Klaus Heuner and Christiane Albert-Weissenberger
Abstract
Legionella is a ubiquitous inhabitant of aquatic habitats, and it is believed that motility is an important feature in the life cycle of L. pneumophila in its environment. To survive, L. pneumophila must be able to respond to different environmental factors and to modulate gene expression. Furthermore, very early after the first report of Legionnaires' disease, researchers discovered that most Legionella strains are flagellated and that regulation of the flagellum and the expression of virulence traits are linked. Expression of the flagellum is modulated by various environmental factors. The flagellar regulon, comprising genes encoding the structural components of the flagellum, is controlled by a cascade of regulators. This chapter will focus on the composition of the flagellar complex and the regulation of flagellar operons, including the virulence traits of L. pneumophila that are coordinately expressed with the flagellar genes. The contribution of the flagellar system to the virulence of L. pneumophila will also be discussed. read more ...
Introduction
It is fascinating to watch amoebae infected to see how fast L. pneumophila move circularly in the host cell at the end of the infection cycle, due to their flagella. Since Legionella is a ubiquitous inhabitant of aquatic habitats and replicates within different protozoa, it is believed that motility is needed to find a new host cell after its release from the exhausted host. If Legionella is successful, a next round of intracellular multiplication starts. If not, Legionella is well prepared to survive for a long time in its environment, perhaps within biofilm communities, until the next host approaches. Motility may also play a role in biofilm formation and development, as shown for other bacteria (O'Toole et al., 2000; Kirov et al., 2004).
Shortly after the first described outbreak of Legionnaires' disease and the identification of L. pneumophila as the causative agent (for details, see Chapter 1), it was reported that legionellae are flagellated (Fig. 6.1A and B) and that they exhibit pili-like structures on the surface (Chandler et al., 1980a; Rodgers et al., 1980). It was also postulated very early that Legionella flagellar expression and virulence are associated. Motility may contribute to the rapid spread of bacteria within the lung, since flagellated bacteria are apparent in the alveolar space of diseased patients (Chandler et al., 1980b). Later it was shown that flagellar expression is linked to the developmental stage of Legionella. Rowbotham (1986) described the differentiation of Legionella from a non-motile replicating form to a non-dividing, motile form. This differentiation occurs at the end of intracellular replication, before the host cell is lysed and flagellated legionellae are released from the host (Fig. 6.1C and D). As these bacteria were found to be more infectious for amoebae than the non-motile forms, it was speculated that the expression of the virulent form of Legionella is linked to the expression of the flagellum (Rowbotham, 1986). For further details of Legionella differentiation, see Chapter 4. A regulatory link between the flagellum and the expression of the virulent phenotype of Legionella was subsequently confirmed by several molecular genetic studies (Pruckler et al., 1995; Bosshardt et al., 1997; Gao et al., 1997; Byrne and Swanson, 1998; Swanson and Hammer, 2000; Hammer et al., 2002; Heuner et al., 2002a; Molofsky and Swanson, 2004; Molofsky et al., 2005; Heuner et al., 2006).
A correlation of motility and virulence has been described for various bacterial pathogens. The flagellum is involved in the adhesion to and invasion of host surfaces, successful colonization of hosts, and biofilm formation (Drake and Montie, 1988; Liu et al., 1988; Allison et al., 1992; Grant et al., 1993; Yao et al., 1994; Mobley et al., 1996; O'Toole et al., 2000; Young et al., 2000; Kirov et al., 2004; Goon et al., 2006). Furthermore, the flagellar secretion system is a type III-like secretion system that can also secrete non-flagellar proteins and virulence factors (Macnab, 1999; Young et al., 1999; Aizawa, 2001; Young and Young, 2002; Konkel et al., 2004).
On the other hand, the flagellin protein is highly antigenic, and its N- and C-terminal regions are well conserved. Mammalian hosts recognize these domains of the flagellin monomer by membrane-bound Toll-like receptors (TLR5) and cytosolic NOD-like receptors (Naip5), which trigger proinflammatory and adaptive immune responses (Gewirtz et al., 2001a,b; Hayashi et al., 2001; Smith and Ozinsky, 2002, Hawn et al., 2003; Smith et al., 2003; Molofsky et al., 2006; Amer et al., 2006; Ren et al., 2006; Akamine et al., 2007). For details see Chapter 5.
Since the flagellar system is composed of more than 50 genes, its expression represents a high metabolic cost for the bacterial cell, making tight regulation of flagellar gene expression necessary (Chilcott and Hughes, 2000). As discussed below, the expression of the flagellum is modulated by different environmental factors, including growth phase, temperature, and the nutrient supply (Heuner et al., 1999). Down-regulation of flagellin expression inside a host cell may also contribute to L. pneumophila virulence for mammals by enabling intracellular bacteria to avoid host surveillance mechanisms. Thus, its ability to modulate flagellar gene expression equips L. pneumophila to survive not only in its natural environment but also within mammalian hosts.
Since a number of L. pneumophila virulence traits are co-regulated with the flagellar system, identifying and characterizing the motility and the flagellar regulon of this bacterium is an important area of research. This chapter will focus on the composition of the flagellar complex, the regulation of the flagellar operons and the link between flagellar expression and virulence of L. pneumophila. read more ...
The flagellum of legionellae
The flagellum of Gram-negative bacteria
In general, the flagellar system is very complex. In most bacteria, it is composed of more than 50 genes (Chilcott and Hughes, 2000; Aldridge and Hughes, 2002; McCarter, 2006). The flagellum consists of a basal body, the hook structure (FlgE, FlgD) and the filament (FlaA, FliD, FlgK, FlgL; Fig. 6.2A). For details see recent review articles (Bardy et al., 2003; Macnab, 2003; Pallen et al., 2005). Typically, the filament consists of thousands (~20,000) of copies of one single protein, flagellin (FlaA or FliC), and it is about 20 nm in diameter (Minamino and Namba, 2004). The cap protein FliD is needed to assemble the FlaA protein into the filament. The basal structure consists of a rod, various rings (MS, P, L), the flagellum-specific export apparatus, the switch complex and the Mot proteins (Fig. 6.2A). The basal body anchors the flagellum to both membranes (MS ring, cytoplasmic membrane; P ring, peptidoglycan; L ring, outer membrane). The Mot proteins (MotA, MotB) generate the power for rotating the flagellum. For many species of bacteria, the switch proteins (FliG, FliM, FliN) allow microbes to switch rotation in response to attractants or repellents recognized by methyl-accepting chemotaxis proteins (MCPs) through a complex chemotaxis system (for details see Blair, 1995; Manson et al., 1998; Szurmant and Ordal, 2004; Baker et al., 2006). Cells run when the flagellar filaments rotate counterclockwise and tumble when the flagella turn clockwise.
The growth of the filament is unusual, because the flagellin monomers are exported through the filament structure and are assembled at the distal tip of the filament. FliD is needed for this process, since fliD mutant strains release flagellin into the medium and do not assemble a flagellum (Shimizu et al., 2003; Minamino and Namba, 2004). The flagellum-specific export apparatus is a type III-like secretion system and has become a model system to study type III secretion (Minamino and Macnab, 1999; Aizawa, 2001; Macnab, 2004). In general, type III secretion systems (T3SSs) secrete proteins (virulence-associated effectors) directly into the target host cell. The flagellar T3SS is responsible for the export of the hook and rod forming proteins, the flagellum forming proteins and the cap protein. But non-flagellar proteins and virulence-associated factors are also secreted by the secretion system of the flagellum. For example, the phospholipase (YplA) of Yersinia enterocolitica, which can cause gastrointestinal illnesses in humans, is secreted by the T3SS of the flagellum. YplA is not required for motility but affects survival of Y. enterocolitica during infection (Young and Young, 2002; Young et al., 1999). Bacillus thuringensis, which sporadically causes severe human infections, secretes virulence-associated factors (haemolysin BL, phospholipase C) by a mechanism that depends on its flagellar secretion system (Ghelardi et al., 2002). Salmonella typhimurium and Campylobacter jejuni also secrete non-flagellar proteins by the flagellar T3SS.
The signal for the secretion via T3SSs is still unclear. Proteins secreted by T3SSs lack a cleavable signal sequence and do not enter the general secretory pathway (Sec-translocon). Recently it was shown that flagellin secretion is facilitated by the 173-bp untranslated DNA fragment upstream of fliC, supporting a model in which the mRNA directs the secretion of newly synthesized proteins via T3SSs. However, there are also signals within the flagellin protein that direct its T3SS-dependent secretion (Majander et al., 2005). read more ...
The flagellum of Legionella
L. pneumophila exhibits a single monopolar flagellum (Fig. 6.1A). The flagellum is composed of the flagellin subunit FlaA, a protein of approximately 48 kDa (Chandler et al., 1980a; Elliott and Johnson, 1981; Heuner et al., 1995). The flagellin-encoding gene (flaA) is found in nearly all Legionella strains tested so far (Heuner et al., 1995), and most Legionella species produce flagella (Elliott and Johnson, 1982; Bornstein et al., 1991; Ott et al., 1991; Bangsborg et al., 1995; Heuner et al., 1995). Within these studies there are some conflicting reports, perhaps because culture conditions in some studies were not optimal for flagellum expression. For example, Legionella grown on supplemented Mueller-Hinton (SMH) agar plates become non-virulent and non-motile (Chandler et al., 1980a; Elliott and Johnson 1981). In one study, using a flaA specific DNA probe for Southern blot analysis, L. oakridgensis, L. longbeachae SG1 and L. israeliensis did not exhibit a flaA-homologous gene, they also were non-flagellated and did not produce a flagellin protein (Heuner et al. 1995). Interestingly, L. oakridgensis is also unique among the legionellae in its ability to grow on agar plates without supplementation with L-cysteine (Orrison et al., 1983). As we will see below, the amino acid supply influences flagellum expression.
In a first report about the flagellar regulon, some flagellar genes (flgAMN, motY, fliK, flhB') were not detected (Heuner and Steinert, 2003). Now that the genomes of four L. pneumophila species are sequenced (strains Philadelphia, Lens, Paris and Corby; Chien et al., 2004; Cazalet et al., 2004 and Glöckner et al., 2007), more genes predicted to belong to the flagellar system of L. pneumophila were identified using in silico methods and in vivo transcriptome analysis (Brüggemann et al., 2006; Heuner et al., 2006). The overall operon structure is similar to that of Pseudomonas (Fig. 6.2B; Dasgupta et al., 2003; Heuner and Steinert, 2003). Experimental data are available for several structural components of the flagellum, including flagellin, the flagellar ATPase FliI, the secretion system component FlhB, the motor components MotA1B1 and the cap protein FliD; together these data support the known general features of these structural proteins (Heuner et al., 1995; Merriam et al., 1997; Dietrich et al., 2001; Molofsky et al., 2005; Brüggemann et al., 2006). Experiments revealed that motA1 and motA1B1 mutants of L. pneumophila are non-motile. Interestingly, in motA1B1 mutants of L. pneumophila the flagellum is not stable and therefore flagellated mutant cells are rarely detected (Molofsky et al., 2005; K. Heuner, unpublished). Since L. pneumophila exhibits a second MotAB encoding operon (motA2B2), the experiments suggest that motA2B2 is not expressed and/or cannot complement the motA1B1 mutant under conditions tested so far. The MotA1B1 proteins are homologous to the MotAB proteins of Pseudomonas, whereas the MotA2B2 proteins are more homologous to the MotAB proteins of Nitrosomonas, Nitrosospira, Ralstonia and Methylobacillus. Yet, the function of the MotA2B2 proteins in Legionella is unknown. The recently (in silico) identified homologue of MotY indicates that the flagellum of L. pneumophila seems to exhibit a sodium-driven motor composed of at least MotA1B1 and MotY.
A number of regulator proteins involved in flagellum expression have already been identified in different bacteria (Aldridge and Hughes, 2002). In L. pneumophila, the master regulator is FleQ, an activator protein of the alternative sigma factor σ54, which acts together with σ54 and σ28 to control expression of the flagellar regulon (details, see below). Most of bacteria also encode chemotaxis proteins. However, in L. pneumophila, so far no homologues of chemotaxis genes (cheWA, cheZYBR, cheB) have been identified, with the exception of CheY-like proteins. Their absence may explain why chemotaxis by Legionella has not been described (Rodgers et al., 1980; Heuner and Steinert, 2003). But since specific conditions are needed to observe swarming or chemotaxis activities, it also seems possible that the conditions needed to induce these activities have not yet been identified. read more ...
The flagellar regulon
Regulation of flagellum expression
Expression of the flagellum varies during the life cycle of Legionella and depends on its growth phase. Legionella switches from a non-flagellated replicating form to a flagellated non-replicating form (Rowbotham, 1986). Furthermore, flagellum expression is an early transmissive trait that starts when Legionella enter stationary phase, a period in broth cultures that corresponds to the end of the infection cycle in host cells (Robotham, 1986; Byrne and Swanson, 1998; Heuner et al., 1999; Brüggemann et al., 2006). The same pattern occurs when bacteria replicate within macrophages (Hammer and Swanson, 1999) and amoebae host cells. For example, in Fig. 6.1C and D bacterial flagellin was detected during the intracellular replication period using an anti-FlaA antiserum. Whereas no flagellin could be detected at early times (Fig. 6.1C), the bacteria are well stained at later stages of infection before lysis of the host cells (Fig. 6.1D).
Other kinds of switches also influence the expression of the flagellum. Growth on SMH agar leads to a conversion of virulent L. pneumophila to a non-flagellated and non-virulent form. There are conflicting reports about the ability of this phase variation to revert (McDade and Shepard, 1979; Elliott and Johnson, 1982; Catrenich and Johnson, 1988). Another phase variation has been described that is reversible and also leads to a non-virulent, non-flagellated form of L. pneumophila. In this case, the phase variable expression of an LPS epitope and the flagellum corresponds to the excision of a 30-kb genetic element. In the non-flagellated cell form, the element replicates as a high-copy plasmid (Lüneberg et al., 1998, 2000, 2001).
Since L. pneumophila can replicate within host cells in very different aquatic habitats, it must modulate gene expression to respond to different environmental factors, such as temperature, osmolarity, viscosity and nutrition. Therefore, it is not surprising that a variety of conditions also influence expression of the flagellum (Fig. 6.3). Increased temperatures negatively regulate expression of the flagellum. Flagellin expression is induced by high osmolarity or low nutrient supply, whereas expression is repressed during exponential growth, increased temperatures, high viscosity, and abundant nutrients, specifically amino acids (Ott et al., 1991; Mauchline et al., 1992; Heuner et al., 1995; Byrne and Swanson, 1998; Heuner et al., 1999). In addition, the results shown in Fig. 6.3 demonstrate that flaA expression is repressed by the addition of amino acids (serine, threonine) while the addition of glucose had no effect. These results nicely demonstrate the fact that indeed amino acids are the major energy source of L. pneumophila (see chapter 11), since motility is generally suppressed under high nutrition conditions. However, the ability of Legionella to swarm or swim on agar plates has not yet been reported (Rodgers et al., 1980; Heuner and Steinert, 2003). In the following section, we discuss the molecular mechanisms and regulators of flagellum expression. read more ...
The cascade of flagellin gene regulation
As mentioned above, genes that are predicted to belong to the flagellar regulon have been identified experimentally by molecular genetic and in silico methods (Fig. 6.2B). To obtain information about the cascade of flagellar gene regulation, the regulators involved were identified and specific mutant strains were generated and then analysed for flagellar gene expression (Heuner et al., 1995, 1997, 2000, 2002a,b; Jacobi et al., 2004; Molofsky et al., 2005; Brüggemann et al., 2006). Results of those studies allowed the construction of a hierarchical cascade of the flagellar regulon (Fig. 6.4).
FlaA (a class IV gene) encodes the flagellin and is directly regulated by the alternative σ28 factor (FliA). FliA was cloned by its ability to bind to the E. coli core RNA polymerase and direct transcription initiation from the flaA-specific promoter that was cloned upstream of a luxAB reporter gene. FliA of L. pneumophila partially restores the swarming ability of an E. coli fliA mutant strain. A fliA mutant of L. pneumophila does not express the flagellin and therefore is non-flagellated (Heuner et al., 1997, 2002a). Later it was discovered that L. pneumophila also need FliA to express some of its virulence traits (see below). An additional gene contributes to flaA expression, named flaR, which encodes a protein belonging to the LysR family (Schell, 1993) of transcriptional regulators (Heuner et al., 2000). FlaR binds to the flaA promoter. Therefore it seems possible that L. pneumophila relies on FlaR to repress FlaA expression during periods when FliA is needed to express virulence-associated traits, but FlaA expression would interfere with intracellular replication by consuming metabolic resources. However, further research is needed to verify this hypothesis.
The FliA-dependent expressed class IV genes are involved in assembly of the flagellum (flaA, fliD) and in motility (motY). So far little is known about fliA gene expression itself, but it is likely that RpoS and CsrA are involved (see below). Salmonella typhimurium control the activity of FliA by a post-transcriptional mechanism. By analogy, FlgM, an anti-σ28 factor, may bind to FliA to prevent activation of FliA-dependent gene expression (Kutsukake et al., 1994; Aldridge et al., 2006). Additionally, FlgN (the chaperone protein of FlgKL) may activate flgM mRNA in the absence of FlgKL. In other bacteria, once the flagellum hook-basal body structure is built, FlgM is exported and repression of FliA relieved so that the flagellum can be completed.
FleQ is the master regulator of the flagellar regulon (Jacobi et al., 2004). FleQ is a σ54 activator protein which, together with the σ54 factor (RpoN), positively regulates the expression of class II genes. Class II genes encode proteins of the basal body and hook structure. In both fleQ and rpoN mutant strains, the flagellin gene is still transcribed due to the activity of FliA. However, the amount of FlaA in both mutants is reduced compared with the wild type strain, whereas fliA gene expression is only slightly affected by a mutation in fleQ of strain Corby (Jacobi et al., 2004 and K. Heuner, unpublished). On the other hand, results of microarray analysis of fliA expression in L. pneumophila Paris revealed that fliA may be part of the FleQ regulon (Brüggemann et al., 2006, and C. Albert-Weissenberger, C. Buchrieser and K. Heuner, unpublished). Together theses results indicate that certain aspects of fliA regulation may vary in different strains. Nevertheless, because of the predicted absence of a basal body, bacteria that lack FleQ and RpoN cannot assemble the expressed flagellin into a flagellar filament. Therefore, flagellin is degraded and its translation is repressed.
Within the class II genes there is another σ54 activator protein, FleR (Jacobi et al., 2004), which seems to be a positive regulator of class III genes. But this hypothesis has to be tested by construction and analysis of a specific fleR mutant strain of L. pneumophila. However, FleS is the sensor protein of the putative FleSR two-component system, and a fleS mutant strain of L. pneumophila (strain AA100) was identified in a signature-tagged mutagenesis analysis. In particular, the fleS mutant strain was shown to be less invasive for host cells, but its flagellation was not investigated (Polesky et al., 2001).
The class III genes encode proteins involved in the formation of the motor and regulation of further flagellum assembly. So far, details of FleQ expression are not known, but primer extension analysis indicated that the promoter seems to depend on sigma factor RpoD (σ70). Furthermore, a Vfr-like (putative E. coli cAMP receptor protein homolog) DNA consensus sequence was identified upstream of its transcriptional start site (Jacobi et al., 2004). Experiments are ongoing to define further details of this cascade by analysing different mutant strains by transcriptome analysis and phenotypic characterization. A link between FleQ- and FliA-dependent expression of the flagellum may be mediated through the anti-sigma factor FlgM, which is present in each of the Legionella genomes sequenced so far. It is also possible that LetA and/or CsrA are involved in this process (see below).
The regulatory cascade for the flagellum is comparable to the cascade of Pseudomonas aeruginosa (Dasgupta et al., 2002, 2003; Jyot et al., 2002). In both bacteria, FleQ is the master regulator, and FleQ interacts with the σ54 factor to induce the expression of class II and class III genes. Regulation of fleQ expression seems to be σ70 factor-dependent and another regulatory protein (Vfr-like) is also involved. Furthermore, in both bacteria FliA, which induces expression of class IV genes, seems to be regulated independently of FleQ. On the other hand, there are also some differences between the species. Preliminary results from a DNA microarray study of fleR expression in wild type and a fleQ mutant strain of L. pneumophila Paris in our lab suggest that regulation via FleR may be different to that found in Pseudomonas (C. Albert-Weissenberger, K. Heuner and C. Buchrieser, unpublished). However, further experiments are needed to clarify this.
Some additional regulators contribute to optimal expression of the flagellin (Fig. 6.4), but, for the most part, the mechanisms involved are unknown and are proposed to be indirect. Hence, in all these studies, the expression of FliA and/or FlaA was investigated, but the expression of FleQ was not. Nevertheless, we will briefly present the results available so far: RelA, a guanine 3′,5′-bispyrophosphate (ppGpp) synthetase, and its product ppGpp influences flaA expression. Accumulation of ppGpp seems to trigger the initiation of flagellar biosynthesis (Hammer and Swanson, 1999). In a relA mutant strain, the flagellum is still expressed, but flaA expression is reduced (Zusman et al., 2002); the contribution of the second (ppGpp) synthetase, SpoT, remains to be addressed. The two-component regulator protein LetA (and LetS) positively influences flaA expression and motility. A more modest contribution is made by LetE, a regulatory protein of unknown function (Bachman and Swanson, 2004b, Gal-Mor and Segal, 2003). The homologueof the carbon storage regulator protein A (CsrA) of L. pneumophila negatively influences the expression of flaA and motility, probably by influencing fliA expression or by binding to flagellar gene transcripts (Fettes et al., 2001; Molofsky and Swanson 2003; Forsbach-Birk et al., 2004). Furthermore, the alternative sigma factor RpoS also positively influences the expression of fliA and thus also flaA (Bachman and Swanson, 2001, 2004a, Chapter 4). An in silico screen of the genome sequence of L. pneumophila for open reading frames encoding proteins with sequence similarities to members of the LuxR family of transcriptional regulators identified three proteins (LpnR2, LpnR3 and LetA) that may influence flaA expression (Lebeau et al., 2004). But until now there are no further details reported regarding the LpnR2- or LpnR3-dependent flagellin gene expression. read more ...
The FliA regulon
As mentioned above, FliA is not only the direct regulator of flagellin expression, but also contributes to expression of virulence traits of L. pneumophila (see below). Therefore, it is of interest to identify all FliA-dependent genes of L. pneumophila, the so-called FliA regulon, to determine which of these factors promote virulence. FliA recognizes a specific nucleotide sequence (σ28 consensus sequence, TAAA-N15-TCCGATAA) in the promoter region (Heuner et al., 1995, 1997). This knowledge was then used to screen the genome of L. pneumophila strain Paris for putative σ28-dependent promoters. Nearly 90 genes were identified; 30 genes exhibit very good matches, having a very similar consensus sequence plus a spacer distance equal to that found in the flaA promoter (C. Buchrieser, unpublished). However, in a recent transcriptome based analysis, only 10 genes were identified that appear to belong to the FliA regulon of intracellular replicating bacteria (Brüggemann et al., 2006). Five of these ten genes were not found in the previous 'in silico' screen. Only three of the ten genes identified belong to the flagellar regulon (flaA, fliD, motY). Of the seven other genes, three encode for proteins exhibiting a signal peptide (lpp0972, lpp1290, lpp2260), four encode proteins of unknown function (lpp1841, lpp2260, lpp2282, lpp2998), two for EnhA-like proteins (lpp0972, lpp1290) and one gene (lpp0952) encodes a putative regulator protein. Further analysis of these genes may help to elucidate the role of FliA for expression of virulence traits in L. pneumophila.
Contribution of the flagellar system to the virulence of Legionella
Role of the flagellum
Early it was shown that motile transmissive bacteria are more infectious for amoebae then the non-motile replicative phase, and it was hypothesized that flagellum expression and virulence are genetically linked (Rowbotham, 1986; Pruckler et al., 1995; Bosshardt et al., 1997; Gao et al., 1997; Byrne and Swanson, 1998; Hammer and Swanson, 1999; Heuner et al., 2002a). Furthermore, flagellated bacteria have been detected inside the human lung of Legionnaires' disease patients (Chandler et al., 1980b). Later Tn10 insertion mutants were described which are defective in flagellum expression and compromised in their ability to infect host cells. But some of these mutants multiply as well as the wild-type strain (Pruckler et al., 1995). By generating and analysing a specific flaA mutant strain, the ability of flagellin to enhance the invasion capacity and thus the infectivity of L. pneumophila was discovered. In contrast, intracellular replication was not significantly affected (Dietrich et al., 2001). L. pneumophila motAB mutant strains are less flagellated and have severely reduced motility compared to wild type. Once again, a motAB mutant strain is less invasive but its intracellular multiplication is not significantly affected. Thus, motility but not flagellin appears to promote contact with host cells and increase infectivity (Molofsky et al., 2005).
Legionella is able to form biofilms, and biofilm formation is believed to be a virulence factor for various bacteria. But the flagellum of L. pneumophila does not contribute to biofilm formation under the conditions tested so far (Mampel et al., 2006). We will also mention here that flagellin does affect host resistance to Legionnaires' disease (Hawn et al., 2003). Furthermore, loss of either bacterial flagellin or the macrophage cytosolic Naip5 (Birc1e) receptor makes mice susceptible to L. pneumophila infection, in part by impairing a pro-inflammatory cell death pathway of mouse macrophages (Swanson and Molofsky, 2005; Molofsky et al., 2006; Amer et al., 2006; Ren et al., 2006). However, addition of recombinant IFN-γ to the infected macrophages restricted replication of the flaA mutant within macrophages (Akamine et al., 2007). Furthermore, the flagellin of L. pneumophila is strongly immunogenic and capable of inducing a protective immunity in a A/J mouse model (Ricci et al., 2005). For details, see Chapter 5. read more ...
Role of the flagellar regulon
As mentioned above, the flagellum and motility contributes to the invasion capacity of L. pneumophila into host cells, but it seems not to be necessary for intracellular replication or biofilm formation. However, motility probably equips Legionella to reach the next host cell for further rounds of multiplication. But which regulators are involved in the proposed link between flagellum and virulence traits expression in L. pneumophila? When we inactivated the regulator of flaA expression, the fliA gene, this mutant strain does not replicate in Dictyostelium discoideum, whereas the flaA mutant or wild type strain does. In addition, the flaA mutant strain is less haemolytic than the wild-type strain (Heuner et al., 2002a, Heuner et al., 2006). These results clearly show that the flagellar regulon is involved in the virulence of L. pneumophila. It further suggests that in addition to flaA more genes regulated by FliA are needed for full fitness and virulence of this bacterium; together these are now called the FliA regulon (see above). At the same time, another group showed that a fliA mutant strain of L. pneumophila is non-cytotoxic and less infectious for macrophages. The same group also reported that FliA is involved in lysosome evasion and regulates binding to the dye crystal violet (Hammer et al., 2002; Molofsky et al., 2005). Recently, another group showed that a fliA mutant is reduced in its ability to form biofilms (Mampel et al., 2006). Taken together, all these results clearly confirm the hypothesis that the flagellar regulon is linked to the expression of virulence traits in L. pneumophila and that the FliA sigma factor plays a major role in this linkage.
On the other hand, the FliA regulon is not the only pathway involved in expression of flagellar regulon-dependent virulence traits in L. pneumophila. The FleQ/RpoN regulon and the FleSR two component system also affect L. pneumophila virulence (Polesky et al., 2001; Jacobi et al., 2004, Heuner et al., 2006). In particular, a FleS mutant strain is less invasive for host cells, and fleQ and rpoN mutant strains of L. pneumophila Corby exhibit less haemolytic activity against red blood cells than does the wild-type strain. Similar results were obtained by analysis of fleQ and rpoN mutants of strain Paris (C. Albert-Weissenberger and K. Heuner, unpublished). Therefore, the FleQ regulon is also involved in the expression of virulence traits of L. pneumophila. read more ...
Regulation of virulence factors in Legionella
The infectious, transmissive phase of L. pneumophila
Virulence of L. pneumophila seems to be multifactorial, involving a complex interplay between the bacteria and the host cell. The virulence factors acquired for replication in protozoan host cells also enable L. pneumophila to infect human cells, and regulation of the virulence gene expression in response to the growth phase is essential to ensure a successful infection. The life cycle of L. pneumophila consists of at least two distinguishable phenotypes: post-exponential phase bacteria escape from the host cell and are flagellated, highly motile, cytotoxic and virulent; replicative bacteria that reside in the host vacuole are sodium resistant, unflagellated, and non-cytotoxic (Byrne and Swanson, 1998; Molofsky and Swanson, 2004). For details, see Chapter 4.
The global regulatory genes csrA, letA/letS/letE and rpoS control maintenance and induction of either the replicative-focused genes or the genes necessary for the post-exponential phenotype (Fig. 6.5). The small RNA-binding protein CsrA represses the expression of transmissive traits in L. pneumophila (Molofsky and Swanson, 2003). By contributing to csrA RNA stability, Hfq is involved in maintenance of the exponential phase. Expression of hfq in exponential phase is induced by RpoS. In stationary phase, hfq transcription is repressed either directly or indirectly by LetA (McNealy et al., 2005). Besides, during nutrient deprivation (presumably signalled by ppGpp) the two-component regulator LetA/S inhibits CsrA activity, probably mediated by not yet identified regulatory RNAs (Hammer and Swanson 1999, Molofsky and Swanson 2003). This transition is enhanced by LetE (Bachman and Swanson, 2004b). The letE gene promoter region exhibits a RpoN consensus sequence and seems to be expressed in a RpoN-dependent manner (Brüggemann et al., 2006; C. Albert-Weissenberger, C. Buchrieser and K. Heuner, unpublished).
RpoS is also involved in the regulation of transmissive phase traits expression in Legionella, but its role is somewhat controversial. RpoS appears to be necessary for intracellular survival and is involved in regulating the replicative phase (Bachman and Swanson, 2001; 2004a; Abu-Zant et al., 2006). On the other hand, RpoS expression increases during stationary phase and coordinates the expression of several transmissive phase traits of L. pneumophila, but not stress resistance (Bachman and Swanson, 2001; 2004a; Abu-Zant et al., 2006). By analysing relA mutant bacteria, Abu-Zant and colleagues were not able to identify a link between the stringent response and RpoS. Therefore, further experiments are necessary to clarify the contributions of the stringent response and RpoS to the intracellular life cycle.
Temperature not only influences motility (as discussed above), but also the expression of virulence traits of L. pneumophila. L. pneumophila serogroup 1 multiplies within Acanthamoeba palestinensis at 35°C but not at 20°C. At 20°C, the amoebae instead digest most of the Legionella (Anand et al., 1983). Similarly, in A. castellanii L. pneumophila Philadelphia 1 grow intracellularly at 37°C but not at room temperature (Moffat and Tompkins, 1992). However, there are conflicting reports about the influence of temperature on the virulence of L. pneumophila. When grown at 37°C in chemostat culture, Legionella are significantly more virulent than those grown at 24°C, as judged by aerosol infection of guinea pigs (Mauchline et al., 1994). In a guinea pig model of lung infection, Edelstein and colleagues showed that bacteria grown in buffered yeast extract broth in batch culture at 25°C were more virulent then those grown at 41°C (Edelstein et al., 1987). However, different strains and different routes of infection were used in the studies, which may explain the differences observed. Furthermore, since neither study analysed bacteria cultured to the transmissive virulent phase, it is difficult to interpret these results. Presumably genetic differences between different L. pneumophila strains, e.g. the existence of the Lvh T4ASS (Samrakandi et al., 2002), also influences the ability to infect host cells at different temperatures. In fact, Ridenour and colleagues found that the lvh-encoded T4ASS positively influences the infectivity of L. pneumophila grown at lower temperatures (Ridenour et al., 2003). read more ...
Regulation of secretion systems
Pathogenicity of L. pneumophila depends greatly on the ability to secrete virulence factors that are displayed on the bacterial cell surface, secreted into the extracellular milieu or injected directly into the host cell. Mainly associated with the virulence of L. pneumophila are the Dot/Icm T4BSS and the Lsp T2SS, but also the twin-arginine translocation pathway (Tat) and the lvh encoded T4ASS. For details, see Chapters 8-10.
The Lsp secretion system promotes the ability of L. pneumophila to infect protozoan and macrophage hosts, to grow in the mammalian lung and to grow at low temperatures (Rossier and Cianciotto, 2001; Soderberg et al., 2004, DebRoy et al., 2006b). It is involved in secretion of various proteins including many virulence-associated proteins (Hales and Shuman, 1999; Chapter 8). The T2SS of L. pneumophila is dependent on the pilBCD locus that is also involved in the biogenesis of type IV pili and on the Legionella secretion pathway, encoded by the lspFGHIJK and lspDE loci. Expression of pilBCD is influenced by temperature, increasing at temperatures below 30°C. However, irrespective of temperature, the pilD expression and T2SS-dependent hydrolytic activity increases from exponential to stationary phase (Liles et al., 1998; Liles, et al., 1999; Flieger et al., 2001). The regulation of pilD has similarity with that of the L. pneumophila flagellin (flaA) gene. Like expression of flaA, pilD expression is regulated not only by temperature but also growth phase- (Ott et al., 1991; Pope et al., 1996; Heuner et al., 1997; Heuner et al., 1999; Heuner et al., 2002; Brüggemann et al., 2006). When type II-dependent enzymatic activities were monitored, L. pneumophila rpoS and letA mutants exhibited a significant reduction of secreted phospholipase A, glycerophospholipid:cholesterol acyltransferase, protease, phosphatase, decreased p-NPPC hydrolase activities and increased lysophospholipase A and lipase activities. In letA and rpoS mutant strains, the plaC mRNA level was reduced compared to the wild-type (Flieger et al., 2002; Broich et al., 2006).
Whether LetA and RpoS influence the hydrolytic enzyme activities directly, via transcriptional regulation of structural gene, or indirectly, by controlling the expression of activators, cofactors, or components of the secretion apparatus, still needs to be elucidated by future studies. Although the activities of these hydrolytic enzymes depend on the T2SS, the type II-dependent RNase activity was not affected, indicating that LetA and RpoS do not regulate the type II machinery directly (Broich et al., 2006).
L. pneumophila requires the Dot/Icm-T4SS to replicate in both protozoa and human macrophage hosts (Andrews et al., 1998; Edelstein et al., 1999; Segal and Shuman, 1999). To establish a replicative niche after invading alveolar macrophages, this secretion system is proposed to translocate bacterial protein substrates into eukaryotic host cells during infection (Segal and Shuman, 1999; Dumenil and Isberg, 2001). The apparatus also mediates an anti-apototic signalling cascade in human macrophages (Abu-Zant et al., 2007). Vincent and colleagues have conducted a comprehensive biochemical and genetic examination of the Dot/Icm secretion system (Vincent et al., 2006b; Chapter 9). At present, a large number of substrates of the Dot/Icm-T4SS have been identified and their activities partially analysed.
Regulation of the genes encoding the Dot/Icm-T4SS seems to be complex and is probably directed by several regulators. Although some studies have been performed, there is still only a diffuse picture of Dot/Icm-T4SS regulation. The sigma factor RpoS, the ppGpp synthetase RelA, and the two-component regulatory system LetA/S affect expression of genes encoding the Dot/Icm secretion system (Zusman et al., 2002; Gal-Mor and Segal, 2003; Lynch et al., 2003; Shi et al., 2006). CpxR also plays a role in regulation of various dot/icm genes (Gal-Mor et al., 2002; Gal-Mor and Segal, 2003; Vincent et al., 2006a). The promoter region of some of the icm genes exhibit an RpoD-dependent -10 promoter consensus sequence and a conserved CpxR binding site ([T]GTAAA-box). Nevertheless, a cpxR mutant showed no growth defect inside A. castellanii and HL-60-derived human macrophages (Gal-Mor and Segal, 2003), indicating that other regulators must also contribute to expression of the Dot/Icm-T4SS. Indeed, the fir genes of Legionella, encoding chaperones of IcmQ, exhibit both a CpxR-binding site (TGTAAANAANNNGAAAG) and a PmrA-binding site (CTTAAGNTTNNCTTAAG/TNNT) that positively influences the expression of the fir genes (Feldman and Segal, 2007). The PrmA binding site was not present in the icmR (fir) gene of L. pneumophila. A similar PmrA box (cTTAATaTT) was identified in the promoter region of several Dot/Icm substrates and genes coregulated with these effector genes (Ceg genes) (Zusmann et al., 2007).
Expression of almost all of the reported substrates of the Dot/Icm secretion system, such as RalF, SidC and members of the SidE family, are induced in stationary phase (Nagai et al., 2002; Conover et al., 2003; Luo and Isberg, 2004; Bardill et al., 2005, Brüggemann et al., 2006; Zusman et al., 2007). LetA is necessary for full transcriptional expression of the ralF gene and RpoS for the sidG gene (Shi et al., 2006; Zusman et al., 2007). These observations are consistent with the enhanced virulence of L. pneumophila in stationary phase. The response regulator PmrA is also involved in the expression of Dot/Icm-translocated substrates. A mutant strain of pmrA does not replicate within A. castellanii (Zusman et al., 2007). In contrast, SidJ expression is constant during the entire cell cycle of L. pneumophila (Liu and Luo, 2007). Similarly, LidA expression is only slightly induced in stationary phase bacteria, and the protein is secreted into the host cell throughout the replication cycle (Derre and Isberg, 2005). Expression of at least some Dot/Icm secretion system substrates, such as SidJ, RalF, and SdeC, is independent of a functional Dot/Icm transporter (Vincent et al., 2006a; Liu and Luo, 2007). The expression pattern of the substrates presumably reflects the temporal requirements of each effector protein during the cell cycle. read more ...
Regulation of other virulence factors
A large number of factors that contribute to virulence have been described, including katA, katB, prpCD, milA, iraA, pmiA, nudA, lpnE, lvgA, rtxA, the ccm locus, ligA, feoAB, rcp and laiA (Bandyopadhyay and Steinman, 1998; Stone et al., 1999; Bandyopadhyay and Steinman, 2000; Fettes et al., 2000; Harb and Abu Kwaik, 2000; Viswanathan et al., 2000; Robey et al., 2001; Cirillo et al., 2002; Robey and Cianciotto, 2002; Edelstein et al., 2003; Chang et al., 2005; Edelstein et al., 2005; Miyake et al., 2005; Newton et al., 2006). However, little is known about the mechanisms that regulate their expression. Some of these factors are discussed below.
The macrophage infectivity potentiator (Mip)
Mip belongs to the enzyme family of FK-506 binding proteins that exhibit peptidyl prolyl cis/trans isomerase (PPIase) activity. It promotes the presence of a p-NPPC hydrolase activity in culture supernatants and is involved in the infection of eukaryotic host cells (Wintermeyer et al., 1995; Cianciotto, 2001; Helbig et al., 2001; Helbig et al., 2003; Kohler et al., 2003; DebRoy et al., 2006a, Wagner et al., 2007). Although the Mip protein is constitutively expressed in both free-living and intracellularly replicating L. pneumophila (Kohler et al., 2000), the transcriptional activity of the mip promoter is repressed after invasion of the monocytic cell line MonoMac 6, and then regains activity after 24 h of intracellular replication (Wieland et al., 2002). Thus, its pattern of expression is typical of flagellin and other factors thought to be dedicated to transmission but dispensable for replication (Molofsky and Swanson, 2004). Secretion of Mip into the supernatant apparently is positively regulated by LetA, since a letA mutation dramatically decreases the secretion of Mip but has no effect on membrane-associated Mip (Shi et al., 2006).
The major cell-associated phospholipase A (PlaB)
PlaB contributes to the cytotoxicity of L. pneumophila, as it possesses phospholipase A and lysophospholipase A activity. Although the plaB mutant multiplies like wild type in U937 macrophages and in A. castellanii amoeba (Flieger et al., 2004), plaB probably promotes dissemination of L. pneumophila in tissues (K. Heuner, E. Schunder, and F. Higa, unpublished). plaB is expressed especially before entry into the late logarithmic growth phase, and PlaB is responsible for the major lipolytic activity present during host cell infection. LetA and RpoS presumably promote PlaB expression or activity, since the cell-associated phospholipase A and lysophospholipase A activities are significantly reduced in letA and rpoS mutants (Broich et al., 2006). read more ...
The major secreted zinc metalloproteinase A (MspA)
MspA is encoded by the proA gene. Although Blander and colleagues (1990) initially reported that MspA is not a virulence factor in guinea pigs, later studies showed that a proA mutant was attenuated in a guinea pig model. Furthermore, the proA gene was identified in a screen for virulence genes that used signature-tagged mutagenesis (Moffat et al., 1994; Edelstein et al., 1999). The proA gene is expressed during intracellular growth in amoebae (Moffat et al., 1994). Likewise, MspA is produced in the lungs of guinea pigs during infection (Williams et al., 1987; Conlan et al., 1988). Purified protease can cause lesions like those of Legionnaires' disease in guinea pig and human lung (Baskerville et al., 1986; Conlan et al., 1986). Furthermore, guinea pigs immunized with MspA develop a protective immunity against lethal aerosol challenge with L. pneumophila (Blander and Horwitz, 1989, 1991), confirming its expression during infection and its contribution to disease. MspA presumably alters human phagocyte function (Sahney et al., 2001). Overall, the data indicate that MspA may play a role in the pathogenesis of human Legionnaires' disease. LetA and RpoS may positively regulate the activity of secreted MspA (Broich et al., 2006). As an aside, MspA itself is necessary for activation of PlaC (Banerji et al., 2005).
The global stress protein (GspA)
GspA is needed for expression of full stress resistance of L. pneumophila. The sigma factors σ7° (RpoD) and σ32 (RpoH) seem to be involved in gspA expression (Abu Kwaik and Engelberg, 1994; Abu Kwaik et al., 1997). The gspA gene is induced during intracellular infection and by heat shock, oxidative stress, acid- and osmotic shock.
The heat shock protein (Hsp60)
Another protein involved in stress response is the 60 kDa heat shock protein (Hsp60), encoded by the htpAB operon (Hoffman et al., 1990). Hsp60 is necessary for full invasion of L. pneumophila. The protein is found in both the cytoplasm and the cell envelope, and it is also expressed and secreted during infection (Fernandez et al., 1996; Garduno et al., 1998a, b). Furthermore, the Dot/Icm secretion system seems to be involved in its secretion (Chong et al., 2006). Hsp60 may also alter the pathogen-host interaction by interfering with signalling pathways, as suggested by the developmental changes initiated by a Saccharomyces cerevisiae that express L. pneumophila Hsp60. Upstream of htpA, an RpoH-like promoter element was identified (Hoffman et al., 1990), consistent with the observation that the htpAB promoter seems to be under the control of RpoH. The authors proposed that RpoH is essential for Legionella, because they were unsuccessful in generating a rpoH mutant strain (Ewann and Hoffman, 2006). However, further research is necessary to elucidate the role of RpoH in L. pneumophila. read more ...
Conclusions
Thirty years after the identification of L. pneumophila as the causative agent of Legionnaires' disease, various factors and mechanisms responsible for the pathogenicity of L. pneumophila have been identified. But there is still little information about the mechanisms that regulate the virulence of L. pneumophila. We are well on the way to understanding the mechanisms of gene regulation of the flagellum, or the flagellar regulon, in more detail. However, there are still open questions regarding the link between the expression of the virulent phenotype and the expression of the flagellum. Except for the flagellin, further virulence factors regulated by the flagellar regulon (e.g. by FliA) are still waiting to be investigated. With the identification and characterization of FliA and LetA, we will probably be able to answer many remaining questions in the near future. The completion of the genome sequences of four different L. pneumophila strains which now provides the opportunity to perform whole genome microarray-based studies will be especially helpful in efforts to identify and characterize both additional regulators and their target genes.
The flagellum and the flagellar regulon remains a promising field to study not only Legionella virulence and protein secretion (T3SS), but also to investigate pathogen-host interactions, since flagellin recognition by Toll-like receptor 5 or Naip5 receptor activates the innate immune system. Together all these aspects indicate that the flagellar system is a rich area for further research in the L. pneumophila pathogenesis field. read more ...
Acknowledgements
Own work was supported by grants from the Deutsche Forschungsgemeinschaft (grants GRK 587/2, HE2845/2-1/2, HE2845/4-1 and HE2845/5-1) and by the Bavarian Research Foundation (Bayerische Forschungsstiftung).
from Legionella: Molecular Microbiology
See also: Pili and Flagella: Current Research and Future Trends
Microbiology news and viewsMolecular biology news and viewsPCR news and viewsMolecular biology conferences
Microbiology conferencesMolecular biology conferences
Microbiology journalsCurrent Issues in Molecular Biology
Open Access Biology Home | PCR | Microbiology Books | Molecular Biology Books | Molecular Biology Gateway | PCR Protocols | Real-Time PCR |