 13 harbours three neighbouring open reading frames showing protein sequence similarities to the proteins encoded from the Bacillus subtilis subsp. subtilis 168 gerA operon, GerAA, GerAB and GerAC. In B. subtilis, these proteins are assumed to form a germinant receptor involved in spore germination induced by the amino acid L-alanine. Results In this study we show that disruption of the gerAA gene in B. licheniformis MW3 hamper L-alanine and casein hydrolysate-triggered spore germination, measured by absorbance at 600 nm and confirmed by phase contrast microscopy. This ability was restored by complementation with a plasmid-borne copy of the gerA locus. Addition of D-alanine in the casein hydrolysate germination assay abolished germination of both B. licheniformis MW3 and the complementation mutant. Germination of both B. licheniformis MW3 and the gerA disruption mutant was induced by the non-nutrient germinant Ca2+-Dipicolinic acid. Conclusions These results demonstrate that the B. licheniformis MW3 gerA locus is involved in germination induced by L-alanine and potentially other components present in casein hydrolysate.


Background
Germination of dormant Bacillus spores and subsequent outgrowth can be induced by various nutrients (amino acids, purine nucleosides, sugars, ions and combinations of these) recognised by receptor proteins encoded by the gerA family operons [1][2][3] and located in the inner membrane of the spore [4][5][6][7]. One or several germination receptor operons have been detected in the genomes of almost all spore formers, and supported by studies of different mutants it has been concluded that spores respond to germinants via receptors diverged from common ancestor(s) ( [6] and references therein). Studies of receptor/germinant interactions have so far mainly been focusing on species belonging to Bacillus cereus, Bacillus subtilis, Bacillus megaterium and Bacillus anthracis [3,[8][9][10][11][12][13][14][15][16]. Bacillus licheniformis, another Gram-positive, spore forming soil bacterium closely related to B. subtilis [17], has on the other hand gained much less attention. B. licheniformis is a frequent contaminant of foods, and is a common spoilage organism of dairy products [18][19][20], bread [21,22], packaged meats [23] and canned goods [24]. It has previously been considered non-pathogenic, and has been widely used in the industry for production of enzymes, antibiotics and biochemicals [25][26][27]. However, B. licheniformis-associated bovine abortion [28,29], implant infection [30], corneal ulcer [31], bacteraemia sepsis [32] and food poisoning [33,34] raise the question of its pathogenic potential. Some strains of B. licheniformis associated with human disease are capable of producing lichenysin A, a surfactin-like toxin [34,35]. Due to its association with food-borne illness and spoilage, and its ability to undergo sporulation, [17,[36][37][38], extended knowledge about the germination apparatus of B. licheniformis is of general interest. To ensure microbiological safe food production of durable foods produced by relatively mild heat treatment, there is an obvious need for more information on spore forming bacteria.
Based on existing literature, B. subtilis could be considered as the model organism for germinant receptor studies. It was through early studies of germination defective mutants, that the theory of a L-alanineinduced germinant receptor was proposed [8]. Later studies identified the gerA locus as a tricistronic operon weakly expressed during sporulation, and that the polypeptide products of gerA probably formed a membrane associated complex [39][40][41]. The products of each of the three genes of gerA were later named GerAA, GerAB and GerAC, and were demonstrated to be simultaneously required for the spore to respond to L-alanine as sole germinant [2]. Genome sequence analysis and germination experiments of different mutants further identified four other tricistronic gerA homologs for B. subtilis; gerB, gerK, yndDEF and yfkQRT [10]. Receptors encoded by two of these operons, gerB and gerK, are confirmed functional when acting cooperatively with each other or with gerA [10,15].
Homologous genes of germinant receptors belonging to the gerA family have been found in most spore formers, although the exact number, organisation and corresponding response germinant may vary for different species and even strains [3,42,43]. B. licheniformis ATCC 14580 is also predicted to possess potential germinant receptor proteins belonging to both the GerA and the GerK clades [44]. The GerAA, GerAB and GerAC protein sequences of B. licheniformis ATCC14580 are closely related to the protein sequences of the corresponding germinant receptor subunits of Bacillus subtilis subsp. subtilis 168. These are in B. subtilis encoded by the gerA operon, gerAA, gerAB and gerAC. Since B. subtilis gerA germination is triggered by L-alanine [2,15], it is plausible that the B. licheniformis gerA operon also is involved in L-alanine germination. It has earlier been documented that spores of B. licheniformis from different strains actually respond to L-alanine as germinant [45][46][47], but to our knowledge, there are no functional studies of receptor/germinant interactions of strains belonging to B. licheniformis.
Mutational studies of B. licheniformis, including the fully sequenced B. licheniformis ATCC 14580/DSM 13 strain [48,49], have long been a challenge, most likely due to their possession of a restriction apparatus destroying foreign DNA [48][49][50]. The construction of a more easily transformable mutant, B. licheniformis MW3, has largely overcome this challenge [50].
In order to facilitate the understanding of germinant/ receptor interactions in B. licheniformis, we have constructed disruption and complementation mutants of the gerAA locus in B. licheniformis MW3. Spores of these mutants have been studied in germination assays with L-alanine, casein hydrolysate and the non-nutrient germinant Ca 2+ Dipicolinic acid (Ca 2+ DPA). These studies reveal that gerA is a main germinant receptor complex of B. licheniformis recognising amino acid(s), and supports the view that L-alanine is an important nutrient-germinant for this species.

Construction of the disruption and complementation mutants
To elucidate the role of the hypothetical GerA proteins during spore germination, a disruption mutant of the gerAA locus in B. licheniformis MW3 was constructed. B. licheniformis MW3 was used as target strain due to its superior transformability compared to its fully sequenced parent strain DSM 13 [50]. The gerAA mutant, NVH-1307, was constructed so that a part of the gerAA gene was substituted with a spectinomycin resistance cassette. This will cause the mutant to acquire spectinomycin resistance, and in addition, affect a potential phenotype related to the disrupted gene. If the target gene is part of an operon, which is the case of gerAA, downstream transcripted genes will also be affected, and the receptor non functional. Sequence analysis showed that in addition to harbouring the spectinomycin cassette in the gerAA locus, NVH-1307 also harboured two additional mutations (one base substitution and one base deletion) in the gerAA locus. These mutations were most likely acquired during PCR amplification of the fragments used to construct the disruption vector (pMAD_Sp R ΔgerAA). These mutations were "accepted" (not corrected) due to their location in the gene targeted for disruption. However, in construction of the plasmid used for gerAA complementation, a polymerase with a higher expected fidelity was applied to limit the risk of such mutations. Sequence analysis of the complementation plasmid pHT315_MW3gerA revealed no mutations in the amplified gerA operon when compared to the sequence of Veith et al. [48].
Genetic modification studies have shown that the germination rates could be significantly increased when specific germinant receptors are over-expressed in B. subtilis [51]. Thus, expression of germinant receptors is apparently not optimised for maximal spore germination, forwarded as a possible evolutionary strategy to prevent premature germination at nutrient conditions inadequate for sustained vegetative growth [3]. Very high levels of receptor expression could on the other hand have a negative effect on the sporulation process [51]. In such cells, the forespore lyses during the process of sporulation, perhaps as a result of premature forespore germination [51]. An appropriate evolutionary adaptation of germinant receptor expression/regulation is thus crucial to allow the cyclic transition between sporulation and germination upon environmental changes. In the construction of the complementation mutants in our study, certain precautions were therefore taken to avoid extensive over-expression of the complemented germinant receptor genes. By including some of the flanking regions of the gerAA, gerAB and gerAC fragment in the complementation plasmid, we wanted to maintain the native regulatory elements of this locus. In addition, a shuttle-vector with an expected low or moderate copy number was sought as a basis for the complementation plasmid. To our knowledge, there is no shuttle-vector available for B. licheniformis where the copy number is demonstrated to be low or moderate. However, Arantes and Lereclus [52] have constructed the pHT315 E. coli/B. thuringiensis shuttle-vector, with a copy number of~15 per equivalent B. thuringiensis chromosome. This vector has successfully been used in germinant receptor complementation studies in B. megaterium [53], and was thus considered as a reasonable choice for B. licheniformis. Despite that this vector has shown to be stably maintained in B. thuringiensis and B. megaterium without a selective pressure [52,54], the antibiotic erythromycin had to be included to ensure persistence of the complementation plasmid during sporulation of the B. licheniformis complementation mutant NVH-1311. This could be due to a different segregation stability of the vector in B. licheniformis. Another possibility is that there is a potential elevated risk of plasmid curing due to sporulation at a high temperature. Sporulation of B. licheniformis MW3, NVH-1307 and NVH-1311 were performed at 50°C since a pilot study showed that sporulation at this temperature was faster, yielded more stable spores (less spontaneous germination) and a higher percentage of phase bright spores (results not shown).
Disruption of gerAA abolish L-alanine and casein hydrolysate induced germination Decrease in absorbance at~600 nm (A 600 ) is used as a convenient method to monitor and compare germination of different spore populations [55,56]. A fall in absorbance reflects a change in the refractive index (light scattering) of the multiple individual spores in a suspension, associated with germination events such as the excretion of spore's depot of Ca 2+ -DPA, followed by water influx, cortex degradation and core swelling [51,[56][57][58][59]. Figure 1 shows a representative experiment where different strains of heat activated (65°C 20 min) spores (in Phosphate buffer) are supplemented with the germinant L-alanine. At these conditions, a clear change in absorbance was observed for spores of wild type (MW3) and wild type complementation mutant (NVH-1311) supplemented with L-alanine. Less than a 5%/h decrease in absorbance was observed for spores of the disruption mutant (NVH-1307). Phase-contrast images captured at the end of the germination assay ( Figure 2), a technique where germinated and dormant spores appear dark and white/bright respectively [56,59,60], supports the absorbance measurements showing that spores of the wild type (MW3) and wild type complementation mutant (NVH-1311) germinate with L-alanine. No apparent increase in number of phase dark spores was observed for spores of the deletion mutant (NVH-1307) supplemented with L-alanine, or the negative controls. Together with the absorbance measurements, this shows that the introduced disruption of the gerAA gene abolishes the ability of B. licheniformis MW3 to use L-alanine as a germinant. The fact that the NVH-1311 complementation mutant showed a similar L-alanine triggered germination phenotype as the wild type spores, supports the hypothesis that an undisrupted copy of the gerAA, gerAB and gerAC genes, with flanking elements, are required for normal germination of B. licheniformis MW3 at these conditions. These findings were also supported by experiments performed with an alternative germination buffer; 50 mM Tris HCl pH 7.4 10 mM KCl (E. Klufterud, C. From; unpublished results).
An earlier study where germination in seven strains of B. licheniformis was investigated, showed that out of 24 amino acids tested, only L-alanine, L-cysteine and Lvaline markedly stimulated germination [46]. In general, a greater germination response with L-alanine than with L-cysteine and L-valine was observed [46]. To assay the germination response of MW3, NVH-1307 and NVH-1311 to several amino acids, casein hydrolysate was used. Casein hydrolysate consists of a mixture of amino acids made from acid hydrolyzation of the milk protein casein and has been used as a germinant for Clostridium bifermentans and B. cereus in earlier studies [61][62][63]. In our study, casein hydrolysate proved to be a potent germinant for B. licheniformis, giving a rapid germination response (~70% phase dark spores as visualised by phase contrast microscopy) both for the wild type MW3 and the complementation mutant NVH-1311. The mutant NVH-1307, which most likely lacks a gerA receptor, did not show any germination response in casein hydrolysate ( Figure 3). D-alanine is a well-known inhibitor of L-alanine germination of B. subtilis and B. licheniformis [64,65,46,15,66]. D-alanine has also been shown to reduce L-valine induced germination of B. subtilis [15,66], but we are not aware of studies reporting the effect of D-alanine on L-valine induced germination of B. licheniformis. In order to abolish germination by Lalanine present in the casein hydrolysate, we added Dalanine in some of the above experiments. In these experiments, the germination response of both MW3 and NVH-1311 was hardly measurable (results not shown), indicating that L-alanine through its triggering of the gerA receptor is an important germinant of B.
licheniformis. The contribution to germination of the remaining amino acids in the casein hydrolysate when D-alanine was present, appear to be minimal. Although one can not rule out that D-alanine also inhibits the effect of other amino acids present in casein hydrolysate (e.g. L-valine), all the findings support the view that gerA and L-alanine constitute one of the main germination pathways of B. licheniformis.

Germination of B. licheniformis with Ca 2+ -DPA
In order to by-pass the spore germination receptor apparatus, experiments using exogenous Ca 2+ -DPA to trigger germination of spores of B. licheniformis MW3 and the mutant strain NVH-1307 were performed. In B. subtilis spores, Ca 2+ -DPA induced germination is believed to act through activation of the cortex lytic enzyme CwlJ, without any requirement of functional germinant receptors [10,67]. Bioinformatic analysis of complete genomes of different spore formers has shown that also B. licheniformis contains a B. subtilis homologous cwlJ gene [43]. If the germination apparatus of B. licheniformis spores is similar to that of its close relative B. subtilis, the wild type and disruption mutant of B. licheniformis should exhibit a similar germination response as B. subtilis to exogenous Ca 2+ -DPA. The DPA concentration needed to trigger germination in B. subtilis is~20 -60 mM, supplemented together with equal (or excess) amounts of Ca 2+ (allowing formation of a 1:1 chelate of calcium and dipicolinic acid) [10]. Also spores of B. cereus and B. megaterium germinate when exposed to Ca 2+ -DPA [68,69]. For B. cereus it has been shown that a final level of 60 mM Ca 2+ -DPA is sufficient to ensure germination [69]. In our experiments (

Conclusions
As demonstrated by genetic mutation and complementation analysis, this study reveals that the gerAA gene in B. licheniformis MW3 has a fundamental role in germination triggered by L-alanine and casein hydrolysate. We also show that D-alanine is an important inhibitor in B. licheniformis amino acid-induced germination. Further, both wild type and the gerAA disruption mutant germinated effectively when exposed to appropriate levels of the non-nutrient germinant Ca 2+ -DPA which by-pass the spore receptor apparatus. However, effective germination with Ca 2+ -DPA seems both strain and species specific. In order to understand and potentially control the germination behaviour of B. licheniformis spores, disclosure of factors involved in the transition from a dormant spore to a metabolically active proliferating cell is of prime importance. Although complete elucidation of the function and cooperation of the different germinant receptors are rather laborious, this study has taken a step in the direction of obtaining more knowledge about this less studied species. The availability of both disruption and complementation mutants will facilitate further research on the function of the GerA receptor of B. licheniformis MW3, as well as its potential involvement in germination triggered by alternative nutrients and cooperation with other germinant receptors.    MQ. Following 15 min centrifugation at 4.080 × g, cells were resuspended in 0.5-1 ml of the PEG6000/MQ solution, aliquoted (100 µl) and stored at -80°C.

Construction of B. licheniformis MW3ΔgerAA::spc
The shuttle vector used for construction of a spectinomycin resistant (Sp R ) insertion deletion in the gerAA was pMAD_Sp R . This vector has been modified from pMAD [75] by insertion of a (Sp R )-cassette in the restriction site SalI [76]. As selective antibiotics for the presence of pMAD_Sp R or its derivative constructs, 100 µg/ml ampicillin and 100 µg/ml spectinomycin was used for E. coli TOP10 growth, and 3 µg/ml erythromycin and 250-300 µg/ml spectinomycin for B. licheniformis growth. This vector carries a constitutively expressed βgalactosidase gene, allowing blue-white screening on plates spread with X-Gal (40 µl 40 mg/ml 5-bromo-4chloro-3-indolyl-β-D-galactopyranoside, VWR, BDH Prolabo). This screening was, however, not always unambiguous following long incubations of plates with B. licheniformis MW3 transformants, probably due to the natural precence of β-galactosidase in B. licheniformis DSM 13 [77]. To construct the gene replacement vector, primers (Table 2) were designed to amplify two DNA fragments, one homologous to upstream (709 bp) and one to downstream (696 bp) regions of the deletion target (567 bp) in the gerAA. Platinum Taq DNA Polymerase High Fidelity kit (Invitrogen) was used for PCR amplification with the following amplification procedure: initial denaturation for 2 min at 94°C, 30 cycles of 30 s at 94°C, 30 s at 50°C and 1 min at 68°C, and final extension at 68°C for 10 min. Primers of the upstream and downstream amplicons contained restriction sites BamHI and EcoRI respectively (Table 2), allowing a two_step ligation into the corresponding restriction sites on either side of the (Sp R )-cassette in pMAD_Sp R . The resulting gene replacement plasmid, pMAD_Sp R ΔgerAA, was controlled for correct orientation of the upstream and downstream fragments by PCR. pMAD_Sp R ΔgerAA was introduced into B. licheniformis MW3 by electroporation, and allelic exchange of internal parts of gerAA (567 bp) with the (Sp R )-cassette of pMAD_Sp R ΔgerAA was allowed by double crossover. The protocol was performed as described by Arnaud et al. [75], except using growth temperatures of 37°C following initial transformation, an incubation temperature of 45°C and spectinomycin present during plasmid curing, and an incubation temperature of 37°C when screening for the double crossover phenotype (spectinomycin resistant and erythromycin sensitive colonies). Chromosomal DNA was purified from a candidate colony and used in PCR amplifications (as described above) with primers hybridizing outside the cloned DNA fragment and inside the spectinomycin cassette (Table 2) to verify the deletion and insertion by sequencing. The disruption mutant was named B. licheniformis MW3ΔgerAA::spc (NVH-1307) and used in the following complementation, sporulation and germination assays.

Construction of gerA complementation mutant
The shuttle vector used as base for trans complementation of ΔgerAA::spc was pHT315 [52]. As selective antibiotics for presence of pHT315 constructs, 100 µg/ml ampicillin was used for E. coli TOP10 growth, while 1 µg/ml erythromycin for B. licheniformis growth. Spectinomycin (250 µg/ml) was also supplemented for the chromosomal antibiotic resistance of NVH-1307 mutants carrying pHT315 derivatives (Table 1). To construct pHT315 complementation plasmids that harbour the gerA operon, DNA isolated from B. licheniformis MW3 was used. Primers, with SalI and XbaI restriction sites ( Table 2) were designed to amplify a 3982 bp fragment spanning from 151 bp upstream to 177 bp downstream gerA (gerAA, gerAB, gerAC). To ensure high fidelity, PCR amplification was performed with Phusion Hot Start II High-Fidelity DNA Polymerase kit (Finnzymes). The amplification protocol was as follows; initial denaturation for 30s at 98°C, 30 cycles of 10 s at 98°C, 30 s at 58°C and 2 min at 72°C, and final extension at 72°C for 10 min. The amplified fragments were cloned into the SalI/XbaI restriction site of pHT315, giving the complementation plasmid pHT315_MW3gerA. The purified plasmid was controlled by sequencing using primers hybridizing to pHT315 and internal gerA. The verified plasmid was introduced into the disruption mutant (NVH-1307) by electroporation as described earlier, giving the strain B. licheniformis MW3ΔgerAA:: spcpHT315_MW3gerA (NVH-1311). The strain was used in sporulation and germination assays. Filter sterilised Ca(NO 3 ) 2 ·4H 2 O, MnSO 4 ·4H 2 O and FeSO 4 ·7H 2 O were added to the medium after it had been autoclaved. pH was adjusted to 7.6 before autoclaving, and the pH of the final sporulation medium was 7.2. Sporulation medium of NVH-1311 was supplemented with 1 µg/ml erythromycin. The cultures were incubated with agitation (225 rpm) at 50°C for 1 to 2 days for B. licheniformis strains MW3, NVH-1307 and NVH-1311, or for 2 days at 30°C for B. subtilis B252 and B. cereus ATCC 14579 until ≥90% phase bright spores as judged by phase contrast microscopy. Spores were harvested by centrifugation for 10 min at 6000 × g at 4°C, and resuspended in 10 ml cold autoclaved MQ. Washing of spores was done by centrifugation and resuspension in MQ a total of ten times. The resulting spore crops, < 10% germinated spores, were stored refrigerated in MQ. When used in the following germination studies, spore crops were between 2 and 7 months old.

Germination assays
The spore suspension was routinely washed by centrifugation at 16.100 × g for 3 min prior to each experiment [78]. Spores were heat activated in MQ at 65°C for 20 min, chilled on ice, centrifuged (16.100 × g for 3 min) and resuspended in 2 × germination buffer (100 mM Kphosphate buffer pH 7.2) for L-alanine germination or 1 × germination buffer (   pHT315R 5´-GGAGAAAATACCGCATCAGG-3´Verification/sequencing of pHT315gerA_MW3 e a The restriction sites are underlined and may differ in sequence from the template genome of Veith et. al. [48]. b The primer is complement to the genome of Veith et. al. [48]. c The primer is complement to the Sp R cassette of pMAD_Sp R . d The gerA fragment amplified contains a sequence 151 bp upstream of gerAA, gerAA, gerAB, gerAC and 177 bp downstream of gerAC relative to the genome of Veith et. al. [48].
e The primer is complement to the pHT315 vector.
Germination was followed as described by Hornstra et al. [13] by monitoring the reduction in absorbance at A 600 as spores turn from phase-bright to phase dark at 30°C in a 96-well microplate in a plate reader (Tecan Intinite M200, Grödig, Austria). The spore suspension was adjusted to an initial A 600 of~2 (Shimadzu UV-VIS 160A, Shimadzu Europa GMBH) prior to addition of germinant. Germinant (filter sterilised L-alanine dissolved in MQ or casein hydrolysate dissolved in 50 mM Tris HCl pH 7.4 10 mM KCl) or negative control (MQ for L-alanine germination and 50 mM Tris HCl pH 7.4 10 mM KCl for casein hydrolysate germination) was automatically injected, and the plate was shaken for 10 s prior to the first reading. A 600 was recorded every 30 s for 142 to 170 min, with 10 s shaking in-between each measurement. The final concentration of germination buffer was 50 mM phosphatebuffer pH 7.2 or 50 mM Tris HCl pH 7.4 10 mM KCl, and final concentration of germinant was 100 mM L-alanine or 1% (w/v) casein hydrolysate. The final concentration of spores gave an initial A 600 of~0.7-0.8. To inhibit germination with Lalanine and potential other amino acids in the casein hydrolysate germination assay, 0.2% D-alanine (w/v, final concentration) was in some experiments added to each test well. The germination progress was described as the percentage of the initial A 600 (% A 600i ) for each measurement point [13]. All experiments were performed in duplicates on two individual spore batches and repeated at least twice. Germination was routinely controlled by phase-contrast microscopy (Olympus BX51, Hamburg, Germany) [13]. Spore germination in Ca 2+ -DPA was performed as follows; spores were washed in cold autoclaved MQ and resuspended in germination buffer (125-250 mM Tris base, 25-100 mM DPA (2,6-Pyridinedicarboxylic acid 99%, Sigma-Aldrich, Steinheim, Germany) pH~8) [79]. Germination was initiated by addition of excess CaCl 2 ·2H 2 O (Riedel de Häen AG, Seelze, Germany), followed by incubation for 3 h with shaking at room temperature (~20°C). The final concentrations of Ca 2 + -DPA in the assay were 20-80 mM DPA 30-100 mM CaCl 2 , and the final concentration of spores gave an initial A 600 of~0.6-0.8. Germination was described as an approximate percentage of phase dark spores after screening of microscopic slides by phase contrast microscopy (100 x). Experiments were performed in duplicate on two individual spore batches and repeated at least twice.

DNA sequencing and bioinformatics
DNA sequencing was performed by GATC Biotech (Konstanz, Germany) or Source BioScience (Nottingham, United Kingdom). The genomic sequence of B.