Bacterial endospores are one of the most resilient cell-types, surviving in harsh environments without nutrition for a prolonged time. Endospores are formed during the process of sporulation, which is a survival strategy when spore-forming bacteria encounter severe environmental stresses or starvation. In Bacillus subtilis, there are seven stages of sporulation, with each step controlled by specific regulatory cascades in a coordinated temporal pattern (Ryter et al., 1966; Piggot and Coote, 1976). Early steps in this process involve the replication of the chromosome, asymmetric septum formation and the coordinated packaging of the DNA into the smaller forespore compartment. In the later stages of sporulation, the developing spore is engulfed and the outer layers of the cortex and coat proteins are assembled. Finally, mother cell lysis occurs and the mature spore is released (Higgins and Dworkin, 2012; Cho and Chung, 2020).

Spores have little to no metabolism and are more resistant than vegetative cells to stresses such as heat, desiccation, radiation, and oxidizing agents (Mason and Setlow, 1986; Banks et al., 1988; Russell, 1990; Setlow and Setlow, 1995; Loshon et al., 1999; Setlow, 2006). Spores can accumulate damage to proteins and DNA during extended periods of dormancy (Popham and Bernhards, 2015), and small amounts of accrued DNA damage is repaired once the spore germinates (Setlow et al., 2003; Moir and Cooper, 2015). However, accumulation of excessive damage is overwhelming and leads to reduced survival during germination (Setlow, 1995; Setlow and Setlow, 1996). Thus, there are many mechanisms to protect important biomolecules from damage during dormancy. To protect the DNA, the cell produces the α/β-type small acid-soluble proteins (SASPs). The SASPs are abundant spore proteins that are synthesized in the forespore and coat the chromosome to protect the DNA from environmental insults, such as ultraviolet (UV) radiation (Setlow et al., 1982; Mason and Setlow, 1986). SASP binding to DNA induces positive supercoils and increases persistence length (Nicholson et al., 1990; Griffith et al., 1994). Additionally, SASPs bound to DNA induce changes in UV photochemistry, namely by forming a spore photoproduct when exposed to UV radiation rather than thymine dimers, which are bulky and destabilizing lesions (Setlow and Setlow, 1987; Nicholson et al., 1991).

Clearly, chromosomal compaction is necessary to physically package the DNA into the smaller forespore compartment. The bacterial chromosome is compacted and organized by the action of the nucleoid-associated proteins (Dillon and Dorman, 2010). The essential, histone-like protein HBsu is a member of the widely conserved HU-family of proteins, which are largely responsible for chromosomal compaction and coordination of DNA transactions (Micka and Marahiel, 1992; Klein and Marahiel, 2002). While lacking sequence or structural homology, the HU-family is generally considered to represent functional homologs of eukaryotic histones. HBsu is present in mature spores and is involved in the organization of the SASPs (Ross and Setlow, 2000). In developing spores, HBsu colocalizes with the SASPs in the nucleoid region and modulates SASP-mediated properties. These modulatory effects, such as amelioration of SASP-mediated increases in DNA persistence length and opposing negative DNA supercoiling, suggested that HBsu is a primary modulator of SASP-mediated DNA alterations (Ross and Setlow, 2000). Taken together, these data indicate that HBsu may be a critical regulator during the process of sporulation.

N^ε-lysine acetylation is a ubiquitous, regulatory post-translational modification (PTM) that influences a variety of biological processes in bacteria (Carabetta and Cristea, 2017; Christensen et al., 2019a,b; VanDrisse and Escalante-Semerena, 2019). Acetylation of lysine residues may alter protein–DNA interactions, inhibit enzymatic function, or change subcellular localization (Gu and Roeder, 1997; Glozak et al., 2005; Ishfaq et al., 2012; Li et al., 2012). Recently, we characterized the B. subtilis acetylome and found that the histone-like protein HBsu is acetylated at seven sites in vivo (Carabetta et al., 2016). Using substitution mutations that mimic the unacetylated and acetylated forms of HBsu, we showed that the inability to acetylate key lysine residues resulted in a more compacted nucleoid. Additionally, we showed that acetylation reduces the DNA binding affinity of HBsu, providing a potential regulatory mechanism (Carabetta et al., 2019). Acetylation of HBsu occurs enzymatically, which is carried out by acetyltransferases (KATs). The GCN5-like N-acetyltransferases (GNATs) are prevalent in bacteria, and catalyze the transfer of an acetyl group from acetyl-CoA to a target primary amine, such as on a lysine residue sidechain. Following a screen of predicted GNAT domain-containing proteins for defects in nucleoid compaction, we found that two GNATs, YfmK, and YdgE acetylate HBsu (Carabetta et al., 2019). In B. subtilis, there are two known lysine deacetylases (KDACs): SrtN of the NAD⁺-dependent sirtuin family and AcuC of the NAD⁺-independent, Zn²⁺-dependent family (Gardner and Escalante-Semerena, 2009). However, we currently do not know the mechanism of deacetylation for HBsu.

HBsu has been proposed to modulate the properties of the SASPs to ensure proper packaging of the chromosome into the spore. As HBsu is acetylated, and these PTMs are important for its chromosomal compaction function, we reasoned that acetylation of HBsu may also regulate sporulation. To test this hypothesis, we utilized substitution mutations that mimic the acetylated (K → Q) and unacetylated (K → R) forms of HBsu, and determined the influence of acetylation on overall sporulation frequency, and spore resistance properties. We found that specific HBsu acetylation patterns are required for proper sporulation. This suggests that HBsu is important for chromosomal dynamics during the early stages of spore formation, and that acetylation at specific sites regulates this process. As the SASPs are essential to protect the DNA from environmental insults, we predicted that altering HBsu acetylation might lead to improper chromosomal packaging, and therefore, to increased susceptibility to heat, UV and chemical stresses. To test this idea, we isolated B. subtilis spores from wild-type and mutant strains, and determined survival to exposure of each agent. We found that specific mutations led to decreases in survival to each stress. By using opposite mutations, we were able to ascertain if each site was more likely to be acetylated to some extent or unacetylated in the mature spore. Taken together, these data indicate that HBsu acetylation is required for proper protection of the spore chromosome, and this further supports the idea that HBsu and the SASPs work together to complete this task. Further work is required to fully elucidate the exact mechanism by which HBsu and the SASPs work together, and how acetylation modulates this process. We propose that specific acetylation patterns change the DNA-binding capacity or protein–protein interactions of HBsu, which is the underlying mechanism that regulates the process of sporulation.

HBsu is an abundant, nucleoid-associated protein that can be modified by acetylation (Carabetta et al., 2016). Previously, we demonstrated that one physiological function of acetylation of HBsu is to decondense the chromosome, and a likely mechanism by which this occurs is that acetylation reduces its DNA binding affinity (Carabetta et al., 2019). We have now demonstrated that acetylation of HBsu at specific sites is an additional regulatory mechanism of the sporulation process in B. subtilis. We found that K3 was likely not acetylated during this process and that some level of acetylation is needed at K18, K37, K75, K80, and K86 (Figure 1). The acetylation status of K41 was unimportant. The mostly likely stages of sporulation that could be affected by HBsu, given its known role in chromosome dynamics, are stages I and II. In stage I, the duplicated chromosome is stretched across the long axis of the cell, arranged in an ori-ter-ori orientation (Webb et al., 1997; Wang et al., 2014). As HBsu is one of the only nucleoid-associated proteins present in B. subtilis (Ohniwa et al., 2011), it is likely that it plays a role in axial filament formation, and could be responsible for arranging the chromosome in this manner. To test this idea, super resolution microscopy with fluorescently labeled chromosomal markers (Wang et al., 2014) and a spoIIIE deletion could be used. SpoIIIE is the translocase that moves the remainder of the spore chromosome across the asymmetric septum during stage II of sporulation, and deletion of this protein causes the cells to become blocked at axial filament formation (Wu and Errington, 1994, 1998). We would expect to see random organization of chromosomal loci in the hbs mutants, if HBsu acetylation was influencing this process. HBsu may also be important during or immediately following the chromosomal translocation process. SpoIIIE strips the DNA of RNA polymerase, transcription factors and chromosome remodeling proteins, like HBsu, during the translocation step (Marquis et al., 2008), so that the remaining two thirds of the DNA is naked when it enters the spore. However, the amount of HBsu present in the forespore is equivalent to that in the mother cell (Ross and Setlow, 2000). Therefore, HBsu must be reestablished at some point on the forespore DNA molecule, and perhaps acetylation regulates the timing or location of binding, which might influence SASP binding properties. Visualizing chromosomal structure in wild-type, ssp and hbs mutant spores by techniques such as Hi-C could address this possibility.

These observations raise further interesting questions. Currently, we do not know if different acetylated species of any protein exist in the cell. In other words, among the pool of HBsu in the cell, are there only singly acetylated species, or are there some species with two, three, or more modifications? Moreover, we do not have any information about the stoichiometry of acetylation at each individual site or the relative abundance of each species, if they exist. The stoichiometry at any particular site is likely dependent upon the growth phase and environmental conditions. We are currently developing a new mass spectrometry-based workflow to make such determinations on a single protein. As of now, multiple published mass spectrometry-based methods exist for quantification of the stoichiometry of the entire proteome (Baeza et al., 2014; Huang et al., 2015; Meyer et al., 2016; Weinert et al., 2017). It is of interest to determine how acetylation patterns and stoichiometry change during the process of sporulation. Another question worth exploring is how acetylation at specific sites regulates the sporulation process. It is likely that acetylation of lysine residues that directly contact the DNA will weaken the binding affinity. This could explain why deacetylation at K3, a site that likely contacts the DNA (Carabetta et al., 2019), is important. HU-family proteins are known to have a preference for structured DNA, such as bends and kinks (Pontiggia et al., 1993; Balandina et al., 2002). Perhaps SASP binding creates specific DNA structures that can only bind specific acetylated forms of HBsu. Another possibility is that these sites lie in protein–protein interaction interfaces. This may be especially true for those modification sites which likely do not contact the DNA, like K37 and possibly K75. Acetylation of a residue in a binding interface my cause a steric hindrance, and interfere with the interaction. The identities of such interacting proteins, if they exist, are not currently known, but a possibility could be the SASPs themselves. The SASPs increase persistence length (stiffness) of DNA, which is incompatible with the organization and compaction necessary to fit the DNA into the small spore compartment (Ross and Setlow, 2000). Perhaps HBsu–SASP complexes have different activities when bound together to DNA, and therefore acetylation might be a regulatory mechanism to control when these complexes form and therefore alter SASP properties. Finally, as HBsu is a histone-like protein, it is still possible that the acetylated forms of HBsu regulate gene expression in either the mother cell or the spore compartment. This possibility is suggested from the spore wet heat resistance experiments (Figure 2 and Supplementary Figure 1B), as multiple hbs mutants displayed larger defects than the ssp mutants. For example, alterations in the lipid composition of the highly impermeable spore inner membrane results in increased sensitivity to wet heat and oxidizing agents (Griffiths and Setlow, 2009; Leggett et al., 2012). If HBsu acetylation altered the expression of the lipid synthesis genes, such as ugtP [diacylglycerol (Jorasch et al., 1998)], pssA [phosphatidylserine (Okada et al., 1994)], or ynwE [cardiolipin (de Mendoza et al., 1993)], this could change the composition of lipids in the inner membrane, resulting in increased permeability and could explain the observed defects for both heat and formaldehyde resistance. Further work is required to address this possibility.

Once the mature spore is released, the DNA should be protected from extreme environmental insults. We wondered if the spores that were formed in these various mutants, even if the overall number was reduced, were normal. We thought it likely that the DNA was not properly packaged and condensed in the hbs mutants and we determined that this was true for wet heat, UV and formaldehyde resistance (Figures 2–4 and Supplementary Figures 1–3). As the short-term resistance properties of the mutant spores were altered, it will be interesting to assess their long-term survival capabilities, and specifically their ability to germinate. The most interesting observation of this data is that the acetylation patterns are different from those observed for the sporulation process. For example, K3 is likely unacetylated during the process of sporulation, and interestingly requires some level of acetylation for stress resistance (Figures 2–4 and Supplementary Figures 1–3). When these changes in acetylation patterns occur during this process is unclear, as they could happen during axial filament formation, during spore packaging, or at a later step. An interesting possibility is that there is expression of a KAT(s) or KDAC(s) that occurs exclusively in the spore compartment, perhaps under the control of a late sporulation sigma factor, like σ^G (Fimlaid and Shen, 2015). This would allow for regulation of HBsu by acetylation to occur during different stages of the sporulation process. The ydgE and yfmK and double mutant spores were similar to wild-type, or had mild phenotypes in all conditions (Figures 2C, 3C, 4C). These mutant spores did show a significant reduction in sporulation frequency (Figure 1), suggesting that they might set the HBsu acetylation patterns early during the process, but are not the KATs responsible for changing the patterns later on. However, we previously identified three more uncharacterized KATs that might acetylate HBsu, YdhI, YokD, and YjbC (Carabetta et al., 2019). As little is currently known about the expression of these enzymes, it is possible that they are induced specifically in the spore compartment during the sporulation process. It is also possible that a deacetylase is induced in the spore compartment, which could also change the stoichiometry of acetylation. The two known B. subtilis deacetylases, SrtN and AcuC deacetylate AMP-intermediate forming enzymes, such as acetyl-CoA synthetase (Gardner and Escalante-Semerena, 2009). The exact mechanism of HBsu deacetylation remains unknown; however, in Escherichia coli, the sirtuin CobB deacetylates the HBsu orthologs HupA and HupB (AbouElfetouh et al., 2015). It is possible that SrtN is an HBsu deacetylase, but this must be confirmed. CobB has at least 50 substrates (AbouElfetouh et al., 2015), but the substrates of both SrtN and AcuC have not been determined. We found that srtN and acuC mutant spores showed significant decreases in formaldehyde survival, and acuC mutants also had decreased wet heat resistance (Figures 2, 3 and Supplementary Tables 3, 4). One possibility is that the decreases are due to changes in HBsu acetylation patterns. However, it is also possible that SrtN and AcuC have other substrates that are important for the sporulation process to proceed normally. These possibilities are not mutually exclusive. Further work is necessary to determine the substrates of SrtN and AcuC and understand the roles these enzymes play during the process of sporulation.

HBsu is part of the most widely conserved HU-family, and these proteins are generally considered to represent functional homologs of eukaryotic histones. Histone proteins contain long, unstructured N-terminal tails that are the sites of various PTMs, including lysine acetylation. The combination of these PTMs on histone tails is regulatory in nature, and constitutes the histone code (Bannister and Kouzarides, 2011; Rothbart and Strahl, 2014). These PTM combinations are hypothesized to regulate the interaction of the histone with DNA and other protein machines, and regulate most DNA processes, such as gene expression (Strahl and Allis, 2000). HBsu is acetylated at seven lysine residues, and we previously proposed that these modifications represent part of an analogous histone-like code in bacteria (Carabetta et al., 2019). Exploration of characterized acetylomes of multiple bacterial species revealed that this modification is evolutionarily conserved, even among distantly related bacteria (Carabetta, 2021). We previously determined that some level of acetylation at K3, K18, K41, K75, K80, and K86 was required for proper chromosomal compaction (Carabetta et al., 2019). If this truly represents a histone-like code, it is likely that most of the biological processes that HBsu influences would also be impacted by acetylation. We have now added sporulation to the list. It is possible that different acetylated species exist, and these influence which regions HBsu binds to on the chromosome. This could be an important regulatory mechanism to regulate SASP binding, and control the stiffness of the chromosome during spore packaging. In support of this idea, it was shown that both HBsu and the SASPs are present in the mature spore, and can bind simultaneously to the same DNA molecule in vitro (Ross and Setlow, 2000). It is possible that changing the ratios of acetylated species in the spore changes the HBsu binding locations on the chromosome, which could reduce the overall compaction or saturation with SASPs. This would explain our observed phenotypes with increased sensitivity to heat, UV and formaldehyde stresses. All together, our data lends further support that a bacterial version of a histone-like code exists.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

JL, CM, OS, and HG performed all experiments, and analyzed and interpreted the data. MB performed the statistical analyses. MB and CM made the figures. VC designed the study, and analyzed and interpreted the data. All authors were involved in the writing and revisions of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.