Chaperone protein DnaK-mediated regulation of spore coat assembly in Bacillus subtilis

Summary

Bacterial spores pose a significant challenge to the food industry due to their ability to germinate from dormancy and their extreme resistance. Spores are encased by a multilayered proteinaceous structure termed the spore coat. Previous studies on spore coat assembly have predominantly focused on the interactions among coat proteins, while the involvement of broader protein quality control systems remains poorly understood. Here, we demonstrate that the molecular chaperone DnaK is essential for the proper assembly of the inner spore coat in Bacillus subtilis. Deletion of dnaK leads to mislocalization of key morphogenetic protein SafA, reduced accumulation of full-length SafA, altered transcriptional dynamics of spoIIID and sigK, and ultimately impaired coat integrity and function. Our findings reveal that DnaK facilitates spore coat assembly through both transcriptional and post-translational mechanisms, uncovering a previously unrecognized regulatory layer in spore coat morphogenesis.

Graphical abstract

Highlights

• •Molecular chaperone DnaK influences sporulation in Bacillus subtilis
• •DnaK deletion alters resistance properties of mature spores
• •DnaK contributes to proper spore coat assembly in Bacillus subtilis
• •DnaK deletion reduces full-length SafA accumulation

Biological sciences; Microbiology

Introduction

Spores formed by bacteria of the Bacillus and Clostridium genera are major agents of spoilage and foodborne disease, posing a significant challenge to the food industry. This is largely due to their exceptional resistance to a wide range of inactivation treatments.^1,2 Moreover, once favorable conditions are restored, these spores can rapidly germinate into vegetative cells.²

Spore resistance largely stems from their complex multilayered architecture, which includes a dehydrated core containing DNA-protective small acid-soluble proteins (SASPs), surrounded by an inner membrane (IM), a peptidoglycan-like cortex and a multi-layered proteinaceous coat. The low water content of the spore core is a key determinant of resistance to various environmental stresses, as it contributes to macromolecular stabilization and limits chemical reactivity. Expansion of the spore cortex against the rigid coat has been proposed as a mechanism contributing to core water extrusion.³ In addition, the spore IM is maintained in a relatively stable and rigid state, and its low permeability and reduced fluidity play important roles in both spore germination and resistance.^3,4 In Bacillus subtilis, the model spore-forming organism, the outermost structure is the spore coat. The coat is composed primarily of spore-specific proteins, many of which are cross-linked to form a relatively rigid structure, and it plays important roles in spore resistance by serving as a critical protective barrier against physical and chemical stresses.⁵

Numerous studies have shown that the spore coat is composed of more than 80 proteins that are precisely regulated in both time and space during sporulation.^6,7,8 These proteins form four concentric layers: the basement layer, inner coat, outer coat, and crust. While the composition of the coat has been increasingly well characterized, many molecular events underlying the assembly process remain unclear.^5,9,10 Proper assembly of the coat requires a complex network of protein interactions involving transcriptional, translational, and post-translational regulatory mechanisms.

Expression of coat genes is orchestrated by a hierarchical mother cell-specific transcriptional cascade involving two RNA polymerase sigma factors and three positively and/or negatively acting transcription factors, namely, σ^E, SpoIIID, GerR, σ^K, and GerE, which sequentially regulate coat gene expression.^11,12 In addition to transcriptional regulation, coat assembly is also governed by morphogenetic proteins. Assembly proceeds in two main stages: initial targeting and encasement.^13,14 Initial targeting to the outer forespore membrane depends on the coat anchor protein SpoIVA, which localizes in coordination with SpoVM and subsequently recruits the encasement protein SpoVID. At the level of individual coat layers, inner layer assembly is dependent on SafA, outer layer assembly is dependent on CotE, and crust formation requires CotX, CotY, and CotZ on top of the outer coat. The recruitment of SafA and CotE is dependent on SpoIVA and SpoVID, whereas the recruitment of CotX, CotY, and CotZ is dependent on CotE.^15,16,17 The organization of the coat interaction network is therefore highly hierarchical, and these morphogenetic proteins are essential for ensuring the proper spatial organization and full encapsulation of the spore by the coat layers. Additionally, self-assembly mechanisms and post-translational modifications also contribute to coat maturation.⁸ For example, the phosphorylation of CotG and CotB by CotH has been shown to be essential for proper spore coat organization.¹⁸ These findings demonstrate that the assembly of the spore coat is a precise process that depends on the high fidelity of protein synthesis and folding. Although protein quality control systems have been well characterized in vegetative cells, their roles during sporulation remain largely unexplored. We hypothesize that the precise regulation of coat assembly during sporulation is closely linked to protein quality control mechanisms.

In prokaryotes, an evolutionarily conserved regulatory network, comprising molecular chaperones and protease systems, maintains a dynamic equilibrium between protein synthesis, folding, post-translational modification, transport, and degradation.^19,20 Properly folded, high-quality proteins are essential for effective cellular function. DnaK is a well-characterized stress protein in Bacillus subtilis. Several recent studies have found that the core chaperone system of B. subtilis, DnaK, GroEL, and trigger factors, performs some unique protein quality control functions in addition to a conserved response to stress.^20,21,22 The deletion of dnaK alone does not cause significant growth or viability defects in B. subtilis under standard conditions.²³ However, the deletion of both dnaK and tig in B. subtilis affects cell wall integrity.²⁴ Moreover, high-throughput genetic screens during sporulation have identified dnaK as a gene potentially involved in spore development.²⁵ Together, these findings suggest that DnaK may fulfill previously underappreciated roles during spore development. However, its physiological relevance under non-stress conditions, particularly in regulating spore formation and resistance, remains incompletely understood.

In this study, we identify a previously unrecognized role for molecular chaperone DnaK in regulating spore coat protein assembly. Through microscopic observation and fluorescence localization analysis, we found that deletion of dnaK leads to aberrant localization of inner coat proteins, thereby compromising the barrier function of the spore coat. Protein analysis further showed a marked reduction in the full-length (FL) form of SafA, suggesting defects in its accumulation. Further transcriptional analysis revealed that the deletion of dnaK affects transcriptional regulation as well as the post-translational control of previously reported transcriptional regulators involved in spore coat assembly. This work reveals a new regulatory layer in spore coat morphogenesis and expands our understanding of how molecular chaperones participate in spore protein quality control.

Results

Deletion of dnaK increases sensitivity to lysozyme and hypochlorite stress

To investigate the contribution of DnaK to spore resistance, we first constructed dnaK mutant strains and examined their survival under various treatment conditions. In a screen for resistance to chemical agents, dnaK mutant spores displayed markedly increased sensitivity to lysozyme and sodium hypochlorite. This phenotype was fully rescued in the complemented strain ΔdnaK-lacA::dnaK, restoring resistance to wild-type levels (Figures 1A and 1B). Interestingly, in contrast to the heightened sensitivity to chemical agents, dnaK mutant spores exhibited increased resistance under high-pressure thermal (HPT) conditions (200 MPa/80°C), which is highly effective for spore inactivation.²⁶ Importantly, spores from the hrcA mutant, which constitutively overexpresses DnaK,²⁷ showed no significant difference in HPT treatment compared with the wild type (Figure 1C). This indicates that the enhanced HPT resistance in the dnaK mutant is not simply due to the absence of DnaK protein per se, but likely stems from downstream consequences of its deletion. By contrast, dnaK deletion had no effect on wet heat resistance, and hrcA mutants likewise showed no significant change, consistent with previous findings (Figure 1D).²⁷

Figure 1.

Analysis of spore resistance under diverse conditions

(A and B) Chemical resistance measures were (A) lysozyme resistance and (B) sodium hypochlorite resistance. Each percentage is the mean of results from three replicate experiments, each performed with a different prepared spore suspension.

(C) Measurement of resistance to combined pressure and thermal treatment at 200 MPa and 80°C.

(D) Different mutant spores were incubated at 80°C, 85°C, and 90°C for 30 min to calculate their resistance to wet heat.

Data are presented as mean ± SD from three independent biological replicates. Statistical significance between two strains was determined using an unpaired two-tailed Student’s t test. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001; ns, not significant.

The spore coat acts as a protective barrier, preventing enzymes and oxidants (e.g., lysozyme and sodium hypochlorite) from accessing the cortex layer.^5,28 These findings indicate that the absence of DnaK specifically compromises spore resistance to lysozyme and hypochlorite, suggesting a defect in the protective barrier function of the spore coat. However, the absence of a significant change in wet heat resistance in the mutant suggests that the core properties and relatively rigid IM are not substantially affected.

DnaK affects spore physiological functions

We evaluated the effect of dnaK deletion on sporulation. Previous studies have reported different observations regarding the effect of dnaK deletion on sporulation efficiency, which may be attributable to differences in sporulation conditions.^24,29 Our data showed no shift in the timing of sporulation curve transitions, but sporulation efficiency differed between strains, suggesting that there is no effect on the process of sporulation but a possible effect on the sporulation efficiency (Figure 2A). Two sporulation conditions with different oxygen contents were designed.³⁰ The dnaK mutant exhibited slightly reduced sporulation efficiency compared to the wild type under oxygen-rich conditions, but still produced a large number of spores (Figure S1).

Figure 2.

Analysis of the physiological functions of dnaK mutant spores

(A) Growth curves in LB medium and sporulation curves in DSM medium.

(B) Relative spore hydrophobicity determined using the BATH assay.

(C–F) The germination was induced using L-Ala (C and D) or dodecylamine (E and F) and measured as the percentage of loss of the OD₆₀₀ (C), percentage of DPA release (D). (F) Is monitored by time-lapse microscopy. Scale bars: 2 μm. Data are shown as mean ± SD from three biological replicates. Statistical significance between two strains was determined using an unpaired two-tailed Student’s t test. ∗∗∗p < 0.001.

Spore surface hydrophobicity, which reflects the functionality of the outer structural layers, was assessed using the BATH (bacterial adhesion to hydrocarbons) assay. The results showed that around 40% of the dnaK spores partitioned into the hexadecane phase compared to only 10% of wild-type spores (Figure 2B). As expected, the surface properties of dnaK spores were significantly altered.

Chaperone proteins have been implicated in early germination events,³¹ and the Tig has been shown to modulate spore translation activity, thereby affecting germination.³² As dodecylamine (DDA) must cross the spore’s outer layers to activate IM channels,³³ this faster response likely reflects increased permeability (Figures 2E and 2F). Further testing was conducted on the germination of dnaK spores under different temperature conditions. Interestingly, at 37°C, only slight differences were observed between wild-type and dnaK mutant spores (Figures 2C and 2D). However, at higher temperatures, dnaK spores showed significantly reduced germination efficiency and failed to germinate effectively (Figure S2). At 60 °C, AGFK failed to trigger germination, suggesting that germinant receptors such as GerK or GerB may be more heat-sensitive (Figure S2C). These germination results are consistent with the phenotype of coat-defective spores.

Together, these results indicate that the absence of DnaK can affect the characteristics of the spore coat.

DnaK deletion leads to structural defects in the spore coat

To directly examine changes in spore coat structure, purified spores from different strains were analyzed by thin-section transmission electron microscopy (TEM). Wild-type spores showed a dense, multilayered outer coat and a defined, lamellar inner coat (Figures 3A and 3D). In contrast, dnaK mutant spores exhibited distinct morphological abnormalities. The spore coat appeared less electron-dense, loosely packed, and showed a granular texture (Figures 3B and 3E). Notably, the boundary between the inner and outer layers was often indistinct, and the lamellar inner coat was completely absent. From the photos, the coat was also less electron-dense than that of the wild type. Furthermore, unlike spoVID mutant spores, which display severe incomplete coat encasement, most dnaK spores retained a complete coat structure (Figures 3C and 3F). Additional representative TEM images are provided in Figure S3, further illustrating the indistinct layering of the spore coat in dnaK mutant spores. Unlike spoVID mutant spores, dnaK spores retain a continuous and overall intact coat structure. Despite this, measurements from over 15 micrographs revealed no significant thinning of the coat compared to wild-type spores.

Figure 3.

Thin-section electron microscopic analysis of spores

Spores from wild-type (WT) strain B. subtilis PY79 (A and D), ΔdnaK (B and E), and ΔspoVID (C and F) were fixed overnight at 4°C in 2.5% glutaraldehyde, washed, and immediately fixed before further processing. In (A)–(C), a section through the short axis of the spore is shown. In (D)–(F), the cross section is through the long axis. The cortex (Cx), inner coat (IC), and outer coat (OC) are indicated. The scale bars are on the lower right of images.

Since TEM could not resolve internal structures such as the cortex or IM, we assessed potential defects using prolonged staining with high concentrations of propidium iodide (PI) and SYTO 16. PI binds DNA and emits red fluorescence only when the IM is compromised, as it cannot penetrate intact spore IM.¹⁴ In addition, SYTO 16 can penetrate the intact spore IM if the spore cortex is hydrolyzed or disrupted.¹⁵ The dye did not enter the core of the dnaK spore (Figure S4A), which can serve as an indirect indicator of cortex and membrane integrity and permeability. Furthermore, we measured the core dipicolinic acid (DPA) content (Figure S4B) and found no significant difference between the two strains, indicating that the degree of core mineralization remains unchanged.

These morphological detection results further indicate that the spore coat structure has undergone significant changes.

DnaK is involved in sporulation

Previous studies suggested that heat shock proteins may alter spore structure during sporulation.³⁴ We next investigated the dynamic localization of DnaK during sporulation by constructing a DnaK-YFP fusion strain (Figure 4). In vegetative cells (T2.5), DnaK-YFP exhibited a diffuse distribution throughout the cytoplasm, consistent with previous observations.³⁵ According to the growth curve, spore formation began approximately 5 h after induction (Figure 2A). At this stage (T5), DnaK-YFP started to accumulate at one pole of the cell. As cells progressed through the engulfment sequence, DnaK-YFP became enriched at both spore poles (T11 and T14, red arrows). Subsequently, as the forespore became dehydrated and phase-bright, DnaK fluorescence in the mother cell decreased or disappeared, whereas the fluorescence intensity within the spore increased (yellow arrows). In mature spores (T24), the fluorescence appeared concentrated at the core region of the spore. These observations indicate that DnaK localization changes dynamically throughout sporulation and that DnaK remains present in mature spores. However, the relationship between the loss of DnaK fluorescence in the mother cell and the strong accumulation of fluorescence within the forespore remains unclear. It is possible that DnaK is newly synthesized within the forespore while being degraded in the mother cell, or that pre-existing DnaK protein is transferred from the mother cell into the forespore. Further investigation is required to distinguish between these possibilities.

Figure 4.

Phase-contrast and fluorescence microscopy analysis of strains carrying the fusion protein DnaK-YFP at different times of sporulation

Time is calculated from inoculation into the sporulation medium. T2.5, cell growth; T5, sporulation start; T11, spore coat formation; T14, sporulation later stage; and T24, spore maturation. Red arrows indicate preparation for sporulation, yellow arrows indicate forming spores, and white arrows indicate mature spores. Scale bars: 50 μm.

DnaK deficiency does not affect the expression profile of spore coat proteins

To further investigate the composition of the spore coat in different strains, coat proteins were extracted using SDS-DTT boiling and separated by 15% SDS-PAGE as previously described [29]. This method solubilizes a subset of coat proteins, while some insoluble components may remain undetected. As shown in Figure 5A, no major differences were observed in the protein banding patterns among the strains. Mass spectrometry analysis of protein extracts from wild-type, ΔdnaK, and ΔhrcA strains revealed highly similar spectral profiles overall. We extracted the MS2 spectra of two characteristic proteins, CotB and CotG, for further comparison (Figures 5B and S5). The mass spectra and identified modification sites were consistent across all strains. A total of 45 known spore coat proteins were detected in the wild-type strain (Table S2). Peptide-level comparison confirmed that all of these proteins were also present in the mutant strains, indicating that the spore coat proteome remained unaltered. Phosphorylation mediated by CotH has been recognized as a key determinant for proper outer coat assembly. Analysis of phosphorylation sites showed extensive phosphorylation of CotB and CotG in both mutant strains, indicating that CotH-dependent phosphorylation remained functional (Table S3). All phosphorylation sites identified in wild-type coat proteins were retained in the mutants, with no observable loss. Notably, a novel phosphorylation site was identified on the YtxO protein in all strains that has not been reported previously.

Figure 5.

Detection of spore coat proteins

(A) 15% SDS-PAGE of the proteins extracted from the coat of WT and ΔdnaK spores. Lane M is 100 kDa protein marker.

(B) Extracted MS2 spectrum of phosphorylated CotG peptides from three samples. Phosphorylation sites are indicated in red within the peptide sequence. b- and y-ions are labeled in green and orange, respectively, confirming peptide sequence assignment and phosphorylation localization.

Although minor variations in mass spectrometry detection may arise due to methodological sensitivity, every spore coat protein identified in the wild type was also detected in the dnaK and hrcA strains. Together, these results indicate that the absence of DnaK does not cause significant loss of spore coat components or affect their phosphorylation status.

DnaK deletion affects the assembly of the spore coat protein

Since no substantial loss of coat proteins was observed in dnaK spores, yet structural disorder was evident, we hypothesized that DnaK deletion impairs spore coat assembly. To test this, we examined the localization of key morphogenetic proteins via fluorescence microscopy (Figure 6). CotZ, a principal morphogenetic protein for crust formation,³⁶ and CotE, essential for outer coat assembly, were fused to GFP and visualized in both wild-type and dnaK mutant spores. Both spores exhibited normal and similar localization, fully encircling the spore surface without detectable abnormalities (Figure 6A).¹³ However, the morphogenetic protein SafA, which directs the assembly of the inner spore coat,^37,38,39 exhibited abnormal localization in mutant strains. In the wild-type spores, SafA-GFP fluorescence was evenly distributed and formed a smooth encasement around the spore, whereas in the mutant spores, the fluorescence appeared aggregated, with some signals forming punctate foci on one side or pole of the spore (Figure 6A). To further investigate the effect of DnaK on SafA localization dynamics, we examined SafA-GFP fluorescence in wild-type and ΔdnaK during sporulation (Figure 6B). In the wild-type strain, SafA-GFP displayed a well-defined localization pattern, initially appearing at the mother cell-forespore interface around 7 h after the onset of sporulation. As sporulation progressed (9–15 h), SafA gradually encased the forespore, forming a continuous fluorescent ring consistent with its role in inner coat assembly. In contrast, ΔdnaK cells exhibited strikingly abnormal SafA-GFP localization. Instead of forming a uniform encasement, fluorescence signals appeared fragmented, aggregated, or restricted to one side of the developing spore. SafA appeared to wander irregularly within the mother cell, as if unable to recognize its proper assembly site. To quantify these differences, SafA-GFP localization patterns were classified into four categories (caps, ring, aberrant, and cytoplasmic) based on fluorescence distribution, and their proportions were analyzed over time (Figure 6B). These results indicate that the absence of DnaK leads to disorganized SafA localization, ultimately resulting in defective assembly of the inner spore coat.

Figure 6.

Fluorescence localization and protein expression detection of morphogenetic proteins

(A) Fluorescence microscopy analysis of spores carrying the fusion protein. The presence (+) or absence (−) of DnaK is indicated. Localization of morphogenetic proteins of B. subtilis representing different spore coat layers is shown. Scale bars: 1 μm.

(B) Subcellular localization of SafA during sporulation in B. subtilis. Fluorescence microscopy analysis of SafA-GFP localization in WT and ΔdnaK cells during sporulation. Samples were collected at indicated time points (7, 9, 11, 13, and 15 h) after inoculation into sporulation medium. Scale bars: 1 μm. For each image, 100 cells were scored and classified into four categories based on fluorescence distribution: caps (polar signal), ring (complete ring around the forespore), aberrant (irregular or asymmetric distribution), and cytoplasmic (diffuse signal in the mother cell cytoplasm without forespore association).

(C) Quantitative analysis of fluorescence intensity of morphogenetic proteins by flow cytometry. Histograms represent fluorescence intensity distributions of the GFP-fused proteins in wild-type and ΔdnaK strains. The mean fluorescence intensity was further quantified from microscopy images using ImageJ; at least five independent fields were analyzed per strain. ∗p < 0.05.

(D) Western blot analysis of SafA-GFP in wild-type and ΔdnaK strains. Coat proteins were extracted from purified spores and probed using anti-GFP antibody. Equal protein loading was confirmed.

We measured fluorescence intensity using flow cytometry and fluorescence microscopy to evaluate the expression levels of morphogenetic proteins. In the mutant strains, a slight but statistically significant decrease in SafA fluorescence was observed, while SpoVID showed a minor reduction in intensity that did not reach statistical significance (Figure 6C). Therefore, consistent with the above interpretation, the observed mislocalization is likely not a consequence of insufficient protein production, but rather results from disruptions in the assembly and localization processes.

To further examine inner coat organization, we assessed CotD, an inner coat protein whose localization depends on SafA.⁴⁰ In the dnaK mutant, CotD exhibits a punctate aggregation pattern resembling that of SafA (Figure 7A). Since early coat proteins typically form a cap at the mother cell proximal (MCP) pole,³⁸ the punctate aggregation pattern likely reflects the failure of SafA to recognize its correct localization, leading to its aggregation at the cell poles due to defects in subsequent encasement steps.

Figure 7.

Location and expression level of SafA-dependent protein CotD

(A) Analysis of CotD localization. The fluorescence photographs are partially enlarged by 5 times. Scale bars: 5 μm.

(B) Flow cytometry analysis of the fluorescence intensity of CotD in different strains.

To determine the cause of SafA mislocalization, we analyzed its expression profile via Western blot. Sporulating cells typically produce two forms of SafA: FL SafA (SafA^FL; 45 kDa) and C30 (30 kDa).⁴¹ Coat protein extracts from GFP-fused spores were probed with an anti-GFP antibody. Due to the lack of a suitable internal loading-control antibody, equal amounts of total spore coat proteins were loaded for Western blotting. Equal protein loading was independently verified by Coomassie Brilliant Blue staining of a parallel SDS-PAGE gel using the same protein samples and loading amounts as those used for the immunoblot shown in Figures 6D and S6. Densitometric analysis using ImageJ showed comparable total lane intensities between the two samples. Protein inputs were kept at relatively low levels to minimize non-specific banding (Figure S6). In ΔdnaK spores, only a single SafA band was detected, corresponding to the C30 fragment based on molecular weight, whereas SafA^FL was absent (Figure 6D). Although overproduction of C30 has been reported to interfere with spore formation,⁴² we did not observe overaccumulation of this fragment in our system (Figure 6D). However, the absence of SafA^FL likely explains its failure to localize correctly. This suggests that SafA^FL cannot be accumulated normally in the absence of DnaK.

In summary, DnaK deletion reduces the accumulation of SafA^FL and thereby impairs its proper localization. The mislocalization of SafA, a key morphogenetic protein, contributes to defective inner coat assembly.

Transcriptional regulation of spore coat assembly is impaired

Spore coat protein expression is governed by a regulatory cascade during sporulation. To elucidate how dnaK deletion perturbs coat assembly, we analyzed the expression of key transcriptional regulators.

We performed quantitative reverse transcription PCR (RT-qPCR) analysis on four mother cell-specific transcription factors—sigE, spoIIID, sigK, and gerE—using RNA extracted 5 h after sporulation induction, when the level of SpoIIID reaches a maximum.⁴³ Compared to the wild-type strain, the DnaK mutant exhibited markedly reduced transcript levels of spoIIID and spoIVCB (sigKN), whereas the expression of sigE and gerE remained largely unchanged. Conversely, the hrcA mutant led to upregulation of sigE and spoIIID, with no significant effect on spoIVCB or gerE (Figure 8A). Furthermore, disruption of the regulatory network also led to impairments in spore coat structure.⁴⁴ We further monitored the transcriptional activity using lacZ fusion strains over time. Consistent with qRT-PCR results, spoIIID and spoIVCB transcriptional activities were significantly lower in the dnaK mutant compared to the wild-type (Figure 8B). Notably, sigK activity exhibited a delayed decrease during late sporulation in the dnaK mutant, suggesting disruption in the feedback regulation of sigK. This transcriptional alteration appears specific rather than global, as alkaline phosphatase transcription was unaffected (Figure S7).^45,46

Figure 8.

DnaK influences the transcriptional regulation of sporulation regulators

(A) Quantitative RT-PCR analysis of sigE, spoIIID, spoIVCB, and gerE transcript levels in wild-type, ΔdnaK and ΔhrcA strains during sporulation. Expression levels are presented as fold changes relative to the wild type. Data are presented as mean ± SD from three biological replicates (n = 3). Statistical significance was determined by one-way ANOVA followed by Tukey’s multiple-comparison test. ∗∗p < 0.01, ∗∗∗p < 0.001; ns, not significant.

(B) β-Galactosidase activity of the sigE::lacZ, spoIIID::lacZ, sigK::lacZ, and gerE::lacZ reporter strains in wild-type (green circles) and ΔdnaK (yellow squares) backgrounds during sporulation. The schematic (top) illustrates integration of the promoter-lacZ fusion at the amyE locus. Activity is shown in Miller units over the sporulation time course. Data are presented as mean ± SD from three independent biological replicates.

These findings indicate that DnaK influences the transcriptional levels of key regulators, particularly sigK and spoIIID, thereby affecting the assembly of the spore coat. Since cotD expression is known to be positively regulated by sigK and gerE but repressed by spoIIID.^47,48 Meanwhile, CotD-GFP fluorescence intensity was reduced in the dnaK mutant, compared to the wild type. Furthermore, although the hrcA mutant showed elevated spoIIID expression (Figure 8A), it did not repress CotD expression (Figure 7B). These observations suggest that sigK could be the central regulatory node in this pathway.

Discussion

In this study, we demonstrate that DnaK is essential for proper spore coat assembly in Bacillus subtilis. Deletion of dnaK primarily disrupted the assembly of the spore coat’s inner layer. Fluorescence microscopy revealed mislocalization of inner coat morphogenetic protein SafA in the dnaK mutant. This defect appears to be associated with the reduced accumulation of FL SafA. Moreover, the absence of DnaK impaired the activation of key transcription factors spoIIID and sigK, which are required for coat encasement and maturation.

Our results indicate that dnaK deletion directly compromises several of these functions. dnaK mutant spores showed increased sensitivity to lysozyme and sodium hypochlorite—a phenotype atypical for general coat gene inactivation but reminiscent of morphogenetic protein mutants.^5,17 Although heat resistance remained unaffected unlike some coat mutants,⁴⁹ this is consistent with the understanding that the spore coat is not the primary determinant of wet heat resistance.³⁰ Moreover, while DnaK is a heat shock protein, dormant spores are not dependent on stress-induced responses. Interestingly, we observed a significant reduction in germination efficiency at elevated temperatures in dnaK spores. Although the mechanism remains unclear, it may be related to altered coat structure, which could affect the regulation of germinant access. The enhanced hydrophobicity of dnaK mutant spores represents a notable change in spore surface properties. Although the localization of the crust morphogenetic protein CotZ did not show an obvious difference in the dnaK mutant, the coat of Bacillus subtilis contains enzymes involved in polysaccharide synthesis,⁵⁰ and it is likely that sps-related proteins are also affected by the loss of DnaK, which may contribute to the observed changes in surface hydrophobicity. Spore surface hydrophobicity is an important property for the biotechnological applications of spores, such as their use as carriers or probiotic delivery platforms.⁵¹ In addition, chaperone proteins can facilitate the proper folding of heterologous proteins displayed on the spore surface. Together, these findings suggest that DnaK exerts macroscopic control over spore surface properties and may serve as a potential target for changing the surface properties.

During spore coat assembly, the initial step involves the formation of a scaffold surrounding the forespore by SpoIVA and SpoVM. The proper assembly of SpoIVA further requires SpoVID. These three proteins are synthesized under the control of σ^E and spore encasement is finished shortly after engulfment.⁵² In our study, the transcriptional activity of sigE remains intact in the dnaK mutant. Moreover, SpoIVA localization is essential for cortex formation.⁵³ Since no cortex-related defects were observed and spoIVA mutants show dispersed coat accumulation as cytoplasmic swirls of partially structured material in the mother cell cytoplasm,^5,54 we infer that the basement layer is not disrupted in the dnaK mutant. Among SpoVID protein interactors, SafA is particularly relevant. safA mutant spores lack the inner coat but retain an apparently unaffected outer coat,⁵⁵ whereas cotE mutant spores lack the outer coat but retain an inner coat,¹⁵ indicating independent assembly of the two layers. Additionally, we observed correct localization of CotE or its downstream targets, but SafA and its dependent protein CotD were mislocalized. Together with TEM data, these results strongly support that specific inner coat assembly is defective in dnaK spores, likely due to the reduced accumulation of FL SafA in the dnaK mutant. However, whether this result from a failure in translation or SafA^FL degraded after translation remains unclear. Furthermore, SpoVID is essential for encasement.¹⁶ Although SpoVID localization in the dnaK mutant did not display a significantly distinct new pattern, its fluorescence intensity showed only a minor, non-significant reduction relative to the wild type. The failure of inner coat proteins to form a complete shell suggests that the assembly defect in dnaK spores likely occurs at the encasement stage.

The molecular mechanism underlying encasement remains poorly understood but relies on the precise temporal regulation of gene expression in the mother cell. For instance, spoIVA is transiently expressed under the control of σ^E and later repressed by SpoIIID, while cotYZ, encoding proteins involved in the final crust layer, is only highly transcribed during late sporulation.⁸ Coat encasement occurs after engulfment and temporally coincides with the late morphogenesis regulated by spoIIID and sigK.¹⁶ Thus, the downregulation of spoIIID and spoIVCB observed in the dnaK mutant likely disrupts the process of coat encasement. In this regulatory cascade, GerR acts as an intermediate regulator whose expression depends on both σ^E and SpoIIID; given that SpoIIID is already downregulated in the dnaK mutant, the expression of GerR is likely also indirectly affected, further compounding the disruption of the transcriptional program governing coat assembly. Moreover, the delayed decline in spoIVCB activity may further perturb the transcriptional cascade necessary for orderly coat assembly, potentially leading to structural defects.⁴⁴ Previous work by Ogura et al. showed that DnaK interacts with SpoIVCA, affecting the maturation of SigK and sporulation quality.²⁹ Although we did not observe a complete loss of σ^K-dependent coat proteins in the dnaK mutant, our findings reinforce the role of DnaK in regulating coat assembly at the post-translational level. Importantly, we also provide evidence that DnaK affects sigK transcription. In contrast, DnaK does not influence phosphorylation of spore coat proteins.⁵⁶ Taken together, we propose that the primary site of DnaK action is the mother cell compartment, where it contributes to the accumulation of FL SafA. In the dnaK mutant, the accumulation of FL SafA is reduced, and through an additional mechanism, the expression of regulators governing the encasement process is also disrupted, collectively leading to the structural defects in spore coat assembly described in this study. Thus, the involvement of molecular chaperones in spore coat encasement represents a previously overlooked regulatory layer in spore morphogenesis. We conclude that the assembly of the spore coat requires coordination with protein quality control systems, including molecular chaperones such as DnaK operating in the mother cell.

Global transcriptional regulation by DnaK is not uncommon in bacteria. In Escherichia coli, for example, DnaKJ interacts with the transcription factor σ³² and represses its activity, thereby limiting widespread transcriptional reprogramming under non-stress conditions.⁵⁷ Similar repression of transcriptional regulators by DnaK has also been observed in Caulobacter crescentus.⁵⁸ However, in our study, alkaline phosphatase activity assays suggested that dnaK deletion did not affect the expression of all sigma factor-regulated genes, indicating that DnaK-mediated regulation during sporulation is likely specific rather than global. Nevertheless, identifying direct targets of DnaK remains challenging due to its weak and transient interactions with diverse substrates. As a result, it remains unclear whether the effects on spoIIID and sigK expression are direct or indirect.

Molecular chaperones and proteases constitute the two core components of cellular protein quality control. In recent years, increasing evidence has highlighted the significance of regulated proteolysis during sporulation. For instance, the CmpA-ClpXP complex targets SpoIVA for degradation, ensuring the elimination of incompletely assembled spore envelope structures under the control of σ^K.⁵⁹ Similarly, the adaptor protein MdfA facilitates ClpCP-dependent degradation of numerous metabolic enzymes and key transcription factors derived from the forespore during sporulation.⁶⁰ These findings underscore the critical importance of precise protein quality regulation throughout spore development. While proteolysis has gained increasing recognition as a pivotal regulatory mechanism during sporulation, the role of molecular chaperones—despite their fundamental importance in bacterial physiology—remains comparatively unexplored. Our study provides direct phenotypic evidence that the chaperone DnaK plays a previously underappreciated role in regulating spore coat assembly and developmental gene expression. Collectively, these findings enhance our understanding of the molecular mechanisms underlying spore coat assembly and illustrate how protein quality control networks coordinate both the structural and regulatory programs in bacterial sporulation.

Limitations of the study

Although this study demonstrates that DnaK is involved in spore coat assembly and remains present during sporulation, the dynamics of DnaK during sporulation require further investigation, particularly the fate of DnaK in the mother cell and the origin of DnaK detected in the forespore. Although deletion of dnaK impairs SafA localization and the accumulation of FL SafA protein, whether DnaK directly interacts with SafA remains unclear. Further biochemical and genetic analyses will be needed to clarify how DnaK regulates SafA localization and the assembly of inner coat proteins. In addition, although reduced transcriptional activity of sporulation regulators was observed in the ΔdnaK mutant, the mechanism underlying this change remains unclear.

Resource availability

Lead contact

Requests for further information, resources, and reagents should be directed to and will be fulfilled by the lead contact, Li Dong (li_dong127@163.com).

Materials availability

Bacterial strains and plasmids generated in this study are available from the lead contact without restriction for research purposes.

Data and code availability

• •The mass spectrometry proteomics data generated in this study have been deposited in the ProteomeXchange Consortium via the iProX partner repository and are publicly available under accession number PXD082055 (iProX project ID: IPX0018770000).
• •This paper does not report original code.
• •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

This work was supported by the National Key Research and Development Program of China (2022YFD2100404) and the National Natural Science Foundation of China (31601563).

Author contributions

K.M., L.D., and X.H. designed the study; K.M., C.L., and T.C. designed and performed the experiments; K.M., L.D., and Z.Z. analyzed the data; C.F., X.H., and L.D. acquired funding and supervised the project; K.M. and L.D. wrote the manuscript.

Declaration of interests

The authors declare no competing interests.

Declaration of generative AI and AI-assisted technologies in the writing process

The authors used ChatGPT-5.5 for language editing purposes only. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Anti-GFP (Green Fluorescent Protein) pAb	MBL Life Science	Cat#: 598; RRID: AB_591819
HRP-conjugated secondary antibody	Proteintech	Cat#: RGAR301; RRID: AB_3712952
Bacterial and virus strains
PY79 (B. subtilis wild type)	Youngman et al.61	N/A
E. coli strain DH5α	Vazyme	Cat#: C502
dnaK::kan, KmR	This work	N/A
dnaK::kan lacA::Pxyl-dnaK::erm, KmR, EmR	This work	N/A
hrcA::kan, KmR	This work	N/A
spoVID::kan, KmR	This work	N/A
cotE::kan, KmR	Lab storage	N/A
dnaK:gfp, SpR	This work	N/A
spoVID::gfp, SpR	This work	N/A
dnaK::kan spoVID::gfp, KmR, SpR	This work	N/A
hrcA::kan spoVID::gfp, KmR, SpR	This work	N/A
safA::gfp, SpR	This work	N/A
dnaK::kan safA::gfp, KmR, SpR	This work	N/A
hrcA::kan safA::gfp, KmR, SpR	This work	N/A
cotE::gfp, SpR	This work	N/A
dnaK::kan cotE::gfp, KmR, SpR	This work	N/A
cotZ::gfp, SpR	This work	N/A
dnaK::kan cotZ::gfp, KmR, SpR	This work	N/A
cotD::gfp, SpR	This work	N/A
dnaK::kan cotD::gfp, KmR, SpR	This work	N/A
hrcA::kan cotD::gfp, KmR, SpR	This work	N/A
amyE::PsigE-lacZ, CmR	This work	N/A
dnaK::kan amyE::PsigE-lacZ, KmR, CmR	This work	N/A
amyE::PspoIIID-lacZ, CmR	This work	N/A
dnaK::kan amyE::PspoIIID-lacZ, KmR, CmR	This work	N/A
hrcA::kan amyE::PspoIIID-lacZ, KmR, CmR	This work	N/A
amyE::PsigK-lacZ, CmR	This work	N/A
dnaK::kan amyE::PsigK-lacZ, KmR, CmR	This work	N/A
amyE::PgerE-lacZ, CmR	This work	N/A
dnaK::kan amyE::PgerE-lacZ, KmR, CmR	This work	N/A
Chemicals, peptides, and recombinant proteins
Kanamycin	Solarbio	Cat#: K1030
Spectinomycin	Solarbio	Cat#: S8040
Ampicillin	Solarbio	Cat#: A1170
Chloramphenicol	Solarbio	Cat#: L1311
Erythromycin	Solarbio	Cat#: E8100
Histodenz	Sigma-Aldrich	Cat#: D2158-100G
Sodium hypochlorite	Macklin	Cat#: S817439
Terbium(III) chloride hexahydrate	Sigma-Aldrich	Cat#: 212903
L-alanine	Sigma-Aldrich	Cat#: 100287
Dodecylamine	Aladdin	Cat#: D113938
Lysozyme	Solarbio	Cat#: L8120
O-nitrophenyl-β-D-galactopyranoside	Solarbio	Cat#: O8040
Difco™ Nutrient Broth	BD Biosciences	Cat#: 234000
HEPES	Sigma-Aldrich	CAS#: 7365–45–9
Luria-Bertani (LB) Broth	Oxoid	Cat#: CM0996B
Critical commercial assays
SYTO™ 16 Green Fluorescent Nucleic Acid Stains	Thermo Fisher Scientific	Cat#: S7578
ClonExpress Ultra One Step Cloning Kit	Vazyme	Cat#: C115-01
HiScript III RT SuperMix for qPCR	Vazyme	Cat#: R323-01
iTaq Universal SYBR Green Supermix	Bio-Rad	Cat#: 1725124
ECL substrate	Vazyme	Cat#: E412
Phanta Ultra Master Mix	Vazyme	Cat#: P518
Deposited data
The mass spectrometry proteomics data	This study	ProteomeXchange: PXD082055; iProX project ID: IPX0018770000
Oligonucleotides
Primers used in this study are listed in Table S1	Sangon Biotech (Shanghai, China)	N/A
Recombinant DNA
pAX-dnaK (lacA::xylR-dnaK)	This study	N/A
pAR-dnaK (dnaK-gfp)	This study	N/A
pAR-safA (safA-gfp)	This study	N/A
pAR-cotE (cotE-gfp)	This study	N/A
pAR-cotZ (cotZ-gfp)	This study	N/A
pAR-spoVID (spoVID-gfp)	This study	N/A
pAR-cotD (cotD-gfp)	This study	N/A
pBS1C-lacZ	This study	N/A
Software and algorithms
ImageJ	NIH, Bethesda, USA	https://fiji.sc/
GraphPad Prism	Graph Pad Prism Corp., San Diego, USA	https://www.graphpad.com/
pLabel	pFind Studio	https://pfind.ict.ac.cn/se/plabel/
FlowJo	BD Biosciences/Tree Star	https://www.flowjo.com/
VisiView	Visitron Systems	https://www.visitron.de/products/visiviewr-software
MaxQuant	Max Planck Institute of Biochemistry	https://www.maxquant.org/maxquant/

Experimental model and study participant details

Bacterial strains and general methods

All Bacillus subtilis strains used in this study were derivatives of the wild-type strain PY79. The primer sequences are listed in Table S1. The methods for strain and plasmid construction are provided in the supplementary materials. Plasmid amplification for nucleotide sequencing, subcloning experiments, and transformation of competent cells were performed with E. coli strain DH5α (Vazyme, Beijing, China). Luria-Bertani medium was used to grow B. subtilis. Sporulation was induced in Schaeffer’s sporulation medium for 24 h. Spores were purified using a discontinuous Histodenz gradient. Sporulation efficiency was determined by comparing colony-forming units (CFUs) before and after heat treatment (60°C for 20 min) to eliminate vegetative cells, followed by plating on LB agar and incubation at 37°C for 12 h.

Construction of strains and plasmids

All mutant strains were constructed by allelic replacement using homologous recombination as previously described.^32,62 Briefly, 1 kb upstream and downstream flanking regions of the target gene were PCR-amplified from PY79 genomic DNA, while an antibiotic resistance cassette containing a promoter was amplified from plasmid pJOE8999. The three fragments were assembled into a single linear construct by overlap extension PCR. The purified full-length construct was then transformed into the wild-type strain, and successful gene deletion was confirmed by diagnostic PCR.

To integrate the dnaK gene under the xyl-inducible promoter in the lacA locus, the dnaK coding sequence was PCR-amplified and assembled into an enzymatically linearized pAX01 vector (lab storage) via Gibson assembly, obtaining plasmid pAX-dnaK (lacA::xylR-dnaK Em^R). Primers were designed to include 20-bp overlapping sequences homologous to the plasmid ends to facilitate seamless assembly. The strain KYS1101 (ΔdnaK) was transformed to obtain strain KYS1102 (lacA::Pxyl-dnaK).

To construct strains expressing the GFP fusion protein, the target gene was amplified with a flexible GGGS linker sequence appended to the 3′-end via primer design, ensuring proper folding.⁶³ The gene-linker fragment was inserted in-frame upstream of the gfp coding region in the plasmid pAR100 ,³² which confers spectinomycin resistance, obtaining pAR-dnaK, pAR-cotZ, pAR-cotE, pAR-spoVID, pAR-safA and pAR-cotD. The resulting gene-gfp-spc cassettes were then introduced into the Bacillus subtilis chromosome by single-crossover homologous recombination at the respective native loci, thereby enabling in situ expression of GFP-fusion proteins under their endogenous regulatory elements.

To construct strains carrying lacZ transcriptional fusions, plasmid pBS1ClacZ was obtained for further modification.⁶⁴ Promoter regions of interest were amplified from Bacillus subtilis genomic DNA using specific primer pairs (PsigE-F/PsigE-R, PspoIIID-F/PspoIIID-R, PsigK-F/PsigK-R, PgerE-F/PgerE-R).⁴⁴ The PCR products were cloned into pBS1C-lacZ to replace the original rfp cassette. The resulting recombinant plasmids were first introduced into E. coli DH5α for propagation, where positive clones were identified by the loss of red fluorescence. The verified constructs were subsequently transformed into B. subtilis, where integration into the non-essential amyE locus occurred through double-crossover homologous recombination.

All transformations were performed using competent cells prepared in MC medium. Antibiotic selection was conducted at the following concentrations: kanamycin (25 μg/mL), spectinomycin (50 μg/mL), chloramphenicol (10 μg/mL), and erythromycin (10 μg/mL). For double-antibiotic selection, each antibiotic was used at half the standard concentration.⁶⁵

Method details

Spore resistance analysis

Lysozyme resistance and sodium hypochlorite resistance

Purified spores were heat-treated at 60°C for 30 min, then suspended in 10 mM Tris–HCl buffer (pH 7.0) containing 1 mg/mL lysozyme and incubated at 37°C for 3 h.⁴⁹ Spore resistance to sodium hypochlorite was evaluated by incubating spores (adjusted to an OD₆₀₀ of 1) at 25°C in 50 mM KPO4 buffer (pH 7.0) supplemented with a 1/100 dilution of sodium hypochlorite.⁶⁶ At designated time points, aliquots were quenched in 1% sodium thiosulfate for at least 10 min, further diluted with water, and then plated on LB agar. Plates were incubated overnight at 37°C for CFU enumeration.

Heat resistance

Spore suspensions were adjusted to an OD₆₀₀ of 1.0 and subjected to heat treatment at 80 °C, 85 °C, and 90 °C for 30 min to inactivate residual vegetative cells and assess spore heat resistance. The use of temperature gradients allowed for clearer differentiation in heat resistance between strains.

High-pressure thermal resistance

High-pressure thermal treatment was used to evaluate spore resistance under combined heat and pressure stress.⁶⁷ Spores from different strains were treated at 200 MPa and 80 °C for 15 min, following optimized sterilization conditions previously established in laboratory.

All data were performed with at least three independent experiments.

Spore physiological analysis

Growth curves and sporulation curves

For growth curve experiments, a single colony from each strain was inoculated into 5 mL of LB broth and incubated overnight at 37°C with shaking. Subsequently, the cultures were diluted in fresh LB broth to an OD600 of 0.1 to serve as the inoculum. Growth curves were monitored using a fully automated bioreactor (Biosan, Latvia) under the following conditions: 20 mL culture volume, 37°C, and 2000 rpm agitation. Growth data were collected from six independent biological replicates. For sporulation curves, 1% (v/v) of the overnight culture was inoculated into Difco sporulation medium (DSM) and incubated at 37 °C with shaking at 200 rpm. Samples were taken at 1-h intervals to measure OD. Data were performed with three independent experiments.

DPA release

The samples were treated, centrifuged briefly, spore supernatant solution was mixed in the HEPES-TbCl3 solution, which consisted of 50 mM K- HEPES (pH 7.4) buffer and 200 μM TbCl₃, the volume was made to 200 μL. Fluorescence was measured using a microplate reader with an excitation wavelength of 270 nm and an emission wavelength of 545 nm. DPA release was normalized to the fluorescence signal obtained from autoclaved spores, which was defined as 100% DPA release. Data are presented as the percentage of total releasable DPA. Experiments were performed using three independent biological replicates.

Spore hydrophobicity

Spore hydrophobicity was assessed using a modified bacterial adhesion to hydrocarbons (BATH) assay.⁶⁸ Spore suspensions were adjusted to an OD₆₀₀ of 1 in distilled water and combined with n-hexadecane at a 3:1 (aqueous:hydrocarbon) volume ratio. After vortexing for 60 s and standing at room temperature for 20 min to allow phase separation. Hydrophobicity was calculated as the percentage of spores remaining in the aqueous layer: % of spore in aqueous layer = (OD₆₀₀ of the spore suspension in water after phase separation/Initial OD₆₀₀ of the spore suspension in water) × 100%.

Spore germination

Purified spores were heat treated (70°C for 30 min), followed by cooling on ice. Spores were germinated at different temperatures in 50 mM K-HEPES buffer (pH 7.4) with 10 mM L-alanine, or with AGFK (10 mM L-asparagine, 10 mM D-glucose, 10 mM D-fructose and 10 mM KCl). Analysis of spore germination with dodecylamine did not use heat-shocked spores and was carried out with spores at an OD₆₀₀ of ∼1 at 45°C in 20 mM KPO₄ buffer (pH 7.4) and 1 mM dodecylamine. The progress of germination was assessed by measuring the OD₆₀₀ of the cultures and was also checked by phase-contrast microscopy. Germination was assessed either by adding 50 mM TbCl₃ to the germination mix and measuring DPA release as described above. Samples are incubated and read in the MicroplateReader, which can be set to different germination temperatures.

Spore coat proteins extraction and analysis

Samples of purified spores of different strains resuspended in 100 μL of buffer (4% SDS, 10% β-mercaptoethanol, 1 mM dithiothreitol, 10% glycerol), and the samples were boiled for 8 min as described previously.^36,69 The extracted proteins were fractionated by SDS-polyacrylamide gel electrophoresis (PAGE) on a 15% gel. Enzymatic digestion of proteins into peptides followed by mass spectrometry detection.

Transmission electron microscopy

Purified spores were fixed overnight at 4 °C with 2.5% glutaraldehyde and post-fixed in 1% osmium tetroxide. After thorough washing with 0.1 M phosphate buffer (pH 7.0), samples were dehydrated through a graded ethanol series (30%–100%), followed by treatment with acetone. Dehydrated samples were infiltrated and embedded in resin, then polymerized at 70 °C overnight. Ultrathin sections (70–90 nm) were prepared using an ultramicrotome (Leica, Germany), stained with 2% uranyl acetate and lead citrate, and examined under a transmission electron microscope (JEM-1200EX, JEOL, Japan). At least 30 spores were analyzed per condition.

Fluorescence intensity assay

Fluorescence intensity of spores was analyzed by flow cytometry. Purified spores with GFP-fusion were resuspended in water and adjusted to an appropriate concentration (∼10⁶–10⁷ CFU/mL). Flow cytometric analysis was performed using an LSRFortessa flow cytometer (BD Biosciences). Fluorescence signals were excited using a 488 nm laser and collected in the FITC/Alexa Fluor 488 channel. Debris was excluded based on forward scatter (FSC) and side scatter (SSC) gating. For each sample, 100,000 events were recorded. Flow cytometry data were analyzed using FlowJo (Tree Star Inc.), and fluorescence intensity was presented as histograms.

Fluorescence microscopy

To visualize the spore coat protein, it was fused with GFP protein, followed by fixation and subsequent observation using a fluorescence microscope. Fluorescence analyses were carried out using an Olympus IX 83 fluorescence microscope fitted with a 100× objective lens of a microscope, and equipped with Photometrics Prime 95B sCMOS camera and driven by VisiView software. The resulting fluorescent images were enhanced using the deconvolution process of the device. Images were processed with ImageJ for minor adjustments of brightness, contrast and color balance. For the localization images of spore coat proteins, the captured images were processed using a completely consistent method.

To characterize whether cortex and inner membrane permeability were altered, a high concentration of 30 μM PI (previously dissolved in DMSO) and 1 μM SYTO16 dye were used togther for staining. After the addition of the dyes, the spore suspension was gently mixed using a vortex mixer and placed on ice under light-avoiding conditions for 3 h. Samples were subsequently observed and imaged using a Zeiss fluorescence microscope.

Western blot analysis

Spore coat proteins were separated by 12.5% SDS-PAGE and transferred onto PVDF membranes using a conventional wet-transfer system. The membranes were blocked with 5% non-fat milk in TBST for 1 h at room temperature and then incubated with an anti-GFP primary antibody (1:7000) overnight at 4°C. After washing with TBST, the membranes were incubated with an HRP-conjugated secondary antibody for 1 h at room temperature. Immunoreactive bands were detected using an enhanced chemiluminescence (ECL, Vazyme) substrate and imaged using ChemiDoc XRS (Bio-Rad, Hercules, CA, USA). Equal loading of spore coat proteins was verified by Coomassie Brilliant Blue staining of a parallel SDS-PAGE gel (Figure S6) and densitometric analysis using ImageJ showed comparable total lane intensities between the two samples.

Quantitative real-time PCR (qRT-PCR)

Spore cultures were washed three times with PBS and RNA was extracted using Bacteria Total RNA Isolation Kit (Seno, China). Extracted RNA (1 μg) was treated with gDNA wiper and subjected to cDNA synthesis using RT SuperMix. qRT-PCR reactions were conducted using SYBR-green mix, and fluorescence detection was performed using LightCycler 480 (Roche). The housekeeping gene gyrA was used as a reference gene.⁷⁰ Each assay was performed in duplicat with at least two RNA templates prepared from independent biological replicates.

β-Galactosidase analysis

The methods for β-galactosidase analysis have been previously described.⁷¹ Briefly, spores collected at the same sporulation time point were resuspended in 1 mL of Z buffer. About 10 μL of toluene was added and mixed thoroughly to permeabilize the spores. Then, o-nitrophenyl-β-D-galactopyranoside was added as the substrate, and the reaction was incubated at 30 °C for 15 min. The reaction was terminated by adding sodium carbonate, and β-galactosidase activity was quantified by measuring absorbance at 420 nm.

Phosphoproteomic analysis by LC-MS/MS

Spore coat protein extracts were reduced and alkylated, followed by tryptic digestion at 37°C for 16–18 h. The resulting peptides were desalted using C18 StageTips and reconstituted in 0.1% formic acid. Peptides were separated using an Easy-nLC 1200 system equipped with a C18 analytical column (75 μm × 150 mm, 3 μm) over a 60-min gradient and analyzed using a Q Exactive HF-X mass spectrometer (Thermo Fisher Scientific) in data-dependent acquisition mode. Full MS scans were acquired over an m/z range of 350–1800 at a resolution of 60,000, followed by HCD MS/MS analysis of the 20 most intense precursor ions at a resolution of 15,000 with a normalized collision energy of 28.

Raw data were processed using MaxQuant (v1.6.1.0) and searched against the UniProt Bacillus subtilis database (Taxon ID: 224308). Trypsin was specified as the protease with up to two missed cleavages. Precursor mass tolerances were 20 ppm for the first search and 4.5 ppm for the main search, and the MS/MS tolerance was 20 ppm. Carbamidomethylation of cysteine was specified as a fixed modification, while oxidation of methionine, protein N-terminal acetylation, and phosphorylation of serine, threonine, and tyrosine residues were considered variable modifications. Peptide-spectrum match, protein, and phosphorylation-site FDRs were controlled at <1%. Phosphorylation sites with a MaxQuant localization probability ≥0.75 were considered confidently localized. Representative MS2 spectra of phosphopeptides were annotated using pLabel software to visualize assigned b- and y-type fragment ions. Three independently prepared spore coat protein samples were analyzed by LC-MS/MS.

Quantification and statistical analysis

Statistical analyses and data visualization were performed using GraphPad Prism (GraphPad Software, San Diego, CA, USA). Data are presented as mean ± SD unless otherwise indicated. The number of biological replicates (n) and the statistical tests used for each experiment are specified in the corresponding figure legends.