Salt-sensitivity of σ^H and Spo0A prevents sporulation of Bacillus subtilis at high osmolarity avoiding death during cellular differentiation

Summary

The spore-forming bacterium Bacillus subtilis frequently experiences high osmolarity as a result of desiccation in the soil. The formation of a highly desiccation-resistant endospore might serve as a logical osmostress escape route when vegetative growth is no longer possible. However, sporulation efficiency drastically decreases concomitant with an increase in the external salinity. Fluorescence microscopy of sporulation-specific promoter fusions to gfp revealed that high salinity blocks entry into the sporulation pathway at a very early stage. Specifically, we show that both Spo0A- and SigH-dependent transcription are impaired. Furthermore, we demonstrate that the association of SigH with core RNA polymerase is reduced under these conditions. Suppressors that modestly increase sporulation efficiency at high salinity map to the coding region of sigH and in the regulatory region of kinA, encoding one the sensor kinases that activates Spo0A. These findings led us to discover that B. subtilis cells that overproduce KinA can bypass the salt-imposed block in sporulation. Importantly, these cells are impaired in the morphological process of engulfment and late forespore gene expression and frequently undergo lysis. Altogether our data indicate that B. subtilis blocks entry into sporulation in high-salinity environments preventing commitment to a developmental program that it cannot complete.

Introduction

The soil is a challenging habitat for microorganisms. Both, the supply of nutrients and physical parameters like pH, temperature, and osmolarity fluctuate frequently. Bacillus subtilis is a soil-dwelling bacterium (Mandic-Mulec et al., 2015) that is well adapted to life in the upper layers of the soil (Belda et al., 2013; Earl et al., 2008). It counteracts various challenges through the induction of the σ^B-controlled general stress regulon (Hecker et al., 2007; Price, 2011) and the engagement of stress-specific adaptation mechanisms (Sonenshein et al., 2002). Furthermore, in response to nutrient limitation, B. subtilis can differentiate into a dormant and stress-resistant endospore (Errington, 2003; Higgins and Dworkin, 2012; Setlow, 2014). Here, we have investigated how B. subtilis simultaneously responds to two common challenges that it encounters in its native habitat: starvation and high salinity (Bremer, 2002; Sonenshein, 2000; Higgins and Dworkin, 2012; Bremer and Krämer, 2000).

In response to starvation, B. subtilis initiates a highly orchestrated developmental program involving hundreds of genes (Eichenberger et al., 2004; Errington, 2003; Higgins and Dworkin, 2012; Steil et al., 2005; Nicolas et al., 2012). During this process, the cell undergoes an asymmetric division resulting in two cell types, the mother cell and forespore, which follow distinct developmental programs of gene expression driven by stage- and compartment-specific transcription factors. These programs are linked to each other through cell-cell signaling pathways, ensuring that gene expression in one cell is kept in register with gene expression in the other. Upon maturation of the developing forespore, lysis of the mother cell releases a highly stress-resistant endospore into the environment (Errington, 2003; Higgins and Dworkin, 2012; Stragier and Losick, 1996). B. subtilis spores can remain dormant in a given ecosystem for extended periods of time, yet can rapidly germinate in response to specific germinants that signal conditions are conducive for vegetative growth (Setlow, 2014; Sinai et al., 2015; Sturm and Dworkin, 2015).

Two transcription factors govern the entry into the sporulation pathway: the response regulator Spo0A and the alternative sigma factor σ^H. These regulators control the expression of hundreds of genes involved in stationary phase adaptation in addition to those involved in the earliest stages of spore formation. Both transcription factors are under complex and interconnected regulatory control (Hoch, 1993; Grossman, 1995; Phillips and Strauch, 2002; Sonenshein, 2000; Chastanet et al., 2010). A set of sensor kinases (KinA – KinE) trigger activation of Spo0A, the master regulator of sporulation, via a multicomponent phosphorelay (Burbulys et al., 1991; Stragier and Losick, 1996). In most cases, the specific signals sensed by these kinases are incompletely understood. Phosphorylated Spo0A (Spo0A~P) directly and indirectly regulates the transcription of over 400 genes (Molle et al., 2003). Depending on the flux through the phosphorelay, different levels of Spo0A~P are generated allowing a graded adaptive response since Spo0A-controled promoters have different affinities for Spo0A~P (Chung et al., 1994; Fujita et al., 2005). The sporulation genes under Spo0A control have low-affinity Spo0A binding sites and therefore require high concentrations of Spo0A~P for their effective expression (Fujita et al., 2005). One of the genes regulated indirectly by Spo0A~P is sigH (also referred to as spo0H), which encodes the alternative sigma factor σ^H (Errington, 2003; Haldenwang, 1995; Higgins and Dworkin, 2012; Stragier and Losick, 1996). The globally acting repressor AbrB negatively regulates transcription of sigH, and Spo0A, in turn, represses abrB transcription. It does so at low Spo0A~P concentrations thereby resulting in an early increase in σ^H levels (Fujita et al., 2005; Strauch et al., 1990). The elevated cellular level of σ^H then promotes expression of sporulation genes, including spo0A, spo0F (encoding one of the phosphorelay components), and kinA, the gene for the central sporulation-specific sensor kinase KinA. In addition, Spo0A~P activates transcription of spo0A and spo0F genes (Britton et al., 2002; Chastanet and Losick, 2011). These positive feedback loops, in combination with finely-tuned flux through the phosphorelay, increase Spo0A and σ^H levels and activity, resulting in the expression of early sporulation genes and entry into the sporulation pathway (Britton et al., 2002; Chastanet and Losick, 2011; Predich et al., 1992).

In its soil habitat (Earl et al., 2008; Mandic-Mulec et al., 2015), B. subtilis not only encounters starvation (Sonenshein, 2000; Errington, 2003), but desiccation also subjects it to high salinity (osmolarity) (Bremer, 2002; Bremer and Krämer, 2000). The cell responds to such osmotic challenges through increased potassium import (Whatmore et al., 1990; Holtmann et al., 2003), the synthesis of the compatible solutes proline and glycine betaine (Whatmore et al., 1990; Brill et al., 2011; Nau-Wagner et al., 2012), and the uptake of various osmoprotectants (Hoffmann et al., 2013; Bashir et al., 2014; Broy et al., 2015; Zaprasis et al., 2013). These cellular adjustments prevent efflux of water, maintain turgor at physiologically adequate levels, and optimize the solvent properties of the cytoplasm for vital biochemical reactions (Bremer, 2002; Bremer and Krämer, 2000; Wood, 2011). As a result, growth of B. subtilis can continue under otherwise osmotically unfavorable circumstances (Boch et al., 1994).

The pronounced desiccation-resistant properties of the B. subtilis endospores (Setlow, 2014) suggest that spore formation would be an effective escape route from adverse high-salinity conditions (Bremer, 2002). However, contrary to expectations, high salinity efficiently blocks sporulation (Kunst and Rapoport, 1995; Ruzal et al., 1998; Ruzal and Sanchez-Rivas, 1998). Work from Sanchez-Rivas and coworkers indicates that this block is imposed relatively early in the sporulation process (Ruzal et al., 1998). However, the specific stage at which this occurs and the underlying molecular mechanism(s) have remained elusive.

Here, we have addressed these ecologically relevant questions and show that B. subtilis cells that are continuously exposed to high-salt environments are blocked at the earliest possible stage of sporulation. They fail to activate genes under the control of σ^H and Spo0A~P. Synthetic over-production of KinA resulted in a nearly complete bypass of the block in sporulation initiation in salt-stressed cells. However, these cells were impaired in engulfment, late forespore gene expression, and exhibited a high frequency of lysis. Collectively, our data suggest that the molecular block averting entry into sporulation under salt stress is not an accident. Rather, it appears to serve a physiologically important role by preventing osmotically stressed cells from committing to a developmental program they cannot complete. This trait is conserved among domesticated and non-domesticated strains of B. subtilis.

Results

High salinity drastically impairs spore formation

During the course of our studies on how B. subtilis copes with osmotic stress (Bremer, 2002; Bremer and Krämer, 2000), we found that cells grown for more than 36 hours in Difco Sporulation Medium (DSM) containing 1.2 M NaCl failed to produce phase bright spores (Fig. 1A). This was in stark contrast to the same culture conditions in medium lacking salt (0 M NaCl), in which almost all of the cells formed spores (Fig. 1A). To quantitatively assess the negative effect of high salinity on sporulation, we grew cells of the B. subtilis wild-type strain JH642 in DSM in the presence of various concentrations of NaCl and determined the frequency of heat-resistant spores using a standard heat-kill assay. In this assay, the starved cultures are incubated at 80° C for 20 min to kill vegetative and sporulation deficient cells and the number of heat-resistant colony-forming units (CFUs) that result from the outgrowth of the surviving spores are compared to total CFUs (and separately to heat-resistant CFUs in the absence of salt). Increasing salinity reduces growth rate (Boch et al., 1994) (Fig. S1) but the impact was less than 2-fold at the highest concentration of NaCl (1.2 M) used here. Cells grown in DSM in the absence of salt had a sporulation efficiency of ~90%. A systematic increase in the salt concentration (0.2 M, 0.4 M, 0.6 M NaCl) decreased this value to approximately 70%, 40% and 1%, respectively (Fig. 1B). An additional increase in the salinity led to an even further reduction in sporulation efficiency. Only about one in a million cells grown in the presence of 1.2 M NaCl formed a heat-resistant spore (Fig. 1B). For the remainder of this study we focused our analysis on media containing either 0.6 M or 1.2 M NaCl.

Fig. 1.

Sporulation is inhibited by high salinity. (A) Representative phase-contrast images of cells of wild-type B. subtilis cells (JH642) sporulated by nutrient exhaustion for 36 hours in the presence of the indicated concentrations of NaCl. The scale bar indicates 2 μm. (B) Quantitative assessment of increasing salt concentration on the formation of heat-resistant spores. Sporulation efficiency (spores %) indicates that number of heat-resistant CFU in the indicated concentration of NaCl compared to heat-resistant CFU in the absence of NaCl. (C) Bar graph showing the negative effects of 0.6 M and 1.2 M NaCl on sporulation of the indicated domesticated and non-domesticated B. subtilis strains and of B. mojavensis.

To investigate whether the inhibition of sporulation by high salinity was specific to the wild-type strain JH642 (Smith et al., 2014), we examined the sporulation efficiency of both domesticated and undomesticated B. subtilis isolates. Although sporulation efficiency in the presence of salt was somewhat variable, all strains tested displayed salt sensitivity (Fig. 1C). We also investigated the sporulation efficiency of Bacillus mojavensis, a species that is indigenous to the Mojave Desert (Earl et al., 2012). We anticipated that this Bacillus species would have an increased tolerance to high salinity as a result of its adaptation to the drought cycles that it experiences in its desert habitat. However, sporulation of B. mojavensis was even more sensitive to high salinity than B. subtilis, as the presence of 1.2 M NaCl in the DS medium abolished sporulation entirely (Fig. 1C). Altogether, we conclude that the sporulation process is highly sensitive to even modest increases in the salinity of the environment and that this is a common trait of both domesticated and undomesticated B. subtilis strains.

High salinity blocks sporulation at a very early stage

Previous studies using electron microscopy to assess the stage at which sporulation is blocked in high salinity suggested that differentiation is inhibited after polar division (Ruzal et al., 1998). To examine a larger number of cells and multiple time points, we used fluorescence microscopy and visualized the morphological stages of sporulation with the membrane dye TMA-DPH. In the same experiment, we monitored the activity of the first forespore-specific transcription factor σ^F (Higgins and Dworkin, 2012) using a P_spoIIQ-gfp promoter fusion (Londono-Vallejo et al., 1997). We imaged the cells during a sporulation time-course at 30 min intervals after the onset of sporulation (T0) in cultures grown in the presence or absence of NaCl (Fig. 2A and S2). As reported previously, in the absence of salt >60% of the cells had a polar septum by hour 2 (T2) and most had initiated the process of engulfment (Rodrigues et al., 2013). Furthermore, virtually all the sporulating cells exhibited σ^F-dependent GFP fluorescence in the forespore compartment (Fig. 2A). In the presence of 0.6 M NaCl, 21% (n > 700) of the cells had a polar septum and displayed expression of the σ^F reporter. However, in media with 1.2 M NaCl, virtually no cells exhibited asymmetric septation or GFP fluorescence even after 3 hours of sporulation (Fig. 2A and S2). Thus, under our assay conditions, high salinity inhibits sporulation prior to polar division and before the activation of the first compartment-specific transcription factor, σ^F.

Fig. 2.

High salinity inhibits sporulation prior to asymmetric cell division. (A) Representative images of cells (strain BDR1048; Table S1) harboring a transcriptional reporter (P_spoIIQ-gfp) for the first forespore-specific transcription factor σ^F induced to sporulate in the indicated concentrations of NaCl. Membranes (false-colored in red) were stained with TMA-DPH and σ^F-dependent expression is shown in green. Hour 1 (T1), 1.5 (T1.5), and 2 (T2) are shown. Additional time points can be found in Figure S2. Scale bar is 2 μm. (B) Diagram illustrating part of the complex genetic circuitry involved in the activation of the transcriptional regulators σ^H and Spo0A for sporulation initiation.

Spo0A and σ^H activity are impaired in the presence of high salt

Based on our cytological analysis, we sought to investigate the effects of high salinity on steps in the sporulation process that occur prior to the activation of σ^F. Accordingly, we examined the activity of the transcription factors Spo0A and σ^H that together govern entry into sporulation (Fig. 2B). First, we monitored the expression of a Spo0A-responsive reporter (P_spoIIE-gfp). The P_spoIIE promoter is recognized by σ^A but also requires active Spo0A (Spo0A~P) for its expression (Molle et al., 2003; York et al., 1992). In the absence of salt, Spo0A activity could be detected with this reporter during the first hour (T1) of sporulation and it continued to rise during asymmetric cell division and engulfment (Fig. 3A and S3). In media containing 0.6 M NaCl, a subset of cells displayed Spo0A-dependent transcription that was similar to the activity observed in the absence of salt. Furthermore, these “Spo0A high” cells frequently went on to form polar septa (Fig. 3A). However, the majority of cells grown in the presence of 0.6 M NaCl exhibited low Spo0A-dependent transcription (Fig. 3A and S3). In the presence of 1.2 M NaCl, all cells in the field had virtually undetectable Spo0A activity (Fig. 3A and S3) consistent with the absence of polar septa (Fig. S2B and S4). Thus, salt-stressed B. subtilis cells are unable to accumulate sufficient levels of phosphorylated Spo0A to promote the initiation of sporulation. Importantly, high salt did not impair GFP fluorescence as a σ^A-responsive promoter (P_veg) fused to gfp had strong fluorescence in the presence and absence of NaCl during sporulation (Fig. S5). This result also indicates that high salinity does not globally impair transcription in sporulation medium.

Fig. 3.

Spo0A and σ^H activity are impaired in B. subtilis cells induced to sporulate at high salinity. (A) Representative images of cells (strain BDR2128; Table S1) harboring a transcriptional reporter (P_spoIIE-gfp) monitoring Spo0A-responsive activity that were induced to sporulate in the indicated concentrations of NaCl. For each condition and time point (in hours), GFP fluorescence is shown in black/white and a merged image of membranes (red) and GFP (green) is shown below. The relevant transcription factors that act on the P_spoIIE promoter are indicated. (B) Representative images of cells (strain BDR3080; Table S1) harboring a fusion of the spo0A promoter to cfp (P_spo0A-cfp) in the indicated concentrations of NaCl at hour 1 (T1) of sporulation. (C) Representative images of cells (strain BDR3064; Table S1) harboring a transcriptional reporter (P_spoVG-cfp) for σ^H-dependent activity that were induced to sporulate in the indicated concentrations of NaCl. (D) Representative images of cells (strain BDR3090; Table S1) harboring a fusion of the sigH promoter to cfp (P_sigH-cfp) at hour 1 of sporulation in the presence of the indicated concentrations of NaCl. For each reporter, all images were scaled identically. Scale bar indicates 2 μm.

To investigate whether transcription of the spo0A gene itself was reduced in high salinity, we examined the activity of the spo0A promoter using a P_spo0A-cfp reporter (Fig. 3B). spo0A transcription was modestly reduced in the presence 0.6 M NaCl and more severely at 1.2 M. The spo0A gene contains two promoters that are inversely controlled by Spo0A. The first promoter (P_v) is recognized by σ^A and negatively regulated by Spo0A~P while the second promoter (P_s) is recognized by σ^H and activated by Spo0A~P (Chastanet and Losick, 2011; Eymann et al., 2001; Predich et al., 1992; Strauch et al., 1992). Thus, the reduction in spo0A expression could be due to the salt sensitivity of Spo0A or σ^H.

To investigate whether σ^H activity was influenced by high salt, we analyzed a σ^H-dependent promoter fusion (P_spoVG-cfp) (Segall and Losick, 1977; Zuber and Losick, 1987). The P_spoVG promoter is negatively regulated by AbrB, which is itself repressed by Spo0A at the transcriptional level. However, the extent of AbrB-mediated repression is relatively modest compared to the absolute requirement for σ^H (Liu and Zuber, 2000; Zuber and Losick, 1987) (Fig. S6). Accordingly, P_spoVG-cfp serves as a reliable reporter for σ^H activity. In cells grown in the absence of salt, σ^H-dependent gene expression was detectable at hour 1 (T1) of sporulation and increased significantly over the next half hour leading up to the formation of asymmetric septa (Fig. 3C). In the presence of 0.6 M and 1.2 M NaCl, the level of σ^H activity was slightly reduced at hour 0.5 and failed to increase over the next half hour (Fig. 3C). At hour 1.5, σ^H activity in cells grown in the presence of NaCl was significantly lower than the pre- and post-divisional cells at hour 1.5 that were sporulated in the absence of salt. Similar results were obtained using the P_spoIIA promoter fused to GFP (Fig. S7). However, since this promoter is under the control of both σ^H and Spo0A, the impact of high salinity was even more pronounced. Finally, transcription of the gene (sigH) encoding the σ^H sigma factor, which is indirectly controlled by Spo0A, was modestly reduced in the presence 0.6 M NaCl and more severely at 1.2 M (Fig. 3D). Collectively, these data suggest that σ^H and Spo0A activity and/or levels are reduced in the presence of high salinity resulting in a failure of B. subtilis cells to efficiently enter the sporulation pathway.

Spo0A and σ^H levels are differentially affected by high salinity

To directly assess the levels Spo0A, σ^H and σ^A we examined all three proteins by immunoblot analysis in a sporulation time course. Wild-type cells were induced to sporulate in the presence and absence of NaCl and samples were taken every half hour (Fig. 4). The levels of σ^A remained unchanged in all conditions throughout the entire sporulation time course. As previously reported (Fujita and Sadaie, 1998b; Fujita and Sadaie, 1998c; Liu and Zuber, 2000), in the absence of NaCl, σ^H levels began to increase within 30 minutes after the initiation of sporulation and achieved maximum levels 30–60 minutes later (Fig. 4). By hour 2, σ^H had returned to pre-sporulation levels. In the absence of salt, Spo0A levels closely followed the increase in σ^H, consistent with the requirement of σ^H for spo0A expression during sporulation (Predich et al., 1992; Strauch et al., 1992). However, unlike σ^H, Spo0A protein levels remained high for the 3-hour time course. This is in line with the observation that Spo0A remains active in the mother cell at later times during the sporulation process (Fujita and Losick, 2003; Fujita and Losick, 2005) (Fig. S3). In the presence of 0.6 M NaCl, σ^H levels increased in a manner similar to the no-salt condition (Fig. 4). This result and our analysis of σ^H activity using P_spoVG-cfp (Fig. 3C) suggest that σ^H activity is sensitive to high salinity. Interestingly, σ^H levels remained high for the rest of the time course in the presence of 0.6 M NaCl. The nature of this stabilization is currently unknown. However, comparing σ^H activity (assessed by P_spoVG-cfp fluorescence) and σ^H protein level in the presence and absence of 0.6 M NaCl at hour 2 reinforces the conclusion that σ^H activity is impaired in high salinity (Fig. S8). In contrast to σ^H, Spo0A protein levels increased more slowly in cells grown in the presence of 0.6 M NaCl, and never achieved as high levels as the no salt growth condition (Fig. 4). This is consistent with the salt-sensitivity of σ^H and the reduced activity Spo0A (Fig. 3A and S3). A comparison of spo0A transcription (Fig. 3B) and Spo0A accumulation (Fig. 4) suggests that Spo0A stability is also influenced by high salinity. Finally, both σ^H and Spo0A levels were significantly reduced and failed to accumulate in medium containing 1.2 M NaCl. Collectively, our data indicate that σ^H activity and Spo0A accumulation and activity are sensitive to high salinity and prevent cells from committing to the sporulation pathway.

Fig. 4.

Spo0A and σ^H protein levels are affected by high salinity. Immunoblot analysis assessing the levels of Spo0A, σ^H, and σ^A in cells of the wild-type strain PY79 that were induced to sporulate in the presence of the indicated concentrations of NaCl. Forespore-specific σ^F activity was assessed by the accumulation of SpoIIQ. A sigH null mutant (strain BDR3057; Table S1) was included to show the requirement for σ^H activity on the accumulation of Spo0A.

The interaction between σ^H and core RNA polymerase is impaired in high salinity

We investigated whether the salt-sensitivity of σ^H in vivo was related to its inability to stably associate with core RNA-polymerase. To do this we took advantage of a strain harboring a functional hexa-histidine fusion to the beta prime subunit of RNA-polymerase (Fujita and Sadaie, 1998b; Fujita and Sadaie, 1998c). Cells were sporulated by resuspension in the presence and absence of 0.6 M NaCl. At hour 1.5 the cells were harvested by centrifugation and lysed in the presence of 300 mM NaCl followed by Ni²⁺-NTA affinity purification of RNA-polymerase (see Materials and Methods). The amount of co-purified σ^H was then determined by immunoblot analysis (Fig. 5). In support of the idea that σ^H is impaired in its ability to associate with core RNA-polymerase in cells grown at high salinity, the amount of co-purified σ^H was reduced in cells sporulated in medium containing 0.6 M NaCl compared to cells sporulated in the absence of salt. Importantly, the amount of core RNA-polymerase purified from the two extracts was similar and the level of σ^H in the lysate from the salt-sporulated cells was higher than from the no-salt cells (Fig. 5). We conclude from this set of experiments that the association of σ^H with core RNA polymerase is impaired during sporulation under high salinity conditions.

Fig. 5.

The association of σ^H with core-RNA-polymerase is affected by high-salinity. Co-purification of σ^H with core RNA polymerase from cells induced to sporulate in the indicated concentrations of NaCl. Top panel: Ni²⁺-chelate affinity purification of β and β′-his6 (RNAP-His6) subunits from sporulating cells at hour 1.5. Whole-cell lysate [C], crude extract [Ex], load [L], flow-through [FT], and eluate [E] from the wild-type (WT) strain PY79 and from strain BDR485 (Table S1) (RNAP-His6) were separated by SDS-polyacrylamide gel electrophoresis and stained with Coomassie brilliant blue. Bottom panel: Immunoblot analysis of the same fractions used in the top panel using anti-core and anti-σ^H antibodies. Arrows highlight the relevant elution fractions for comparison.

Enrichment for suppressors with increased sporulation at high salinity identifies mutations in kinA and sigH

On the basis of the results described above, we wondered whether it might be possible to isolate B. subtilis mutants that bypass the salt-imposed block in sporulation resulting in increased sporulation efficiency at high salinity. To this end, we used a genetic suppressor enrichment strategy. The wild-type B. subtilis strain JH642 was sporulated by nutrient exhaustion in DSM containing 1.2 M NaCl. After eliminating vegetative and sporulation-impaired cells by heat-kill (80°C for 20 min), serial dilutions were plated on LB-agar plates. The few heat-resistant colonies that emerged were then re-inoculated into fresh DSM containing 1.2 M NaCl and the cycle was repeated. Mutants showing an increased sporulation phenotype in comparison to the JH642 wild type were isolated. Although the sporulation efficiencies of the eleven independently isolated suppressor strains were only moderately (3- to 8-fold) increased (Tables 1 and 2), we mapped the mutations in two of them by whole genome re-sequencing. Strikingly, in each of these two suppressors only a single nucleotide polymorphism was observed in comparison to the JH642 reference genome (Smith et al., 2014). In strain NWB6, we identified a point mutation in the promoter region of kinA (Table 1), and in strain NWB7 a point mutation in the coding region of the sigH gene was present that resulted in an amino acid substitution (E146A) in the σ^H protein (Table 2). With this information in hand, we PCR-amplified and sequenced the kinA, sigH, and spo0A genes from the remaining nine suppressor strains. This analysis identified eight additional mutant alleles of the kinA promoter (Table 1) and one further allele of the sigH coding region (Table 2). None of the 11 suppressor strains harbored a mutation in spo0A. We also sequenced kinA and sigH (and their promoters) from B. subtilis 168, PY79, 3610, and DK1042 to investigate whether the strain-to-strain variation in sporulation efficiency at high salinity (Fig. 1C) could be explained by polymorphisms in these loci. Both genes and their promoter regions were identical to those in JH642, indicating that other differences among these strains account for the observed variability.

Table 1.

Suppressor strains harboring mutations in the kinA promoter region.

Strain	kinA promoter region sequence	Spores (%)	Fold increase
WT	TAGAAGGAGAATACTCATTTTCTAGCGAATCATACTAGGTAAAAGTCAATCTGTATATGTCGAAA - 14 nt – AAAGGAGGGATTCTGTG …	0.00016	-
NWB6	TAGAAGGAGAATACTCATTTTCTAGCGAATCATTCTAGGTAAAAGTCAATCTGTATATGTCGAAA - 14 nt – AAAGGAGGGATTCTGTG …	0.00086	5
NWB11	TAGAAGGAGATTACTCATTTTCTAGCGAATCATACTAGGTAAAAGTCAATCTGTATATGTCGAAA - 14 nt - AAAGGAGGGATTCTGTG …	0.00132	8
NWB13	TAGAAGGAGAAAACTCATTTTCTAGCGAATCATACTAGGTAAAAGTCAATCTGTATATGTCGAAA - 14 nt - AAAGGAGGGATTCTGTG …	0.00104	7
NWB16	TAGAAGGAGATTACTCATTTTCTAGCGAATCATACTAGGTAAAAGTCAATCTGTATATGTCGAAA - 14 nt - AAAGGAGGGATTCTGTG …	0.00129	8
NWB17	TAGAAGGAGAATACTCATTTTCTA– CGAATCATACTAGGTAAAAGTCAATCTGTATATGTCGAAA - 14 nt - AAAGGAGGGATTCTGTG …	0.00078	5
NWB19	TAGAAGGAGAATACTCATTTTCTAGCGAATCATACTAGGTAAAAGTCAATCTGTATTTGTCGAAA - 14 nt - AAAGGAGGGATTCTGTG …	0.00097	6
NWB22	TAGAAGGAGAATACTCATTTTCTAGCGAATCAAACTAGGTAAAAGTCAATCTGTATATGTCGAAA - 14 nt - AAAGGAGGGATTCTGTG …	0.00065	4
NWB24	TAGAAGGAGAATACTCATTTTCT– GCGAATCATACTAGGTAAAAGTCAATCTGTATATGTCGAAA - 14 nt - AAAGGAGGGATTCTGTG …	0.00081	5
NWB25	TAGAAGGA TAATACTCATTTTCTAGCGAATCATACTAGGTAAAAGTCAATCTGTATATGTCGAAA - 14 nt - AAAGGAGGGATTCTGTG …	0.00113	7

Suppressor strains that partially overcome the strong sporulation defect exhibited by the B. subtilis wild-type strain JH642 carry mutations in the promoter region of kinA. Mutations are highlighted in red. Predicted σ^H- (grey) and Spo0A- (black) recognition sites are indicated. Bold letters represent the start codon (GTG) of the kinA coding sequence and the ribosome-binding site is underlined. The spore titers and fold increases in sporulation efficiency of the suppressor strains are indicated.

Table 2.

Suppressor strains harboring mutations in the coding region of sigH.

Strain	Mutation	Spores (%)	Fold increase
WT	-	0.00016	-
NWB7	E146A [GAA→GCA]	0.00055	3
NWB21	I149F [ATT→TTT]	0.00067	4

Suppressor strains that partially overcome the strong sporulation defect exhibited by the B. subtilis wild-type strain JH642 carrying mutations in the coding region of sigH, the structural gene for the σ^H sporulation-specific sigma factor. The two mutations resulting in single amino acid substitutions in σ^H are indicated. The spore titers and fold increases in sporulation efficiency of the suppressor strains are indicated.

Overexpression of kinA upon entry into sporulation bypasses the early salt-stress imposed block

The identification of suppressor mutations in the promoter of kinA that modestly increased sporulation under high salinity (1.2 M NaCl) conditions, prompted us to investigate whether higher levels of KinA could overcome the salt-sensitive block. σ^H is required for the expression of both spo0A and kinA (see Fig. 2B) and thus it seemed possible that the overexpression of kinA could trigger enhanced sporulation in hypertonic medium. Accordingly, we used a strain in which kinA is expressed under the control of a strong IPTG-inducible promoter (P_hyperspank). Previous work from Fujita and co-workers has shown that increased expression of kinA using this expression system can bypass the need for starvation signals and induce sporulation during growth in rich medium (Eswaramoorthy et al., 2010; Eswaramoorthy et al., 2009; Fujita and Losick, 2005). We induced sporulation in the presence of 0, 0.6 and 1.2 M NaCl and monitored Spo0A activity using the P_spoIIE-gfp reporter and followed sporulation with the fluorescent membrane dye TMA-DPH (Fig. 6A and S9A). Overexpression of kinA in all three conditions resulted in similar levels of KinA protein (Fig. S9B). Strikingly, KinA overproduction at 0.6 M NaCl resulted in high Spo0A activity and a statistically significant increase (>79%, n>430) in the number of cells with polar divisions and engulfing forespores at hour 2 (T2) (Fig. 6A and S9). Even in the presence of 1.2 M NaCl, asymmetric septa and engulfing forespores were readily detectable, albeit with some delay compared to 0.6 M NaCl and the no-salt condition (Fig. S9). Thus, these results indicate that increasing the level of phosphorylated Spo0A can bypass the salt-sensitivity of σ^H and any other possible early blocks to sporulation.

Fig. 6.

Overexpression of kinA bypasses the early salt-sensitive block in sporulation. (A) Representative images of cells harboring a Spo0A activity reporter (P_spoIIE-gfp) induced to sporulate in the presence of 0.6 M NaCl. Wild-type (WT; strain BDR2128) (Table S1) and a strain (BDR3087; Table S1) harboring an IPTG-inducible allele of kinA were imaged at the indicated times after the initiation of sporulation. A 5-hour sporulation time course of cells overexpressing kinA (BDR3087) comparing 0 M, 0.6 M and 1.2 M NaCl can be found in Figure S9. Scale bar indicates 2 μm. (B) Bar graphs of sporulation efficiency in the wild-type (WT) and the kinA overexpression strain induced to sporulate in the presence of the indicated concentration of NaCl.

Importantly, over-expressing kinA from the P_hyperspank promoter significantly increased the number of heat-resistant spores in the presence of 0.6 M and 1.2 M NaCl (Fig. 6B). However, despite the near complete bypass of the early morphological events in 0.6 M NaCl (Fig. 6A and S9), the sporulation efficiency was still considerably reduced (by ~100-fold) compared to the no salt condition in which about 90% of the cells formed heat resistant spores (Fig. 6B). These data suggest that high salinity also impairs additional, later steps in the sporulation pathway.

High salinity impairs engulfment, σ^G activity and causes lysis

To define additional steps in the sporulation program that are sensitive to high salinity, we took advantage of a strain that harbors fluorescent reporters for all of the sporulation-specific sigma factors (Meeske et al., 2015). In this strain, the reporter for the first compartment-specific transcription factor σ^F was a promoter (P_spoIIQ) fused to yfp (P_spoIIQ-yfp) (Londono-Vallejo et al., 1997). The P_spoIID promoter that is recognized by the first mother-cell-specific transcription factor σ^E (Rong et al., 1986) was fused to mCherry (P_spoIID-mCherry), and the reporter for the late forespore transcription factor σ^G was a promoter (P_sspB) fused to cfp (Nicholson et al., 1989). Finally, the P_gerE promoter, recognized by the late mother-cell transcription factor σ^K (Cutting et al., 1989), was fused to yfp (P_gerE-yfp). σ^F-dependent YFP fluorescence is first observed in the forespore at hour 2. By the time σ^K is activated in the mother cell (hour 4), σ^F activity is largely absent and YFP fluorescence is significantly reduced. Accordingly, we could use this fluorescent protein twice.

We introduced the P_hyperspank-kinA construct into this multi-reporter strain and induced sporulation in the presence of 0.6 M NaCl and IPTG. Over-expression of kinA resulted in activation of σ^F in the forespore and σ^E in the mother cell in almost all the sporulating cells (Fig. 7A and S10A). Interestingly, many cells exhibited defects in the morphological process of engulfment: the migration of the mother-cell membranes around the forespore was uncoordinated and the engulfed forespores were smaller and contained membrane aberrations (Fig. 6A, 7A and B, S9A and S10A). Furthermore, although many sporulating cells displayed forespore-specific σ^G activity after the completion of engulfment, a subset of cells appeared to have reduced or undetectable levels (Fig. 7A and B and S10A). These phenotypes were also apparent in the 0.6 M NaCl condition where kinA was not overexpressed (Fig. 7A and B). Phase contrast images at later time points revealed a second and more striking phenotype: a reduction in phase grey spores and a concomitant increase in lysed cells (Fig. 7C and S10). Altogether, our results indicate that membrane remodeling and σ^G activity in the forespore display salt-sensitivity but the principal defect in spore formation in the presence of high salt is lysis of the sporulating cell prior to the maturation of the dormant spore.

Fig. 7.

High salinity impairs engulfment and late forespore gene expression and causes cell lysis. (A) Representative images of cells harboring transcriptional reporters for all four stage- and compartment-specific sigma factors induced to sporulate in the presence of the indicated concentrations of NaCl. The membranes and reporters for σ^F (P_spoIIQ-yfp), σ^E (P_spoIID-mCherry), σ^G (P_sspB-cfp) and σ^K (P_gerE-yfp) are shown in black/white or as a merged image with membranes (red) and σ^G activity (green). Wild-type (WT) sporulating cells (strain BCR1071; Table S1) and a strain overexpressing kinA (BCR1274; Table S1) were imaged at hour 4 of sporulation. Additional time points can be found in Figure S10A. (B) Representative images highlighting impaired engulfment and the defect in σ^G activation in a subset of sporulating cells grown in the presence of 0.6 M NaCl. Forespores with reduced σ^G activity are indicated (yellow arrow heads). Images show membranes (top) and an overlay of the σ^G activity reporter (green) and the σ^E activity reporter (red) (C) Representative phase contrast images from hour 5 of sporulation highlighting the increased frequency of cell lysis and reduction in phase-grey and phase-bright spores when cells bypass the early salt-sensitive block to sporulation. Additional time points can be found in Figure S10B. Lysed cells are highlighted by red arrow heads. Scale bar indicates 2 μm.

Discussion

Building on previous observations (Kunst and Rapoport, 1995; Ruzal et al., 1998), we report here that spore formation in B. subtilis is highly sensitive to even modest increases in the external salinity, an environmental challenge that leads to an almost exponential drop in sporulation efficiency once a threshold of 0.6 M NaCl is achieved (Fig. 1B). Interestingly, at this level of external salinity, growth of B. subtilis begins to be impaired (Boch et al., 1994) and the cellular osmostress protection mechanisms of vegetative cells (e.g. the import of osmostress protectants and the synthesis of the compatible solutes) are already active (Hoffmann et al., 2013; Brill et al., 2011; Nau-Wagner et al., 2012). Under extreme saline conditions (1.2 M NaCl) where growth of B. subtilis is more significantly retarded (Boch et al., 1994), only one out of a million vegetative cells can form a mature, heat-resistant endospore (Fig. 1B). The goal of our study was to define the molecular basis for the salt-stress-imposed block in sporulation. Data reported by Sánchez-Rivas and co-workers (Ruzal et al., 1998; Ruzal and Sanchez-Rivas, 1998) had already indicated that this happened early in the sporulation process but the underlying molecular mechanism(s) and the precise stage at which the salt-stress-imposed block occurred remained ill defined. Our data clarify these issues.

Our study shows that entry into sporulation by the B. subtilis cell is already impaired prior to polar division and the activation of the forespore-specific transcription factor σ^F (Fig. 2A, S2 and S4). Since cells that have formed the asymmetric septum and activated σ^F are fully committed to the sporulation process (Errington, 2003; Higgins and Dworkin, 2012; Dworkin and Losick, 2005), the high-salinity-imposed block occurs prior to the “point of no return”. Consistent with these findings, we further show that gene expression under the control of the master regulator of sporulation, Spo0A, and of the earliest acting sporulation-specific sigma factor σ^H, was strongly affected by high salinity (Fig. 3). Our data suggest that the molecular underpinnings for these defects rest on an insufficient cellular level of active Spo0A (Fig. 4), and the inability of σ^H to associate effectively with core RNA-polymerase (Fig. 5).

Previous work from Hecker and co-workers (Reder et al., 2012b; Reder et al., 2012a) demonstrated that the sigma factor (σ^B) that controls the general stress regulon (Price, 2011; Hecker et al., 2007) activates spo0E encoding a phosphatase that attenuates flux through the phosphorelay. Although salt-shocks are among the strongest inducers of the general stress response in B. subtilis (Nannapaneni et al., 2012), induction of σ^B-controlled genes by this environmental cue is only short-lived (Young et al., 2013). Importantly, in cells experiencing chronic exposure to high salinity, as was used in our experiments, the general stress regulon is not induced (Spiegelhalter and Bremer, 1998). Accordingly, it is unlikely that the σ^B-dependent down-regulation of Spo0A~P makes a significant contribution to the early block in sporulation observed here. In support of this idea, the reduction in sporulation efficiency resulting from high salinity was indistinguishable between wild-type and an isogenic sigB mutant (Fig. S11).

The genetic control of spo0A transcription is complex and involves just-in-time regulatory events (Chastanet and Losick, 2011; Chastanet et al., 2010) that ensure that the cellular level of Spo0A~P increases in such a fashion that promoters with different affinities for this transcription factor become active at defined time points during starvation (Fujita et al., 2005; Fujita and Losick, 2005). To initiate spore-formation, high threshold levels of Spo0A~P and proper activation dynamics are required (Fujita et al., 2005; Grossman, 1995; Molle et al., 2003; Vishnoi et al., 2013). Our immunoblot analysis assessing Spo0A levels indicate that reduced amounts of Spo0A are present in cells cultured in the presence of 0.6 M NaCl and they are even further decreased when the salinity is increased to 1.2 M NaCl (Fig. 4). Hence, the inability of B. subtilis cells to sporulate efficiently at high salinity is likely explained by the insufficient accumulation of the master regulator of sporulation, Spo0A. Our data suggest that Spo0A protein is itself sensitive to high salinity (Fig. 3B and 4) although the molecular mechanism underlying this instability remains unknown. However, a second salt-sensitive event that contributes to the drop in Spo0A levels is the reduced association of σ^H with core RNA-polymerase (Fig. 5). Since the σ^H-dependent promoter of spo0A is a key genetic control element in setting cellular levels of Spo0A (Chibazakura et al., 1995; Eymann et al., 2001; Predich et al., 1992), insufficient amounts of the σ^H-RNA-polymerase holoenzyme will contribute to a failure to achieve sporulation-promoting levels of Spo0A (Britton et al., 2002; Hoch, 1991; Predich et al., 1992).

We do not know the reasons why σ^H associates ineffectively with core RNA-polymerase (Fig. 5). One possibility is that an increase in ECF sigma factors in response to salt stress competes for core. However, we favor the idea that an altered composition of the cytoplasmic ion and solute pools is the culprit. Indeed, a B. subtilis mutant defective in the major Na⁺-sodium extrusion system Mrp (Ito et al., 1999; Swartz et al., 2005) not only becomes highly sensitive to Na⁺ ions but also exhibits a severe sporulation defect (Kosono et al., 2000). B. subtilis typically maintains a very low cytoplasmic Na⁺ level. As assessed by ²³Na-nuclear magnetic resonance spectroscopy, the intracellular Na⁺ pool in an mrp mutant increases from practically non-measurable levels in the wild-type strain to approximately 12 mM under growth conditions where the external NaCl concentration was only 80 mM (Gorecki et al., 2014). Since the Mrp system functions as a Na⁺/H⁺ antiporter, its operation is also connected to pH homeostasis (Krulwich et al., 2011; Swartz et al., 2005). It is therefore of interest to consider data reported by Zuber and co-workers who found, as reported here for salt-stressed cells, a defect in the association of σ^H with core RNA-polymerase in B. subtilis cells that were subjected to low pH stress (Liu et al., 1999). The ion pool, in particular that of potassium, is certainly different in osmotically stressed cells from those that are not subjected to this challenge (Whatmore et al., 1990). Furthermore, transient changes in the intracellular Na⁺ content of high-salinity-exposed B. subtilis cells might occur as a consequence of the operation of osmotically induced and Na⁺-driven osmostress protectant importers for glycine betaine (OpuD) and proline (OpuE) (Kappes et al., 1996; von Blohn et al., 1997).

Suppressor mutants with single nucleotide polymorphisms that map either in sigH or in the promoter region of kinA partially bypassed the salt-stress-imposed block in sporulation (Table 1 and 2). Although these mutations only modestly enhanced sporulation under high salinity, they provided additional support for the idea that Spo0A and σ^H activities are impaired under these conditions. Consistent with our findings that the interaction between σ^H and core RNA polymerase is reduced in high salt, the two amino acid substitutions in σ^H (E146A and I149F) both reside in the putative helix-turn-helix located within sub-region 3.1, which is associated with sigma binding to core RNA polymerase (Murakami and Darst, 2003). All the other suppressor mutations map in the regulatory region of kinA (Table 1), a gene that is dependent on σ^H and Spo0A for its expression (Fujita and Sadaie, 1998a; Predich et al., 1992). Since these mutations likely increase kinA transcription, we tested a strain in which kinA could be overexpressed. Strikingly, KinA overproduction almost completely bypassed the block to entering the sporulation pathway in 0.6 M NaCl (Fig. 6A and S9). It also significantly increased the frequency of mature, heat-resistant spores; however, sporulation efficiency was still 100-fold lower than that observed in the absence of salt stress (Fig. 6B). This difference can be explained, in part, by impaired engulfment and reduced σ^G activity but is principally due to cell lysis prior to maturation of the spore (Fig. 7, S9 and S10).

At first glance, the presence of a salt-stress imposed block to sporulation runs contrary to the idea that highly desiccation-resistant endospores (Nicholson et al., 2000; Setlow, 1995; Setlow, 2014) might serve as an effective osmostress escape route when vegetative growth is no longer possible (Boch et al., 1994). One possible explanation for this block is that B. subtilis has evolved an alternative strategy to survive starvation at high salinity that is more compatible with lower activities of σ^H and Spo0A. Indeed, we find that wild-type cells retain nearly complete viability in 1.2 M NaCl for at least 24 hours after they have exhausted their nutrients (Fig. S12). After 24 hours, the viability of these starved and salt-stressed cells steadily declines.

An alternative view places the strong and early block to sporulation by high salinity in a wider cellular context that takes the physiological status of starving (Eichenberger et al., 2004; Errington, 2003; Higgins and Dworkin, 2012) and salt-stressed cells (Bremer, 2002; Bremer and Krämer, 2000) into account. The commitment to sporulate is a measure of last resort after the cell has run out of other options (Vlamakis et al., 2013; Lopez et al., 2009). It is a time-consuming and energy demanding process and involves the coordinated transcription of almost a quarter of the genome (Eichenberger et al., 2004; Fawcett et al., 2000; Nicolas et al., 2012; Steil et al., 2005; Molle et al., 2003). Even under favorable laboratory conditions, it takes about seven hours to complete (Eichenberger et al., 2004), and after polar division and the first compartment-specific transcription factors are activated, a point of no return in this developmental program is reached (Errington, 2003; Higgins and Dworkin, 2012). Hence, the B. subtilis cell must assure that the dwindling nutritional and energetic resources available to it, and the environmental circumstances, will allow completion of the sporulation process. Otherwise, all things are lost: the mother cell will lyse and no spore will be formed.

High salinity environments impose considerable constraints on the physiology and growth of the B. subtilis cell (Boch et al., 1994; Bremer, 2002; Bremer and Krämer, 2000). A key event in its defense against osmotic stress is the de novo synthesis and high-level accumulation of the compatible solute proline to maintain physiologically adequate levels of cellular hydration and turgor (Brill et al., 2011; Whatmore et al., 1990). The size of the proline pool is linearly dependent on the degree of the environmentally imposed osmotic stress and reaches about 500 mM in B. subtilis cells exposed to very high (1.2 M NaCl) saline growth conditions (Brill et al., 2011; Hoffmann et al., 2013; Zaprasis et al., 2013). Genetic disruption of the osmostress-adaptive proline synthesis route makes B. subtilis highly salt sensitive, highlighting the important role of the pool of this compatible solute for growth under osmotically unfavorable circumstances (Brill et al., 2011). The synthesis of a single proline molecule requires the expenditure of ~20 high-energy phosphate bonds (Akashi and Gojobori, 2002). Hence, the maintenance of proline up to levels of 500 mM (Hoffmann et al., 2013; Zaprasis et al., 2013) is an enormous physiological task for starving B. subtilis cells that are faced with entering the sporulation pathway while simultaneously experiencing severe osmotic stress. The data presented here suggest that the block to sporulation under salt stress is not an accident; rather we propose it serves as a safeguard to avert osmotically stressed B. subtilis cells from committing to a developmental program they cannot complete and would die trying.

Experimental procedures

General methods

B. subtilis strains were derived from 168, JH642 or PY79. Unless otherwise indicated, cells were grown in LB (Lennox) or CH (Sterlini-Mandelstam) media at 37°C. Sporulation was induced by re-suspension at 37°C according to the method of Sterlini-Mandelstam or by nutrient exhaustion in supplemented DS medium (DSM complete) (Harwood and Cutting, 1990). Sporulation efficiency was determined in cultures grown for 24–36 hours as the total number of heat-resistant (80°C for 20 min) colony-forming units (CFUs) compared with the total CFUs or heat-resistant CFUs of cells sporulated without NaCl. Tables of strains, plasmids and oligonucleotide primers and a description of strain and plasmid constructions can be found online as supplementary data (Table S1, S2 and S3).

Fluorescence microscopy

Sporulating cells were concentrated by centrifugation at 8, 000 rpm for 1 min and immobilized on 2% agarose pads. Fluorescence microscopy was performed using an Olympus BX61 microscope equipped with a UplanF1 100X phase contrast objective lens and a CoolSnapHQ digital camera (Photometrics) or a Nikon TE2000 inverted microscope with a Nikon CFI Plan Apo VC 100X objective lens. Images were acquired using Metamorph software (Molecular DEVICES, Sunnyvale, CA, USA). Membranes were stained with TMA-DPH (50μM) (Molecular Probes; ThermoFisher Scientific) and fission of mother cell membranes was assessed as previously described (Doan et al., 2013). Image analysis and processing was performed in Metamorph.

Immunoblot analysis

Whole-cell lysates from sporulating cells were prepared as previously described (Doan and Rudner, 2007). Samples were heated for 5 min at 65°C prior to loading. Equivalent loading of proteins was based on the OD₆₀₀ of the cell cultures at the time of harvest. Samples were separated on a 12.5% SDS-polyacrylamide gel and transferred to a methanol-activated PVDF membrane. Membranes were blocked in 5% nonfat milk with 0.5% Tween-20 for 1 hour. Blocked membranes were probed with anti-σ^H (diluted 1:2,500), anti-Spo0A (diluted 1:5,000) or anti-SpoIIQ (diluted 1:10,000) (Doan et al., 2009), anti-σ^A (diluted 1:10,000), anti-KinA (diluted 1:5,000), and anti-core RNA-polymerase (diluted 1:5,000). These primary antibodies were diluted into PBS with 0.05% Tween-20 and incubations were carried out at 4°C overnight. Primary antibodies were detected with horseradish-peroxidase conjugated anti-mouse or anti-rabbit antibodies and detected with Western Lightning ECL reagent as described by the manufacturer (PerkinElmer, Waltham, MA, USA).

Protein Pull-down assay

25-ml cultures of strain BDR485 (Table S1) harboring a functional rpoC-his6 fusion were grown in CH medium in the presence and absence of 0.6 M NaCl. At an OD₆₀₀ of ~0.5, sporulation was induced by resuspension in the presence and absence of 0.6 M NaCl. 1.5 hours later, 20 ml from each culture were harvested by centrifugation (5000 rpm for 10 min at room temperature), washed two times in 10 ml cold lysis buffer (20 mM Tris pH7.5, 300 mM NaCl, 20 mM imidazole, 5 mM MgCl₂, 5 mM β-mercaptoethanol, 1 mM PMSF, 1 μg ml⁻¹ leupeptin and 1 μg ml⁻¹ pepstatin) and the pellets were flash frozen and stored at −80° C. The cell pellets were thawed and re-suspended in 1 ml cold lysis buffer with 2mg/ml lysozyme and incubated at 4°C for 20 min. The cell suspension was lysed by sonication followed by the addition of 10 μl of 1 mg/ml DNase (Worthington, Lakewood, NJ, USA), 10 μl of 10 mg/ml RNaseA, and 20 μl of 100 mM PMSF. The lysate was clarified by centrifugation at 13,000 rpm at 4°C for 5 minutes and the supernatant was incubated with 50 μl of Ni²⁺-agarose (Qiagen Hilden, Germany) for 1 h at 4° C. The resin was washed 5 times with 1.5 ml wash buffer (20 mM Tris pH 7.5, 300 mM NaCl, 20 mM imidazole, and 5 mM β-mercaptoethanol) and then re-suspended in 100 μl SDS sample buffer and heated at 80° C for 10 min. Protein fractions were separated on a 12.5% SDS-polyacrylamide gel and visualized with Coomassie brilliant blue or by immunoblot.

Suppressor enrichment

To enrich for mutants that could form heat-resistant spores in the presence of high salinity, cells of the wild-type B. subtilis strain JH642 (Smith et al., 2014) were sporulated (30 h after reaching stationary phase) in DS medium containing 1.2 M NaCl. After eliminating vegetative cells by pasteurization (80 °C, 20 min), serial dilutions were plated onto LB agar plates and the spore titer was determined. Colonies arising from spores were then re-inoculated in fresh DS medium containing 1.2 M NaCl. After 30 h of growth after reaching stationary phase, the spore titer of the cultures was again determined. Cultures that showed an increased spore titer in comparison to that of the JH642 wild-type strain were analyzed for mutations either by whole genome re-sequencing, or by targeted sequencing of PCR products derived from the spo0A, kinA and sigH genes. The DNA primers used to amplify these genes and their flanking regions can be found in Table S3.

Mapping Suppressors by whole genome sequencing and PCR analysis

Chromosomal DNA from B. subtilis was isolated using the DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany). To identify the mutation(s) in the suppressor mutant strains NWB6 and NWB7 (Table 1), the genomic DNA was subjected to whole-genome sequencing. Library preparation and sequencing analysis were performed by the Göttingen Genomics Laboratory (Göttingen, Germany). The reads were mapped on the reference genome of the B. subtilis JH642 strain (GenBank accession number: CP_007800) (Smith et al., 2014). Mapping of the reads was performed as previously described (Zaprasis et al., 2014) using the Geneious software package (Biomatters Ltd., New Zealand). Single nucleotide polymorphisms (SNPs) were considered as significant when the total coverage depth exceeded 25 reads with a frequency variance of >90%. For the molecular analysis of all other suppressor strains, the promoter and coding regions of kinA, sigH and spo0A were PCR-amplified and Sanger sequenced (MWG Eurofins, Ebersberg, Germany).