All B. subtilis strains are listed in Table 1. The hbs point mutations were previously constructed at the native locus, as previously described (Carabetta et al., 2019). LB (Luria-Bertani) agar and Schaeffer’s sporulation media (DSM) were prepared as described previously (Albano et al., 1987). Strains were streaked out on LB plates and incubated overnight at 30 °C. The following morning, cells were inoculated into 20 mL of DSM and grown at 37 °C with aeration for 6 h, with growth monitored by Klett colorimetry. 1 mL of cells were harvested after 2, 3, 4, and 6 h, corresponding to the entry into stationary phase (T₀), and hours after (T₁, T₂, and T₄ time points), which correlates to stages of sporulation. The samples were centrifuged at 13,000 x g for 3 min. The resulting cell pellets were processed further either for RT-qPCR, western blot, or microscopy.

Samples collected at each timepoint were processed with using the FastRNA™ Pro Blue Kit (MP Biomedicals) to extract total RNA as per manufacturer’s instructions, with modifications. Briefly, the cell pellets were resuspended in 1 mL RNApro solution and lysed in tubes containing Lysing Matrix B provided in the kit. The sample tubes were processed using the FastPrep-25 5G Instrument (MP Biomedicals) for three cycles to ensure lysis, 45 s for the first two, then one 30 s cycle. The concentration of total RNA was measured using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific). Average RNA concentrations obtained varied from 800 to 3,500 ng/μL, with an A260: A280 ratio between 1.8 and 2.1. Next, RNA samples were treated using the Turbo DNA-free™ Kit (Invitrogen) according to the manufacturer’s instructions, to remove potential contaminating genomic DNA. Reverse transcription was carried out using the SuperScript™ IV VILO™ kit (Invitrogen) as per manufacturer’s instructions, with the following modifications. The annealing time was extended to 20 min at 25 °C and the reverse transcription time to 20 min at 50 °C. The resulting cDNA was purified using the MinElute PCR Purification Kit (Qiagen) according to the manufacturer’s protocol with the following modification. To elute DNA, 12 μL of elution buffer was added to the spin column containing sample, incubated for 5 min at 37 °C, and then centrifuged at 12,000 x g for 2 min. The cDNA samples were diluted 1:5 in nuclease-free water (1X concentration) and the concentrations measured using the NanoDrop 2000. For standard curves, the cDNA was diluted 5-fold six times. Five mL of 1X, 0.2X, 0.04X, 0.008X, 0.0016X, and 0.00032X concentrations of cDNA for standards and 5 mL of 0.1X for samples, were added to a MicroAmp Optical 96-well plate (Applied Biosystems) 0.12.5 μL of PowerTrack SYBR Green (Applied Biosystems) and 0.25 μL of each primer (10 μM) were added to each well, for a 25 μL total volume. All primers used in this study are listed in Supplementary Table S1. The primers were designed using Primer3 (version 4.1.0), and synthesized by Eton Biosciences (Union, NJ). Plates were run on QuantStudio 7 Pro Real-Time PCR instrument (Applied Biosystems) and analyzed using the relative standard curve method by the built in QuantStudio Design and Analysis software (Applied Biosystems). A melt curve analysis was performed after each run. All reactions were run in technical duplicates, and each experiment was repeated at least three independent times. The rpoD transcript was used as an internal control to adjust for differing amounts of input cDNA. All data was first normalized to the corresponding strain and time point input RNA value determined for rpoD and then the mutants compared to the wildtype normalized RNA value of the corresponding time point (Larionov et al., 2005). A one-way ANOVA for each gene at each timepoint was analyzed to identify significant differences among strains, followed by a post hoc Benjamini–Hochberg correction and Dunnett’s test to control for multiple comparisons. A p-value of < 0.05 was considered significant. Figures displaying normalized fold gene expression were created using GraphPad Prism (Version 9.4.0).

One mL of cells was collected by centrifugation after 4 and 6 h of growth in DSM at 37 °C. The cell pellets were resuspended in STM buffer (0.71 mM MgCl₂, 7.14 mM NaCl, 73 mM Sucrose, 35.7 mM Tris, pH8) + 350 μg/mL lysozyme with volumes normalized by 1/10th of the measured Klett value and incubated at 37 °C for 5 min. Samples were boiled 1X cracking buffer (0.225 M Tris–HCl, pH 6.8, 50% glycerol, 5% sodium dodecyl sulfate [SDS], 0.05% bromophenol blue, 1% β-mercaptoethanol) for 10 min and an equal amount of protein loaded on a 15% Tris-Glycine gel. Proteins were transferred to a nitrocellulose membrane using the iBlot 3 Western Blot Transfer System (Invitrogen), according to the manufacturer’s instructions. Membranes were blocked in 5% milk in Tris-buffered saline containing 1% Tween-20 (TBS-T) for 1 h and then incubated overnight at 4 °C with the indicated primary antibodies, at a 1:5,000 dilution for anti-SigF, anti-SigK, or anti-SpoIIAB, and a 1:10,000 dilution for anti-SigG. A 1:5,000 dilution of anti-EFG was included as a loading control. The anti-SigF and anti-SigG antibodies were kindly provided by D. Rudner (Harvard Medical School) and the anti-SigK and anti-SpoIIAB antibodies were kindly provided by N. Bradshaw (Brandeis University). The following day, the membranes were washed four times in TBS-T, then incubated in anti-rabbit secondary antibodies conjugated to horseradish peroxidase (Invitrogen) at a1:5000 dilution for 1 h at room temperature and washed in TBS-T as before. Bands were visualized using the ECL Prime Western Blotting Detection Reagent (Cytiva) according to the manufacturer’s instructions and imaged using the ChemiDoc™ MP Imaging System (BIO-RAD). All western blots were quantified using Image J (NIH), and all signal intensities were normalized EF-G. To calculate fold changes compared to wildtype, EF-G normalized intensities in the mutant strains were divided by the EF-G normalized intensities in wildtype.

The cell cultures of BD630 and mutant strains were grown to T₀, T₁, T₂, and T₄ as described above. Five mL of culture was harvested and centrifuged at 11,000 x g for 3 min. The supernatant was discarded, and the cell pellets resuspended in 5 mL of phosphate buffered saline (PBS). Cells were washed once with 1 mL of PBS, centrifuged at 11,000 x g for 1 min, and fixed in 1 mL of 4% formaldehyde for 15 min at room temperature. Following incubation, pellets were washed twice with 1 mL of PBS, and permeabilized for 10 min in 500 μL of 25 μg/mL (final concentration) lysozyme, and then immediately washed three times with 1 mL of PBS. Fixed and permeabilized cells were stained for 15 min with 10 μg/mL of wheat germ agglutinin conjugated with Alexa Flour 488 (WGA), washed with 500 μL of PBS, for a total of three times, and then stained for 5 min with 2 μg/mL of Sytox to stain DNA. Following staining, cells were washed with PBS, resuspended in 500 μL of PBS, and 1 μL of cells was placed on 1% agarose pads, and imaged on a Nanoimager super-resolution microscope (ONI). Mature, phase-bright spores were assessed in transillumination mode. Sporulating cells were imaged using 488 nm laser. The laser power was set at 4–5% for the 488 nm wavelength for Sytox and 19–20% for WGA. Samples were analyzed using NimOS (version 1.19.4). Staging was determined using biological replicates, analyzing at least 200 cells in each sample.

The sporulation frequency was performed as previously described (Carabetta et al., 2013) with the following modifications. Samples were collected after 2 h of incubation in DSM (T₀). The sporulation frequency was calculated as heat-resistant colony-forming units (CFUs) per ml/total viable cells (CFUs/ml). The experiments were carried out at least three independent times. The two-tail Student’s t-test was used to compare the sporulation frequency between samples. A p-value < 0.05 was considered significant.

We hypothesized that one way acetylation of HBsu influences the process of sporulation is through the regulation of gene expression. To test this idea, a quantitative real-time PCR (RT-qPCR) assay was developed and optimized. HBsu is acetylated at seven sites throughout the protein, at K3, K18, K37, K41, K75, K80, and K86 (Carabetta et al., 2016). To examine the influence of acetylation, we utilized our collection of point mutations at the native hbs locus that mimic the acetylated state [glutamine substitutions (Carabetta et al., 2019)]. We examined four different house-keeping genes to serve as an internal control, rpoA, rrnA-16S, rrnA-5S, and rpoD. The efficiency of each set of primers (Supplementary Table S2) was determined and rpoD was ultimately selected as the internal control for normalization of differing amounts of input cDNA. Based upon the expression patterns of sporulation-specific genes, 25 genes were selected for analysis (Stragier and Losick, 1996). The selected genes were representative of various early and late stages of sporulation. Samples were collected at 4 time points, T₀–T₄. The T₀ time point corresponds to the beginning of sporulation (stages 0–I), when the cells “decide” to sporulate. The T₁ time point corresponds to stages II–III of sporulation, characterized by significant changes in the cell morphology, including asymmetric septation. The T₂ time-point is later in the process and corresponds to stages IV–VI of sporulation, while T₄ corresponds to stages V–VII. As the sporulation-specific sigma factors are important master regulators driving this process, various regulon members were included as reporters of sigma factor activity (Supplementary Table S3 and Figure 1A). Additionally, we analyzed the asynchronous cultures at the selected time-points to determine and confirm the staging of sporulation (Figure 1B). As expected, there were cells at different stages concomitantly within the same population (Supplementary Figure S1).

The hbsK41Q mutant and wild-type strains were grown in DSM and analyzed throughout sporulation from T₀–T₄. First, we examined the activity of the housekeeping sigma factor σ^A and the stationary phase sigma factor σ^H, which are responsible for the entry into the sporulation process, by examining the expression of the phosphorelay proteins (kinA, kinB, spo0A, spo0B, and spo0F). At T₀ and T₁, representing early sporulation when these genes are likely important, we did not observe significant differences in expression compared to the wild type (Figure 2). At T₂, there was a 2.6-fold increase in spo0A expression (adj. p-value = 0.001). At T_4, there were reduced levels of spo0F, spo0A, kinA, and kinB, but these differences were not statistically significant and would likely not biologically relevant as phosphorylated Spo0A is not required at this late stage.

Next, we examined the expression of genes that are dependent upon Spo0A ~ P, including spoIIE, and the spoIIA and spoIIG operons, with the proteins required during stage II of sporulation (Figure 3). Expression of these genes typically occurs before asymmetric septation and they are all involved in the activation of σ^F. spoIIAB was 18.4-fold (adj. p-value = 0.0074) increased at T₂. Despite this increase in transcription, a corresponding increase in protein levels was not observed (Supplementary Figure S2). Similarly, there was a 1.86-fold overexpression of spoIIGA at T₁ (adj. p-value = 0.0071) and 41.8-fold overexpression at T₂. spoIIE was significantly overexpressed 34.1-fold (adj. p-value = 0.0018) at T₂. This suggests that the increase in spo0A transcription observed at T₂ leads to more Spo0A ~ P in the cell, which increases expression of Spo0A-dependent promoters, especially at T₂. In agreement, bursts in spo0A transcription have been observed to result in increases in Spo0A ~ P levels (Mirouze et al., 2011).

σ^F and σ^E are compartment-specific sigma factors that are active in the forespore and mother cell, respectively. Engulfment is the morphological process following σ^E activation, which occurs under the transcriptional control of both σ^F and σ^E. The coordinated transcription between the distinct cellular compartments leads to the completion of engulfment, generating a double-membrane organelle within the mother cell. We found that sigF transcription was significantly increased at T₂ (16.83-fold, adj. p-value = 0.0011), and decreased at T₄ (0.29-fold difference, adj. p-value = 0.0071, Figure 4). We analyzed expression levels of σ^F by Western blot, and found that in agreement with RT-qPCR, σ^F was overexpressed 6 and 7.6-fold at T₂ and T₄, respectively, even though transcription was reduced at T₄ (Figure 5A). The increased protein levels, coupled with wild-type levels of the anti-sigma factor SpoIIAB (Supplementary Figure S2), implied that σ^F activity will also be increased, so σ^F-dependent genes were analyzed, including spoIIQ and sigG. For spoIIQ, there was a 31.7-fold overexpression at T₂, although not statistically significant, due to the presence of outliers among replicates. At T₂ there was an overexpression of sigG (35.1-fold, adj. p-value = 0.0071) and at T₄ there was a 3.69-fold increase (adj. p-value = 0.00006) in expression. In agreement, there was a 10.3-fold increase in protein at levels at T₄ (Figure 5B). Together, these results suggest that overproduction of σ^F at T₂ leads to inappropriately increased expression of the σ^F regulon.

To assay the mother-cell specific σ^E activity, asnO, spoIID, spoIVCA, and spoVK were analyzed. There was a large increase in spoIID levels (26.6-fold), which is only dependent upon σ^E for activation, suggesting increased sigma factor activity. asnO was significantly increased across the entire course, with a 1.79, 2.83, and 66-fold increase at T₁, T₂, and T₄, respectively (adj. p-value = 0.031, 0.0027, and 0.0474, respectively). spoVK and spoIVCA were both overexpressed at the later time points, while not statistically significant (Figure 6). These results suggest that σ^E activity is also increased when K41 is acetylated.

The completion of engulfment stage dramatically changes the transcriptional program within both compartments, with the activation of σ^G in the forespore. To assay σ^G activity, we examined the expression of spoVAA, gpr, spoVT and sspB (Figure 7). At the T₄ time point spoVT was overexpressed 7.87-fold (adj. p-value = 0.002) and sspB, which is solely dependent on σ^G for expression, was overexpressed at 2.72-fold, 4.47-fold, and 61.6-fold at T₁, T₂, and T₄, while not statistically significant. The presence of a large outlier in one replicate (>120-fold increase, compared to ~30-fold for all replicates) is the reason statistical significance was not reached T₄, despite the large increase in expression. spoVAA was overexpressed at T₂ (6.59-fold, adj. p-value = 0.000009) and T₄ (26.8-fold, adj. p-value = 0.0071), whereas gpr did not exhibit large differences in expression.

To analyze the late stages of sporulation, we examined the expression of the coat proteins. The expression of cotE and cotH is under the control of σ^E and σ^K, and cotD is under the control of only σ^K. We observed a significant increase in the expression level of cotE (19.9-fold, adj. p-value = 0.0003) and cotH (18.3-fold, adj. p-value = 0.0002) at T₄ (Figure 8). cotD and lipC expression were not different from wildtype, suggesting that σ^K activity was not altered. The large increase in cotE and cotH expression likely resulted from the increase in σ^E activity. As we observed a gradual increase in the expression of spoIVCA in the hbsK41Q mutant, which encodes a protein essential for posttranslational processing of pre-σ^K, we examined σ^K levels by western blot (Figure 9). σ^K was overproduced at T₄ compared to wildtype. σ^K is produced as a longer, inactive form that must be cleaved N-terminally for activation and both bands were detected. The higher band is the inactive protein before cleavage, and the smaller band represents the active protein. σ^K is overproduced at T₄ compared to wildtype, with significantly more processing (92% compared to 56%).

Since we observed many changes associated with the hbsK41Q mutants, we analyzed the opposite, deacetylated form with the hbsK41R mutant strain. At T₂, for most genes analyzed, similar levels to wildtype were observed (Table 2). asnO, spo0F, spoVK and spoVG were increased 2- to 6-fold in the hbsK41Q strain but displayed reduced expression in the hbsK41R strain. This suggests that many of these genes may be regulated by HBsu and acetylation at K41 is required for gene expression at T₂. At T₄, roughly one third of the transcripts that were significantly increased in the hbsK41Q strain were altered to the same extent in the hbsK41R strain. For example, cotE and cotH were both highly overexpressed in the hbsK41Q mutant (19.9-fold and 18.3-fold, respectively) and were overexpressed to a similar extent in the hbsK41R mutant (27.4-fold, adj. p-value = 0.0003 and 26.8-fold, adj. p-value = 0.0002, respectively). This might be due to the influence of other transcription factors and suggests that HBsu K41 acetylation is not the primary regulatory determinant for expression of these genes at T₄.

Since most important sporulation genes were overexpressed and sigma factor activity was increased, we hypothesized that hbsK41Q strains sporulate earlier than the wildtype. To test this idea, we examined sporulation frequencies at T₀ (Supplementary Figure S4), a time when the decision to enter sporulation usually occurs, and few or no intact spores should be observed. At T₀, the wild-type and hbsK41R strains sporulated at a very low frequency of 0.04 and 0.07%, respectively, as expected. However, there was a 12.5-fold increase in sporulation frequency of hbsK41Q strains (0.55%), which agrees with the idea that increased gene expression leads to earlier sporulation. Given that hbsK41Q is more susceptible to heat than the wildtype, and the sporulation frequency assay utilizes a 30-min heat exposure to eliminate vegetative cells, we are likely underestimating the actual sporulation frequency. In agreement with these findings, 2-fold more hbsK41Q cells entered sporulation at T₀ compared to wildtype, and by T₁, 3-fold more cells made it to late sporulation (Supplementary Figure S1). This suggests that increased gene expression of the early sporulation genes increases the likelihood of transition into and progression through the sporulation program.

Previously, we found significant reductions in sporulation frequency if the acetylation status was altered at K3, K18, K37, K75, K80, and K86 (Luu et al., 2021). For each of these sites, we observed a reduction in the expression of key genes throughout the sporulation cycle in acetylation mutants, exactly opposite to that seen in the hbsK41Q strain. The following results were statistically significant in univariate analyses but did not remain significant after correction for multiple testing. Nonetheless, we consider these findings biologically relevant and have chosen to report them (Supplementary Table S4). In the hbsK3Q strain at T₄, the expression of two key late sporulation proteins, cotD and sspB were reduced (0.32-fold and 0.34-fold, Figure 10A and Supplementary Table S4). The hbsK3Q mutant has a reduced sporulation frequency, higher susceptibility to heat and formaldehyde exposure (Luu et al., 2021). Together, our data suggests when K3 is acetylated, the late sporulation genes are not turned on at the proper time, which leads to the decreased heat and chemical stress resistance. Similarly, cotD and sspB were reduced at the T₄ in the hbsK37Q mutant (Figure 10B and Supplementary Table S4). The hbsK37Q strain was significantly more susceptible to heat exposure, with a survival rate below 10% after 30 min (Luu et al., 2021). hbsK86Q spores had a reduction in sporulation frequency (<50%) and were significantly more susceptible to heat exposure (Luu et al., 2021). There was also a decrease in the late sporulation genes cotD and sspB at T₄ (Figure 10C and Supplementary Table S4). In addition, there was a decrease in spoIID expression at T₄ (Figure 10D and Supplementary Table S4). This suggests that the acetylation of K3, K37, and K86 turn off gene expression of late sporulation genes. In agreement with these findings, we observed reduction in mature spores compared to wildtype at T₄ (Supplementary Figure S1). In agreement, we observed a 2.5- to 6-fold reduction in mature spores produced at T₄ compared to the wildtype (Supplementary Figure S1), which may indicate a defect in spore assembly and maturation.

The hbsK75Q mutants have a reduced sporulation frequency, but the resulting spores were resistant to heat and UV stress. The only noted defect was reduced survival to formaldehyde after 40 min of exposure (Luu et al., 2021). Some early sporulation genes were modestly reduced, including kinA, sigF, spo0A, spoIIQ, and spoIIAB (Figure 11A), with many of the effects observed at T₄, when these genes would not be expected to be important. These changes likely do not explain the observed defects in sporulation frequency in this mutant. σ^F levels were reduced by about 50% at T₂ compared to the wildtype (Figure 5A), while σ^G was reduced to 0.04 and 0.4 at T₂ and T₄, respectively (Figure 5B). σ^K levels were also reduced at T₄, to 0.05-fold compared to wildtype (Figure 9). In addition, some late sporulation genes were reduced at T₄, including cotE, cotH, and sspB (Figures 11B,C and Supplementary Table S4), which might explain reduced sporulation frequency and survival to formaldehyde, as well as a decreased production of mature spores compared to the wildtype (3% vs. 13%) at T₄ (Supplementary Figure S1). In support of this, there were decreased mature spores produced compared to wildtype at T₄ (Supplementary Figure S1).

The hbsK80Q mutant had the second most alterations in gene expression, with most genes, especially the mother-cell specific ones, downregulated. Previously, we found that the sporulation frequency of hbsK80Q strain was decreased by more than 50%, and that they were more susceptible to formaldehyde and UV radiation (Luu et al., 2021). The mature spores produced at T₄ were reduced compared to the wildtype (2% vs. 13%), although, like the wildtype, the majority of cells proceeded to late sporulation (Supplementary Figure S1). Some forespore-specific early sporulation genes were downregulated at the T₁ and T₂ time points (i.e., kinA, sigF, spoIIQ, spoIIAB, spoIIE), which could, at least in part, explain the reduction in sporulation frequency of this mutant (Figure 12A and Supplementary Table S4). There were also reductions in σ^E- and σ^K-dependent genes. Most of these genes had decreased expression by more than 2-fold compared to the wildtype at the T₄ time point (Figures 12B,C and Supplementary Table S4). Given that these late genes are involved in forming protective layers of the spores, the observed decrease in expression might explain the higher susceptibility to chemical and UV stresses. As there were multiple under-expressed genes, we next examined the levels of the sigma factors. There were no differences in σ^G or σ^F levels compared to the wildtype (Figures 5A,B). However, we were unable to detect σ^K (Figure 9), which might explain the reduction in the transcription of sspB and the cot genes. As acetylation of HBsu at K80 leads to decreased gene expression, this suggests that deacetylation of K80 is required for proper late gene expression.