Control of the Diadenylate Cyclase CdaS in Bacillus subtilis

Background: Bacillus subtilis CdaS is a sporulation-specific diadenylate cyclase.

Results: Activity of CdaS is regulated by its N-terminal autoinhibitory domain.

Conclusion: The synthesis of c-di-AMP is under tight control in B. subtilis.

Significance: The activity of CdaS is governed by a hexamer/dimer transition.

Abstract

The Gram-positive bacterium Bacillus subtilis encodes three diadenylate cyclases that synthesize the essential signaling nucleotide cyclic di-AMP. The activities of the vegetative enzymes DisA and CdaA are controlled by protein-protein interactions with their conserved partner proteins. Here, we have analyzed the regulation of the unique sporulation-specific diadenylate cyclase CdaS. Very low expression of CdaS as the single diadenylate cyclase resulted in the appearance of spontaneous suppressor mutations. Several of these mutations in the cdaS gene affected the N-terminal domain of CdaS. The corresponding CdaS mutant proteins exhibited a significantly increased enzymatic activity. The N-terminal domain of CdaS consists of two α-helices and is attached to the C-terminal catalytically active diadenylate cyclase (DAC) domain. Deletion of the first or both helices resulted also in strongly increased activity indicating that the N-terminal domain serves to limit the enzyme activity of the DAC domain. The structure of YojJ, a protein highly similar to CdaS, indicates that the protein forms hexamers that are incompatible with enzymatic activity of the DAC domains. In contrast, the mutations and the deletions of the N-terminal domain result in conformational changes that lead to highly increased enzymatic activity. Although the full-length CdaS protein was found to form hexamers, a truncated version with a deletion of the first N-terminal helix formed dimers with high enzyme activity. To assess the role of CdaS in sporulation, we assayed the germination of wild type and cdaS mutant spores. The results indicate that cyclic di-AMP formed by CdaS is required for efficient germination.

Introduction

For intracellular signal transduction, both prokaryotic and eukaryotic cells rely on the use of second messenger signaling nucleotides. Among these messengers, 3′,5′-cyclic AMP (cAMP) has been the most intensively studied. In bacteria, this nucleotide is implicated in a wide range of biological processes including carbon catabolite repression and virulence (1–3). In addition, the stringent-response factor (p)ppGpp is involved in switching off cellular household activities in response to nutrient limitation (4). Moreover, the cyclic dinucleotides c-di-GMP3³ and c-di-AMP were found to be implicated in controlling lifestyle decision-making between virulence and cell division (5–7). Although c-di-GMP is widespread in many bacterial lineages, c-di-AMP is common in Gram-positive bacteria (both Actinobacteria and Firmicutes) and occurs elsewhere only in bacteria of the Bacteroidetes and in δ-proteobacteria. Very recently, the hybrid cyclic nucleotide 3′5′-3′5′ cyclic GMP-AMP (cGAMP) was discovered in Vibrio cholerae and shown to be implicated in intestinal colonization (8).

The Gram-positive model bacterium Bacillus subtilis produces the signaling nucleotides ppGpp, c-di-GMP, and c-di-AMP. As in other bacteria, ppGpp mediates the stringent response in B. subtilis (9). Little is known about the function of c-di-GMP in this bacterium. Unlike many other species, B. subtilis does not rely on c-di-GMP signaling for biofilm formation. The evidence available to date suggests that c-di-GMP is involved in the control of motility (10, 11).

The existence of c-di-AMP was first discovered when the crystal structure of the B. subtilis DisA protein was solved (12). This protein has two interdependent activities: (i) it is a diadenylate cyclase that synthesizes c-di-AMP, and (ii) it binds DNA via its RuvA-like C-terminal DNA binding domain and scans DNA integrity. Upon detection of branched DNA molecules as found in e.g. Holliday junctions, the DisA catalytic activity is inhibited, and this reduction in c-di-AMP concentration was reported to result in the delay of sporulation (12, 13). More recently, two additional diadenylate cyclases, CdaA and CdaS, were discovered in B. subtilis (14, 15). A common feature of all diadenylate cyclases is the presence of the conserved diadenylate cyclase (DAC) domain (previously referred to as domain of unknown function DUF147) (12, 15, 16).

The presence of three diadenylate cyclases is rather unusual in bacteria and raises the questions, For what purpose do the individual enzymes serve and how are their particular activities controlled? In B. subtilis, DisA and CdaA are constitutively expressed, whereas CdaS is expressed only during sporulation (17), suggesting a sporulation-related function for CdaS. c-di-AMP formation in vegetative cells was reported to be implicated in the control of cell division and cell wall biosynthesis, two intimately inter-linked cellular processes (14, 15, 18). In addition to their expression, the activity of the diadenylate cyclases is also subject to regulation. The activity of B. subtilis DisA was reported to be controlled both by the integrity of the chromosomal DNA (13) and by a regulatory protein-protein interaction with the RadA recombination protein (19). Similarly, the activity of CdaA is modulated by a protein-protein interaction with the CdaR protein (15). For both couples of diadenylate cyclases and their regulators, the two proteins form highly conserved gene clusters. For CdaS, the mechanism by which the activity of this enzyme is controlled has not yet been elucidated even though a mutation in the N-terminal domain of CdaS results in strongly increased activity (15). Importantly, the three diadenylate cyclases of B. subtilis contain different types of domains in addition to the enzymatically active DAC domain. It is, therefore, tempting to speculate that these different domains serve to control the enzymatic activities of the enzymes.

None of the individual diadenylate cyclases of B. subtilis is essential for the growth of the bacterium; however, the bacteria are unable to survive if none of them is active in the cell, indicating that c-di-AMP is an essential second messenger (14, 15). This conclusion is supported by the essentiality of the unique diadenylate cyclase (corresponding to CdaA) in low-GC Gram-positive pathogens including Listeria monocytogenes, Staphylococcus aureus, and Streptococcus pneumoniae (20–22).

In this work we have characterized the diadenylate cyclase CdaS. Our results indicate that the activity of CdaS is modulated by its N-terminal domain. CdaS forms a hexamer as a trimer of dimers with interaction surfaces provided by the DAC domains on one hand and the N-terminal domains on the other. We demonstrate that disruption of hexamerization by deleting the N-terminal domain results in strongly increased activity of the enzyme. Moreover, we investigated the importance of CdaS expressed in the forespore during late-stage sporulation and provide evidence for an effect of depletion of CdaS-derived c-di-AMP on spore germination.

EXPERIMENTAL PROCEDURES

Bacterial Strains and Growth Conditions

The B. subtilis strains were derived from the laboratory strain 168 (trpC2), and they are listed in Table 1. Escherichia coli XL1-Blue (Stratagene) and BL21 (DE3) (23) were used for cloning experiments and protein overproduction, respectively.

TABLE 1.

B. subtilis strains used in this study

Strain	Genotype	Source
168	trpC2	Laboratory collection
GP983	trpC2 ΔcdaS::ermC	15
GP1173	trpC2 lacA:: (pxylA-yfp aphA3)	45
GP1327	trpC2 lacA:: (pxylAcdaS aphA3) ΔdisA::tet ΔcdaS::ermC ΔcdaA::cat	15
GP1334	trpC2 lacA:: (pxylAcdaSL44F aphA3) ΔdisA::tet ΔcdaS::ermC ΔcdaA::cat	15
GP1350	trpC2 lacA:: (pxylAcdaSA61V aphA3) ΔdisA::tet ΔcdaS::ermC ΔcdaA::cat	Suppressor mutant picked from GP1327
GP1353	trpC2 lacA:: (pxylAcdaSA76V aphA3) ΔdisA::tet ΔcdaS::ermC ΔcdaA::cat	Suppressor mutant picked from GP1327
GP1354	trpC2 lacA:: (pxylAcdaSE46K aphA3) ΔdisA::tet ΔcdaS::ermC ΔcdaA::cat	Suppressor mutant picked from GP1327
GP1355	trpC2 lacA:: (pxylAcdaSP201 aphA3) ΔdisA::tet ΔcdaS::ermC ΔcdaA::cat	Suppressor mutant picked from GP1327

E. coli was grown in LB medium. B. subtilis was grown in LB medium or in C minimal medium (24) supplemented with carbon sources and auxotrophic requirements (at 50 mg l⁻¹) as indicated. CSE medium is C medium supplemented with potassium succinate and potassium glutamate (6 and 8 g/liter, respectively). LB plates were prepared by the addition of 17 g of Bacto agar/liter (Difco) to LB (23).

When required, media were supplemented with antibiotics at the following concentrations: ampicillin (100 μg ml⁻¹), kanamycin (50 μg ml⁻¹) (for E. coli); spectinomycin (150 μg ml⁻¹), kanamycin (10 μg ml⁻¹), tetracycline (12.5 μg ml⁻¹), chloramphenicol (5 μg ml⁻¹), and erythromycin (2 μg ml⁻¹) plus lincomycin (25 μg ml⁻¹) (for B. subtilis).

DNA Manipulation and Transformation

Transformation of E. coli and plasmid DNA extraction were performed using standard procedures (23). All commercially available plasmids, restriction enzymes, T4 DNA ligase, and DNA polymerases were used as recommended by the manufacturers. DNA fragments were purified from agarose gels using the QIAquick PCR purification kit (Qiagen, Hilden, Germany). DNA sequences were determined using the dideoxy chain termination method (23). Chromosomal DNA of B. subtilis was isolated as described (25).

Standard procedures were used to transform E. coli (23), and transformants were selected on LB plates containing ampicillin (100 μg/ml). B. subtilis was transformed with plasmid DNA according to the two-step protocol (25). Transformants were selected on sporulation plates containing the appropriate antibiotics.

Expression of Recombinant Diadenylate Cyclases in E. coli

To assess the biochemical activity of the diadenylate cyclase variants, the enzymes were expressed in E. coli BL21(DE3). This organism does not produce c-di-AMP and is, therefore, well suited to analyze the production of the nucleotide (18). For this purpose, the cdaS alleles were amplified from chromosomal DNA of the respective mutants using the primer pair FX71/FX72 (15) and cloned between the NdeI and BamHI sites of the expression vector pET28a (Novagen). The resulting plasmids and the corresponding mutations are listed in Table 3. To generate truncated forms of CdaS, the corresponding gene fragments were amplified by using the primer pairs FX98 (AAAMCATATGGACCAGTGTCTCCTTTGTGAATTAGATGATTTG)/FX72 and FX99 (AAACATATGCCGGCATTTATTGAGCTGGCCAAGG)/FX72. The corresponding plasmids were pGP1993 (ΔH1) and pGP1994 (ΔH2).

TABLE 3.

Activity of CdaS variants

Intracellular c-di-AMP was determined from E. coli cultures harboring the indicated expression vectors (pET28a derivates). Cultures were grown in LB medium, and c-di-AMP amounts were determined using HPLC/MS as described under “Experimental Procedures.” Each value is the mean of three independent experiments. S.D. are given in parentheses.

Vector	Mutation	ng c-di-AMP/mg protein
pET28a	Empty vector	0 (0)
pGP1974	WT	623 (35)
pGP1975	L44F	56,400 (11,480)
pGP2553	E46K	12,196 (140)
pGP1995	A61V	7,250 (2,525)
pGP1998	A76V	2,700 (463)
pGP2552	P201Q	7,970 (463)
pGP1993	ΔH1	12,500 (1,195)
pGP1994	ΔH2	10,500 (1,053)

Protein Purification

E. coli BL21 (DE3) was used as the host for the overexpression of recombinant proteins carrying a His tag for affinity purification. The cultures were grown in 1 liter of LB medium at 37 °C. Expression was induced by the addition of isopropyl 1-thio-β-d-galactopyranoside (final concentration 1 mm) to logarithmically growing cultures (A₆₀₀ of 0.5), and cultivation was continued at 16 °C for 18 h. Cells were harvested, and the pellets were resuspended in 30 ml of disruption buffer (50 mm Tris/HCl, 200 ml NaCl, 10 mm imidazole, 10% (v/v) glycerol, pH 7.5). The cells were disrupted by using a French press (18,000 p.s.i., 138,000 kilopascals; Spectronic Instruments). The crude extracts were passed over a 2.5-ml nickel-nitrilotriacetic acid column. The column was washed with 30 ml of disruption buffer followed by stepwise elution with imidazole. Five different concentrations of imidazole were used: 10, 50, 75, 150, and 200 mm imidazole in 5 ml of disruption buffer, pH 7.5. The Bio-Rad dye binding assay was used to determine protein concentrations. Bovine serum albumin was used as the standard.

Determination of the Oligomerization State of the CdaS Wild Type and Variants

For analytical gel filtration a Superdex 200 10/300 GL column (GE Healthcare) connected to an Äkta Purifier system was used. The column was equilibrated with 50 mm Tris/HCl, 200 mm NaCl, 150 mm imidazole, 10% glycerol, pH 7.5. Purified proteins were applied onto the column, and molecular weights were calculated in comparison to a gel filtration standard (Bio-Rad).

Blue native-PAGE was performed as previously described (26). Purified protein was applied onto a 5–24% non-denaturing polyacrylamide gel. The gel run was performed at 10 °C and 20 V/cm for 10 h.

Western Blotting

For Western blot analysis, proteins were transferred from native gels onto polyvinylidene difluoride (PVDF) membranes (Bio-Rad) by electroblotting. Rabbit anti-His-tag polyclonal antibodies (Antibodies-Online, Aachen; 1:1,000) served as primary antibodies. The antibodies were visualized by using anti-rabbit immunoglobulin alkaline phosphatase secondary antibodies (Promega) and the CDP-Star detection system (Roche Diagnostics), as described previously (24).

Analysis of the Cyclic Dinucleotide Pools

The concentration of cyclic di-AMP was determined by a liquid chromatography-coupled tandem mass spectrometry method essentially as described previously (27). Briefly, E. coli cultures were grown in LB medium at 37 °C. At an A₆₀₀ of 0.5–0.7 1 mm isopropyl 1-thio-β-d-galactopyranoside was added to induce the expression of the cdaS alleles and further incubated for 3 h. The culture (10 ml) was harvested by quick centrifugation at 4 °C. An additional sample was taken for the determination of total protein amount for normalization purposes. The cell pellet was resuspended in 300 μl of extraction buffer (acetonitrile/methanol/water mixture, 40/40/20, v/v/v) and shock-frozen in liquid nitrogen. Samples were then treated by a heating step for 10 min at 95 °C followed by centrifugation for 10 min at 20,800 × g at 4 °C. The extraction of the resulting pellet was repeated twice with 200 μl of extraction mixture at 4 °C omitting the heating step. Supernatants were pooled and stored at −20 °C overnight. After centrifugation for 10 min at 20,800 × g at 4 °C, the supernatant was removed and dried in a SpeedVac. The dried supernatants were solved with 200 μl of H₂O. After repeated centrifugation and addition of the internal standard ¹³C,¹⁵N-c-di-AMP, part of the supernatants was analyzed by LC-MS/MS.

Quantification of c-di-AMP by MS/MS

The chromatographic separation was performed on a Series 200 HPLC system (PerkinElmer Life Sciences) equipped with a binary pump system and a 200-μl sample loop. A combination of column saver (2.0 μm filter, Supelco Analytical), security guard cartridge (C18, 4 × 2 mm) in an analytical guard holder (Phenomenex), and an analytical NUCLEODUR C18 Pyramid RP column (50 × 3 mm, 3 μm particle size, Macherey-Nagel) temperature controlled (Series 200 Peltier column oven, PerkinElmer Life Sciences) at 30 °C was used. Eluent A consisted of 10 mm ammonium acetate and 0.1% (v/v) acetic acid in water, and eluent B was methanol. The injection volume was 50 μl, and the flow rate was 0.4 ml/min throughout the chromatographic run. 100% A was used from 0 to 5 min followed by a linear gradient from 100% A to 70% A until 9 min. Re-equilibration of the column was achieved by constantly running 100% A from 9 to 13 min. The internal standard ¹³C,¹⁵N-c-di-AMP and c-di-AMP were eluted with identical retention times of 9.1 min (18). The analyte detection was performed on an API 3000 triple-quadrupole mass spectrometer equipped with an electrospray ionization source (AB SCIEX) using selected reaction monitoring (SRM) analysis in positive ionization mode. The following SRM transitions using a dwell time of 40 ms were detected: ¹³C,¹⁵N-c-di-AMP (+689/146 (quantifier), +689/345 (identifier)) and c-di-AMP (+659/136 (quantifier), +659/330 (identifier), and +659/524 (identifier)). The SRM transitions labeled as “quantifier” were used to quantify the compound of interest, whereas “identifier” SRM transitions were monitored as confirmatory signals. The quantifier SRM transitions were most intense and were, therefore, used for quantification. The mass spectrometer parameters were as follows: ion source voltage, 5500 V; temperature, 350 °C; nebulizer gas, 6 p.s.i.; curtain gas, 15 p.s.i., MS/MS was performed using nitrogen as collision gas. The following collision energies were applied: 61 eV (+689/146), 29 eV (+689/345), 63 eV (+659/136), 29 eV (+659/330), 33 eV (+659/524).

Spore Preparation

Spores of strains GP1173 and GP983 were prepared at 37 °C on Schaeffer's agar plates (28) for 2 days. Plates were then incubated at room temperature for 4 weeks to ensure full maturation of the spores (29). Spores from all plates were harvested and washed 4 times in 40 ml of ice-cold sterile demineralized Milli-Q water (6000 rpm for 15 min at 4 °C). This washing procedure was repeated for 10 days until pure spore crops (>95% phase bright spores) were obtained. Spores were stored in water at 4 °C and immediately used for germination assays.

Spore Germination Assays

Spores at an A₆₀₀ of 10 were heat-treated at 70 °C for 30 min, cooled on ice, washed, and germinated at an A₆₀₀ of 0.5–0.7 in LB medium at 37 °C. As a negative control, 25 mm Tris-HCl, pH 8.2, was used. The optical density (600 nm) was measured in a spectrophotometer at 10, 20, 30, 60, 120, 180, 240, and 360 min after dilution in LB. The percentage of the drop in A₆₀₀ was calculated relative to the start OD and plotted using the Excel software. Alternatively, germination was followed using phase contrast microscopy at 60, 120, and 240 min.

Germination Time-lapse Microscopy

Spores were heat-treated at 70 °C for 30 min, cooled on ice, and washed. Spores at an A₆₀₀ of 10 were then resuspended in 150 μl of 25 mm Tris-HCl, pH 8.2, after which 1.5 μl was loaded on LB-agarose microscopy slides. Slide preparation and time-lapse microscopy settings used were as previously described (30) with some minor alterations. In short, 75 mg (1.5%) of high resolution low melting agarose (Sigma) was dissolved in 5 ml of LB medium instead of minimal medium. After application of the coverslip, the slide was placed in the pre-warmed microscopy chamber (30 °C) and then immediately used for detecting spores. The following equipment (provided by DeltaVision) was used for the time-lapse microscopy experiments: a Deltavision Elite (GE Healthcare) Microscope with Trulight illumination and a PCO sCMOS camera used in combination with a 100× phase contrast objective (1.3 NA). To preserve focus in the z axis, software-based autofocus and hardware-based ultimate focus were performed before imaging. Software-based autofocus using diascopic light was programmed for 0.06-μm steps and a total range of 1.2 μm. Snapshots for movies were taken at intervals of 2 min for 6 h using 32% white LED light and 0.05-s exposure for bright field pictures. Raw data were stored using softWoRx 3.6.0 (Applied Precision).

RESULTS

Isolation of Mutant CdaS Variants with Increased Activity

As reported previously, c-di-AMP is essential for the viability of B. subtilis (14, 15). Depletion of the c-di-AMP pool due to regulated expression of CdaS as the single diadenylate cyclase resulted in the appearance of suppressor mutants. The corresponding mutations affected either the promoter region, thus leading to constitutive expression of the ectopic cdaS allele, or the cdaS structural gene and led to the hyperactive enzyme variant CdaS_L44F (15). To get further insights into the control of CdaS activity, we isolated a series of suppressor mutants that allowed growth of B. subtilis GP1327 (ΔdisA ΔcdaA ΔcdaS lacA::(p_xylAP_xycdaS)) in the absence of the inducer xylose. Such mutants were readily obtained, and the known potential mutation sites, the p_xylA regulatory region, and the cdaS gene were analyzed by DNA sequencing. The vast majority of the suppressor mutants harbored a mutation within the p_xylA promoter-operator region of the ectopic cdaS gene. As observed previously, these mutations resulted in the destruction of the XylR binding site and thus constitutive cdaS expression even in the absence of the inducer xylose (15). However, in five mutant strains we identified single-site mutations within the ectopic cdaS allele (see Table 2).

TABLE 2.

Screening of suppressor mutants reveals mutations within cdaS

GP1327 was cultured without xylose, which resulted in the formation of suppressor colonies. More than 80 suppressor colonies were isolated and sequenced for their P_xylcdaS regions.

Strain	Site of mutation	Mutation	Consequence
>70 suppressors without no.	xyl promoter	Alteration of XylR binding site	Constitutive cdaS expression
	cdaS allele	C121A	Q41K
GP1334	cdaS allele	C130T	L44F
GP1354	cdaS allele	G136A	E46K
GP1350	cdaS allele	C182T	A61V
GP1353	cdaS allele	C227T	A76V
GP1355	cdaS allele	C602A	P201Q

The crystal structure of the orthologous YojJ protein of Bacillus cereus (PDB code 2FB5), which shares 47% sequence identity overall to CdaS, reveals that the conserved DAC domain, which is responsible for c-di-AMP synthesis, is flanked by an N-terminal domain consisting of two α-helices (Fig. 1A) (15). Four of the five mutations affected residues in the N-terminal domain of CdaS, whereas the fifth mutation affects a proline residue close to the C terminus of the protein.

FIGURE 1.

A model of the three-dimensional structure of CdaS. A, top, the model is based on the known structure of the corresponding YojJ protein of B. cereus (PDB code 2FB5) using the SWISS-MODEL homology-modeling server (46). The DAC domain and the N-terminal helices are shown in light blue and dark blue, respectively. The positions of the amino acid exchanges resulting in hyperactive CdaS variants are highlighted in red. A, bottom, schematic view of the CdaS architecture. N-terminal helices are termed H1 and H2. Numbers indicate the amino acid position of the respective elements. B, three-dimensional model of the CdaS hexamer. This model is assembled from the monomeric units of A. Each monomer is colored in yellow, magenta, blue, green, pink, and gray, respectively. The structure is stabilized by interactions via the dimerization interfaces at the N and C terminus.

To test the idea that these mutations result in a hyperactive enzyme variants that can counteract the extremely low CdaS expression in B. subtilis GP1327 in the absence of the inducer xylose (due to the slightly leaky inducible promoter), we cloned the mutant alleles into the expression vector pET28a (except cdaS_Q41K, which could not be isolated from chromosomal DNA). The production of c-di-AMP was assayed in the heterologous host E. coli. Compared with wild type CdaS, the expression of the mutant alleles led to an increase in c-di-AMP formation by a factor of 4 to almost 100, as shown in Table 3. Thus, all mutations indeed led to hyperactive CdaS variants. This finding led to the conclusion that the two helices of the N-terminal domain may be important to control the activity of the DAC domain and thus c-di-AMP production by CdaS.

The Role of the N-terminal Helices as an Autoinhibitory Domain

To test the hypothesis that the helices of the N-terminal domain regulate the activity of the DAC domain, we constructed two N-terminally truncated CdaS variants. In CdaS_ΔH1 (pGP1993), the first helix was deleted, whereas CdaS_ΔH2 (pGP1994) had a deletion of both helices (i.e. the whole N-terminal domain) resulting in a protein consisting exclusively of the DAC domain. Expression of the corresponding plasmids in E. coli and quantification of c-di-AMP amounts resulted in a 20- and 17-fold increase in comparison to the full-length protein (pGP1974), respectively (see Table 3).

Thus, the N-terminal helices of CdaS are indeed important for regulation of the enzymatic activity of the DAC domain. The increased activities in the absence of one or both of the N-terminal helices suggest a role as an autoinhibitory domain, which allows the full activity only in the presence of an as yet unknown stimulus.

Analysis of the Oligomeric State of CdaS

To understand the molecular mechanism of regulation of the diadenylate cyclase CdaS in more detail, we used the known structure of B. cereus YojJ as a guide (Fig. 1B). Analysis of the crystal structure of YojJ suggests that it forms hexamers, comprising three pairs of independent dimer interfaces. The first interface (between chains A and B, C and D, and E and F) buries hydrophobic residues in the N-terminal helices of YojJ that are separated by heptads (Ile-18 [Leu-22; CdaS equivalent residues are given in brackets], Ala-21 [Ile-25], Leu-25 [Ala-29], Met-28 [Ile-32], Ile-32 [Leu-36] from helix 1 and Ile-39 [Leu-43], Met-43 [Leu-47], Ile-46 [Leu-50], Leu-50 [Phe-54], Val-53 [Met-57], Ala-57 [Ala-61], and Tyr-61 [Tyr-65] from helix 2) as expected in the packing of pairs of helices into anti-parallel coiled-coils. Leu-40 [Lue-44] also forms part of an apolar contact between helix 2 of chain A and the DAC domain of chain B, specifically residues Phe-190 [Phe-193], Tyr-197 [Tyr-200], Ile-199 [Ley-202], a contact that is repeated in chain pair C and D and pair E and F. In total, about 1500 Ų of solvent-accessible area is buried in the A-B interface, about 13% of the total surface area. A second interface is formed between pairs of DAC domains (chains A and E, B and C, and D and F); residues Thr-108 [Thr-112]-His-113 [Glu-117] pack against one another, akin to an anti-parallel pair of β-strands, and Pro-117 [Ser-121], Leu-118 [Leu-122], Glu-120 [Glu-124], Ser-121 [Ser-125], Ile-122 [Ile-126], Tyr-124 [Phe-128]-Leu-129 [Leu-133] make up the other part of this interface, burying some 700 Ų of solvent-accessible surface area, ∼6% of the total. The interface residues in YojJ are maintained almost universally in CdaS, suggesting strongly that it also forms a stable hexamer.

To obtain experimental support for the hexamerization of CdaS, we purified the wild type protein and analyzed its oligomeric state by blue native-PAGE (Fig. 2). We observed a weak signal corresponding most likely to the less stable DAC-DAC dimer that migrated just below the 45-kDa marker (the molecular mass of the CdaS monomer is 23 kDa). A more intense band is seen above the 68-kDa marker that is likely to represent the more stable, elongated coiled-coil dimer that, therefore, migrates aberrantly. The most slowly migrating species above the 158-kDa marker is likely to be the hexamer as suggested by the structural findings of the related YojJ. Beneath, the weaker signal near the 158-kDa (second band from the top) presumably represents a degradation product, as the CdaS protein is highly unstable.

FIGURE 2.

Analysis of the CdaS oligomeric state via blue native-PAGE. A, SDS-PAGE of purified CdaS wild type and variants. The plasmids pGP1974 (wt), pGP1995 (A61V) und pGP1993 (ΔH1) were transformed in E. coli. Transformants were cultivated in LB medium, and expression of the recombinant enzymes was induced by the addition of isopropyl 1-thio-β-d-galactopyranoside at 16 °C for 18 h. The following purification of the proteins was performed with nickel-nitrilotriacetic acid in 50 mm Tris-HCl, pH 7.5, 200 mm NaCl 10 mm imidazole 10% (v/v) glycerol. Elution fractions were separated by a 12% gel and visualized with Coomassie staining. B, blue native-PAGE of the CdaS variants. 1 μg of samples wt and ΔH1 and 3 μg of sample A61V (shown in A) were applied on a 5–24% non-denaturing gel and separated by blue native-PAGE as described in under “Experimental Procedures.” Proteins were blotted onto a PVDF membrane and visualized using an antibody against the His tag. The theoretical molecular masses of the monomeric units are 23 kDa (WT and A61V) and 18 kDa (ΔH1). The dashed line indicates the analysis of respective samples on different gels.

To further test whether the coiled-coils contribute to hexamer assembly in CdaS, we also analyzed the CdaS_ΔH1 and CdaS_A61V variants by blue native-PAGE (Fig. 2B). For CdaS_ΔH1 CdaS_ΔH1, which lacks residues 1–39, corresponding to the whole of helix 1, a single species was observed that migrated close to the 45-kDa marker, most likely representing a DAC-DAC dimer. Higher order species are not observed because the N-terminal coiled-coil helix bundle cannot form in the absence of helix 1. For the CdaS_A61V variant, where replacement of the alanine by valine is expected to be tolerated by the protein even though Ala-61 is part of the dimer interface, the protein behaved like wild type, with bands observed that likely represent the hexamer, the degradation product, and the two different dimer forms (see Fig. 2B).

We further analyzed the oligomeric state of wild type CdaS and the two variants by analytical gel filtration (Fig. 3). For the wild type protein, we found peaks corresponding to 167, 37, and 17 kDa, likely representing hexamers, dimers, and monomers of CdaS. By contrast, there was a major peak at 164 kDa corresponding to a hexamer for CdaS_A61V. The analysis of CdaS_ΔH1 resulted as expected in a single peak for 36 kDa, corresponding to a DAC-DAC dimer.

FIGURE 3.

Analysis of the CdaS oligomeric state via analytical gel filtration. Protein samples of the CdaS variants shown in Fig. 2A as well as a gel filtration standard (Bio-Rad) were separated using a Superdex™ 200 10/300 GL column. Molecular weights of the standard proteins are indicated. A₂₈₀ = absorption at 280 nm; mAU, milliabsorbance units; V_E = elution volume.

In summary, wild type CdaS is capable of forming a hexamer, whereas CdaS_ΔH1 forms only dimers, consistent with the analysis of the crystal packing of the orthologous YojJ, which forms hexamers with three sets of two independent dimer interfaces. Loss of one of these interfaces, the truncation of helix 1, results in a protein that can only form DAC-DAC dimers.

Absence of CdaS Affects Spore Germination in B. subtilis

A remaining question is why B. subtilis contains a spore-specific diadenylate cyclase. Previously, it has been shown that the cdaS gene is expressed in the forespore under the control of σ^G (17). One hypothesis is that a spore-specific depot of c-di-AMP is required to enable proper germination of the resulting spore. To investigate the effect of a cdaS deletion on spore germination behavior, wild type and ΔcdaS spores were germinated in LB medium, and the subsequent germination events were monitored using spectroscopic and microscopic analysis as described under “Experimental Procedures.” As a response to the germination triggers, dormant spores release dipicolinic acid, which is present in the spore core, and simultaneously rehydrate due to the uptake of water molecules. This transition can be visualized by phase-contrast microscopy as the spore turns from phase bright to phase dark. Alternatively, this event can be measured via spectroscopy in which a drop in optical density signifies germination efficiency (31). As shown in Fig. 4A, the germination efficiency of ΔcdaS spores was reduced about 2-fold as compared with wild type spores (compare the decrease of the A₆₀₀ for the cdaS mutant to the absorbance decrease for the wild type; A₆₀₀ decrease of 15% versus 30%, respectively). This observation was confirmed by phase contrast microscopy, indicating that CdaS activity during sporulation is indeed important for the spore ability to germinate. Interestingly, the absence of CdaS does not seem to affect the timing of germination and outgrowth (Fig. 4B and data not shown) but rather affects a subpopulation of individual spores that are unable to germinate under the conditions used.

FIGURE 4.

Germination response of heat-activated spores in LB. A, the change in optical density of wild type spores in Tris (open circles) or LB (black circles) and of ΔcdaS spores in Tris (open squares) or LB (black squares) was measured as described under “Experimental Procedures” and plotted in Excel. The percentage of the initial optical density (at 600 nm) is indicated on the y axis, with the optical density at the start of the experiment (T0) set at 100%. Phase contrast microscopy pictures of both strains in LB were taken after 60 min (germination) and 120 min (outgrowth). The scale in μm is indicated by the scale bar in the lower right corner. B, snapshot pictures taken from a movie for wild type spores and ΔcdaS mutant spores on LB agarose slides. Time is indicated as hours and minutes.

DISCUSSION

B. subtilis and related species of the Bacillaceae are unique in possessing the sporulation-specific diadenylate cyclase CdaS. This enzyme is not encoded even in related spore-formers of the genus Clostridium. In this work we provide evidence for a function of CdaS in spore germination, and we demonstrate that the enzymatic activity is subject to autoinhibition by the N-terminal domain.

The synthesis of c-di-AMP in B. subtilis is tightly controlled at the levels of expression and activity of the diadenylate cyclases. This control allows the cell to produce precisely the required amounts of c-di-AMP under all conditions. As shown previously, both elevated levels of c-di-AMP and a lack of the nucleotide interfere with growth of B. subtilis (15). CdaA is constitutively expressed; however, the enzymatic activity is stimulated by a protein-protein interaction with the CdaR regulatory protein (15, 17). The second vegetative diadenylate cyclase DisA depends on the alternative sigma factor, σ^M, for expression (32). In addition, the activity of DisA is inhibited upon interaction with the RadA recombination protein (19).

The essentiality of c-di-AMP for B. subtilis makes it difficult to assign specific roles to the c-di-AMP pools formed by the three diadenylate cyclases. However, both the conservation of the radA disA gene cluster in bacteria that possess DisA, the DNA scanning activity of DisA, and disA mutant phenotypes (14, 33, 34) suggest that c-di-AMP formed by this enzyme is involved in linking DNA integrity and cell division. Similarly, the conservation of CdaR and CdaA in cell wall-forming Firmicutes (and their absence from wall-less Mycoplasma) and the implication of c-di-AMP in cell wall metabolism in bacteria containing CdaA as the only diadenylate cyclase (18, 35–37) suggest a role for CdaA-derived c-di-AMP in cell wall homeostasis.

For CdaS, the control of its activity has so far not been studied. Of all diadenylate cyclases in B. subtilis, CdaS is subject to the strictest control at the level of gene expression (17), and expression is limited to the forespore compartment of sporulating cells and spores. In contrast to CdaA and DisA, CdaS is not encoded within a conserved gene cluster that might contain a protein regulator of the diadenylate cyclase activity as observed for the two other enzymes. Instead, CdaS contains an N-terminal domain (also referred to as YojJ domain) that is composed of two α-helices. As shown in this study, both a variety of point mutations in these helices as well as the deletion of one or both helices results in a strongly increased enzymatic activity of CdaS. Thus, the N-terminal YojJ domain likely serves as an autoinhibitory domain that helps to restrict c-di-AMP production by CdaS.

The presence of autoinhibitory domains in a protein is quite common. In Mycobacterium tuberculosis, the adenylate cyclase Rv1264 is subject to autoinhibition. In this case the N-terminal domain inhibits the enzyme activity by an interaction with the catalytic domain (38, 39). The intrinsic inhibition of this adenylate cyclase depends on the pH value (39). In the alternative sigma factor σ⁵⁴, the N-terminal domain prevents an interaction of the sigma factor with the RNA polymerase core enzyme, thus preventing the formation of a functional holoenzyme (40). An autoinhibitory role has also been demonstrated for the C-terminal membrane binding domain of E. coli pyruvate oxidase, which inhibits enzyme activity until it adheres to the membrane and liberates the active site (41). A regulatory autoinhibition was also observed for the B. subtilis transcription factor RocR. This protein activates σ⁵⁴-containing RNA polymerase in the presence of arginine or ornithine. The response to the substrate is achieved by preventing the N-terminal autoinhibitory domain from interfering with transcription activation (42, 43). In all those proteins with autoinhibitory domains, the inhibition can be overruled under specific conditions that result in activation of the proteins. This suggests that the N-terminal domain of CdaS exerts a dynamic control of the enzymatic activity of the protein.

Our analysis of the CdaS structure and of its oligomerization state allows some conclusions to be drawn about the potential mechanism by which the N-terminal domain inhibits the enzyme. The crystal structure, the native gel electrophoresis, and the gel filtration experiments indicate that wild type CdaS forms hexamers. However, to produce c-di-AMP from two molecules of ATP, the DAC domains of two subunits have to be appropriately positioned with respect to each other to form a symmetric active site, as seen in DisA, for example (12). An enzymatically competent structure cannot be adopted by the full-length, wild type protein as observed by x-ray crystallography (Fig. 1B), which may thus explain the relatively weak enzyme activity. Strikingly, the deletion of just one helix of the N-terminal domain is sufficient to prevent hexamer formation. With this truncation, an enzymatically competent form of the protein would be formed by the association of the DAC domains into the same type of DAC-DAC dimer seen in DisA. The DAC-DAC dimer interface seen in the structure of YojJ is not extensive (700 Ų, ∼6% of the total surface area) and is, therefore, susceptible to disassociation and re-association into an alternative dimer, whose formation is likely to be stabilized by the presence of either substrate or product. The DAC domains of YojJ can be superimposed on the DAC domains in DisA with 1.9 Å root mean square deviation, and there are no steric clashes between protomers in this alternative dimer. In the absence of nucleotide, this rearrangement would bury about 400 Ų, ∼6% of the total surface area. The conserved residues involved include glycines (Gly-126 [Gly-130], Gly-132 [Gly-136], Gly-161 [Gly-164]) to provide protein backbone flexibility and to wrap around the phosphates; polar, positively, and negatively charged residues (Gln-105 [Gln-109], Thr-162 [Thr-165]; His-130 [His-134], Ser-153 [Lys-156], Arg-165 [Arg-168]; Glu-159 [His-162], Glu-184 [Glu-187]) to interact specifically with nucleotidyl functional groups and to neutralize the negative charge of the phosphates. Asp-131 [Asp-135], the equivalent of which in DisA (Asp-77) has been determined to be essential for cyclase activity (12) is also maintained. The DisA-like DAC-DAC dimer is compatible with high enzyme activity. Presumably the missense mutations that produce high enzyme activity affect the switch between inactive and active states, which must require a significant conformational change to occur, to proceed from an inactive hexamer to a dimer that is not maintained in the hexameric assembly. The binding of ATP or the free energy released on ATP hydrolysis presumably drives the conformational change. Other alternatives to driving the conformational change include post-translational modifications (e.g. phosphorylation) or the binding of allosteric effectors. However, there is no experimental indication for either of these possibilities (44). The hydrophobic nature of the contacts between the helical domains affords the possibility that the helices slide over one another during the change in the relative positions of the helical and DAC domains within a CdaS monomer, and perhaps the A61V mutation can stimulate this movement. Alternatively, the binding or hydrolysis of ATP ensures that the relative positions of the two domains in CdaS are maintained, but the less extensive and less stable YojJ crystallographic DAC-DAC dimer is disassociated to permit the formation of a DisA-like DAC-DAC dimer. Both domains in this dimer contribute to the binding of each molecule of ATP, thereby contributing significantly to DAC-DAC dimer stabilization. Indeed, Ala-61 opposes Leu-44, which is in the interface between helical and DAC domains, and clearly changes to Leu-44 and presumably its immediate environment have a big effect on enzymatic activity, consistent with the idea that changes in domain organization occur in the full-length protein to result in the formation of the catalytically competent DisA-like DAC-DAC dimer.

Concerning the significance of having a forespore-specific c-di-AMP cyclase in B. subtilis, we have demonstrated that the absence of CdaS in sporulating cells affects the ability of a significant portion of spores to germinate in rich media. It is possible that the levels of c-di-AMP in these spores do not reach the threshold concentration required for successful germination. Interestingly, not all cdaS mutant spores are affected in germination. We hypothesize that the spores that do germinate and out-grow contain sufficient c-di-AMP derived from the vegetative DisA and CdaA cyclases before engulfment of the prespore.

The function of CdaS in germination suggests that a compound present in germinating spores might trigger a repositioning of the inhibitory N-terminal domain of the protein to trigger c-di-AMP production. The identity of this effector as well as the exact role of c-di-AMP in spore germination remain matters of further investigation. Experiments to address these questions are under way in our laboratories.