SUMMARY
SUMMARY Sigma factors provide promoter recognition specificity to RNA polymerase holoenzyme, contribute to DNA strand separation, and then dissociate from the core enzyme following transcription initiation. As the regulon of a single sigma factor can be composed of hundreds of genes, sigma factors can provide effective mechanisms for simultaneously regulating expression of large numbers of prokaryotic genes. One newly emerging field is identification of the specific roles of alternative sigma factors in regulating expression of virulence genes and virulence-associated genes in bacterial pathogens. Virulence genes encode proteins whose functions are essential for the bacterium to effectively establish an infection in a host organism. In contrast, virulence-associated genes can contribute to bacterial survival in the environment and therefore may enhance the capacity of the bacterium to spread to new individuals or to survive passage through a host organism. As alternative sigma factors have been shown to regulate expression of both virulence and virulence-associated genes, these proteins can contribute both directly and indirectly to bacterial virulence. Sigma factors are classified into two structurally unrelated families, the σ70 and the σ54 families. The σ70 family includes primary sigma factors (e.g., Bacillus subtilis σA) as well as related alternative sigma factors; σ54 forms a distinct subfamily of sigma factors referred to as σN in almost all species for which these proteins have been characterized to date. We present several examples of alternative sigma factors that have been shown to contribute to virulence in at least one organism. For each sigma factor, when applicable, examples are drawn from multiple species.
INTRODUCTION
Sigma factors are a class of proteins constituting essential dissociable subunits of prokaryotic RNA polymerase. The association of appropriate alternative sigma factors with core RNA polymerase provides a mechanism for cellular responses mediated through redirection of transcription initiation. Sigma factors provide promoter recognition specificity to the polymerase and contribute to DNA strand separation; they then dissociate from RNA polymerase core enzyme following transcription initiation (16). As the regulon of a single sigma factor can be comprised of hundreds of genes, sigma factors provide effective mechanisms for simultaneously regulating large numbers of prokaryotic genes. In some cases, the genes comprising a sigma factor regulon have a clearly defined primary function (e.g., genes regulated by the sporulation sigma factors in Bacillus subtilis [171]); in others, the genes comprising a regulon contribute to multiple functions (e.g., the stationary-phase and general stress response genes regulated by σB in Listeria monocytogenes [100]). One newly emerging field is identification of the specific roles of alternative sigma factors in regulating expression of virulence genes and virulence-associated genes in bacterial pathogens.
Virulence and virulence-associated genes are those that contribute to at least one aspect of bacterial disease transmission and infection processes. Specifically, virulence genes encode proteins whose functions are essential for the bacterium to effectively establish an infection in a host organism. Examples of virulence genes are L. monocytogenes inlA, which encodes the internalin-A protein important for invasion of nonprofessional phagocytes (129), and the spv gene cluster of Salmonella enterica, which allows for bacterial growth inside macrophages (128). In contrast, virulence-associated genes can contribute to bacterial survival in the environment (e.g., the ica operon of Staphylococcus aureus, which produces an adhesin important for biofilm formation on plastic surfaces such as those on indwelling medical devices [141]) or to survival in the host (such as bsh of L. monocytogenes, encoding bile salt hydrolase, which enhances bacterial survival in the intestinal environment prior to intracellular infection [48]). Therefore, activation of virulence-associated genes may enhance the capacity of the bacterium to spread to new individuals or to survive passage through a host organism. As alternative sigma factors have been shown to regulate expression of both virulence and virulence-associated genes, these sigma factors can contribute both directly and indirectly to bacterial virulence.
Virulence factor expression appears to be tightly regulated in bacterial pathogens. In some cases, pathogens have a “master regulator” of virulence gene expression, such as the positive regulatory factor A (PrfA) in L. monocytogenes. PrfA, a transcriptional activator, is required for expression of the majority of recognized L. monocytogenes virulence genes. Alternative sigma factors often function to regulate expression of virulence and virulence-associated genes in response to particular stimuli. Alternative sigma factors may regulate a small number of genes, each of which may be critical to infection (e.g., PvdS of Pseudomonas aeruginosa [discussed below] [147]), or they may regulate functions that contribute to virulence but also have additional physiological roles in the cell. For example, Salmonella enterica serovar Typhimurium σE regulates genes that provide resistance to oxidative stress, which also aids bacterial survival in macrophages (82). This review focuses on both direct and indirect roles of selected alternative sigma factors in regulating virulence of bacterial pathogens of plants and animals.
Sigma factors can be classified into two structurally unrelated families: the σ70 and the σ54 families. Table 1 lists sigma factors in both the σ70 and the σ54 families that are currently recognized as contributing, either directly or indirectly, to bacterial virulence. For several alternative sigma factors, nomenclature in the literature has been inconsistent. In this document, in general, we refer to sigma factor families by number (e.g., the σ54 family) and to specific sigma factors by letter (e.g., P. aeruginosa σN). For certain sigma factors, we use the predominant designation from the literature instead (e.g., FliA).
Alternative sigma factors involved in virulence
The σ70 family includes primary sigma factors (e.g., Bacillus subtilis σA) as well as related alternative sigma factors (145, 164). Alternative sigma factors within the σ70 family are further categorized by the physiological processes they control, e.g., stress response. In general, these groupings by function also correlate with phylogenetic relationships among the protein sequences (164). Within the σ70 family of sigma factors is a large, phylogenetically distinct subfamily called the extracytoplasmic function (ECF) factors. These sigma factors are responsible for regulating a wide range of functions, all involved in sensing and reacting to conditions in the membrane, periplasm, or extracellular environment (70). Structurally, σ70 family factors have four major regions, with the highest levels of conservation in regions 2 and 4. Subregions within region 2 are involved in promoter melting (region 2.3) and −10 sequence recognition (region 2.4). Region 4.2 is involved in −35 recognition. For a recent review on the σ70 family of sigma factors, see reference 164.
Although no sequence conservation exists between σ54 and σ70-like family members, both types bind to core RNA polymerase. However, the holoenzyme formed with σ54 sigma factors has different properties than the σ70 holoenzyme. While the C terminus (region III) of σ54 enables DNA binding, all σ54 species require a separate activator protein along with the core RNA polymerase (RNAP) to form an open promoter complex. The σ54 N terminus, which inhibits isomerization in the absence of the appropriate activator, stimulates initiation upon activation (19). Further, promoter structures recognized by σ54-RNAP differ from those recognized by σ70-RNAP. σ54 promoters are highly conserved, short sequences that are located at positions −24 and −12 upstream of the transcription initiation site, whereas σ70 promoter sites are typically located at −35 and −10 upstream. σ54 promoters, which are called −24/−12 promoters, are almost completely invariant at the −24/−12 positions (GG and GC, respectively) and in their spacing in both gram-negative and gram-positive bacteria. For reviews on the structure-function relationships of σ54, see references 19 and 142.
We present several examples of alternative sigma factors that have been shown to contribute to virulence in at least one organism. The text is organized by sigma factor to include the three subfamilies (stress response, σ28, and ECF) within the σ70 family, as well as those within the σ54 family. For each sigma factor, when applicable, examples will be drawn from multiple bacterial species.
STRESS RESPONSE ALTERNATIVE SIGMA FACTORS
The ability to reproduce, or simply survive, under a wide variety of environmental conditions contributes to a microbial pathogen's potential for transmission by various routes. For example, to establish a food-borne infection in a human host, a bacterium first must survive transit in a contaminated food. Following ingestion, the bacterium then must survive exposure to rapid and dramatic changes in environmental conditions, including the acidic pH within the stomach, followed by vastly differing conditions during intestinal passage and/or infection (e.g., exposure to bile, vacuolar stresses, etc.) Survival under these extreme and rapidly changing conditions requires timely and appropriate alterations in gene expression and protein activity that occur in a bacterial cell in response to stimuli signaling these new environmental conditions. At the transcriptional level, these alterations are often controlled by changes in associations between different alternative sigma factors and core RNA polymerase, which essentially reprogram promoter recognition specificities of the enzyme to allow expression of new sets of target genes.
The general stress-responsive alternative sigma factors σS (RpoS) and σB transcribe genes contributing to bacterial survival under conditions of environmental stress in gram-negative and in gram-positive bacteria, respectively (Table 2). σS was identified in both Escherichia coli and S. enterica serovar Typhimurium as an alternative sigma factor that activates the expression of numerous genes required to maintain cell viability during stationary phase (51, 119). σS also plays a key role in protecting E. coli and S. enterica serovar Typhimurium from different environmental stress conditions, including starvation, hyperosmolarity, oxidative damage, and reduced pH (51, 119). Since its initial discovery, the presence of σS and its role in the stress response has been confirmed in many gram-negative bacterial species, including P. aeruginosa, Borrelia burgdorferi, and Vibrio cholerae (49, 93, 227). Through enhancing environmental survival, as well as by directly activating virulence genes, σB and σS have both direct and indirect roles in bacterial pathogenesis.
Virulence genes and virulence-associated genes regulated by stress response sigma factors σB and σS and phenotypes of sigma factor null mutants in selected bacterial species
Sigma BσB (initially called σ37) of Bacillus subtilis was among the first bacterial alternative sigma factors identified (65, 66). In B. subtilis and related species such as L. monocytogenes and S. aureus, σB activity increases in response to numerous environmental stresses, including exposure to acid, ethanol, and heat (12, 22, 53). The σB regulon in B. subtilis contains at least 127 genes, including those with functions in stress resistance, transcriptional regulation, and membrane transport (169, 174). In B. subtilis and L. monocytogenes, sigB, which encodes σB, is the seventh open reading frame in an operon containing eight genes involved in σB regulation (rsbR, rsbS, rsbT, rsbU, rsbV, rsbW, sigB, and rsbX) (Fig. 1A) (54, 223). All eight genes, including sigB, are cotranscribed from a housekeeping sigma factor (σA)-dependent promoter (PA) located upstream of rsbR. A σB-dependent promoter (PB), located upstream of rsbV, is responsible for enhanced transcription of the four downstream genes in the sigB operon (rsbV, rsbW, sigB, and rsbX) under conditions that stimulate σB activity (11, 12, 95).
(A) sigB operon structures in various gram-positive bacteria. Promoter sites are marked by arrows. PA promoters are transcribed with RNAP-σA, and PB promoters are transcribed with RNAP-σB. (B) Posttranslational regulation of σB activity via a partner-switching mechanism. Proteins shown are encoded by the B. subtilis and L. monocytogenes sigB operons depicted in panel A. Arrows indicate activation of protein activity, and T-bars indicate repression of protein activity. “P” represents a phosphate group. The proteins indicated by dark gray are encoded in all bacteria listed in panel A, whereas RsbU (light gray) is absent in the pathogenic Bacillus spp. and the proteins indicated by white are encoded only in L. monocytogenes and B. subtilis.
While regulation of σB activity involves both transcriptional and posttranslational control, predominant regulation occurs via the Rsb proteins. Bacterial species differ in the numbers and identities of rsb genes carried in their genomes (Fig. 1A), suggesting divergent evolution of the sigB operon, both in its overall components and in the sequences of its individual proteins, even among closely related bacterial species (54). We hypothesize that differential evolution of the σB stress response system among various genera has enabled different bacteria to optimize cellular response and survival strategies that are appropriate for highly specific niches.
Although all seven Rsb proteins identified in B. subtilis and L. monocytogenes are not conserved among all bacterial species bearing σB, two key proteins (RsbV and RsbW) are conserved among all species examined to date and thus appear to be minimally essential for regulating σB activity (54). Specifically, in log phase, nonstressed B. subtilis cells, σB is inactivated by its association with the anti-σB protein, RsbW (i.e., the “anti-sigma factor”). In stressed cells, however, the unphosphorylated form of the anti-σB antagonist protein, RsbV, (i.e., the “anti-anti-sigma factor”) competes for binding to RsbW. As the relative concentration of the RsbW-RsbV complex increases, the concentration of free σB also increases, thus allowing σB to bind to core RNA polymerase (47). In B. subtilis, both environmental and energy stress signals induce dephosphorylation of RsbV. Environmental stress signals specifically activate the B. subtilis RsbU serine phosphatase through involvement of RsbR, RsbS, RsbT, and RsbX (211, 212, 223). In addition to its role in partner-switching regulation under environmental stress, B. subtilis RsbX also functions as a feedback regulator for σB activity (Fig. 1B) (12). While both energy and environmental stresses have been shown to activate L. monocytogenes σB (25), specific interactions among the Rsb proteins have not yet been investigated. To date, specific activation mechanisms have been most extensively reported for B. subtilis σB.
Pathogenic Bacillus species.At least two pathogenic species of Bacillus encode σB (55, 208). In Bacillus anthracis, only σB, RsbV, and RsbW are encoded in the sigB operon (Fig. 1A). A third rsb gene, rsbY, encodes a protein with low similarity to B. subtilis RsbP. rsbY is located in close proximity to, but not within, the B. anthracis sigB operon (55). As in B. subtilis and L. monocytogenes, the sigB operon is autoregulated by σB and is induced by heat shock and entry into stationary phase (55). A B. anthracis sigB mutant strain is virulence attenuated, producing less than half the mortality of the parent strain in the mouse model of anthrax (55). More-detailed studies are needed to determine if virulence attenuation is due to direct or indirect effects.
The organization of the Bacillus cereus sigB operon is identical to that of B. anthracis sigB, and likewise, the sigB operon is autoregulated by σB and is induced by heat shock and entry into stationary phase (208). Through use of a combination of two-dimensional gels and Northern hybridizations, 15 B. cereus genes and proteins were determined to be σB dependent, including RsbV and the KatE catalase (209). The activity of currently recognized B. cereus virulence factors, including protease, lecithinase, and hemolytic activity, as well as production of nonhemolytic enterotoxin, was not affected by disruption of sigB (208), suggesting that σB does not directly contribute to B. cereus virulence.
Staphylococcus species. Staphylococcus aureus was the first pathogenic bacterium in which sigB was identified. (115, 224) (Fig. 1A). In S. aureus, the sigB operon is comprised of four genes, which are homologous to B. subtilis rsbU, rsbV, rsbW, and sigB. As in B. subtilis, all genes in the operon are expressed during exponential growth, presumably from the σA-dependent promoter. The internal PB promoter was confirmed as σB dependent through in vitro transcription analyses (44). Transcriptional regulation of the S. aureus sigB operon is complex, generating multiple transcripts that appear to include a bicistronic sigB-rsbW transcript as well as a sigB monocistronic transcript. In support of an autoregulatory role for S. aureus σB under conditions of environmental stress, an rsbV-W-sigB transcript was induced following exposure of cells to either 4% ethanol or a 48°C heat shock (115).
S. aureus σB activity is regulated posttranslationally by Rsb proteins. The open reading frame immediately upstream of S. aureus sigB encodes the anti-sigma factor, RsbW (146). As in B. subtilis, S. aureus σB also is activated via an RsbU pathway (62). An 11-bp deletion in rsbU in the NCTC8325 strain generated some phenotypic characteristics similar to those of a ΔsigB strain, (e.g., decreased H2O2 resistance [114]). Giachino et al. (62) confirmed that NCTC8325 does not produce a functional RsbU and that complementation of this strain with a complete rsbU allele restored phenotypes to those of the rsbU+ Newman wild-type strain. However, some NCTC8325 phenotypes were identical to those of other rsbU+ strains (e.g., lipase production [see below]), suggesting the existence of multiple S.aureus σB activation pathways, including at least one that is RsbU independent (114). As with RsbU, loss of RsbV results in a dramatic decrease, although not complete loss, of S. aureus σB activity (165).
Through application of full-genome microarray screens for σB-dependent genes in three S. aureus strains, as many as 251 genes have been identified as being σB regulated (14), including several genes encoding proteins involved in synthesis of capsular polysaccharides. A number of adhesins, which are involved in Staphylococcus virulence, are upregulated by σB. Multiple genes encoding exoenzymes and toxins (e.g., hla and nuc) are downregulated as σB is activated (14), which may reflect σB's role in controlling expression of S. aureus virulence gene regulators. For example, a number of the exoenzymes and toxins that are downregulated by σB depend on an effector RNA produced from the agr locus (RNAIII) for heightened expression (204). RNAIII levels are reduced when σB activity increases. The mechanism responsible for this phenomenon remains unclear (13, 14) but may involve the regulator SarA (8, 44, 79).
Multiple groups have described S. aureus ΔsigB mutants as having pigment loss and decreased peroxide resistance, but higher α-hemolysin, coagulase, clumping factor, and lipase activity, compared to the wild type (27, 62, 79, 114, 152). These characteristics have all been associated with S. aureus virulence (61, 67, 98, 131, 150, 184). It is likely that optimal levels of virulence factor expression and activity are required for efficient S. aureus infection and that too much or too little activity or expression at the wrong time is detrimental for the infection process. These hypotheses remain to be rigorously tested. In various animal models, wild-type S. aureus and an otherwise isogenic ΔsigB strain showed no difference in virulence (22, 152). In additional, conflicting studies, Horsburgh et al. (79) found no difference in virulence between rsbU+ and rsbU mutant strains in a murine skin abscess model, while Jonsson et al. (92) showed that both rsbU and sigB mutant strains displayed decreased virulence phenotypes compared to the wild-type strain in murine septic arthritis, including reduced mutant persistence in kidneys and reduced mouse mortality, weight loss, arthritis, and interleukin-6 production.
The contradictory evidence surrounding the role of σB in S. aureus virulence suggests that σB contributions to virulence may be indirect or not detectable in some model systems. For example, σB may contribute indirectly to S. aureus virulence through regulation of biofilm formation. Biofilm formation can be a prerequisite for establishing infection by staphylococci, and σB has been shown to enhance microcolony and biofilm formation in Staphylococcus species (5, 177). Two studies have shown induction of S. aureus biofilm formation in a σB-dependent fashion (5, 177), although another showed that a ΔsigB strain formed biofilms and produced PIA, the polysaccharide adhesin encoded by the ica operon, equally as well as the wild type (207). S. aureus σB contributions to biofilm formation likely occur through σB-dependent transcription of the ica operon, which encodes essential elements of biofilm biosynthesis (177).
Staphylococcus epidermidis also encodes σB. The sigB operon of S. epidermidis is similar to that of S. aureus (Fig. 1A); however, σB serves different functions in the two species. Processing of lipase, a virulence factor, is dependent on σB in S. epidermidis (102), while in S. aureus, lipase production is higher in a sigB mutant than in the wild-type strain (114). Multiple studies of σB and S. epidermidis virulence suggest that σB's effects are mediated primarily through its influence on biofilm formation in this organism (33, 106, 107). Stress induction of σB in S. epidermidis increases biofilm formation and synthesis of the adhesin PIA, but an rsbU mutant does not form biofilms or produce PIA (106). As in S. aureus, S. epidermidis σB contributes to biofilm formation via regulating expression of ica genes. By downregulating the icaR repressor, active σB causes an increase in icaA expression and a biofilm-positive phenotype (107).
Listeria monocytogenes.σB has also been extensively studied in the gram-positive pathogen Listeria monocytogenes. While the sigB operon structures are identical in L. monocytogenes and in B. subtilis (11, 219) (Fig. 1A), signal transduction pathways differ in the two organisms. In B. subtilis, environmental and energy stresses are conveyed to σB through two interconnected but separate pathways. The environmental stimulus pathway is transmitted by regulatory proteins encoded in the sigB operon (RsbT, RsbU, RsbV, and RsbW). In addition to requiring RsbV and RsbW, the B. subtilis energy stress pathway also requires proteins encoded in a two-gene operon (rsbQ-rsbP) that is physically separate from the sigB operon (18). This operon is not present in L. monocytogenes. Instead, both energy stress and environmental stress activation of σB in L. monocytogenes occurs through a single pathway, which includes RsbT, RsbU, RsbV, and RsbW (25).
A genome-wide search for predicted σB-dependent promoters by using a hidden Markov model followed by application of a specialized, partial microarray identified 54 genes under positive control of σB in L. monocytogenes, although the full regulon is likely to be as large as that of B. subtilis (100). σB regulates expression of virulence and virulence-associated genes in L. monocytogenes (Fig. 2A; Table 2). bsh encodes a bile salt hydrolase that is important for virulence of L. monocytogenes (48) and is directly regulated by σB (100, 199). Another recently identified σB-dependent virulence-associated gene is hfq, which encodes an RNA-binding regulatory protein (29). Deletion of the σB-dependent opuC (57), which encodes an osmotransporter, also negatively affects L. monocytogenes virulence (194, 217).
Examples of regulatory networks involving sigma factors and other transcriptional regulators or multiple sigma factors. (A) The σB-PrfA network of L. monocytogenes. Some genes are activated solely by σB (e.g., hfq and opuCA), some solely by PrfA (e.g., actA and hly), and some by both factors (e.g., inlA and bsh). (B) The short sigma factor cascade regulating type III secretion in P. syringae. σN mediates transcription of hrpL, which encodes a sigma factor responsible for transcription of the hrp and avr genes of the type II secretion system, as well as other virulence genes. All transcription depicted in panels A and B is the result of direct activity by the sigma factor or regulator at the promoter sites. (C) The complex interaction of several sigma factors that affect virulence in M. tuberculosis. Multiple sigma factors activate expression of other sigma factors and of virulence-associated genes (white). The interactions depicted here were deduced from global expression profiles and may be the result of either direct or indirect regulation by the sigma factor(s).
As in S. aureus, L. monocytogenes σB also controls expression of virulence gene regulators (Fig. 2A). Of the two promoters directly upstream of the gene encoding positive regulatory factor A (PrfA), P2prfA is σB dependent. Dual deletion of sigB and the σA-dependent P1prfA promoter (leaving only the σB-dependent P2prfA) reduced hemolytic activity and intracellular growth to the same low levels as deletion of both prfA promoters (151). σB activity at the P2prfA promoter was also directly confirmed, both by quantitative reverse transcription-PCR (101) and with β-glucuronidase reporter fusions of prfA promoters, which demonstrated σB- and growth phase-dependent expression from P2prfA (191).
Several PrfA-regulated genes are also σB dependent, suggesting interplay between the two regulators (143) (Fig. 2A). For example, expression of the PrfA-regulated inlA gene, which encodes the cell surface protein internalin-A, is also at least partially σB dependent (100, 200). Internalins are cell wall-anchored proteins with important roles in the intracellular pathogenesis of L. monocytogenes, and several members of the internalin gene family show reduced expression in a sigB mutant compared to in the wild type (100). Internalin-A, specifically, is responsible for invasion of nonprofessional phagocytes (129). Loss of σB reduced invasiveness of the mutant strain compared to that of wild-type L. monocytogenes in two intestinal epithelial cell lines, Henle-407 and Caco-2 (103). In addition, inlA transcription was greatly reduced in the ΔsigB strain, and internalin-A was undetectable by Western blotting (103). None of the effects of the sigB deletion on inlA were mediated through loss of σB-dependent transcription of PrfA, however, as a ΔP2prfA strain had the same levels of invasiveness, inlA transcription, and internalin-A concentration as the wild-type strain.
Wiedmann et al. (219) tested the effect of a sigB deletion on virulence in a mouse model and found a small, but significant, decrease in spread of the mutant strain to the liver compared with that of the wild-type strain. Mouse infection experiments have been widely used to evaluate virulence characteristics of L. monocytogenes, including the preliminary evaluations of the ΔsigB mutant (151, 219). In recent years, however, increasing evidence suggests that the murine model does not appropriately represent human L. monocytogenes infection by the oral route (121, 122). The gastric pH of mice is higher than the pH of the human stomach (97); thus, the role of L. monocytogenes acid tolerance is likely to be less important in mouse infection than in human infection. More importantly, in the human, L. monocytogenes has the ability to cross the intestinal barrier, the blood-brain barrier, and the fetoplacental barrier. Human E-cadherin acts as an L. monocytogenes internalin-A receptor, and the interaction between the receptor and the internalin-A surface protein contributes to the ability of L. monocytogenes to target and cross human intestinal and placental barriers (123). Murine E-cadherin, which differs in amino acid sequence from human E-cadherin, does not interact effectively with L. monocytogenes internalin-A, and hence mice show limited susceptibility to intragastric L. monocytogenes infection (121). In fact, in the mouse, translocation of L. monocytogenes across the intestinal barrier is typically no greater than that of the nonpathogenic Listeria innocua. Further, L. monocytogenes also does not appear to target the murine brain stem or the fetoplacental unit, even following intravenous injection (121, 122). L. monocytogenes strains do vary in their ability to cause systemic infection in intragastrically infected mice (38), and some strains of mice (A/J) are also more susceptible than others (C57BL/6) to intragastric infection (39). However, as a consequence of the biological differences in murine and human L. monocytogenes translocation across the intestinal barrier, data from mouse infection experiments may underestimate a given strain's human virulence following oral infection.
The guinea pig has emerged as a more appropriate model than the mouse for studying oral L. monocytogenes infection (121, 122). Like humans, guinea pigs exhibit gastroenteritis following L. monocytogenes infection by the oral route (122). Cultured guinea pig epithelial cells allow internalin-A-dependent L. monocytogenes entry, and both guinea pig and human E-cadherins bear a proline at critical amino acid position 16. When guinea pigs were inoculated orally with L. monocytogenes strain EGD or an otherwise isogenic ΔinlA strain, significantly higher numbers of the wild type than of the ΔinlA strain were recovered from guinea pig liver and spleen. In contrast, in the mouse model, low, statistically indistinguishable wild-type and ΔinlA numbers were recovered from mouse organs (121). Lecuit et al. (121) also demonstrated that transgenic mice expressing human E-cadherin enable bacterial invasion of host cells. Taken together, these results illustrate the importance of appropriate internalin-A/E-cadherin interactions in the development of systemic listeriosis following oral infection with L. monocytogenes.
Mycobacterium tuberculosis σB and σF. Mycobacterium tuberculosis, a high-GC-content bacterium, has 13 sigma factors (for a review, see reference 137). Two of these 13, σB and σF, appear to share an evolutionary origin (54). M. tuberculosis σF appears more similar to σB of the low-GC gram-positive bacteria than to σF of B. subtilis, which is a sporulation factor. Specifically, M. tuberculosis σF is antigenically closer to B. subtilis σB (43) and has the same consensus promoter recognition sequence (10, 60), and expression patterns for its encoding gene are similar to those of B. subtilis sigB (42, 133). As with B. subtilis σB, M. tuberculosis σF is regulated posttranslationally by an anti-sigma factor and anti-anti-sigma factor partner-switching mechanism (10). The gene encoding M. tuberculosis σF is immediately downstream of the gene encoding its anti-sigma factor, UsfX, as is the case with B. subtilis σB and its anti-sigma factor, RsbW. M. tuberculosis sigB, on the other hand, is located 3 kb downstream of the gene for the primary sigma factor, σA, and is not flanked by genes encoding sigma factor regulatory proteins (45). The sigB genes in M. tuberculosis and in L. monocytogenes also share some characteristics. For example, expression of M. tuberculosis sigB is growth phase dependent, as is expression of sigB in other species (42, 80). The same studies also showed that sigB transcription is induced under a variety of stresses, including peroxide stress, heat shock, and cold shock. In spite of these observations on σB stress induction, no studies on contributions of this protein to either M. tuberculosis stress resistance or virulence have been reported.
Microarray analysis of the M. tuberculosis σF regulon identified ahpC, a gene implicated in virulence, as greatly reduced in expression in a ΔsigF mutant (60). In addition, another sigma factor, sigC, which is required for M. tuberculosis lethality in mice (202), is also σF dependent (Table 3). Several studies have linked M. tuberculosis σF with virulence. Mice infected with a ΔsigF strain displayed a longer time to death than mice infected with the wild-type strain, and the weight loss caused by wild-type M. tuberculosis did not occur in mice infected with the mutant strain (26). In a separate study, CFU counts recovered from the lungs and spleens of infected mice were approximately 40 times higher for the wild type than for the ΔsigF strain. Histopathological analyses showed that the ΔsigF mutant caused fewer, smaller granulomas and less inflammation than the wild type (60) after 12 weeks. In summary, multiple lines of evidence support direct and indirect roles for σF in M. tuberculosis virulence.
Genes regulated by mycobacterial alternative sigma factors and phenotypes of sigma factor null mutants
Sigma S (RpoS)In gram-negative bacteria, RpoS (σS) is functionally similar to σB in that it is responsible for stationary-phase and stress response gene expression. The chromosomal organizations of the rpoS and sigB loci, as well as the transcriptional and posttranscriptional regulatory mechanisms for these genes and proteins, are distinctly different, however. Regulation of σS expression and activity is extremely complex, relying on transcriptional, translational, and posttranslational mechanisms (for a thorough review, see reference 75). Further, a sequence comparison of 31 σ70 family sigma factors groups Escherichia coli σS separately from B. subtilis σB, indicating that while σB and σS may have similar functions, they are not highly homologous proteins (132).
Escherichia coli.A few reports have examined associations between σS and E. coli virulence, but little direct evidence of a link exists. Wang and Kim (214) demonstrated that E. coli K1 invasion of brain microvascular endothelial cells was higher for stationary-phase cells than for exponentially growing cells, possibly due to stationary-phase activity of σS. Indeed, complementation of rpoS into an rpoS mutant significantly increased invasion for one E. coli isolate but not for another (214). σS is not essential for murine urinary tract colonization (37) and actually appeared to be detrimental during competitive colonization experiments in the mouse intestine (113). It is also possible that the lack of an appropriate animal model for investigating all aspects of E. coli pathogenesis (e.g., the absence of an appropriate model for studying hemolytic uremic syndrome infections caused by enterohemorrhagic E. coli [193]) has impeded identification of a direct role for σS in E. coli pathogenesis.
It is likely that σS contributes indirectly to E. coli pathogenesis. E. coli O161:H7 strains tend to be acid resistant, and rpoS mutants show decreased acid resistance and fecal shedding in mice and cattle (175). Several studies have shown that rpoS transcription and σS activity are induced under stress conditions such as osmotic shock, heat, and low pH and that survival of rpoS mutants is reduced under these same conditions (3, 37, 59, 74, 215). Thus, in addition to enabling survival in high-acid and high-salt foods, σS may enhance E. coli host survival and transmission.
Salmonella species. S. enterica serovar Typhimurium σS is highly similar to E. coli σS, in both function and regulation. In contrast with E. coli, however, numerous studies have shown the unequivocal dependence on σS for full virulence of S. enterica serovar Typhimurium. For example, the plasmid-borne spv gene cluster is required for S. enterica serovar Typhimurium virulence, and several studies have demonstrated that transcription of this gene cluster is σS dependent (51, 68, 112, 157) (Table 2). In fact, an rpoS mutant is up to 10-fold less virulent than an rpoS+, plasmid cured (spv-negative) strain, and the levels of plasmid-cured rpoS+ bacteria in the intestine were significantly higher than those of plasmid-cured rpoS mutants (51, 153), indicating that the effect of σS on virulence is likely due to its role in regulating expression of chromosomal genes in addition to its effects on the plasmid-borne spv locus. In addition, mouse-based virulence assays show that in comparison to the wild-type strain, the rpoS mutant has a 3- to 4.5-log-unit higher 50% lethal dose (LD50) (35, 51, 153). Similarly, an rpoS aroA strain was more virulence attenuated than an aroA strain, which has been used in vaccine candidate trials, as determined by spleen bacterial counts and time-to-death analyses (35). σS does not contribute to levels of S. enterica serovar Typhimurium adherence, invasion, or intracellular survival, however (153).
Further evidence for the role of σS in Salmonella virulence was obtained through analysis of rpoS alleles from recognized avirulent or virulence-attenuated strains. For example, the Salmonella enterica serovar Typhi vaccine strain Ty21a contains an rpoS sequence that generates a nonfunctional σS (181). Virulence attenuation in the Salmonella enterica serovar Typhimurium LT2 strain may be a consequence of low levels of rpoS mRNA translation due to the presence of a rare UUG start codon on the transcript (124, 203). As in laboratory-generated rpoS mutants, the LT2 strain is greatly decreased in its ability to reach the spleen and liver of mice (221).
Pseudomonas aeruginosa. P. aeruginosa produces many exotoxins that contribute to its pathogenesis. σS appears to have multiple regulatory roles in P. aeruginosa. In some cases, σS positively regulates P. aeruginosa toxin expression; in others, it negatively regulates expression; and in still others, it appears to have no effect at all. For example, in an rpoS mutant, both exotoxin A and alginate production are approximately 50% of that of the wild type (196, 201) (Table 2). However, both reduced rpoS expression (109) and loss of σS (196, 201) resulted in increased expression of pyocyanin, an antibiotic that also inhibits lymphocyte proliferation. In two studies, loss of σS was shown to have little to no effect on production of elastase or LasA protease (196, 201). Some of the phenotypic effects on P. aeruginosa virulence factor production that are associated with loss of σS may be indirect, for example, resulting from reduced expression of quorum-sensing systems (Table 2). σS contributes to expression of members of the P. aeruginosa rhl and las quorum-sensing systems (168, 218). These quorum-sensing gene products are responsible for regulating production of several virulence factors, including lectins (190, 222); aminopeptidase, endoproteinase, and lipase (158); and rhamnolipid (166, 232). Several studies have shown quorum-sensing mutants to be avirulent or less virulent than the wild-type strain in mouse (167, 185, 195, 232), amoeba (34), and rat models (126). Finally, the role of σS in P. aeruginosa virulence is highly dependent on the model system in which it is assessed. For example, while an rpoS mutant was as virulent as the wild type in a rat chronic lung model (201), it was approximately half as virulent as the wild-type strain in a Galleria mellonella (silk moth) larva model (196).
SIGMA 28 SUBFAMILY
σ28 is a subfamily of the σ70-like sigma factors. Members of this subfamily are structurally and functionally related and span many genera of both gram-positive and gram-negative bacteria. While the primary regulatory role of σ28 in many bacterial species is to transcribe genes required for flagellar synthesis and bacterial motility (69, 144), it also contributes to other functions. For example, in the nonmotile Streptomyces coelicolor, σ28 contributes to expression of a diverse set of genes, including those responsible for sporulation and agarase production (96). Examples of σ28 factors are FliA of enteric bacteria and σD of B. subtilis.
Salmonella enterica serovar Typhimurium.As in many enteric bacteria, the genes for flagellar biosynthesis and function in S. enterica serovar Typhimurium are divided into three hierarchical classes based on their temporal order of transcription (116). One operon, flhDC, is categorized into class I. The flhDC operon encodes activators required for transcription of the class II operons, including fliA, which encodes the σ28 subfamily sigma factor responsible for expression of the class III genes (84, 161), and flgM, which encodes the FlgM anti-sigma factor that regulates activity of FliA (63, 162). The remaining class II genes encode proteins responsible for formation of the flagellar basal body and hook apparatus. Through an additional posttranslational regulatory mechanism, following formation of the flagellar structure, FlgM is secreted through the basal body/hook assembly, which enables derepression of FliA and allows subsequent transcription of the class III genes (81, 117). Inactivation of any of the class II genes interrupts complete formation of the flagellum, and the accumulated FlgM prevents further flagellar filament formation. Loss of FlgM results in an approximately sixfold increase in transcription of the FliA-dependent class III genes (118). Interestingly, while flgM mutants are virulence attenuated, an additional mutation that inactivates FliA function restores virulence to the strain (187). The mechanism for this phenomenon is still unknown. Many studies have shown the importance of flagella for virulence of S. enterica serovars (87, 182, 188, 197), although the specific aspect of flagellar function that contributes to virulence remains unclear.
Other species.Regulation of flagellar gene expression in Yersinia enterocolitica is similar to that in S. enterica serovar Typhimurium. fliA encodes a σ28 factor responsible for motility of the bacterium, and the master regulators FlhC and FlhD are required for expression of all genes encoding proteins active in subsequent flagellar synthesis (85). Motility is also required for full Y. enterocolitica invasion efficiency (228). However, another group of enteric bacteria has a different strategy for regulating flagellar gene expression. Helicobacter pylori, Campylobacter jejuni, and Vibrio cholerae do not have the flhDC master operon. Instead, early flagellar gene expression is carried out via a σ54 factor, while later genes are transcribed by FliA (89, 105, 154). These species also encode one or more σ54 activator proteins, such as FlgR. Virulence in these species is linked to production of flagella. In C. jejuni, virulence proteins are secreted through the flagella, and full virulence requires a complete flagellar export apparatus (110). Multiple studies have shown H. pylori virulence to be dependent both on expression of flagellin proteins and on flagellar motility (94, 140).
ECF SIGMA FACTORS
Members of the extracytoplasmic function subfamily of σ70 sigma factors regulate functions related to sensing and responding to changes in the bacterial periplasm and extracellular environment. These sigma factors are conserved in both gram-positive and gram-negative species. The first ECF sigma factor identified was E. coli σE, which was recognized as a second heat shock sigma factor in this organism (213). Although σE does not appear to affect virulence in E. coli, other ECF sigma factors contribute to regulation of virulence genes and virulence-associated genes in a number of bacteria, including S. enterica serovar Typhimurium, Pseudomonas aeruginosa, and Mycobacterium tuberculosis. A recent review covers some aspects of ECF sigma factors and their involvement in pathogenesis (4).
Sigma E (RpoE) rpoE contributes to oxidative stress resistance in S. enterica serovar Typhimurium. To illustrate, inactivation of rpoE diminishes bacterial survival and growth inside host macrophages (21, 82). Further, while an rpoE mutant is severely attenuated in virulence in a mouse model of infection (82, 205), an rpoE mutant strain appears to be fully virulent in gp91phox−/− mice, which are defective in phagocyte oxidative burst (205). Expression of htrA, a gene required for oxidative stress resistance, macrophage survival, and S. enterica serovar Typhimurium virulence (7, 23, 91), is dependent on σE (50, 130). However, the survival and virulence defects in rpoE mutants are not entirely due to loss of htrA expression, because the attenuated virulence phenotype of an htrA mutant is less severe than that of the rpoE mutant (82).
σE also appears to contribute to oxidative stress resistance in other gram-negative pathogens. In Haemophilus influenzae, rpoE expression was discovered to increase 102-fold inside macrophages, and survival of an rpoE mutant was reduced relative to that of the wild type in the macrophage (36). Vibrio cholerae rpoE mutants are virulence attenuated, exhibiting a reduced ability to colonize the mouse intestine and an LD50 that is 3 log units higher than that of the wild type (111). Although σE is not essential for growth in H. influenzae or V. cholerae, interestingly, it is essential for growth in Yersinia enterocolitica (76). σE in Y. enterocolitica also appears to regulate htrA, which is important for virulence in this bacterium as well (77, 127).
AlgU of P. aeruginosa (also called AlgT) is homologous and functionally equivalent to σE of E. coli (139, 229). As in the pathogens mentioned above, P. aeruginosa algU mutants have increased sensitivity to oxidative stress (139) and reduced survival in macrophages and neutrophils (230). Additionally, AlgU regulates biosynthesis of alginate, a major virulence factor in P. aeruginosa infections of cystic fibrosis patients. Expression of the major alginate biosynthesis gene algD and production of alginate are dependent on algU (138, 189). Thus, while regulation of oxidative stress resistance is functionally conserved between P. aeruginosa AlgU and σE in multiple bacterial species, AlgU also has an additional role in P. aeruginosa virulence through the regulation of alginate production.
PvdS and FpvIIn P. aeruginosa, secretion of the siderophore pyoverdine, a virulence factor, is required for in vivo growth and virulence. Pyoverdine is released when cells experience iron-limiting conditions, which is common during host infection. Pyoverdine enables P. aeruginosa to sequester iron from the environment. The secreted pyoverdine chelates extracellular iron, and the resulting ferri-pyoverdine complex is transported back into the bacterial cell (210), as described below.
The genes involved in pyoverdine synthesis are located in three clusters on the P. aeruginosa chromosome, with the major genes comprising the pvd locus. Among these genes is pvdS, which encodes an alternative sigma factor. PvdS appears to be predominantly responsible for regulating genes in the pvd locus as well as other pyoverdine synthesis genes (147, 160, 198). The binding of iron by pyoverdine, which occurs outside of the cell, initiates a signaling cascade that leads to enhanced expression of pvd genes and additional secretion of pyoverdine and other virulence factors. Upon forming a complex with iron, pyoverdine binds to the FpvA cell surface receptor protein. FpvA is responsible for transporting the pyoverdine into the cell, but it also triggers a signal cascade to the membrane-bound anti-sigma factor, FpvR, which releases PvdS and allows it to transcribe the pvd genes. FpvR also controls the activity of another sigma factor, FpvI (9). The signal from bound pyoverdine also results in release (and hence activation) of this factor, which is responsible for expression of fpvA.
In addition to increasing pyoverdine synthesis and secretion, free PvdS also activates transcription of genes encoding two more virulence factors, those encoding exotoxin A and PrpL endoprotease. Expression of genes responsible for pyoverdine, exotoxin A, and PrpL production is also controlled by the regulator PtxR; expression of ptxR is also controlled by PvdS. A pvdS deletion mutant generates less PrpL (220) and only 5% of the exotoxin A produced by a wild-type strain (159).
Loss of PvdS results in decreased P. aeruginosa virulence in a rabbit aortic endocarditis model (226). The PrpL endoprotease contributes to the ability of P. aeruginosa to persist in a rat chronic pulmonary infection model (220). PvdS is required for virulence and appears to regulate only virulence-related genes.
Mycobacterial ECF Sigma Factors Mycobacterium tuberculosis has 13 recognized sigma factors; among these, 10 are ECF sigma factors. At least six M. tuberculosis sigma factors affect virulence, including the primary sigma factor (31), σF, and four ECF sigma factors, σC, σD, σE, and σH (Table 3). The regulons of many of these sigma factors (σC, σD, σE, σF, and σH) have been identified through application of M. tuberculosis genome arrays (20, 60, 99, 134, 135, 179, 202). M. tuberculosis ECF sigma factors do not appear to control many currently characterized virulence genes. For example, σD does not appear to directly regulate any virulence-associated genes (20, 179), although it does control the putative transcriptional regulator Rv1856. It is possible that this putative regulator is responsible for direct control of virulence gene expression, but no evidence currently exists to support this hypothesis (179). σC, σE, and σH each control a relatively small number of virulence or virulence-associated genes, as well as some regulatory genes that may influence expression of other virulence genes. Several other ECF sigma factors also regulate a number of known or putative regulatory genes (Table 3). Interestingly, in some cases, this group of sigma factors contributes to regulation of other sigma factors within the group. For example, sigB expression is affected by σC, σE, and σH (99, 134, 135, 202). σC activates expression of hspX, mtrA, and senX3 (202), three genes shown to be required for virulence. mtrA and senX3 are examples of two-component system response regulators. Other virulence-associated genes regulated by M. tuberculosis ECF sigma factors include genes for heat shock proteins and oxidative stress response proteins. For example, the heat shock genes hsp and htpX are σE dependent (134), and hsp, dnaK, and clpB are regulated by σH (99, 135). A number of putative thioredoxins and other oxidative stress genes are controlled by σH (99) (Table 3), and sodA, encoding the superoxide dismutase, is regulated by σE (134). The contributions of these ECF sigma factors to expression of oxidative stress resistance genes may explain reduced survival of the respective null mutant strains under oxidative stress conditions or inside macrophages (134, 135).
Recently, deletion of sigC was shown to render M. tuberculosis unable to cause death in infected mice (202). Deletion of another sigma factor gene, sigH, also produced a nonlethal strain (99). Interestingly, despite the inability to cause fatalities, both sigC and sigH mutants grew to wild-type numbers in macrophages and murine tissues (99, 135, 202). Although the reasons for the similar phenotypes in the two different mutant strains are unknown, it is possible that a subset of virulence-associated genes are regulated by both factors. Alternatively, σC and σH may provide similar contributions to M. tuberculosis, but through different mechanisms. σE appears to affect M. tuberculosis virulence differently than σC and σH. As with sigH, sigE expression is induced inside macrophages (64, 90). Loss of σE, however, does result in decreased strain survival in macrophages and a greater susceptibility to killing by activated macrophages (134). In mouse infection models, the sigE mutant is delayed in its ability to cause lethality but is not completely compromised, as with the sigC and sigH mutant strains (2, 136). Manganelli et al. (136) reported a lower number of sigE mutants in the lungs compared to the wild type, while Ando et al. (2) reported no difference. This discrepancy may be due to differences in mouse strains used in the two studies.
Multiple studies suggest that σD also contributes to M. tuberculosis virulence. Deletion studies of sigD show the mutant to be less virulent than the wild type in BALB/c and C3H:HeJ mouse infections, allowing substantially longer mouse survival (20, 179). The ΔsigD strain did not show a difference in time to death in SCID mice, which lack T and B cells (20), suggesting that σD regulates pathogenicity in a manner that is dependent on cell-mediated immunity. In addition, loss of σD resulted in much milder tissue damage and granuloma formation in lung tissue histopathology in BALB/c mice (179).
Several alternative sigma factors present in M. tuberculosis affect virulence, whether through direct, indirect, or both types of strategies. In addition, some alternative sigma factors of M. tuberculosis autoregulate transcription of their own genes. Many sigma factors also activate transcription of other alternative sigma factors (Fig. 2C). In all, M. tuberculosis appears to have control over expression of its virulence genes via a complex network of multiple alternative sigma factors.
HrpL Pseudomonas syringae is a plant pathogen with several pathovars that display selective host specificity. Infection of a plant by a specific pathovar will cause disease in susceptible host species, while eliciting a programmed cell death termed the hypersensitive response (HR) in resistant plants. The groups of genes responsible for both of these reactions have been termed hrp and avr. These gene products encode either the type III secretion machinery that translocates proteins into host plant cells or the effector proteins that are delivered and that interact with host elements. Most of the hrp genes are regulated by the alternative sigma factor HrpL, which has been shown by microarray analysis to be almost exclusively responsible for virulence functions (56). Strains with mutations in hrp genes cannot elicit disease or HR in plants (for a review, see reference 120). Likewise, inactivation of HrpL decreases P. syringae pv. Phaseolicola growth in leaves (178).
Another important phytopathogen, Erwinia amylovora, also utilizes an hrp-encoded type III secretion system. As in P. syringae, E. amylovora HrpL is an alternative sigma factor that directs transcription of several hrp genes (104). Inactivation of HrpL prevents E. amylovora from causing disease in susceptible plant species or HR in resistant plants (216). E. amylovora also has a dsp, or “disease-specific,” gene cluster which is homologous to the avr genes of P. syringae (15). dspA is dependent on HrpL for expression and is required for virulence (58). In addition to E. amylovora, several other members of the Erwinia genus carry hrpL and other hrp genes, including the tumorigenic pathogen Erwinia herbicola (149, 155, 156) and the soft-rot pathogens Erwinia carotovora (24, 125, 180) and Erwinia chrysanthemi (6). The hrp-encoded type III secretion system is thus a common virulence mechanism among plant pathogens and is widespread among several types of pathogens, including tumorigenic, macerating, and soft-rot-causing species.
SIGMA 54
σ54 forms a distinct subfamily of sigma factors, apart from the σ70-like family. In almost all species, the σ54 factor is called σN. σN has been identified in many species, spanning a diverse phylogeny, including Legionella pneumophila (88), Pseudomonas spp. (72, 86, 108), Enterococcus faecalis (40), Campylobacter jejuni (89), and Listeria monocytogenes (183). A physiological theme for σN-dependent genes has not yet emerged, as the regulated genes described to date control a wide diversity of processes (Table 4). Often nitrogen metabolism is controlled by σN, but other functions of σN-dependent genes can be found in several organisms.
Virulence genes regulated by σN in multiple bacterial species
Pseudomonas aeruginosa.Evidence of σN involvement in bacterial pathogenesis and virulence is well documented for P. aeruginosa. Alginate has been identified as a virulence factor that is important in strains colonizing cystic fibrosis patient lungs. algD and algC, two important genes for the biosynthesis of alginate, are controlled by σN (17, 233). In addition, through gene fusion and microarray studies, expression of a large number of flagellar structural genes was shown to be dependent on σN (41).
Flagellar motility and pilus-mediated attachment are established virulence factors in P. aeruginosa (148, 186). Pili are external structures that are responsible for adhesion to host cells and interactions such as internalization. P. aeruginosa rpoN mutants do not produce pilin or form pili (206), and they demonstrate drastic loss of adhesion to multiple cell types (28, 32, 172). Wild-type P. aeruginosa also is internalized by host cells more efficiently than an rpoN mutant (172), suggesting an enhanced capacity of the wild-type strain to invade host cells. Reduced virulence due to loss of flagellar motility is also possible in rpoN-disrupted strains, as mutants are decidedly nonmotile (73, 206). rpoN mutants also do not produce the proteinaceous flagellin subunit or form flagella (206). Several studies have shown that P. aeruginosa strains lacking flagella are severely virulence attenuated (46, 52, 148).
P. aeruginosa rpoN mutants are also less virulent than wild-type strains in multiple infection models. An rpoN mutant strain showed diminished cytotoxicity to Madin-Darby canine kidney (MDCK) cells (32) and reduced virulence in several mouse models specifically developed to study P. aeruginosa pathogenicity; compared to the wild type, rpoN mutants cause lower mortality rates in infected mice (32, 73) and reduced fecal carriage and recovery from gastrointestinal tissues (170). In addition, no pathology was observed following infection with an rpoN mutant in a murine corneal scratch model (173). Cohn et al. (30) reported that rpoN mutants did not readily colonize human tracheal epithelium xenografts implanted in mice, although the difference in bacterial numbers of the mutant and wild-type strains was not statistically significant. In general, the defects associated with the rpoN mutation were greater than with strains that were specifically pilin negative, indicating the existence of an additional, pilus-independent mechanism through which σN also contributes to virulence (28, 32, 170, 172).
Pseudomonas syringae.σN of P. syringae controls hrp gene expression and influences virulence. Regulation occurs via a short regulatory cascade, wherein σN and its enhancer-binding proteins HrpR and HrpS direct transcription of hrpL, the product of which is the alternative sigma factor required for expression of the hrp and avr genes (83) (Fig. 2B). Xiao et al. (225) showed that while expression of hrpL and HrpL-dependent genes requires hrpR and hrpS, constitutive expression of hrpL can provide full expression of HrpL-dependent genes with or without hrpR and hrpS. In addition, avrD, which is transcribed from an HrpL-dependent promoter, requires rpoN, hrpL, and hrpS for its expression (192). Characterization of P. syringae pv. Maculicola rpoN mutants identified a more severe phenotype than in hrpL mutants, however (72). rpoN mutants were nonmotile, displayed nitrogen utilization defects, and were unable to produce the phytotoxin coronatine, cause disease or HR, or induce host defense mRNAs. Complementation of hrpL into this strain partially restored some phenotypes but did not restore coronatine production. Other studies have also shown that σN is required for production of coronatine biosynthetic intermediates (synthesized by the cfl/CMA and cmaABT gene products) as well as hrpL transcription and HrpL-dependent gene expression (1, 71). Thus, σN regulates a range of virulence factors in P. syringae, some via hrpL activation and others by HrpL-independent mechanisms.
Vibrio species.The contributions of σN to virulence in Vibrio species are similar to its contributions in P. aeruginosa. V. cholerae rpoN mutants lack flagella and are completely nonmotile (105). In a competitive infant mouse colonization trial, an rpoN mutant was 10- to 20-fold less able to colonize the intestine than the wild type (105). This defect is not entirely due to lack of flagella, because a flaA mutant, while inhibited in intestinal colonization, was still superior to the rpoN mutant in colonization. Prouty et al. (176) also demonstrated the involvement of σN in expression of several V. cholerae flagellar structural genes. σN is required for flagellin production and motility in the fish pathogen Vibrio anguillarum as well. A mutant lacking σN was also severely impaired in its ability to infect fish immersed in contaminated water but was not virulence attenuated in an intraperitoneal injection model (163).
Other species.σN contributes to virulence in a number of gram-negative pathogens. In addition to the examples provided above, the uropathogen Proteus mirabilis is 1,020-fold less virulent than the wild type when σN is inactivated but remains identical to the wild type with respect to growth, glutamine synthesis, and fimbria production (231). σN does not share a common role among all pathogens, however. For example, the plant pathogens, Pseudomonas syringae, Erwinia carotovora, and Xanthomonas campestris all use type III secretion systems to cause disease in host organisms, and σN has a substantial effect on virulence and hrp gene expression in P. syringae and E. carotovora (24) but is not required for expression of hrp genes or for virulence in X. campestris (78). As with the low-GC bacterial σB, σN appears to be an alternative sigma factor that has evolved to regulate virulence determinants in some species but not in others.
CONCLUSIONS
Bacteria utilize alternative sigma factors to regulate a wide range of physiological processes. In pathogenic bacteria, alternative sigma factors often affect virulence. Virulence effects can be mediated either through direct virulence gene regulation or indirectly, by regulating genes that increase fitness of the bacterium during transmission and infection. Direct effects on virulence genes include σB activation of the L. monocytogenes virulence genes inlA and prfA and σS-dependent expression of the S. enterica serovar Typhimurium spv genes. Indirect effects of sigma factors on virulence may be more difficult to identify, but alternative sigma factors frequently have roles in virulence by regulating virulence-associated genes that aid in a bacterium's survival during infection. For example, σE enhances survival of oxidative stress and hence aids in bacterial survival of the oxidative burst within macrophages. The stress response sigma factors σB and σS contribute to survival of multiple stresses (e.g., acid and osmotic stresses) important for bacterial survival of passage through a host stomach and gastrointestinal tract. In addition, σB and σS contribute to environmental survival, and thus transmission, of food-borne pathogens in foods and food-processing environments. Another alternative sigma factor role that contributes to environmental survival, and has virulence implications, is regulation of biofilm formation, e.g., by σB in S. aureus and S. epidermidis.
Functional roles for alternative sigma factors can be clearly defined and highly specific (e.g., sporulation sigma factors) or multifunctional. While Pseudomonas syringae HrpL's role is predominantly virulence related, most alternative sigma factors contribute to multiple, diverse functions in a cell. In some cases, sigma factors are conserved across pathogenic and nonpathogenic species, with virulence genes constituting a relatively small subset of the total regulon in the pathogenic species. For example, σB is present and contributes to stress resistance in the nonpathogenic B. subtilis and Listeria innocua (S. Raengpradub, unpublished data), both of which are closely related to the pathogenic L. monocytogenes. It is possible that virulence gene incorporation into the L. monocytogenes σB regulon is a relatively recent evolutionary event. Likewise, as no evidence currently supports a direct role for σS in E. coli virulence gene regulation, the inclusion of virulence genes in the regulatory network of S. enterica serovar Typhimurium σS may have occurred after the species divergence of S. enterica serovar Typhimurium and E. coli.
A comparison of homologous sigma factor functions among different bacterial genera reveals that the roles of sigma factors vary greatly among bacterial species, even for closely related species such as E. coli and S. enterica serovar Typhimurium. In some cases, as with M. tuberculosis σF, distinct virulence-related phenotypes have been observed in alternative sigma factor null mutants. For others, such as S. aureus σB, while virulence genes are directly transcribed by the sigma factor, ΔsigB strains are not severely virulence attenuated. Even more apparent are the different roles for σS. σS is required for virulence in S. enterica serovar Typhimurium and yet does not demonstrate a pronounced role in E. coli pathogenesis.
A common mechanism of virulence regulation by alternative sigma factors involves coordinated networks of sigma factors along with other transcriptional regulators. Alternative sigma factors may regulate not only individual genes involved in virulence but also other sigma factors or transcriptional regulators that in turn regulate virulence genes and virulence-associated genes (Fig. 2). For example, σB of L. monocytogenes not only directly regulates bsh and inlA but also contributes to expression of PrfA, which is required for transcription of almost all of the currently recognized L. monocytogenes virulence genes. σB of S. aureus also affects expression of a virulence gene regulator, RNAIII. σN and HrpL of P. syringae present a different type of regulatory network, in which one sigma factor controls expression of another. HrpL also controls expression of the HrpR and HrpS two-component system regulators. Regulatory networks can be very complex, as in the multiple sigma factor interactions of M. tuberculosis.
Finally, to extrapolate bacterial pathogen research findings to ensure relevance in human infection, the importance of identifying and applying suitable model systems that accurately mimic interactions between pathogen and humans is essential. This point is illustrated by the significantly reduced traversal of the intestinal barrier by L. monocytogenes in wild-type versus (human) E-cadherin transgenic mice (121). In addition, pathogens such as P. aeruginosa that can infect a multitude of different hosts are likely to respond differently and to have different virulence requirements depending on the host species. Significant efforts are still needed to identify or develop appropriate model systems for exploration of virulence mechanisms that are important in human infection.
ACKNOWLEDGMENTS
This work was partially supported by National Institutes of Health award no. RO1-AI052151-01A1 (to K.J.B.).
- Copyright © 2005 American Society for Microbiology
REFERENCES
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.↵
- 96.↵
- 97.↵
- 98.↵
- 99.↵
- 100.↵
- 101.↵
- 102.↵
- 103.↵
- 104.↵
- 105.↵
- 106.↵
- 107.↵
- 108.↵
- 109.↵
- 110.↵
- 111.↵
- 112.↵
- 113.↵
- 114.↵
- 115.↵
- 116.↵
- 117.↵
- 118.↵
- 119.↵
- 120.↵
- 121.↵
- 122.↵
- 123.↵
- 124.↵
- 125.↵
- 126.↵
- 127.↵
- 128.↵
- 129.↵
- 130.↵
- 131.↵
- 132.↵
- 133.↵
- 134.↵
- 135.↵
- 136.↵
- 137.↵
- 138.↵
- 139.↵
- 140.↵
- 141.↵
- 142.↵
- 143.↵
- 144.↵
- 145.↵
- 146.↵
- 147.↵
- 148.↵
- 149.↵
- 150.↵
- 151.↵
- 152.↵
- 153.↵
- 154.↵
- 155.↵
- 156.↵
- 157.↵
- 158.↵
- 159.↵
- 160.↵
- 161.↵
- 162.↵
- 163.↵
- 164.↵
- 165.↵
- 166.↵
- 167.↵
- 168.↵
- 169.↵
- 170.↵
- 171.↵
- 172.↵
- 173.↵
- 174.↵
- 175.↵
- 176.↵
- 177.↵
- 178.↵
- 179.↵
- 180.↵
- 181.↵
- 182.↵
- 183.↵
- 184.↵
- 185.↵
- 186.↵
- 187.↵
- 188.↵
- 189.↵
- 190.↵
- 191.↵
- 192.↵
- 193.↵
- 194.↵
- 195.↵
- 196.↵
- 197.↵
- 198.↵
- 199.↵
- 200.↵
- 201.↵
- 202.↵
- 203.↵
- 204.↵
- 205.↵
- 206.↵
- 207.↵
- 208.↵
- 209.↵
- 210.↵
- 211.↵
- 212.↵
- 213.↵
- 214.↵
- 215.↵
- 216.↵
- 217.↵
- 218.↵
- 219.↵
- 220.↵
- 221.↵
- 222.↵
- 223.↵
- 224.↵
- 225.↵
- 226.↵
- 227.↵
- 228.↵
- 229.↵
- 230.↵
- 231.↵
- 232.↵
- 233.↵
