Decoding the functional role of the calcium ATPase YloB in microbially induced calcite precipitation and sporulation in Solibacillus silvestris

Graphical Abstract

YloB in Solibacillus silvestris is not involved in calcium detoxification or biomineralization but is important for sporulation, likely transporting Ca²⁺ into the forespore for Ca-DPA accumulation and spore maturation.

Abstract

Calcium homeostasis is essential for bacterial physiology, yet its regulation and role in specialized processes such as microbially induced calcite precipitation (MICP) and sporulation remain poorly understood. Here, we characterized the sole annotated P-type calcium ATPase in Solibacillus silvestris, which was previously shown to be upregulated under MICP conditions. We named the protein YloB based on homology to the Bacillus subtilis protein. Structural modelling confirmed that YloB possesses the conserved domains and motifs typical of calcium-binding P-type ATPases. Deletion of yloB did not affect growth under elevated calcium levels, nor did it reduce MICP yields, indicating that YloB is dispensable for calcium detoxification and biomineralization. In contrast, ∆yloB cells produced only partially dehydrated spores with severely reduced heat resistance, a defect rescued by complementation. A YloB-mNeonGreen fluorescent fusion localized specifically to the cell membrane surrounding the forespore during early engulfment and was retained at this site throughout the remainder of sporulation. A transcriptional P_yloB-msfGFP reporter demonstrated that yloB transcription was induced in the early stages of sporulation but did not respond to a calcium stimulus. These results show that YloB does not act to detoxify calcium from the cytoplasm of S. silvestris, but rather functions to pump calcium into the forespore, enabling Ca-dipicolinic acid accumulation and proper spore maturation. Our study reveals a specialized role for a bacterial calcium ATPase in sporulation, distinguishing calcium transport for endospore development from biomineralization and global calcium homeostasis.

Data Availability

All numerical data underpinning the graphs shown in the figure are available in Table S3.

Introduction

The calcium concentration in living cells is tightly regulated and plays an important role in signalling [1]. While the function of calcium has been extensively studied in eukaryotes, the regulation of calcium homeostasis and its role in bacterial physiology remain to be fully elucidated. Notably, intracellular bacterial calcium levels are kept in a narrow range between 100 and 300 nM, regardless of extracellular conditions, underscoring that calcium homeostasis is carefully controlled [2]. In response to calcium exposure or depletion, the bacterium Bacillus subtilis was shown to undergo significant alterations in gene expression, supporting a crucial role of Ca²⁺ ion homeostasis in bacterial physiology [3]. It has also been shown that calcium influences various physiological processes in a range of bacteria, such as swarming motility, virulence and sporulation [4,6].

One process in B. subtilis that is well known to require controlled Ca²⁺ levels is sporulation. This developmental cascade is triggered by nutrient depletion and is initiated by phosphorylation of the master regulator Spo0A, which activates genes for asymmetric division of the mother cell to form the forespore. The forespore is subsequently engulfed in a phagocytosis-like process and matures into a highly stress-resistant spore, encased by two membranes, a peptidoglycan layer and a protective coat [7,8]. As the sporulation process reaches its late stages, the spore becomes increasingly dehydrated through the accumulation of dipicolinic acid (DPA), comprising ~10% of dry weight in spores [9]. DPA is chelated in a 1:1 ratio with mainly calcium ions, ensuring the spore’s stress resistance, especially against wet heat [10,11].

Spore germination is triggered by the sensing of specific signal molecules, including certain amino acids, sugars and nucleotides that indicate favourable conditions for cell growth. The commitment of the spore to germinate leads to the release of Ca-DPA and the hydrolysis of cortex peptidoglycan [12]. This leads to a series of events such as complete spore hydration, an increase in membrane permeability and protein restoration which results in the outgrowth into a vegetative cell [12].

Beyond intracellular processes, calcium was also suggested to promote and stabilize biofilms by the formation of calcite lead structures [13,14]. Calcite is a calcium carbonate mineral that can be produced by bacteria in a process called microbially induced calcite precipitation (MICP), which is prevalent throughout the bacterial world [15,16]. In general, there are two pathways through which MICP can occur, the ureolytic and non-ureolytic pathways. Calcite precipitation by ureolytic bacteria involves the hydrolysis of urea to ammonia and carbon dioxide by the enzyme urease [17]. This causes an increase in pH and, in the presence of calcium ions, leads to the precipitation of insoluble calcium carbonate in the surrounding medium. On the other hand, non-ureolytic MICP describes a variety of metabolic processes including, for instance, denitrification, sulphate reduction, photosynthesis and heterotrophic pathways, which result in the precipitation of calcium carbonate, although it is not fully understood how these processes lead to mineral formation [16,18]. We recently showed that a sporulating soil isolate of the species Solibacillus silvestris could perform heterotrophic MICP using catabolism of acetate to provide the necessary carbonate ions from the released CO₂, converting nearly all available calcium into calcium carbonate. This remarkable efficiency and the ability to sporulate establish S. silvestris as an emerging model organism for studying the physiological mechanism of MICP in an application-relevant bacterium [19]. To provide calcium for MICP, it was proposed that ambient calcium passively enters the cell and is actively exported by calcium pumps. Release of calcium ions from the export system would thereby lead to a local enrichment of the ion on or near the cell surface, creating the supersaturated conditions needed to trigger mineral formation on the bacterial cell surface [16,19, 20]. For S. silvestris, this hypothesis was supported by the observation that the sole P-type calcium ATPase annotated in the genome of the organism was upregulated around sixfold upon calcium supplementation in the stationary phase [19].

A different model for MICP has been proposed for B. subtilis, where it was reported that, rather than the mineral being formed on or around the cell, calcite formation starts intracellularly by a subpopulation of cells during biofilm formation [13]. The authors hypothesized that intracellular accumulation of amorphous calcium carbonate granules, facilitated by P-type ATPases such as YloB, serves as a nucleation hub. Nucleated minerals are subsequently released to establish an extracellular calcite scaffold, effectively building a rigid skeleton for the biofilm. The study also proposed that biomineralization is an actively regulated mechanism, coupling calcium homeostasis to colony morphogenesis rather than arising from passive precipitation as an indirect consequence of metabolic activity [13]. Interestingly, B. subtilis YloB shows homology to the P-type ATPase of S. silvestris identified as upregulated during MICP conditions [19].

P-type ATPases are multi-domain, membrane-bound proteins that use ATP to pump ions across biological membranes [21]. They are grouped by the ion they transport; those specific for Ca²⁺ belong to the P2A subclass and actively maintain intracellular calcium homeostasis through ATP-driven phosphorylation–dephosphorylation cycles. All share a conserved core architecture – three cytoplasmic domains (the phosphorylation domain with a DKTG motif, the nucleotide-binding domain and the actuator domain with a TGE motif) plus ten transmembrane helices forming the ion pathway that undergo conformational shifts to move Ca²⁺ across the membrane [22]. To date, only a few calcium P-type ATPases have been shown to be directly implicated in the export of calcium ions from bacterial cells. Two of these calcium ATPases were identified in the human pathogens Streptococcus pneumoniae and Mycobacterium tuberculosis, where they were found to be essential to cope with toxic calcium concentrations in a eukaryotic host [23,24]. Beyond their role in exporting calcium from bacterial cells, calcium-binding P-type ATPases have also been found to act as calcium importers during growth in a calcium-deficient environment or for the development of heat-resistant endospores in B. subtilis [25,26].

To better understand the calcium flux in S. silvestris during the MICP process, here we aimed to characterize the sole calcium P-type ATPase annotated in the genome of the organism, which we named YloB based on its homology to the B. subtilis gene. Generation and testing of a yloB deletion mutant for growth at increased calcium levels and MICP yields showed no differences to the WT. However, cells lacking YloB were found to form mostly immature spores that were not fully dehydrated and lacked heat resistance. Fluorescently labelled YloB clearly localized to the spore membrane during the early steps of spore engulfment, whereas no YloB could be detected in non-sporulating cells. Using a transcriptional reporter assay, we demonstrated that P_yloB was predominantly active during the stationary phase, and activity was specific to cells that had initiated forespore engulfment. Promoter activation upon the addition of calcium was not conclusively observed. These findings indicate that YloB is vital for correct sporulation, most likely by providing the calcium ions required for spore dehydration, but does not play an active role in calcium detoxification or biomineralization by S. silvestris.

Methods

Strains and growth conditions

All strains and plasmids used in this study are listed in Table S1 (available in the online Supplementary Material). For routine growth, Escherichia coli strains were grown in standard lysogeny broth (LB) at 37 °C with shaking (120 r.p.m.). Selection for plasmid uptake and maintenance in E. coli was achieved with 100 µg ml⁻¹ ampicillin. Cultures of the conjugation donor strain were additionally supplemented with 50 µg ml⁻¹ of 5-aminolevulinic acid (5-ALA).

S. silvestris was routinely grown in LB pH 8.2, adjusted with 20 mM Tris-HCl pH 9, at 30 °C with shaking (120 r.p.m.). To select for successfully conjugated clones carrying the plasmid, LB pH 8.2 medium was supplemented with 2 µg ml⁻¹ erythromycin. Where indicated, S. silvestris was cultivated in yeast extract acetate (YA) medium containing 50 mM Tris-HCl pH 7.8, 0.2% w/v yeast extract, 0.5 mM MgSO₄, 0.01 mM MnSO₄ and 200 mM sodium acetate. The precipitation medium YAC (yeast extract, acetate, calcium) was based on YA medium and additionally contained 100 mM calcium nitrate, unless otherwise stated.

MICP assay

Calcium carbonate precipitation was quantified using a previously established methodology that was shown to accurately account for all calcium ions lost from the soluble phase of the culture [19]. In brief, overnight cultures grown in LB medium were used to inoculate 20 ml YAC media to an initial OD₆₀₀ of 0.02. The cultures were incubated for 1–6 days at 30 °C and shaking at 140 r.p.m. The cultures were then harvested by centrifugation at 400 g for 5 min. To remove bacterial cells from the precipitate, the pellet was repeatedly washed with distilled water until the supernatant remained clear. The precipitate was dried at 60 °C until no further change in weight was detectable. The amount of precipitate was determined by calculating the difference in weight of the centrifuge tubes before the experiment and after drying the precipitate. The concentration of precipitated calcium carbonate in the original culture was calculated from the weight of the precipitate, molecular mass of CaCO₃ and the harvested culture volume.

Plasmid construction

Genomic DNA of S. silvestris was used to amplify genomic regions of interest. All generated plasmids are derivatives of pMAD-oriT [19] and were constructed via Gibson Assembly [27] with minor modifications. A 5× isothermal reaction buffer was prepared containing 500 mM Tris-HCl (pH 7.5), 50 mM MgCl₂, 1 mM dNTPs, 50 mM DTT, 25% (w/v) PEG-8000 and 5 mM NAD. The Gibson Assembly Master Mix (1.2 ml total) was set up with 1× isothermal reaction buffer, 6.4 units T5 exonuclease, 40 units Phusion DNA polymerase and 6.4 units Taq DNA ligase and aliquoted into 15 µl portions. Assembly reactions were set up by combining a master mix aliquot and 300 ng linearized vector with a 1:3 molar ratio of insert DNA in a final volume of 20 µl, followed by incubation at 50 °C for 30 min. Subsequently, 10 µl of the reaction was used to transform competent E. coli.

Linearized vector was generated either by restriction digestion via the EcoRI/SalI sites or by PCR using the primer pair SG1680 and SG1681, followed by a DpnI digest to remove the template DNA. Correct assembly was verified by colony PCR using the primer pair SG1540 and SG1541, followed by sequencing of the insert.

pMSMAD03: Plasmid to delete yloB via double homologous recombination. To amplify the upstream fragment, the primers SG1682 and SG1684 were used. To generate the downstream fragment, the primers SG1683 and SG1685 were used. The final insert fragment was obtained via overlap PCR using the primers SG1686 and SG1687 and then used for Gibson Assembly with linear pMAD-oriT.

pMSMAD05: Plasmid for insertion of mNeonGreen as a C-terminal fusion to yloB via double homologous recombination. To amplify the upstream fragment, the primers SG1694 and SG1695 were used. To obtain the mNeonGreen fragment, the primers SG1696 and SG1697 were used using pSHP1 as the template. To generate the downstream fragment, the primers SG1698 and SG1699 were used. The final insert fragment was obtained via overlap PCR using the primers SG1694 and SG1699 and then used for Gibson Assembly with linear pMAD-oriT.

pMSMAD06: Plasmid used for the reporter assay as promoter-less sfGFP control. To amplify sfGFP, the primer pair SG1796 and SG1787 was used, using pHJS105 as a template. The resulting fragment was then used for Gibson Assembly with linear pMAD-oriT.

pMSMAD08: Plasmid used for the reporter assay, containing P_yloB sfGFP. To amplify the promoter of yloB, the primers SG1790 and SG1795 were used. The resulting fragment was used in a Gibson Assembly reaction together with pMSMAD06 that was linearized using the primers SG1793 and SG1681.

pMSMAD11: Plasmid for complementation of yloB-mNG under its native promoter. The fragment P_yloB yloB was obtained using the primers SG1790 and SG1687 and the genomic DNA of SGB1108 as a template. The resulting fragment was then used for Gibson Assembly with linear pMAD-oriT.

pMSMAD12: Plasmid for complementation of yloB-mNG D328N under its native promoter. The fragment P_yloB yloB-mNG D328N was generated through a two-step PCR strategy using genomic DNA from strain SGB1108 as the template: In the first reaction, primers SG1790 and SG2087 were used; in the second, primers SG2088 and SG1687 were employed. The resulting fragments were then used for Gibson Assembly with linear pMAD-oriT.

pMSMAD13: Plasmid for complementation of yloB-mNG D709A under its native promoter. The fragment P_yloB yloB-mNG D709A generated through a two-step PCR strategy using genomic DNA from strain SGB1108 as the template: In the first reaction, primers SG1790 and SG2089 were used; in the second, primers SG2090 and SG1687 were employed. Genomic DNA from strain SGB1108 served as the template for both PCRs. The resulting fragments were then used for Gibson Assembly with linear pMAD-oriT.

Construction of in-frame DNA deletions and insertions in S. silvestris

For the incorporation of constructed vectors into S. silvestris, conjugation was performed as described [19], with the exception of using E. coli ST18 as the donor strain to facilitate counterselection based on this strain’s auxotrophy for 5-ALA [28]. Overnight cultures of S. silvestris and transformed E. coli ST18 were prepared by inoculating 5 ml LB medium at the respective pH values stated above, with the addition of 50 µg ml⁻¹ 5-ALA for ST18. The following day, for each strain, 20 ml of LB medium was inoculated to an initial OD₆₀₀ of 0.1 and incubated until ST18 reached the exponential phase (OD₆₀₀=0.4–0.6). The bacterial cells were pelleted by centrifugation, and ST18 cells were additionally washed in 1 ml LB to remove residual 5-ALA. Both cell pellets were resuspended and combined in a total volume of 10 ml standard LB medium containing 20 mM MgCl₂. After incubation for 1 h at 30 °C, the cells were again pelleted and resuspended in the small amount of medium remaining after decanting the supernatant. The cell suspension was spotted on an LB plate containing 20 mM MgCl₂ and incubated for 24 h at 30 °C. The next day, the cells were scraped off and resuspended in 1 ml LB pH 8.2, serially diluted to 10⁻², and 3×20 µl per dilution was spotted on an LB plate containing erythromycin selection and incubated at 30 °C for up to 2 days until the appearance of colonies.

Deletion of the yloB gene and integration of the yloB-mNG fusion construct were generated by double homologous recombination as described [19]. Briefly, an overnight culture was prepared in LB medium with erythromycin selection and grown at 30 °C. The next day, 10 ml LB with erythromycin selection was inoculated to a starting OD₆₀₀ of 0.1 and incubated at 30 °C for 2 h, at which time the temperature was increased to 42 °C for 5 h. Serial dilutions were then plated onto LB agar with erythromycin selection and incubated at 42 °C for 24 h. Individual colonies were selected and used to inoculate 5 ml LB pH 8.2 without antibiotics and incubated overnight at 30 °C. The temperature was then increased to 42 °C for 3 h. Serial dilutions were plated onto LB pH 8.2 agar without selection and incubated for 24 h at 42 °C. Colonies were replica‐patched onto erythromycin and antibiotic-free plates. Erythromycin-sensitive clones were screened for gene deletion or mNG insertion by PCR using primers SG1842 and SG1843.

Growth rate analysis

The growth rate was determined in YA medium with increasing concentrations of calcium nitrate (0, 20, 50 and 100 mM). Test tubes of 16 mm diameter containing 3 ml medium were inoculated with an overnight culture to an initial OD₆₀₀ of 0.05 and incubated at 30 °C with shaking (120 r.p.m.). The OD was measured in 30-min intervals for 7 h using an Ultrospec 10 cell density meter (Biochrom). The exponential growth phase was identified, and the growth rate was calculated by fitting the data with the exponential growth equation using GraphPad Prism 10.

Quantification of sporulation, spore brightness and heat resistance

To determine the heat resistance of spores in S. silvestris WT and the ∆yloB strain, 5 ml YA was inoculated with an overnight culture to a starting OD₆₀₀ of 0.05. The cultures were incubated at 30 °C for 24 and 48 h with shaking (120 r.p.m.). The cultures were analysed by phase-contrast microscopy using a Leica DMi8 inverted microscope and a Leica K8 Scientific CMOS camera. For each of the three biological replicates, cells from around seven independent microscopic fields of view were counted (total across all replicates n~1,000). Cells were scored as sporulated if they showed a visible spore that was at least partially phase-bright. The fraction of sporulated cells was determined by dividing the number of sporulated bacteria by that of the total cell count.

To quantitatively assess spore brightness, greyscale images were analysed using ImageJ [29]. For each spore, a straight line was manually drawn perpendicular to the longitudinal axis, approximately through the widest part of the spore. The ‘Plot Profile’ function was then used to extract pixel intensity (grey value) data along this line, yielding a 1D intensity profile across the spore’s width. To enable direct comparison across spores and strains, individual profiles of all cells were manually aligned relative to the midcell position. Per strain, a total of 30 cells of different biological replicates were analysed.

To determine the proportion of heat-resistant spores, serial dilutions were prepared and then pasteurized at 80 °C for 20 min. One hundred microlitres of pasteurized and non-pasteurized dilutions was each plated onto LB pH 8.2 plates and incubated overnight at 30 °C. Colony numbers were enumerated for the dilution giving between 50 and 300 colonies, and the c.f.u. ml⁻¹ was determined for both the pasteurized and non-pasteurized cultures. The proportion of heat-resistant spores was determined by dividing the c.f.u. ml⁻¹ of the pasteurized culture by that of the non-pasteurized culture.

Fluorescence microscopy and membrane staining

Microscopy was performed with cells immobilized on Teflon-coated multi-spot microscope slides (Hendley-Essex) with 1.5% (w/v) agarose in H₂O. Per well, 0.5 µl of the respective cell culture was spotted and covered with a cover slip. If required, membranes were stained with Nile Red (NR) (9-diethylamino-5-benzo[α]phenoxazinone). The staining was conducted by incubation of 100 µl culture with 1 µg ml⁻¹ of NR for 5 min at 30 °C, immediately followed by spotting the cells on the agarose-covered slide and microscopy.

For imaging cells in exponential phase, 3 ml YA medium was inoculated from an overnight culture to a starting OD₆₀₀ of 0.05, grown to an OD₆₀₀ of 0.5 and then spotted for microscopy, or supplemented with 100 mM calcium nitrate for 1 h and then spotted. For cells in the stationary phase, 5 ml YA medium was inoculated from an overnight culture to a starting OD₆₀₀ of 0.05 and grown for either 24 or 48 h as indicated. When analysing stationary cells supplemented with calcium, cells were exposed to 100 mM calcium nitrate for 1 h after 24 h of growth.

For imaging, a Leica DMi8 inverted microscope was used, equipped with a Leica HC PL APO 100×/1.40 oil PH3 objective, CoolLED pE300W light source and a Leica K8 Scientific CMOS camera. Fluorescence of mNeonGreen and msfGFP was excited using the CoolLED pE-300w light source in combination with a GFP filter cube (excitation 470/40 nm, dichroic 495 nm and emission 525/50 nm). For mNeonGreen, illumination intensity was set to 20% with an exposure time of 500 ms, while msfGFP was imaged at 2% intensity and 50-ms exposure time. Fluorescence of NR was excited using the CoolLED pE-300w light source together with a TXR filter cube (excitation 560/40 nm, dichroic 585 nm and emission 630/75 nm) at 60% illumination intensity and 50-ms exposure time. Imaging was performed in biological triplicates. All images were analysed using ImageJ [30], while GFP intensity/µm² was calculated from automatically detected cells using a custom Python script.

Flow cytometry

Flow cytometry was used to quantify the GFP fluorescence intensity in reporter cell populations. The strain SGB1114 was analysed in the exponential and stationary growth phase. To reach the exponential phase, 5 ml of YA medium was inoculated in duplicate from an overnight culture in YA medium to a starting OD₆₀₀ of 0.05 and incubated at 30 °C with shaking (120 r.p.m.). When the cultures reached an OD of ~0.4, the culture was split into two 2.5 ml cultures and one batch was exposed to 100 mM calcium nitrate followed by incubation for another hour. The control culture was left untreated, with incubation continuing the same as the challenged culture.

To prepare stationary phase cells, 5 ml of YA medium was inoculated in duplicate from an overnight culture to a starting OD₆₀₀ of 0.05 and incubated at 30 °C with shaking (120 r.p.m.) for 24 h. Then, each culture was split into two 2.5 ml cultures, and one batch was supplemented with 100 mM calcium nitrate followed by 1 h of additional incubation at 30 °C. The other batch was further cultivated without the addition of calcium nitrate.

Prior to flow cytometry analysis, cultures were diluted at a 1:5 ratio with 0.85% (w/v) sodium chloride solution. The measurement was performed using the Attune NxT Flow cytometer (Invitrogen) equipped with a 488-nm laser with a flow rate of 100 µl min⁻¹. Per sample, 50,000 single cells were measured. The GFP signal was detected in BL1 (530/30 nm). To set the threshold to detect GFP-positive cells above any intrinsic fluorescence of S. silvestris cells, a negative control strain was used which carried plasmid pMSMAD06, i.e. containing the gene encoding sfGFP but without a promoter.

Bioinformatic analyses

The genomic information for S. silvestris was obtained from GenBank (accession number JBFEAN000000000), while the protein sequence of YloB from B. subtilis was retrieved from UniProt (accession number O34431). Reciprocal blast was performed using blastp algorithm [31] against the B. subtilis 168 reference genome (RefSeq GCF_000009045.1). Protein and transmembrane domains were determined by HMMER scan [32], and 2D topology was analysed using PROTTER [33] using default parameters. A pairwise alignment of YloB of B. subtilis and S. silvestris was conducted with ClustalW [34], and conserved features were manually marked within the sequence based on the work of Raeymaekers et al. [26]. Protein structures were predicted with AlphaFold 3 using default settings [35]. Chimaera X [36] was used to visualize structures and generate overlays between the predicted structures of S. silvestris and B. subtilis YloB proteins.

Results

In silico analysis of YloB

To investigate the putative function of the candidate calcium-binding ATPase in S. silvestris, encoded in locus AB1K09_00890, we performed a reciprocal blast analysis and found that the direct homologue of the protein was YloB in B. subtilis, sharing a percentage identity of ~43% and a similarity of 64% (Fig. S1). Therefore, we named the S. silvestris protein YloB in accordance with its homologue.

A predicted structural model generated with AlphaFold 3 [35] revealed that YloB of S. silvestris consists of three headpiece domains and ten transmembrane helices and shows the classical five domains of an ATPase: The actuator domain, the nucleotide binding domain responsible for ATP binding and the phosphatase domain are all predicted to be cytoplasmic (Fig. 2a). The first six transmembrane helices harbour the calcium-binding site, composed of residues from two binding motifs, and represent the transport domain, while the remaining four helices spanning the membrane form the support domain (Fig. 1a) [37]. An overlay with the predicted structure of YloB of B. subtilis showed a high degree of similarity, with only minor differences (Fig. S2B–D). Quantitative superpositioning of core Cα atoms yielded a pruned RMSD of 0.95 Å, indicating strong conservation of the ATPase fold, whereas the RMSD across all 858 Cα pairs was 13.19 Å, indicating some flexible regions of the membrane protein (Fig. S2B).

Fig. 1. YloB is a calcium-binding ATPase, but dispensable for growth and calcium carbonate precipitation. (a) Structure prediction was generated by AlphaFold 3 [35] and visualized using ChimeraX (47). The protein’s N- and C-termini are labelled. The orientation of the protein was determined by 2D topology using PROTTER analysis [33]. The three cytoplasmic headpiece domains were identified as the actuator domain (purple), the nucleotide binding domain (dark blue) and the phosphorylation domain (blue), while the transport (orange) and support (red) domains span the membrane [37]. (b) Growth rate during the exponential phase of the indicated strains grown at different calcium nitrate concentrations. (c) Precipitated calcium carbonate of indicated strains. All cultures were grown in YAC medium for the indicated times, when the calcium carbonate precipitate was harvested and quantified. The horizontal dashed line visualizes the theoretical maximum of calcium carbonate precipitation based on 100 mM available calcium nitrate. For panels (b) and (c), the bars depict the mean, error bars are the sd of the mean and individual data points of independent triplicate experiments are shown as black symbols. No significant (ns) differences between the WT and ΔyloB strains were observed in a Student’s t-test.

Next, we conducted a pairwise sequence alignment between the two homologues to check for the presence of the conserved residues known to be important for P-type ATPase function [26,37] (Fig. S1). This revealed that the residues involved in ATP binding as well as the phosphorylation site, D328 in S. silvestris YloB, were conserved between the proteins. In addition, the residues forming the second calcium-binding motif were also identical, with the motifs VAxIxE present in transmembrane helix M4 and NxxxD in helix M6 (V279 and N727 in B. subtilis; V281 and N705 in S. silvestris). However, two residues of the first calcium-binding motif differed. This motif, composed of residues from transmembrane helices M5, M6 and M8, corresponds in B. subtilis to N699/E702, T730/D731 and Q809, while in S. silvestris, the equivalent positions are N677/A680, T708/D709 and R785. These two substitutions in S. silvestris likely result in a net loss of negative charge at the first calcium-binding motif (Fig. S1), the consequences of which are unclear. To conclude, structural similarities and presence of the relevant core motifs for calcium binding and ATP hydrolysis indicate that YloB of S. silvestris likely is a functional calcium-binding ATPase, and we further investigate the roles of the phosphorylation site D328 and the Ca²⁺-binding residue D709 later in this study.

Deletion of yloB does not affect calcium sensitivity of S. silvestris

To investigate the physiological role of YloB, we deleted the gene by double homologous recombination using a temperature-sensitive pMAD-oriT-derived plasmid [19]. We readily obtained the deletion mutant (∆yloB), suggesting that the gene is not essential.

To determine if YloB is involved in calcium homeostasis in S. silvestris, we determined the growth rate of the WT and ∆yloB mutant upon calcium stress, supplementing the growth medium with 0, 20, 50 and 100 mM calcium nitrate. Without any calcium stress, WT cells showed a growth rate of 0.70 h⁻¹. We observed that even the addition of 20 mM calcium nitrate led to a significant decrease in its growth rate to 0.33 h⁻¹, with a further reduction to 0.20 h⁻¹ when cultivated with 50 mM calcium nitrate (Fis 1b and S3). However, increasing the calcium nitrate concentration from 50 to 100 mM did not result in a further decline. In all tested conditions, ∆yloB showed no significant difference to WT growth behaviour (Figs 1b and S3).

YloB is dispensable for biomineralization

In previous studies, it was hypothesized that YloB might be involved in the MICP process in B. subtilis, and it was shown that yloB expression was upregulated in MICP conditions in S. silvestris [13,19]. To determine whether YloB plays an active role in MICP in S. silvestris, WT and ∆yloB cells were cultivated in YAC precipitation medium (100 mM calcium nitrate). Quantification of mineralized calcium carbonate after 1–6 days showed that the amount of mineral formed increased steadily over time (Fig. 1c). After 6 days, 90% of the available soluble calcium had precipitated as mineral in both WT and ∆yloB cultures, with no statistically significant differences observed between strains (Fig. 1c). This indicated that YloB was not required for biomineralization by S. silvestris and therefore likely is not responsible for actively providing Ca-ions for mineral nucleation.

The putative calcium-transporting ATPase YloB affects sporulation in S. silvestris

As YloB in B. subtilis was shown to be important for the dehydration of endospores [26], we analysed sporulation of S. silvestris WT and ∆yloB cells. For this, we imaged the cells by phase-contrast microscopy after 24 and 48 h of growth in YA medium, a minimal medium optimized for S. silvestris growth under MICP-relevant conditions [19]. Preliminary experiments had shown that this medium supported a detectable level of sporulation of this bacterium, while the Difco Sporulation Medium commonly used for B. subtilis did not. Using phase-contrast microscopy, around 3% of both WT and ∆yloB cells showed a visibly developing spore after 24 h (Fig. S4A), which increased to ~6 % after 48 h with no difference observed between the two strains (Fig. 2a). A striking difference was observed in the spore morphology of the two strains, with the WT forming larger spores with stronger phase-brightness in comparison to ∆yloB (Figs 2b and S4E), suggesting that ∆yloB spores might not fully mature. To test whether this morphological difference resulted in a defect of heat resistance, we determined the proportion of viable cells (c.f.u. ml⁻¹) remaining after pasteurization of the cultures. In the WT, 0.9 % of all cells had formed a heat-resistant spore after 48 h (Fig. 2c), showing that the vast majority of those spores visible by phase-contrast microscopy had matured as expected. In contrast, ∆yloB showed an almost 3-log reduction compared to WT, with only 0.001% of all cells having formed a heat-resistant spore (Fig. 2c), even though 6% of cells contained a phase-contrast visible spore (Fig. 2a).

Fig. 2. Cells lacking YloB show a severe sporulation defect. Cells of the WT, ΔyloB strain and the deletion strain complemented with a plasmid to produce YloB-mNG (YloB⁺; amino acid substitutions as indicated) were analysed for spore formation. (a) Proportion of total spores of indicated strains determined by phase-contrast microscopy after 48 h of growth in YA medium. The fraction of sporulated cells was divided by the total cell count. (b) Top: Representative phase-contrast images of sporulated cells of the indicated strains. Bottom: Spore brightness was quantified by plotting grey value intensity profiles across the width of the cell, n=30. Symbols indicate mean grey values, and dotted lines show sd for each strain. (c) Heat resistance of the spores. The ratio of heat-resistant spores was determined by dividing the viable count (c.f.u. ml⁻¹) of the pasteurized culture by that of an untreated sample of the same culture. For each strain, the 100% baseline corresponds to the mean viable count across three biological replicates: WT 2.8×10⁸ c.f.u. ml⁻¹, ΔyloB 2.1×10⁸ c.f.u. ml⁻¹, ΔyloB/YloB⁺ 3.9×10⁸ c.f.u. ml⁻¹, ΔyloB/YloB⁺ D328N 1.5×10⁸ c.f.u. ml⁻¹ and ΔyloB/YloB⁺ D709A 1.4×10⁸ c.f.u. ml⁻¹. In panels (a) and (c), the mean (bar), the sd (error bar) and individual data points (symbols) are shown for biological triplicates. One-way ANOVA was conducted to test for significant differences. ns=not significant, P>0.01 (**) and 0.001 (***). Scale bars 2 µm.

To ensure that the difference in spore morphology and loss of heat resistance was indeed due to the absence of YloB, the ∆yloB strain was complemented by plasmid-borne expression of an intact copy of a C-terminal yloB-mNeonGreen (mNG) fusion under control of its native promoter. The resulting strain yloB/YloB-mNG⁺ was analysed microscopically and showed a spore morphology like WT cells. In addition, the fluorescently tagged protein was detectable in the cells, showing that the protein was successfully produced (Figs 2b, c and S4D). The proportion of heat-resistant spores was indistinguishable from the WT; thus, the YloB-mNG protein was deemed functional. (Fig. 2c). Importantly, spore morphology and yield of heat-resistant spores were indistinguishable from those obtained when the ∆yloB strain was complemented with a plasmid containing the yloB gene without the mNG fusion partner (Fig. S5), showing that the mNG-tag did not impair YloB function. Curiously, the fraction of visibly sporulating cells in the population was significantly increased in the complemented strain (Fig. 2a), likely caused by the additional copies of the gene on the plasmid construct. Nevertheless, as the proportion of heat-resistant spores was not altered compared to the WT strain, the presence of multiple yloB copies did not appear to significantly affect final sporulation outcome of the population. Taken together, the complementation experiment showed that the sporulation defects observed in the deletion strain were specifically due to loss of yloB, strongly supporting a role of the putative Ca-ATPase in spore maturation of S. silvestris.

To further investigate which functional features of YloB were required for proper spore maturation, we also complemented the ∆yloB strain with two mutated variants of YloB-mNG: a substitution in the phosphorylation site, D328N, predicted to abolish ATPase activity, and a substitution in the calcium-binding site, D709A, predicted to prevent Ca²⁺ coordination. Both fusion proteins were readily detectable by fluorescence microscopy and with localization to the forespore membrane, as was seen with the WT construct (Fig. S4D), showing that the substitutions did not negatively affect protein production. Importantly, neither variant restored spore morphology. Instead, both strains predominantly formed phase-dark spores resembling those of the ∆yloB mutant and exhibited a low proportion of heat-resistant spores similar to the deletion strain (Fig. 2b, c). Both mutant constructs also led to the same unexplained increase in sporulating cells in the population (Fig. 2a), suggesting that this effect was not linked to YloB activity. Together, these findings demonstrate that residues D328 and D709 are essential for YloB function and strongly support that YloB functions as a Ca²⁺-dependent P-type ATPase during spore maturation, consistent with the in silico analyses presented above.

YloB is localized in the spore membrane of S. silvestris

Next, we aimed to study the localization of YloB in vegetative cells and during sporulation. To this end, we generated an mNG translational fusion to the C-terminus of YloB. The strain was generated by integrating the fusion into the native gene to circumvent any artefacts from altered gene copy number or changes in gene expression.

The mNG fusion strain was next examined by fluorescence microscopy in both exponential and stationary growth phases. No fluorescent signal was detected in exponentially growing cells. For cells in the stationary phase, most cells did not show a fluorescent signal, but in cells containing a spore, a clear fluorescent signal was seen surrounding the spore (Fig. 3a, b), suggesting that YloB was only produced in sporulating cells and located in the membrane environment around the spore. Consistent with this, fluorescence was also observed in cells in the earlier phases of sporulation, i.e. engulfment of the forespore. These cells showed a bulge at one cell pole in phase-contrast microscopy, without a mature, phase-bright spore being visible. In these cells, the YloB-mNG fusion was clearly localized to the region where the mother cell membrane migrates around the forespore, with some cells showing a small fluorescent crescent at the division septum (likely early phase of engulfment), while others showed fluorescence around the whole forespore (late engulfment) (Fig. 3a). Interestingly, no fluorescent signal was observed in the cell membrane away from the division septum, suggesting that YloB possesses an as yet unidentified mechanism to specifically localize to the membrane region engulfing the forespore. We had previously shown that exposure of stationary phase cells of S. silvestris to 100 mM calcium nitrate resulted in a six-fold increase in gene expression of yloB determined by RT-qPCR [19]. To further explore this, we supplemented both exponential and stationary phase cells with 100 mM calcium nitrate for 1 h and examined whether this treatment altered the overall fluorescence intensity or localization of YloB-mNG. However, no apparent differences were observed compared to untreated controls (Fig. 3b, c).

Fig. 3. YloB exclusively localizes to the spore membrane. Fluorescence microscopic analyses of yloB::yloB-mNeonGreen cells, encoded in the native site. Phase contrast (Ph), mNeonGreen fluorescence (mNG), membrane stain via NR and the overlay are depicted. Scale bars 2 µm. (a) Left: localization of YloB-mNG in the spore membrane at different stages of sporulation including vegetative, septum formation with early engulfment, late stage of engulfment of the forespore and dehydrated spore. Right: quantification of NR and mNG signals, n=30. Symbols indicate mean grey values, and dotted lines show sd for each strain. Both signals show a similar pattern, indicating that YloB-mNG localizes to the membrane. (c) Cells imaged in exponential and stationary phases. (c) Cells exposed to 100 mM calcium nitrate for 1 h during the exponential or stationary phase.

P_yloB is active during sporulation

Given our observation that the mNG-tagged protein was only produced in sporulating cells, we next wanted to test whether the previously observed change in gene expression was indeed specifically due to calcium addition or an unspecific effect caused by the chosen stress condition. To test this, we constructed a transcriptional reporter strain, where the native promoter of yloB (P_yloB) was placed in front of the gene encoding monomeric superfolder GFP (msfGFP). The resulting plasmid construct (pMAD-oriT-P_yloB-msfGFP) was introduced into WT S. silvestris by conjugation. Because the standard growth temperature of this bacterium is within the permissive range for replication of the vector, this allowed stable, extrachromosomal maintenance of the construct.

Flow cytometry analysis revealed that the promoter was highly active in the stationary phase cells, with about 82% of the bacterial cells exhibiting a GFP signal (Figs 4a and S6). In contrast, only ~5% of the cell population in the exponential phase were detected as GFP positive. This indicated that the yloB promoter was predominantly active during the stationary phase and almost completely inactive during exponential phase, consistent with our observations for production of the YloB-mNG protein fusion shown above.

Fig. 4. P_yloB shows growth phase-dependent activation and is repressed by a high calcium level in the stationary phase. (a) Cells were cultivated to the exponential (exp) and stationary (stat) phase, and the proportion of GFP-positive cells was determined using flow cytometry, with a negative control strain harbouring promoter-less msfGFP used to set the cut-offs. The mean and the sd for biological triplicates are represented. (b) Quantification of GFP signal intensities. Cells were cultivated to the exponential or stationary growth phase and challenged with or without 100 mM calcium nitrate for 1 h, as indicated. The proportion of GFP-positive cells was quantified by flow cytometry, with GFP-positive cells further divided into those with high, medium or low signal intensity, shown by coloured bar segments. The total proportion of GFP-positive cells is shown in per cent on top. Data represent the mean of three biological replicates. (c) Microscopic visualization of the GFP-reporter strain in the stationary phase at the indicated stages of sporulation. Before microscopy, cell membranes were stained with NR for 5 min. Sporulation stages were assigned based on cell morphology in phase-contrast (Ph) images and with NR staining. GFP fluorescence was then assessed for each stage. (d) Quantification of GFP intensity per square micrometre per cell over different stages of sporulation as shown in (c). Student’s t-test was conducted to test for significant differences, P>0.01 (*) and P>0.005 (***).

Moreover, the flow cytometric data showed a distinct grouping of the GFP-positive cells into three sub-populations, which we used to group the cells into three categories of low, medium or high fluorescence intensity (Fig. 4b). This highlighted further differences between the growth phases. During the exponential phase, the low proportion of cells for which a GFP signal was detectable predominantly fell into the ‘low intensity’ category, supporting the lack of promoter activity during exponential growth. Cells with medium or high fluorescence intensity were almost exclusively detected during the stationary phase and together accounted for the majority of GFP-positive cells in this growth phase, again consistent with the results obtained for the YloB-mNG protein fusion.

With the functionality of the transcriptional reporter established, we next wanted to see whether exposure to calcium elicited changes in P_yloB activity. Following the challenge with 100 mM calcium nitrate for 1 h, exponential phase cells showed a small but statistically significant increase from around 5% to 8% in GFP-positive cells, with most of these cells still falling into the low fluorescence category (Fig. 4b). For stationary phase cells, we observed a marked decrease in overall GFP-positive cells from ~82% in the control to ~39% after calcium challenge. In addition, the distribution of fluorescence intensity shifted from predominantly medium or high in the untreated culture to a mainly low intensity signal for the culture supplemented with calcium (Fig. 4b). This was surprising given we had previously observed elevated yloB mRNA levels during MICP conditions [19]. One explanation might be that the calcium shock caused a reduction in cells initiating sporulation, thus requiring less P_yloB activity. However, indirect effects due to the increased calcium level, e.g. slowed msfGFP protein folding, might also have caused a reduced fluorescence output; thus, definitive conclusions cannot be drawn at this stage. Nevertheless, the data clearly suggest that P_yloB is not specifically activated by the presence of high calcium concentrations. This, in turn, indicates that the observed change in yloB mRNA detected by RT-qPCR in our earlier study [19] was likely an indirect consequence of the chosen stress conditions.

To understand the basis for the different categories of fluorescence intensities among the GFP-positive cells, we returned to fluorescence microscopy to analyse the P_yloB reporter strain on the single cell level. To allow us to correlate promoter activity with the different sporulation stages, the cells were stained with the membrane dye NR. Based on the NR staining pattern, cells were categorized into five morphological stages: (1) vegetative cells (even NR staining outlining the cell), (2) septum formation (an asymmetric division septum is visible), (3) early engulfment (a crescent of NR-stained membrane partially surrounds the forespore), (4) late engulfment (a bright NR signal surrounds the entire forespore) and (5) cells containing a dehydrated spore (the NR signal surrounds a phase-bright spore) (Fig. 4c). Based on these criteria, each cell was assigned to one category, followed by quantitative analysis of GFP fluorescence per square micrometre of the individual cells (Fig. 4d). This analysis revealed no or minimal GFP intensity in vegetative cells and those displaying an asymmetric septum. A striking increase in GFP intensity was observed in cells in the early engulfment phase, suggesting that P_yloB activity is strongly induced at this stage of the sporulation process. Cells in the later sporulation phases, i.e. undergoing late engulfment or containing a dehydrated spore, exhibited a reduced, although still clearly visible fluorescence intensity (Fig. 4c, d). These results indicate that yloB expression is specifically induced in the mother cell upon initiation of forespore engulfment and likely switched off again soon after, explaining the gradually decreasing GFP intensities in cells of the later sporulation stages. Moreover, comparing the microscopic data to the flow cytometric analyses suggests that the three sub-populations with different GFP intensities shown in Fig. 4(b) likely reflect cells in the different sporulation phases as shown in panels (c) and (d), with the high fluorescence sub-population likely including cells in the forespore engulfment phase, while cells in later sporulation phases or only just beginning forespore engulfment would likely fall into the medium or low intensity category. Taken together, the reporter analyses as well as the YloB-mNG protein fusion show a dynamic regulation of yloB expression that is tightly linked to sporulation in the stationary phase cells, consistent with the marked sporulation defect of the yloB-deleted strain.

Discussion

Calcium is a universal signalling ion that plays vital roles not only in eukaryotes but also in bacteria. Hence, cells need to tightly regulate their calcium homeostasis, but the detailed mechanism by which bacteria achieve this is not understood. In addition, active calcium transport was proposed to be an important factor for biomineralization of calcium carbonate, and it is known to be essential for the production of heat-resistant endospores. S. silvestris CGN12 is an environmental isolate from a high-calcium natural site (limestone rock), capable of growth at high calcium concentrations, sporulation and calcite biomineralization [19,38]. It therefore presented an ideal model to gain deeper insights into the role of calcium transport in these processes. Here, we set out to characterize the sole annotated calcium P-type ATPase in this bacterium.

In silico analyses showed that the calcium ATPase of S. silvestris is a homologue of YloB in B. subtilis. While sequence identity (43%) and similarity (63%) were only moderate, structural predictions showed a high similarity between the two proteins (Figs S1 and S2). This is consistent with the fact that P-type ATPases show high variability in the amino acid sequence, while the overall structure is conserved [37]. All of the key motifs for calcium P-type ATPases were identified in YloB of S. silvestris, with the exception of two differing residues in the first calcium-binding motif, with Glu702 and Gln809 of the B. subtilis protein substituted with Ala680 and Arg785 in S. silvestris. A Glu-to-Arg exchange was also observed in the P-type ATPase LMCA1 in Listeria monocytogenes, where it was found to be responsible for a higher pH optimum of the protein [39]. Given that S. silvestris optimally grows at alkaline pH (>pH 8), it is plausible that YloB carries similar substitutions to ensure a higher pH optimum, which presents an interesting direction for future research. From these results, we concluded that YloB shows all features of a functional calcium ATPase.

To determine the physiological role of YloB in S. silvestris, we deleted the encoding gene by double homologous recombination. Since it was hypothesized that YloB could act as a calcium export system in S. silvestris, it was expected that its gene deletion would result in a higher sensitivity to calcium compared to the WT. However, ∆yloB cells exhibited comparable sensitivity to calcium as the WT (Fig. 1b). Similar results have been reported for the yloB deletion mutant in B. subtilis [13], indicating that YloB does not function as a calcium efflux detoxification mechanism. Next, we wondered whether cells lacking YloB could still precipitate calcium into calcite mineral. This hypothesis was supported by previous findings, where increased yloB mRNA levels were found under MICP conditions [19]. In B. subtilis, the deletion of yloB was reported to abolish calcite mineral formation during biofilm formation and thus cause a defect in biofilm formation and architecture [13]. However, in S. silvestris, the ∆yloB strain exhibited a similar calcite precipitation rate and yield as the WT (Fig. 1c), indicating that it plays no role in biomineralization. Moreover, we did not observe increased activity of P_yloB in the transcriptional reporter assays, nor an obviously increased production or altered localization of YloB-mNG in response to increased calcium supplementation. From these data, we conclude that YloB is not the transporter responsible for calcium efflux in S. silvestris.

Instead, the role of YloB may be more specific to calcium handling during sporulation, where precise calcium delivery is critical for spore maturation. Calcium is required as a chelator for DPA, and the accumulation of Ca-DPA causes partial dehydration of the endospore, which gives the spore heat-resistant properties [9]. S. silvestris cells that lacked YloB produced only partially phase-bright spores, suggesting that they are not fully dehydrated (Fig. 2b). While WT and ∆yloB cells exhibited similar sporulation rates in terms of total number of sporulating cells, the proportion of generated heat-resistant spores was dramatically reduced in ∆yloB (Fig. 2a, c). In B. subtilis ∆yloB, phase-dark spores were shown to have a deficiency in calcium accumulation into the spore core, as calcium levels in the forespore were decreased compared to WT spores [40]. It is therefore plausible that YloB in S. silvestris similarly functions to transport calcium into the forespore.

Complementation of the ∆yloB strain with a plasmid-borne WT copy of the gene led to the curious observation that a markedly higher number of cells showed microscopically visible spore formation than in the WT, but there was no difference in the fraction of heat-resistant cells (Fig. 2a, c). This suggests that in the complemented strain, many of the cells showing at least a partially phase-bright spore did not go on to produce a mature spore. Thus, while the complementation construct restored the ∆yloB strain’s ability to form heat-resistant, fully dehydrated spores, it seemed to have an additional effect on the regulatory cascades of sporulation, leading to more cells entering the sporulation process, but not completing full maturation. Given that the same effect was also observed in the single amino acid variants of YloB, even though these variants did not restore heat resistance and thus likely produced an inactive YloB protein (Fig. 2a, c), it appears that it is the presence of additional copies of the gene and/or its promoter, rather than increased YloB protein, which may be responsible for this effect. Future work addressing this phenomenon may be able to shed new light on the regulatory mechanisms of yloB and how it is linked with the sporulation process.

By studying the activity of the yloB promoter as well as the localization of fluorescently tagged YloB-mNG, we could show that YloB has a major role during sporulation. While P_yloB was active in a small minority of vegetative cells, no YloB-mNG signal could be observed here (Figs3a 4a). This is in agreement with observations in B. subtilis, where YloB could only be found in sporulating but not in vegetative cells [26]. Single-cell measurements by microscopy showed that S. silvestris cells that have started forming an asymmetric spore septum strongly activated P_yloB, with declining promoter activities in the later stages of sporulation (Fig. 4c, d). At the same early stage of sporulation, the YloB-mNG protein appears specifically at the asymmetric division septum and then follows the forespore engulfment process (Fig. 3b). Interestingly, despite the high promoter activity and strong fluorescent signal from the mNG-tagged protein, we could not observe the protein elsewhere in the mother cell membrane (Fig. 3a, b). Such a pattern of protein incorporation has been reported for some early-stage spore coat proteins and for SpoVV, a DPA transporter, which are expressed in the mother cell and located in the outer spore membrane [41,42]. The P_yloB-msfGFP reporter gave a clear signal only in the mother cell cytoplasm, not the spore (Fig. 4c), suggesting promoter induction is mother cell-specific and that therefore YloB is produced by the mother cell. While the sigma factor regulating yloB expression remains unclear, analysis of the promoter sequence did not reveal conserved canonical binding sites. In B. subtilis, yloB regulation is similarly elusive, while some involvement of the forespore-specific SigG was speculated [26]. However, our reporter assay shows expression early in the mother cell, consistent with SigE-dependent regulation, even though a conserved SigE motif could not be identified in the promoter region of yloB in S. silvestris. Our phenotypic observations indicate that yloB expression is likely activated by a mother-cell regulatory process, highlighting the need for further work to clarify the regulatory mechanism.

How the protein specifically localizes to the division septum and the membrane region surrounding the forespore is an interesting question. One potential explanation could be curvature-dependent localization. The membrane growing around the forespore is the only region with positive membrane curvature in the cell, and this has been proposed to target sporulation-relevant proteins to their specific site [43]. An alternative mechanism could be localization through interaction with a partner protein in the inner spore membrane. Such a mechanism has been reported for B. subtilis and Clostridioides difficile SpoIIIAH, which is located in the outer forespore membrane, but specifically interacts with SpoIIQ in the inner forespore membrane to form a channel through both membranes, facilitating communication between the two cells [44,46]. While further research will be required to elucidate the localization mechanism for YloB, the latter model might be attractive, given that the Ca²⁺ ions will have to cross both the outer and inner forespore membrane to reach the forespore, while YloB can only span one lipid bilayer.

Taken together, our findings suggest that YloB plays a central role in transporting calcium from the mother cell into the forespore to enable proper Ca-DPA accumulation and the formation of heat-resistant spores, rather than serving as a calcium efflux pump for detoxification or involvement in MICP.

Abbreviations

5-ALA

5-aminolevulinic acid

DPA

dipicolinic acid

LB

lysogeny broth

MICP

microbially induced calcite precipitation

msfGFP

monomeric superfolder GFP

NR

Nile Red

YA

yeast extract acetate