Resilience in the Face of Uncertainty: Sigma Factor B Fine-Tunes Gene Expression To Support Homeostasis in Gram-Positive Bacteria

Abstract

Gram-positive bacteria are ubiquitous and diverse microorganisms that can survive and sometimes even thrive in continuously changing environments. The key to such resilience is the ability of members of a population to respond and adjust to dynamic conditions in the environment. In bacteria, such responses and adjustments are mediated, at least in part, through appropriate changes in the bacterial transcriptome in response to the conditions encountered. Resilience is important for bacterial survival in diverse, complex, and rapidly changing environments and requires coordinated networks that integrate individual, mechanistic responses to environmental cues to enable overall metabolic homeostasis. In many Gram-positive bacteria, a key transcriptional regulator of the response to changing environmental conditions is the alternative sigma factor σ^B. σ^B has been characterized in a subset of Gram-positive bacteria, including the genera Bacillus, Listeria, and Staphylococcus. Recent insight from next-generation-sequencing results indicates that σ^B-dependent regulation of gene expression contributes to resilience, i.e., the coordination of complex networks responsive to environmental changes. This review explores contributions of σ^B to resilience in Bacillus, Listeria, and Staphylococcus and illustrates recently described regulatory functions of σ^B.

INTRODUCTION

Gram-positive bacteria can thrive in a myriad of environments, ranging from water, soils, and food surfaces to different types of hosts. One key to their survival lies in their ability to respond efficiently to different and rapidly changing environments by shaping their transcriptome in response to environmental conditions. A key regulator contributing to the survival of multiple Gram-positive genera under changing conditions is the alternative sigma factor σ^B, a subunit of the RNA polymerase holoenzyme. σ^B is well established as contributing to the response and survival of Bacillus, Listeria, and Staphylococcus species during exposure to a variety of adverse conditions, such as low pH, bile, and osmotic stress (1–7). One example of a mechanistically simple σ^B-mediated stress response is the σ^B-dependent expression of bsh, the gene encoding bile salt hydrolase in Listeria monocytogenes, upon encountering bile in the host digestive tract (8). σ^B also plays a clear role in L. monocytogenes' acid response; σ^B is activated in response to many types of acid stress (e.g., exposure to acidic pH of 2.5 to 4.5) (9) and directly upregulates the transcription of many genes that encode effector proteins that counteract acid stress (10–14). In this review, we contrast a simple, mechanistic definition of stress response, such as that mediated by increased expression of bile salt hydrolase, with the concept of resilience, which includes but is not limited to stress response. Resilience is defined by the Merriam-Webster dictionary as “an ability to recover from or adjust easily to misfortune or change.” Resilience, therefore, requires not only a successful response to an initial change or stress but also the ability of an organism to continue to respond to or to take advantage of subsequent changes in the environmental conditions. For pathogens, the concept of resilience includes the production of the virulence factors needed to mediate survival within hosts, e.g., for evasion of the host immune system and the ability to obtain nutrients from the host environment.

The enteric pathogen Listeria monocytogenes offers remarkable examples of σ^B-mediated responses to the complex and rapidly changing environmental conditions encountered in animal hosts, which provide this pathogen with the resilience capacity to cause human infection. To illustrate, bacterial exposure to an initial stress condition (e.g., acidic pH in the stomach) not only triggers the expression of acid resistance functions that facilitate survival in this environment but also upregulates the transcription of genes important for survival in subsequent host compartments. For example, the σ^B-mediated response to low pH also enables the invasion of intestinal epithelial cells, which is partially but not solely facilitated by σ^B-dependent contributions to the transcription of invasion proteins (e.g., L. monocytogenes InlA) (15, 16). Thus, σ^B-dependent contributions to resilience go beyond simple upregulation of a specific stress response gene. This review summarizes recent insights into the σ^B regulon in Bacillus, Listeria, and Staphylococcus species, including the complex σ^B-mediated regulatory network interactions that facilitate the integrated and coordinated fine-tuning of physiological functions important for bacterial resilience. Importantly, all three genera include environmentally transmitted and foodborne pathogens, which require the ability to adapt, survive, and grow in diverse and rapidly changing environments for successful transmission to and from their animal hosts.

THE ALTERNATIVE SIGMA FACTOR σ^B

Sigma factors are required for transcription initiation, as they confer promoter specificity to the RNA polymerase holoenzyme and induce helix destabilization to expose the DNA template strand for RNA synthesis. A primary sigma factor, σ^A, in Gram-positive bacteria (17) is required for housekeeping functions essential for cellular growth and reproduction. In general, bacteria possess at least one alternative sigma factor (e.g., σ^B) to mediate specialized functions, such as cell differentiation, biofilm formation, and modulation of an appropriate response to changing environmental conditions (18).

σ^B is one of the most comprehensively studied alternative sigma factors of Gram-positive bacteria. This sigma factor is found in a subset of Gram-positive bacteria and has been identified in Bacillus, Listeria, and Staphylococcus species, as well as in other genera in the order Bacillales (e.g., Oceanobacillus and Paenibacillus) (19–21). σ^B was first identified in 1980 in the Gram-positive model organism Bacillus subtilis (22). While the specific function of σ^B in B. subtilis was initially unknown, subsequent work showed that this alternative sigma factor plays a key role in bacterial survival under adverse conditions, including entry into stationary phase (3–6). In the late 1990s, σ^B was also identified in L. monocytogenes, as well as in other Listeria species (1) and in Staphylococcus species (23). The specific functions initially described for σ^B in L. monocytogenes included bacterial responses to acid (1) and osmotic stress (2). Similarly, the roles of σ^B originally described in Staphylococcus aureus involved heat, acid, and hydrogen peroxide resistance (7).

σ^B activity is tightly controlled both transcriptionally and posttranscriptionally. The multiple levels of regulation allow the bacterial cell to rapidly induce σ^B activity in response to different environmental conditions. Interestingly, the systems regulating σ^B differ considerably among Staphylococcus, Listeria, and Bacillus species (Fig. 1). Both B. subtilis (as well as Bacillus licheniformis, Bacillus halodurans, Bacillus clausii, and Oceanobacillus iheyensis) and L. monocytogenes (as well as Listeria innocua and Listeria welshimeri) possess an 8-gene cluster that includes the rsbVW-sigB-rsbX operon, regulated by a σ^B-dependent promoter, and an upstream rsbRSTU operon, which is transcribed from a σ^A-dependent promoter (1, 24–26). Interestingly, Bacillus cereus (as well as Bacillus anthracis and Bacillus thuringiensis) only has a 4-gene rsbVW-sigB-rsbY operon, which includes separate σ^B promoters upstream from each of the genes rsbV and rsbY, as well as an additional σ^A promoter upstream from rsbY (Fig. 1A) (27–29). S. aureus lacks rsbRST as well as rsbX; the resulting rsbUVW-sigB operon includes a σ^A promoter upstream from rsbU and a σ^B promoter upstream from rsbV (30). Despite the differences in operon structure, all of these organisms incorporate a σ^B-dependent positive-feedback loop that regulates the transcription of the sigB and rsb genes.

FIG 1.

Regulation of SigB expression and activity. (A) Conservation of sigB and rsb genes and operon structures across Gram-positive bacteria. Filled arrows represent open reading frames color coded by homology. Thin bent arrows indicate promoters, and the respective sigma factors are shown. Sequence, organization, conservation, and promoter information was obtained from NCBI (http://www.ncbi.nlm.nih.gov/), BioCyc (http://biocyc.org/), and STRING (http://string-db.org/). (B) Signal transduction leading to SigB activation. See the text for details.

Posttranscriptional regulation of σ^B activity involves a phosphorylation cascade that is catalyzed by the regulation of sigma B (Rsb) proteins. During exponential growth, σ^B is held in an inactive state, bound to the anti-sigma factor RsbW. The activation of σ^B requires binding of the dephosphorylated form of the anti-anti-sigma factor RsbV to anti-sigma factor RsbW, which then releases σ^B, thereby allowing σ^B to bind to RNA polymerase and, thus, to trigger transcription at cognate promoters (Fig. 1B) (19, 31). The control of σ^B activity by RsbVW is highly conserved in species containing σ^B (26); however, the upstream regulation of this sigma/anti-sigma/anti-anti-sigma partner-switching module differs among genera. In B. subtilis and L. monocytogenes, the phosphatases RsbP and RsbU dephosphorylate RsbV in response to energy stress (32–34) and environmental stress (35–40), respectively (reviewed in reference 20; 34, 41). RsbU activity is controlled by the switch kinase RsbT and its antagonist RsbS. Together with RsbR and RsbS, RsbT is part of the “stressosome,” a multiprotein signaling hub that includes multiple copies of each protein (reviewed in reference 42). In contrast to B. subtilis, posttranscriptional regulation of RsbVW-σ^B in B. cereus involves the RsbK multisensory histidine kinase and the RsbY phosphatase (43). Similarly, S. aureus also lacks RsbRST, and RsbU appears to play a different role than in B. subilis (44, 45). Importantly, posttranslational activation of σ^B, which is conserved across organisms, allows for very rapid response of bacterial cells to changes in environmental conditions (induction of σ^B activity takes <5 min) (3, 46), which is critical for resilience under rapidly changing environmental conditions.

DIVERSE ROLES OF σ^B IN BACTERIAL RESILIENCE IN LISTERIA, STAPHYLOCOCCUS, AND BACILLUS SPECIES

The number of genes regulated by σ^B varies between species (Table 1). The σ^B regulon in B. subtilis has been reported to include between 150 and 200 genes, depending on the experimental conditions used in a given study (19, 44, 47, 48). Some of the key roles of σ^B in Bacillus species include resistance to heat, acid, starvation, nitric oxide, and osmotic stress and antibiotics, and σ^B is also involved in integrating the stress adaptation and sporulation pathways (49). Key stress response functions regulated by σ^B in Listeria species include the responses to acid, oxidative, and energy stress (1, 10, 50). Overall, σ^B appears to upregulate ∼200 genes in L. monocytogenes (51–54). The σ^B regulon in S. aureus appears to encompass about 200 genes (52–55), the products of which, in addition to general stress response, are also involved in resistance to several clinically important antibiotics, such as methicillin and vancomycin (55–58).

TABLE 1.

Overview of σ^B functions in B. subtilis, L. monocytogenes, and S. aureus

Organism	Size of σB regulon (references)	Associated function(s) (reference[s])
B. subtilis	∼150 genes (19, 47, 48, 165)	Sporulation (49)
		Antibiotic resistance (61, 166)
		Growth and starvation (167)
		Transitional growth phase (168)
L. monocytogenes	∼130 genes (8, 72, 162)	Osmotic, cold, and acid stress (10, 11, 50, 70, 169)
		Virulence (8, 15, 16)
		Host cell invasion (15)
		Bile resistance (170)
		Attachment community formation (98)
		Antibiotic resistance (62)
S. aureus	∼200 genes (51–54)	Antibiotic resistance (58, 171–174)
		Virulence (175)
		Cell envelope homeostasis (89)
		Persistence (176)
		Biofilm (177)
		Host cell internalization (148)
		Intermediate metabolism (52)
		Membrane transport processes (52)

Despite a number of conserved functions and features of σ^B-dependent regulatory systems, the specific functions regulated by σ^B differ considerably among species. One hypothesis drawn from these observations is that the roles of σ^B and the σ^B regulon may have evolved independently to facilitate bacterial survival under conditions encountered in specific environments. For example, σ^B facilitates survival of acid stress in Bacillus and Listeria species (reviewed in references 19 and 31), whereas the contributions of σ^B to acid stress resistance appear to be minimal in S. aureus (59). Osmotic stress induces σ^B activity in both Listeria and Bacillus species (13, 60), and σ^B plays a role in antibiotic resistance for Bacillus, Listeria, and Staphylococcus species (58, 61, 62). Interestingly, in B. subtilis, the σ^B regulon plays a role in the response to nitric oxide (NO). While the flavohemoglobin encoded by hmp is the main detoxifier of NO (63), a study using reporter fusions to a σ^B-dependent promoter showed that σ^B activity increased upon nitric oxide stress under aerobic conditions in B. subtilis strain PB198 (64). Interestingly, depending on the source of the nitrous stress, the pathway by which σ^B was activated differed: NO gas induced σ^B activity via RsbP and the energy stress pathway, while sodium nitroprusside as a NO donor induced σ^B activity via RsbU and the environmental-stress-dependent pathway (64). The results of a study comparing the responses to the NO donor MAHMA-NONOate at the protein level in B. subtilis and S. aureus confirmed the induction of the σ^B regulon by NO in B. subtilis but not in S. aureus (44). This is not surprising, since S. aureus lacks the RsbP-dependent regulatory pathway to induce σ^B activity (52).

σ^B clearly plays a broad and diverse role in regulating gene expression among different Gram-positive genera. Multiple lines of evidence indicate that σ^B is involved in many functions beyond the regulation of individual genes that facilitate survival of a specific stress. In the following sections, we highlight these broader functions by illustrating recent findings of the involvement of σ^B in bacterial resilience, including σ^B-dependent contributions to (i) metabolism, (ii) cell envelope homeostasis, (iii) biofilms, and (iv) pathogenesis.

σ^B IN METABOLISM OF HARMFUL COMPONENTS AND UTILIZATION OF DIFFERENT CARBON SOURCES

Increasing evidence supports the idea that σ^B contributes to the regulation of metabolic functions, including (i) metabolism of harmful components to allow survival and (ii) adaptation to allow the utilization of different carbon sources (19, 31, 65), as detailed below. Both of these metabolic functions clearly contribute to bacterial resilience, as the abilities to counteract harmful compounds and to nimbly adapt the cell's metabolism to changing energy sources are critical for the growth and survival of bacteria exposed to rapidly changing and complex environments.

One example of σ^B's contributions to bacterial survival in the presence of harmful components is the σ^B-mediated expression of bsh, which encodes a bile salt hydrolase, in L. monocytogenes (8). The expression of bile salt hydrolase allows L. monocytogenes to survive bile encountered in the small intestine. For a foodborne pathogen, exposure to bile occurs shortly after bacterial passage from the stomach to the duodenum. We speculate that the activation of σ^B in the acidic environment of the stomach prepares L. monocytogenes for subsequent survival in the hostile environment of the small intestine, which not only contains bile but also represents a hyperosmotic environment (66, 67). Another example of σ^B-mediated metabolic capabilities relevant to resilience is the consumption of protons by decarboxylation of glutamate to γ-aminobutyrate (GABA) (GAD system), which helps to elevate the intracellular pH and, hence, facilitates bacterial survival in acidic environments; the expression of L. monocytogenes gad genes is, in part, σ^B dependent (68–70).

Increasing evidence also indicates that L. monocytogenes σ^B is involved in complex networks regulating carbohydrate metabolism-related functions, including phosphotransferase systems (PTS) and the metabolism of GlcNAc or glycerol. Oliver et al. identified a σ^B-dependent promoter upstream from the mpo operon (mpoABCD) that encodes a mannose-specific PTS (71). Conversely, a proteomics study showed that σ^B negatively regulates other PTS components, including Lmo1997, Lmo1998, Lmo2002, Lmo0427, Lmo0484, and Lmo2648 (72). A recent study by Wang et al. (73) that used deletion mutants to assess the growth of L. monocytogenes in medium supplied solely with PTS-dependent carbon sources found that growth under these conditions was dependent on coregulation by σ^B and two other alternative sigma factors, σ^L and σ^H, in a complex regulatory network that was influenced by temperature and the respective carbon source (glucose, mannose, cellobiose, or glycerol) (74). Importantly, PTS-mediated carbohydrate uptake is also linked to virulence (75, 76). A number of studies have shown that when PTS systems are active, positive regulatory factor A (PrfA), the predominant transcriptional regulator of virulence genes, is downregulated via several intermediate steps that are not yet fully resolved (77). The currently proposed model involves PTS^Mpo, PTS^Man, and the activator of the man operon, ManR. Glucose uptake through PTS^Mpo activates ManR by dephosphorylation, which in turn upregulates the transcription of the manLMN operon encoding PTS^Man (78, 79). Uptake of glucose through PTS^Man then results in PrfA inhibition via a mechanism that appears to involve dephosphorylation of the EIIAB^Man subunit (77, 78). These data not only indicate an involvement of σ^B in L. monocytogenes carbohydrate metabolism but also link this involvement with regulation of PrfA activity and, therefore, virulence. L. monocytogenes σ^B also regulates an operon involved in glycerol metabolism (80); specifically, proteomics data indicate that σ^B upregulates three proteins (Lmo2695 to Lmo2697) that appear to be subunits of dihydroxyacetone kinase, which is part of the glycerol metabolism pathway. The identification of a σ^B promoter consensus sequence upstream from lmo2695 and observation of the reduced ability of an L. monocytogenes sigB mutant to use glycerol as a sole carbon source further support σ^B's contributions to the regulation of glycerol metabolism (80). Glycerol is used as a non-PTS-dependent alternative carbon source by intracellular L. monocytogenes bacteria. Upregulating glycerol metabolism also upregulates PrfA activity, again highlighting a link between carbohydrate metabolism and virulence functions (81).

L. monocytogenes σ^B also regulates the transcription of the nagABR operon (13, 72), which encodes two deaminases necessary for N-acetylglucosamine (GlcNAc) degradation (NagA and NagB) (82) and NagR, which functions as a transcriptional inhibitor in the absence of GlcNAc (83). Apart from being a vital part of the bacterial cell wall, monomeric GlcNAc is also a major component of chitin; therefore, GlcNAc is among the most abundant carbon sources in the environment. L. monocytogenes metabolizes GlcNAc in a temperature-dependent fashion, but only in the absence of glucose catabolites (84). σ^B-dependent regulation of GlcNAc catabolism thus represents another example of a situation where σ^B contributes to bacterial resilience by facilitating the use of alternative carbohydrate sources in the absence of glucose. The mechanism of catabolite repression that appears to be involved in this regulatory circuit remains to be elucidated.

In S. aureus, σ^B appears to contribute indirectly to the regulation of hyaluronidase, which hydrolyzes hyaluronic acid. Hyaluronic acid is a component of the extracellular matrix present in many tissues that can be used as a carbon source by S. aureus and many other bacteria (85). The transcription of S. aureus hyaluronidase, encoded by hysA, is controlled by the accessory gene regulator (Agr) quorum-sensing system (86, 87). Work by Ibberson et al. (88) suggests that σ^B (as well as CodY) downregulates agr and therefore indirectly and negatively controls hyaluronidase activity in S. aureus. Based on the observation that sigB and codY mutants express higher levels of hyaluronidase activity, hysA is proposed to be under direct positive control by the effectors of Agr and negatively modulated by CodY and σ^B (88). The specific mechanisms affecting hysA transcription levels by altering the balance among Agr, CodY, and σ^B and the question of whether these mechanisms contribute to virulence by providing access to a readily available host carbon source remain to be elucidated.

σ^B IN CELL ENVELOPE HOMEOSTASIS

The cell wall ensures bacterial integrity by maintaining a physical barrier between the cell and its environment and by giving it its shape. Maintenance of cell wall homeostasis is an important factor for bacterial resilience during growth and adaptation to changing conditions. σ^B-dependent gene regulation contributes to such resilience in various ways, including (i) cell envelope homeostasis and modification of cell envelope composition in the absence of stress (12, 89) and (ii) translation of stress signals at the cell envelope into upregulation of virulence factors (90). This section concentrates on L. monocytogenes and S. aureus, as there are no clear data on contributions of σ^B to cell wall homeostasis in Bacillus species.

In L. monocytogenes, via a σ^B-dependent promoter, σ^B positively regulates the transcription of dapE (71), which encodes a key intermediate (mesodiaminopimelate) of the peptidoglycan synthesis pathway (80). DapE is upregulated in stationary-phase bacteria (91) and within host cells and is speculated to mediate the anchoring of surface proteins to the cell wall (92). Abram et al. (12) reported that a sigB mutant showed unusual Gram-staining properties, even in the absence of stress, supporting σ^B-dependent regulation of bacterial cell wall components. In S. aureus, σ^B plays a role in cell wall homeostasis via regulation of asp23 expression. Asp23 has been used in many S. aureus studies as a marker for σ^B activity because its transcription is exclusively regulated by σ^B (93). It is also one of the most abundant proteins in stationary-phase S. aureus, but until recently, its function remained unclear. Muller et al. (89) showed that Asp23 is anchored to the cell wall by AmaP and that improper localization or deletion of Asp23 results in the upregulation of cell wall stress genes. These findings suggest that Asp23 plays a role in cell envelope homeostasis in stationary-phase cells and that the disruption of either the production of Asp23 or its correct localization results in increased stress for the cell. A recent study by Ishii et al. (90) also indicated that the response to environmental stress sensed at the S. aureus cell wall results in σ^B-dependent upregulation of virulence genes. They used RNA-seq to explore the transcriptomic change in S. aureus strain Newman in response to exposure to surfactant, which would be one of the first hurdles encountered by S. aureus upon reaching the lung. According to the Centers for Disease Control and Prevention (CDC), lung infections by S. aureus occur mainly in polymorbid patients and account for 13 to 15% of clinical methicillin-resistant Staphylococcus aureus (MRSA) infections (http://www.cdc.gov/abcs/reports-findings/survreports/mrsa12.pdf). Three key genes were upregulated in S. aureus exposed to surfactant, including essC (encoding a type VII secretion system) (94), hlgB (encoding hemolysin gamma, which induces pores in target cells) (95), and psiA (encoding a protein with similarity to proteins involved in lipid metabolism). This unique response to surfactant as a stimulus was found to be σ^B dependent, as the mRNA levels for essC, hlgB, and psiA did not increase in response to surfactant in a sigB mutant strain. In contrast, none of the general inducers of virulence genes in S. aureus (Agr, ArlS, and Sae) were necessary to elicit the observed regulatory response to surfactant. Mouse infection experiments also showed that mutations in essC, hlgB, and psiA resulted in reduced numbers of bacteria recovered from the lungs and a lower death rate, further supporting a role for σ^B in the production of cell wall components (EssC) and lipid metabolism (PsiA) with a proposed link to virulence. These data provide an example of σ^B-dependent regulation both of cell wall-associated functions and of proteins with potential virulence-related functions that facilitate resilience of S. aureus in a hostile host-associated environment.

σ^B IN BIOFILMS

The formation of either true biofilms with a typical extracellular matrix or bacterial attachment communities (which do not show the accumulation of extracellular matrix typical for traditional biofilms) facilitates bacterial survival in a variety of different environments, such as in food processing facilities or clinical settings, including on foreign objects that may be placed in a patient for extended periods of time (e.g., catheters and implants). As such, biofilms represent resilient bacterial communities. Increasing evidence indicates that σ^B may play a role in biofilm formation or the establishment of attachment communities in Bacillus, Listeria, and Staphylococcus species (96–98).

While the main transcriptional regulator for B. subtilis biofilms is Spo0A (99), σ^B also appears to contribute specific regulatory functions to biofilm formation in this organism. The role of Spo0A-dependent regulation in B. subtilis is complex, and the control of Spo0A levels is not yet completely understood (for a review, see reference 100). Briefly, intermediate levels of Spo0A induce the production of a biofilm matrix via a signaling cascade of repressors and anti-repressors, including SinI, SinR, SlrR, and AbrB, while high levels of Spo0A accumulate in some cells during biofilm maturation and induce sporulation and dispersion from the mature biofilm. Spo0A is active in its phosphorylated form, and the phosphatase Spo0E regulates Spo0A activity by dephosphorylating Spo0A (101). Reder et al. identified a σ^B-dependent promoter upstream from spo0E in silico and experimentally showed σ^B-dependent transcription of spo0E (49). A sigB mutation in B. subtilis strain 168 severely affected biofilm formation and morphology, a phenotype that was shown by Nagorska et al. (96) to involve σ^B-dependent expression of the putative exopolysaccharide (EPS) synthetase yxaB. Using quantitative reverse transcription (qRT)-PCR, yxaB was determined to be transcribed in an operon with yxaA; in silico analysis and experiments using reporter fusions identified a σ^B-dependent promoter upstream from the transcriptional start site of yxaA. The biofilm-deficient phenotype of the sigB mutant could be partially rescued by overexpression of yxaB, which resulted in structured, floating biofilms that were more fragile than the biofilms formed by the parent strain (96).

In L. monocytogenes, many aspects of biofilm formation still remain to be explored (102). A number of studies show that L. monocytogenes has the ability to rapidly and strongly adhere to inanimate surfaces (103, 104). However, attempts to unambiguously show that L. monocytogenes forms true biofilms with the typical accumulation of extracellular matrix find either a lack thereof (105, 106), a very sparse extracellular matrix (107, 108) with strain-dependent formation of extracellular fibrils of unclear function (109), or ball-shaped structures held together by chains of interconnected cells (110). To date, the exact nature of the fibrils in L. monocytogenes biofilms remains to be determined. A study that observed filaments in biofilms formed by certain strains of L. monocytogenes argued that these filaments were probably flagella, because they were not observed in nonmotile strains of L. monocytogenes (111). Some studies (98, 112) suggest that σ^B plays a role in establishing Listeria attachment communities. For example, van der Veen and Abee (98) showed differential sigB expression in L. monocytogenes strain EGD-e during growth on surfaces. While posttranscriptional regulation of σ^B activity may inhibit a proportional increase in σ^B activity in parallel with increased sigB transcript levels, higher sigB transcript levels were found in attachment communities on polystyrene in static medium than in planktonic cells grown in broth, a finding that was even more pronounced under continuous-flow conditions. The cell density of attachment communities on surfaces was significantly lower for a sigB null mutant than for a wild-type strain (98) in both static and continuous flow systems, further supporting a role for σ^B in Listeria attachment to surfaces. In another study, no difference in adherence of the cells to stainless steel surfaces was observed between an L. monocytogenes 10403S sigB null mutant and a wild-type strain (113), suggesting that the involvement of σ^B in the formation of attachment communities and biofilms may be strain specific and/or may depend on environmental conditions.

In S. aureus, a recent study found reduced biofilm formation in sigB and sarA mutants (as well as in atl, codY, and rsbU mutants), with the sarA mutation having the largest effect; as sarA is positively regulated by σ^B, the effect of σ^B on biofilm formation may be due to reduced sarA transcription in the sigB null mutant (114). The same study suggested that impaired biofilm production in sigB and sarA mutants is linked to increased production of extracellular proteases that may degrade the biofilm matrix in these strains. Negative regulatory effects of σ^B and SarA on the production of extracellular proteases (115) are supported by the fact that the biofilm-impaired phenotype of sigB and sarA mutants is further reduced in a protease-deficient S. aureus strain compared to the biofilm formation by strains with the same mutations in a wild-type background (114).

Interestingly, the effects of a sigB null mutation on biofilm formation were influenced by the culture conditions: the addition of human plasma to the growth substrate reversed the sigB mutation phenotype, reflecting the complexity of the regulatory network that fine-tunes biofilm regulation. Additional circumstantial evidence for σ^B involvement in S. aureus biofilm formation has been reported (97); reduced biofilm generation was observed in naturally occurring S. aureus “white variants,” which carry mutations in their sigB coding sequence. However, no experimental data were presented to exclude the possibility that the biofilm formation defect in the white variant strains is due to mutations in genes other than sigB. A possible mechanistic explanation for the reduced cell mass in biofilms formed by the white variant strains is that sigB mutations may relieve the inhibitory effect of σ^B on the Agr regulon, which occurs in the wild type (116, 117). Increased AgR activity, in turn, upregulates extracellular nucleases, proteases, and hemolysins that degrade the extracellular matrix of the biofilm.

A recent study showed that the S. aureus hyaluronidase HysA regulates the amount of host hyaluronic acid that is incorporated into biofilms in vivo. hysA in turn is under indirect negative control by σ^B (118). Interestingly, several studies support a role for σ^B in the biofilm formation of Staphylococcus epidermidis; mutations in sigB or rsbU lead to reduced biofilm production (119, 120), and the stability of biofilms under nutrient limitation is σ^B dependent (121). σ^B-dependent biofilm regulation involves the ica operon in an oxygen-dependent way (122, 123). ica encodes the polysaccharide intercellular adhesin (PIA) that is crucial for intercellular adhesion during biofilm formation. In the model proposed by Knobloch et al. (120), ica transcription is downregulated by IcaR, which is inhibited in turn by σ^B, resulting in a net positive effect of σ^B on PIA production. This conclusion is supported by the reduced biofilm formation phenotype of sigB mutant strains and by the results from Northern blot analyses of ica and icaR transcript levels in various strains: in cells with an intact σ^B operon, icaR transcript levels are low while ica transcript levels increase, and the inverse effects occur in the absence of active σ^B (120).

σ^B IN PATHOGENESIS

Perhaps the most illuminating examples that illustrate the contributions of σ^B to resilience are the roles of this alternative sigma factor in pathogenesis, virulence, and survival in the rapidly changing environments encountered in the host. Specifically, for foodborne and enteric pathogens, the ability to survive the changing conditions encountered along the gastrointestinal tract is essential to host invasion. Bacterial responses to changes in pH, temperature, and bile and salt concentrations are core functions of σ^B, as discussed above. In addition, there is clear evidence that σ^B also regulates functions directly relevant for pathogenesis and the coordinated expression of virulence genes in the host. While preliminary data on the σ^B regulon in the pathogenic Bacillus anthracis suggested that a B. anthracis sigB null mutant may be less virulent in a mouse model (124), further studies on the contributions of σ^B to virulence in pathogenic Bacillus species are still needed. Hence, the section below focuses on Listeria and Staphylococcus.

While PrfA is a key transcriptional regulator of the core virulence genes in L. monocytogenes, σ^B contributes to the larger network that regulates virulence factors; specifically, a number of studies have shown considerable and complex overlaps between the σ^B and PrfA regulons (8, 13, 125, 126), including direct transcriptional regulation of genes by both σ^B and PrfA. Importantly, σ^B also contributes to the transcriptional regulation of prfA itself, through the prfA P2 promoter, one of three prfA promoters. Contributions of σ^B to transcription from the prfA P2 promoter have been confirmed by phenotypic characterization of deletion mutants (127), as well as by experiments with reporter fusions (128) and in vitro transcription assays (129). Interestingly, the prfA P2 promoter is not only σ^B dependent but also contains a PrfA binding site. The specific role of this PrfA binding site for regulating prfA transcription remains to be clearly defined (129–132). A number of studies have also shown that σ^B regulates the transcription of inlA (8, 9, 133) via the σ^B-dependent inlA P4 promoter, one of four inlA promoters (8, 9, 133). inlA encodes internalin A, which is the binding partner of host cell E-cadherin (134). This interaction plays a crucial role in the attachment of L. monocytogenes to host cells and their subsequent internalization by the host cells (135). The importance of σ^B-dependent transcription of inlA has been demonstrated through guinea pig oral infection experiments, which showed clear virulence attenuation for both a sigB null mutant and a mutant with the σ^B-dependent inlA P4 promoter deleted (133). The importance of InlA for host cell invasion may explain why a sigB mutant is less invasive in CaCo-2 cells, which express high levels of E-cadherin (136), and in guinea pig intestinal infection models but not when the gastrointestinal tract was circumvented by inoculating guinea pigs intravenously (133). This study also indicated a role for σ^B in virulence beyond the upregulation of InlA, as further supported by the fact that a sigB null mutant showed more severe virulence attenuation in a guinea pig oral infection model than a strain carrying only a mutation in the σ^B-dependent inlA P4 promoter, indicating a role for σ^B in addition to regulation of inlA through the P4 promoter.

Additional important roles for σ^B in fine-tuning gene expression during L. monocytogenes infection include indirect repression of the flagellar genes (91) and mediation of reduced expression of the PrfA regulon (126). σ^B-dependent downregulation of flagellar genes (91) occurs through a long 5′ untranslated region (UTR) that is transcribed from a σ^B-dependent mogR P1 promoter. This 5′ UTR acts as an antisense RNA to the flagellar genes on the opposite strand. Overexpression of the mogR P1 transcript leads to decreases in motility and flagellar gene transcript levels in L. monocytogenes. This observation also explains previous findings that L. monocytogenes sigB mutants show increased motility at 30°C (13), as the absence of the σ^B-dependent antisense RNA in the 5′ UTR of mogR removes one level of repression. Interestingly, Ollinger et al. (126) showed that, in the presence of high levels of active PrfA, σ^B appears to indirectly downregulate the transcript levels of genes in the PrfA regulon, including hly and actA. These classical, PrfA-dependent L. monocytogenes virulence genes play pivotal roles during pathogenesis by enabling bacterial escape from the phagocytic vacuole (hly) and spread to neighboring host cells (actA). Indirect negative regulation of virulence by σ^B was most prominent in L. monocytogenes strains carrying two mutations: a prfA* mutation (137) that renders the PrfA transcriptional regulator of virulence genes constitutively active, along with a concomitant sigB null mutation. σ^B-dependent modulation of PrfA regulon expression reduced the cytotoxic effects of a PrfA* strain in HepG2 cells, possibly facilitating extended intracellular survival and, hence, reduced exposure to extracellular antimicrobial compounds. While this σ^B-dependent downregulation has been shown at both the transcriptional (through microarray hybridization analysis) and the phenotypic (through hemolysis assays) level (126), the mechanism remains to be elucidated. Overall, the modulating role of σ^B in the expression of flagellar, hly, and actA genes might serve to balance virulence protein levels during the challenging first stages of host infection. For example, downregulation of hly may prevent host cell lysis. This hypothesis is consistent with the findings of Glomski et al. (138), who showed that a strain carrying an hly allele that is more active than that of the wild type was attenuated in virulence. σ^B-dependent downregulation of flagellar gene expression, on the other hand, may serve to attenuate the host immune response. Overall, σ^B-mediated resilience of L. monocytogenes during host infection thus involves many mechanisms, including resistance to key stress conditions encountered by the pathogen. Such resistance can result from direct regulation of a single gene (e.g., bsh) or from modulation of regulatory networks, which together facilitate survival at both the extra- and intracellular stages of infection.

Conflicting data exist on the role of σ^B in S. aureus virulence, in part because an important experimental strain (S. aureus 8325-4) and its derivatives harbor a mutation in rsbU, which codes for a key positive regulator of σ^B activity (93). These rsbU mutant strains can cause human infections (139), but this mutation explains why early studies performed with these strains (7, 140, 141) found no influence of σ^B on virulence. However, results from S. aureus virulence investigations are inconsistent even in strains with an intact σ^B operon. For example, Bischoff et al. (54, 116) and Atwood et al. (114) found that σ^B positively regulates sarA, one of the main transcription factors of virulence genes, in S. aureus strain Newman (54, 116) and USA300 strains (114). These results were not confirmed by Horsburgh et al. (142), who found no effect of a sigB deletion in S. aureus SH100 on SarA at either the transcriptional or the translational level and no difference in pathogenicity between the wild-type strain and its isogenic sigB mutant in a mouse skin abscess model (142). Other in vivo studies found decreased virulence of sigB mutants in models of arthritis (in S. aureus SH100) (143) and metastatic organ infections (144) but no effects of a sigB deletion on pathogenicity in a mouse pyelonephritis and rat osteomyelitis model (S. aureus WCUH29) (145). The roles of SarA in these in vivo studies cannot be compared because only Horsburgh et al. (142) specifically investigated sarA expression levels. It is conceivable that σ^B acts as a modulator of virulence in S. aureus, a conclusion that has been put forward by several authors (54, 146, 147) and is supported by a recent study that evaluated the role of σ^B during intracellular growth of S. aureus in a host cell line (148). This study showed σ^B-dependent changes in the expression of several genes at the mRNA and protein levels. Specifically, σ^B was found to positively regulate the expression of the transcription factor encoded by spoVG and to negatively regulate hemolysin A (hla) expression, most likely through σ^B-dependent regulation of genes encoding other transcription factors (139). SpoVG directly regulates a small subregulon that includes genes involved in virulence (149); therefore, σ^B-dependent transcription of spoVG might be responsible for indirect σ^B-dependent effects on the transcription of genes that lack a direct σ^B-dependent promoter (55). The mechanism of negative regulation of hla by σ^B is unknown. Additionally, σ^B involvement in S. aureus virulence appears to occur through σ^B-dependent expression of adhesins, exoproteins, and toxins (54) and through the expression of virulence genes in small-colony variant (SCV) S. aureus (150). Several studies showed that the role of σ^B-dependent virulence gene expression is more prominent in SCVs than in the parent strains and that σ^B is involved in the emergence of SCV subpopulations (151–153). In contrast, Bui et al. (154) found a single-nucleotide polymorphism in the rsbU gene in a line of S. aureus that stably grew as an SCV, which seems to indicate that a strain with largely inactive σ^B is also able to grow with the SCV phenotype. SCVs are associated with chronic infections (for a review, see reference 155), and their emergence appears to result from adaptation to the host environment. The observation that the σ^B regulon diverges between SCVs and the parent strains illustrates the dynamic nature of gene regulation, including possible consequences of selection pressures that may drive modification of regulatory networks to allow for resilience under different conditions.

In S. aureus, σ^B indirectly and negatively regulates the virulence factor toxic shock syndrome toxin 1 (TSST-1) (156) and the putative virulence factor hyaluronidase (88). TSST-1 is the causative toxin of the potentially fatal toxic shock syndrome and is encoded by tst on a mobile genetic element present in some S. aureus strains. Andrey et al. (156) showed that σ^B indirectly represses tst transcription via a mechanism involving SarA and Agr. While hyaluronidase provides bacteria access to an abundant carbon source within host tissue (88), it is also speculated to play a direct role in pathogenesis by making host tissue more penetrable, although this notion has not been proven experimentally (157–160). Supporting this hypothesis is the fact that a hysA mutant is attenuated in a mouse skin abscess model (86). Incidentally, the parent strain used in these experiments (S. aureus 8325-4) is likely to express increased hyaluronidase activity due to a mutation in rsbU, which constitutively suppresses σ^B activity (142, 161), thereby abolishing the indirect negative effect of σ^B on hysA transcription.

The increasing evidence supporting roles for σ^B and other transcriptional regulators in bacterial resilience, both in extra- and intrahost environments, also provides an opportunity for discovery of new drugs for the treatment or prevention of infections with Gram-positive pathogens such as L. monocytogenes or S. aureus. For example, small molecules that interfere with σ^B activation or the assembly of a σ^B-containing RNA polymerase holoenzyme could represent potential therapeutics. The feasibility of identifying such agents is supported by studies that have identified a small molecule that interferes with σ^B activation in L. monocytogenes and B. subtilis (162), as well as another small molecule that inhibits the binding of the Escherichia coli extracellular stress sigma factor σ^E to the polymerase core enzyme (163). Targeting transcriptional regulators that are broadly involved in resilience may also provide an opportunity to develop drugs with reduced risk for the development of resistance mechanisms that become fixed in a pathogen population, as resistance-conferring mutations will likely reduce an organism's resilience in diverse environments.

CONCLUSIONS

While σ^B has been characterized as a stereotypical stress response alternative sigma factor, emerging evidence suggests a much broader role for this alternative sigma factor in resilience, as supported by contributions of σ^B and the σ^B regulon to metabolism, cell envelope homeostasis, and biofilm formation and to pathogen transmission and virulence across different Gram-positive bacteria. σ^B not only facilitates bacterial survival under one or more stress conditions, including during a series of sequential stress exposures, but also provides critical contributions to the regulation of gene expression in complex and rapidly changing environments, such as during host infection or during growth and survival in extrahost environments, including in foods. σ^B appears to facilitate resilience by inducing the expression of genes that allow for survival and growth under subsequently encountered stress conditions. In the case of Gram-positive pathogens, such as L. monocytogenes, these connections suggest that the extrahost environment can affect infectivity and virulence; for example, through modulating the expression of stress response and virulence genes that play important roles in the initial phases of the pathogen-host interaction. Additionally, the connection between σ^B and pathogenesis further supports mechanistic links between stress response and virulence and opens up interesting avenues to target sigma factors for drug development (162, 164). Importantly, our emerging recognition of the broad contributions of σ^B to bacterial resilience have been critically aided by the emergence of high-throughput methods (e.g., RNA-seq) and an integrative approach to understanding regulatory networks as a whole, in addition to the study of isolated regulons. Future high-resolution single-cell approaches for characterizing gene expression (e.g., single-cell RNA-seq) will likely provide new insights into the roles of σ^B in bacterial resilience. We hypothesize that stochastic gene expression patterns may generate highly resilient bacterial subpopulations. Future studies will also likely identify additional transcriptional regulators that contribute to bacterial resilience, leading to recognition of resilience as a key theme in bacterial physiology.

Biographies

Claudia Guldimann received her veterinary degree from the University of Bern, Switzerland, in 2009. In parallel to her studies, Claudia completed a doctoral thesis at the University of Bern. After a stint in a rural Swiss mountain practice, she proceeded to a postgraduate position working on Listeria monocytogenes central nervous system infections at the University of Bern, receiving a Ph.D. in 2013. She found that the fact that Listeria monocytogenes is relevant to human food safety, as well as to animal health, makes it a fascinating model to study host-pathogen interactions in a one-health approach and proceeded to work on Listeria monocytogenes gene expression at the single-cell level as a postdoctoral associate at Cornell University. Claudia recently returned to Switzerland to a postdoctoral position at the University of Zürich, where she will continue to work on host-pathogen interactions in Listeria monocytogenes.

Kathryn J. Boor is the Ronald P. Lynch Dean of the College of Agriculture and Life Sciences at Cornell University, Ithaca, NY. She joined the Cornell Food Science Department as an Assistant Professor in 1994 and was named Department Chair from 2007 to 2010. Dr. Boor earned a B.S. in Food Science from Cornell University, an M.S. in Food Science from the University of Wisconsin—Madison, and a Ph.D. in Microbiology in 1994 from the University of California, Davis. Her research focuses on biological factors that affect bacterial survival in complex systems, an interest stemming from her work on developing safe food production and handling systems for small-scale farmers in East Africa in the 1980s. Her Ph.D. dissertation work in Chet Price's laboratory at UC Davis provided Dr. Boor with the tools of molecular biology to interrogate bacterial stress responses mediated through RNA polymerase, all with the ultimate goal of improving food safety.

Martin Wiedmann, Dr. med. vet., Ph.D., Gellert Family Professor in Food Safety, Cornell University, received a veterinary degree and a doctorate in Veterinary Medicine from the Ludwig Maximilians University in Munich (1992 and 1994) and a Ph.D. in Food Science from Cornell (1997). He currently is the Gellert Family Professor of Food Safety at Cornell University. His research interests focus on farm-to-table microbial food quality and food safety and the application of molecular tools to study the biology and transmission of foodborne pathogens and spoilage organisms. He is a fellow of the Institute of Food Technologists (IFT), a fellow of the American Academy of Microbiology (AAM), and a member of the International Academy of Food Science and Technology. He also received the Young Scholars award from the American Dairy Science Association (2002), the Samuel Cate Prescott Award from IFT (2003), the International Life Science Institute North America Future Leaders Award (2004), and the American Meat Institute Foundation Scientific Achievement Award (2011).

Veronica Guariglia-Oropeza, who is originally from Caracas, Venezuela, obtained a B.S. in Biology, Genetics and Biochemistry at the Central University of Venezuela (2006). In 2008, she moved to the United States to obtain her Ph.D. in Microbiology at Cornell University under the mentorship of John Helmann, studying the role of extracytoplasmic function sigma factors in the cell envelope stress response (2013). Veronica then joined the Food Safety Laboratory of Kathryn Boor and Martin Wiedmann as a postdoctoral associate, where she has since been promoted to research associate. Veronica leads a group of graduate and undergraduate students with the collective goal of employing cutting-edge technologies to define the regulatory response and key components that enable pathogens to survive under adverse conditions and, thus, have the power to develop improved control interventions.