The Dot/Icm Type IVB Secretion System of Legionella

The Dot/Icm Type IVB Secretion System of Legionella

Carr D. Vincent and Joseph P. Vogel

Abstract

The major virulence system of Legionella pneumophila is a specialized secretion system encoded by 26 dot/icm genes. This secretion machine is classified as a type IVB secretion system due to its homology to the transfer apparatus of IncI conjugative plasmids. In Legionella, the Dot/Icm secretion system is used to export a large number of protein substrates into the cytoplasm of phagocytic cells, thereby allowing Legionella to survive and replicate within these normally bactericidal cells. read more ...

Legionella introduction

Legionella pneumophila is a Gram-negative bacterium that is found in the environment as a parasite of freshwater amoebae (Fields et al., 2002). Unlike most bacteria, which are rapidly killed after being phagocytosed by an amoeba, L. pneumophila cells are able to survive in these eukaryotic hosts and replicate, eventually killing the host cell. When humans inhale aerosolized water droplets containing these bacteria or infected amoebae, the bacteria can cause either a mild disease with flu-like symptoms known as Pontiac fever or a severe form of pneumonia known as Legionnaires' disease.

L. pneumophila is able to cause disease because it can prevent being killed by human alveolar macrophages, immune cells that are normally highly bacteriocidal (Horwitz and Silverstein, 1980). The life cycle of L. pneumophila in macrophages appears to be similar to its life cycle in amoebae. Immediately upon uptake by a macrophage, the Legionella containing vacuole (LCV) avoids fusion with endocytic vesicles and lysosomes (Horwitz, 1983b). Instead, the LCV recruits endoplasmic reticulum (ER)-derived vesicles, acquiring the ER protein markers Rab1, Sec22b, Arf1, and calnexin (Kagan and Roy, 2002; Derre and Isberg, 2004; Kagan et al., 2004). After transiently associating with mitochondria, the LCV becomes surrounded by rough ER, forming a compartment known as the replicative phagosome (Horwitz, 1983a; Swanson and Isberg, 1995). It is in this compartment that the bacteria multiply. During the replication cycle in A/J mouse bone marrow-derived macrophages, the LCV becomes acidified and acquires lysosomal markers (Sturgill-Koszycki and Swanson, 2000). Thereafter, the bacteria lyse the macrophage and spread to uninfected host cells.

Several genetic screens have been performed to identify virulence factors required for L. pneumophila to grow in macrophages (Berger and Isberg, 1993; Sadosky et al., 1993; Vogel et al., 1996; Andrews et al., 1998; Segal et al., 1998; Vogel et al., 1998). These screens primarily identified a set of genes necessary for avoidance of the endocytic pathway that were named dot (defect in organelle trafficking) or icm (intracellular multiplication). These genes are linked in two pathogenicity islands on the L. pneumophila chromosome and were subsequently shown to encode a specialized secretion system called a type IV secretion system (T4SS) that is related to the conjugation system of an IncI plasmid (Segal et al., 1998; Vogel et al., 1998). This specialized secretion system is the focus of this review. read more ...

Type IV secretion systems and type IV coupling proteins

T4SSs include both plasmid transfer systems and adapted conjugation systems utilized by pathogens to deliver proteins and/or DNA into eukaryotic host cells. As plasmid transfer systems are better characterized, it is worth reviewing what is known about these systems prior to discussing adapted conjugation systems. Components involved in plasmid transfer systems can be grouped into three classes of proteins; Mpf, Dtr, and T4CP (Schroder and Lanka, 2005). The Mpf (mating pair formation) proteins are involved in assembling the conjugal pilus and the channel structure that spans the bacterial inner and outer membranes. The Dtr (DNA transfer and replication) proteins are involved in plasmid replication and generation of the protein-DNA substrate that is delivered to recipient cells. Finally, T4CPs (type IV coupling proteins) are inner membrane proteins that mediate interactions between the Mpf and Dtr proteins, 'coupling' the substrates to the export apparatus. All T4CPs contain an ATPase domain that is required for their function, suggesting that these proteins provide energy for substrate export.

The T4CPs perform several critical roles in type IV secretion systems. First, they link substrates to the Mpf export apparatus by functioning as the inner membrane receptors for Mpf complexes. Second, T4CPs play a critical role in determining the specificity of the Mpf complex for a particular substrate or class of substrates. Third, they are thought to play an active role in substrate translocation by functioning as molecular motors, providing energy through ATP hydrolysis to drive export of substrates.

T4CPs have been shown to interact with both their cognate Dtr and Mpf proteins. Direct interactions between several plasmid transfer coupling proteins and their substrates (a relaxase that is covalently attached to the single-stranded DNA substrate) have been experimentally demonstrated (Cabezon et al., 1997; Hamilton et al., 2000; Szpirer et al., 2000; Schroder et al., 2002; Atmakuri et al., 2003). Coupling proteins have also been shown to interact with Mpf proteins. In A. tumefaciens, the coupling protein VirD4 has been shown to interact with the core subcomplex protein VirB10, as well as with the ATPases VirB4 and VirB11 (Atmakuri et al., 2004; de Paz et al., 2005). Likewise, in the plasmid transfer systems R27 and R388, the coupling proteins TraG and TrwB have been shown to interact with the respective VirB10 homologues of these systems, TrhB and TrwE (Gilmour et al., 2003; Llosa et al., 2003). Thus, there is a growing body of evidence that the coupling proteins are the inner membrane receptors for secreted substrates of T4SS, linking substrates to the secretion apparatus.

Several studies indicate that the T4CP is the primary factor that determines the specificity of the secretion system for a particular substrate (Cabezon et al., 1994; Hamilton et al., 2000; Llosa et al., 2003). This has been demonstrated in studies using chimeric secretion systems, consisting of a T4CP from one system expressed with the Mpf complex from another. For example, Cabezon and colleagues investigated the specificities of the R388 and RP4 plasmid transfer systems (Cabezon et al., 1994). The R388 plasmid transfer system can transfer both the R388 plasmid and the broad host range plasmid RSF1010. Similarly, the RP4 transfer system can transfer both the RP4 plasmid and RSF1010. When the coupling protein from RP4 (TraG) was expressed with the Mpf complex from R388, this hybrid T4SS retained activity, as it was able to transfer the RSF1010 plasmid. However, the hybrid T4SS could no longer transfer the native substrate of the R388 Mpf, the R388 plasmid. Thus, the Mpf complex does not confer substrate specificity, nor does it exhibit specificity for a particular T4CP. However, the individual coupling proteins may interact with particular substrates, demonstrating that the specificity of the transfer system is determined solely by the coupling protein.

The third proposed function of T4CPs in substrate secretion is that they serve as molecular motors, providing mechanical energy through hydrolysis of ATP to 'pump' substrates through the complex. In support of this proposed role, T4CPs exhibit limited sequence similarity to the DNA motor proteins FtsK and SpoIIIE, both of which are involved in chromosome segregation (Llosa et al., 2002). These proteins have been shown to translocate along DNA in the presence of ATP, suggesting a similar mechanism for T4CPs in plasmid translocation. T4CPs contain highly conserved Walker box motifs, indicating that ATP hydrolysis is likely to be essential for their function (Walker et al., 1982). DNA dependent ATP hydrolysis activity was recently demonstrated for the R388 coupling protein TrwB and several T4CPs have been shown to require their Walker box motifs for plasmid transfer (Balzer et al., 1994; Moncalian et al., 1999; Kumar and Das, 2002; Tato et al., 2005). Moreover, the crystal structure of TrwB reveals that the protein forms a hexamer with a quaternary structure similar to F₁-ATPase (Gomis-Ruth et al., 2001). The hexamer has a central channel approximately 22 Angstroms in width that constricts to seven Angstroms at the cytoplasmic face of the protein. It has been proposed that this channel may serve as the conduit through which secreted proteins and DNA are pumped by the T4CP across the inner membrane. read more ...

Adapted conjugation systems

In addition to their function as plasmid transfer systems, T4SS contribute to the virulence of a variety of bacterial pathogens, including the plant pathogen Agrobacterium tumefaciens and the mammalian pathogens Bartonella henselae, Bordetella pertussis, Coxiella burnetii, Helicobacter pylori, and Legionella pneumophila (Cascales and Christie, 2003). These T4SS, termed adapted conjugation systems, are used by pathogens to deliver toxins into host cells and are thought to have evolved from plasmid transfer systems, as they have significant sequence homology (Winans et al., 1996; Christie and Vogel, 2000). In the case of the L. pneumophila Dot/Icm T4SS, this evolutionary relationship is clear, as the secretion system retains the ability to transfer the broad host-range RSF1010 plasmid to a bacterial recipient (Segal et al., 1998; Vogel
et al., 1998).

Adapted conjugation systems can be divided into two groups, type IVA and type IVB, based on sequence homology (Christie and Vogel, 2000). Although both classes are thought to be functionally homologous, they share very little sequence similarity and type IVB systems are typically composed of approximately twice as many proteins. The best-characterized type IV secretion systems are the type IVA systems, which include the extensively characterized A. tumefaciens VirB/D4 secretion system and related T4SS. The best-characterized type IVB secretion system is the Legionella pneumophila Dot/Icm system.

The A. tumefaciens system is composed of twelve proteins, VirB1-11 and VirD4 (Christie et al., 2005; Schroder and Lanka, 2005). Much is known about the assembly and structure of the proteins that make up these secretion machines, including protein interactions between subunits, the biochemical functions of individual components, and even the translocation route of substrates as they pass through the secretion apparatus. This secretion system contains three energy transducing proteins: VirB4, VirB11, and the coupling protein VirD4. Each of these proteins contains a conserved Walker box nucleoside triphosphate binding motif that is required for substrate secretion. The 'core' of the secretion complex is composed of the VirB6, VirB7, VirB8, VirB9, and VirB10 proteins. Three of these proteins, VirB6, VirB8, and VirB10 are inner membrane proteins; VirB9 and the lipoprotein VirB7 are found in the outer membrane linked to each other by a disulfide bond. Conformation of VirB10 has been shown to change in response to ATP hydrolysis by the VirB11 and VirD4 proteins. Accordingly, VirB10 is thought to function as an energy 'sensor' that regulates the formation of a bridge between the inner and outer membrane components (Cascales and Christie, 2004). The secretion system also assembles a pilus, composed of VirB2 (the pilin) and a minor component VirB5. The pilus is thought to be attached to the core complex through interactions with the lipoprotein VirB7.

It should be noted that two adapted conjugation systems that lack a coupling protein have been identified, the Brucella spp. Vir T4SS and the Bordetella pertussis Ptl T4SS. Little is known about the Brucella T4SS due to the intractability of this organism to genetic analysis. It is known, however, that the B. pertussis Ptl substrate, pertussis toxin, contains a Sec signal sequence and is transported across the cytoplasmic membrane by the Sec machinery prior to secretion across the outer membrane by the Ptl T4SS (Burns, 2003). Thus, this exceptional case hints that the function of the T4CPs in other systems is similar to the role of the Sec machinery in transporting substrates across the cytoplasmic membrane (Schroder and Lanka, 2005).

In contrast to the well-characterized type IVA system of A. tumefaciens, relatively little is known about the type IVB systems. Type IVB systems include the Dot/Icm T4SS of L. pneumophila and Coxiella burnetii, as well as the plasmid transfer systems R64 and ColIb-P9 (Christie and Vogel, 2000; Sexton and Vogel, 2002). Only three Dot/Icm proteins have detectable homology to the type IVA systems: DotB is homologous to VirB11, the carboxy-terminus of DotG is homologous to VirB10, and DotL has limited homology to the coupling protein VirD4. The remaining Dot/Icm proteins have no homology to the VirB/D4 system, making predictions of protein functions difficult. read more ...

Overview of the Legionella dot/icm genes

Twenty-six dot/icm genes have been identified. Table 9.1 includes a description of the Dot/Icm proteins including their size, predicted localization, experimentally determined localization via fractionation experiments, other relevant information, and key references. Following is a brief description of the Dot/Icm proteins that have been characterized to date.

DotA

DotA was the first Dot protein identified (Berger and Isberg, 1993). Historically, a dotA mutant (Lp03) has served as the 'standard' dot/icm mutant in many laboratories, as this mutant is completely defective for all of the known dot/icm phenotypes including intracellular growth, alteration of the host endocytic pathway, contact dependent cytotoxicity, and plasmid conjugation. DotA is a large (113 kDa) protein with eight transmembrane domains (Roy and Isberg, 1997). Surprisingly, this protein has been shown to be secreted by wild-type L. pneumophila during growth in broth (Nagai and Roy, 2001). Secreted DotA forms ring-shaped structures, and it has been speculated that these structures may form channels in host membranes for the delivery of effectors. However, it is unclear how DotA could function as both an inner membrane component of the Dot/Icm secretion apparatus and as a secreted substrate. Moreover, there is no precedent for secretion of a polytopic inner membrane protein by any known bacterial secretion system. Thus, given the present evidence, it is unlikely that DotA is a secreted substrate of the Dot/Icm T4SS. Instead, release of DotA is likely due to shedding of membrane proteins by L. pneumophila during growth in liquid media.

DotB

One of the most extensively characterized Dot/Icm proteins is DotB, a distant homologueof the A. tumefaciens protein VirB11. In A. tumefaciens, VirB11 is thought to provide energy to the secretion machinery and to be involved in assembly of the secretion apparatus, pilus assembly, and substrate export (Sagulenko et al., 2001; Schroder and Lanka, 2005). DotB is one of three Dot/Icm proteins with a Walker box motif, suggesting that DotB is involved in providing energy to the complex through hydrolysis of ATP. DotB was shown to have ATPase activity, form hexameric rings, and to be peripherally associated with the inner membrane, presumably bound to inner membrane components of the Dot/Icm complex (Sexton et al., 2004b). Although DotB interaction partners have not yet been identified, the DotL protein is a likely candidate based on a report demonstrating interactions between VirB11 and VirD4, a DotL homologue(Atmakuri et al., 2004). Three roles have been proposed for DotB (Sexton et al., 2005). The first role is in assembly of the Dot/Icm complex. Although this has not been definitively demonstrated, this role of DotB seems likely based on the function of its homolog, the VirB11 protein in A. tumefaciens. The second function of DotB is in protein export. This role has been confirmed, as single amino acid changes in DotB conferred defects in the export of a subset of Dot/Icm substrates (Sexton et al., 2005). Third, DotB may have a role in retraction of the Dot/Icm pilus, as DotB has homology to PilT, a protein involved in retraction of pili in type II secretion systems. This activity has not been confirmed, since the Dot/Icm pilus has not been identified. read more ...

DotL

The DotL protein has two predicted transmembrane domains at its amino-terminus and has been experimentally localized to the inner membrane (Buscher et al., 2005). It also contains a highly conserved Walker box motif, suggesting ATP binding and hydrolysis is important for the function of DotL. DotL has limited homology to type IV coupling proteins (Buscher et al., 2005). For example, DotL is similar to a well characterized T4CP, TrwB, from the plasmid R388, although this low level of homology (BLAST score 7e-2) is not surprising given the generally low sequence homology between members of the T4CP family. As the type IV coupling protein, DotL is thus predicted to perform two critical functions in the secretion apparatus: (1) as the receptor for secreted substrates, and (2) to provide energy to the secretion apparatus to drive substrate translocation.

An unusual property of DotL is that it is not only essential for growth inside macrophages, it is also essential for growth on bacteriological media in the wild-type L. pneumophila strain Lp02 (Buscher et al., 2005). This phenotype is not shared by any other known type IV coupling protein. In addition to dotL, two other dot/icm genes are essential, dotM and dotN. Although the functions of these proteins are not known, the conservation of this unusual phenotype suggests that DotL, DotM, and DotN may function together as a subcomplex (discussed further below). Interestingly, although the dotL gene is essential in a wild-type Lp02 strain background, suppressor mutations of the ΔdotL lethality phenotype can easily be obtained. The high frequency of spontaneous suppressor mutations suggests that there are a large number of genes that will suppress the loss of dotL, and it has been demonstrated by our laboratory that mutations in 20 of the 23 non-essential dot/icm genes can suppress the loss of dotL (Buscher et al., 2005). To explain these observations, it was proposed that DotL may regulate the secretion channel and control whether the complex is open or closed. Loss of DotL could result in a constitutively open complex, which would be detrimental to the cell. Suppressor mutations in other dot/icm genes that disrupt the secretion pore would rescue the detrimental effects of the loss of the channel 'gatekeeper.' Thus, the surprising phenotype of ΔdotL lethality distinguishes it from other coupling proteins, but at the same time supports the model that coupling proteins regulate secretion by the T4SSs.

Interestingly, a dotL deletion is not lethal in another wild-type L. pneumophila strain, JR32, although the mutant cells become hypersensitive to low salt concentrations on agar medium (Buscher et al., 2005). Because both Lp02 and JR32 were derived from the same clinical isolate, Philadelphia-1, it is currently unknown how this difference between strains arose. One possible explanation is that during passage in vitro, Lp02 acquired a mutation that confers ΔdotL lethality. Alternatively, explanation JR32 may have acquired a mutation that mediates resistance to ΔdotL lethality.

The ΔdotL lethality phenotype has proven useful in a number of studies. In the Vogel laboratory, a screen to identify suppressors of ΔdotL lethality identified many known dot/icm genes as well as a novel dot/icm gene, dotV (J.A. Sexton and J.P. Vogel, unpublished). In addition, ΔdotL lethality was used as a selection for dominant negative alleles of DotB, aiding in the study of the structure and function of this protein (Sexton et al., 2005). We have also used the ΔdotL lethality phenotype to identify factors necessary for the assembly and/or activation of the Dot/Icm complex (Vincent et al., 2006a). In summary, ΔdotL lethality supports the current model for the role of coupling proteins as regulators of type IV secretion and has proven useful in the identification and characterization of the Dot/Icm proteins. read more ...

DotO

DotO has been proposed to be a functional homologueof the A. tumefaciens VirB4 protein, an inner membrane ATPase that is required for assembly of the secretion apparatus (Schroder and Lanka, 2005; Segal et al., 2005). Like DotB and DotL, DotO contains a Walker box domain, indicating that this protein is likely to provide energy to the Dot/Icm apparatus through hydrolysis of ATP. Although DotO does not contain any clear transmembrane motifs, it localizes exclusively to the inner membrane (Vincent et al., 2006b). Interestingly, DotO and DotH were observed by immunofluorescence microscopy in a fibrous structure on the surface of L. pneumophila after exposure to macrophages or to macrophage-conditioned medium (Watarai et al., 2001). This surface structure was also observed late in the intracellular growth cycle, shortly before the bacteria escape from the host cell. Based on this, it was proposed that this structure might be involved in escape from the host cell or in infection of new host cells. However, an alternate explanation for the exposure of DotO and DotH is that after infection of macrophages, wild-type L. pneumophila cells become susceptible to the fixation conditions used for microscopy, and fixation results in the release of intracellular contents. Unfortunately, it is difficult to test this hypothesis and the reason for the surface exposure of DotO and DotH remains unclear.

DotU/IcmF

DotU and IcmF are somewhat misleadingly named, as these proteins are not thought to be part of the Dot/Icm secretion apparatus itself, but rather function as accessory factors that are required for stabilization of other Dot/Icm proteins (Sexton et al., 2004a; VanRheenen et al., 2004). Supporting this idea, the DotF, DotG, and DotH proteins are destabilized in a ΔdotU ΔicmF mutant. Interestingly, homologues of dotU and icmF are found in a wide variety of bacterial species, where they are frequently associated with genes encoding the recently discovered type VI secretion systems (IAHP loci) (Sexton et al., 2004a; Mougous et al., 2006). Thus, the function of DotU and IcmF as 'macromolecular chaperones' appears to be conserved; once horizontally acquired by L. pneumophila these proteins may have been become adapted to the Dot/Icm type IV secretion system.

IcmQ

IcmQ is found in the bacterial cytoplasm and purified IcmQ can insert into membranes and form pores (Dumenil et al., 2004). Membrane insertion and the pore-forming activity of IcmQ is inhibited in the bacterial cytoplasm by its chaperone, IcmR (Dumenil and Isberg, 2001). In addition, IcmQ can be observed on the surface of L. pneumophila after exposure to macrophages (Dumenil et al., 2004). Based on these findings, it has been proposed that IcmQ may have a role in forming the Dot/Icm secretion channel or in forming pores in the phagosomal membrane. However, several caveats to this model must be considered. Exposure of IcmQ during infection was observed in wild-type and dot/icm mutants, demonstrating that IcmQ exposure is not the result of secretion of IcmQ by the Dot/Icm complex. Like the surface exposure of DotO and DotH, surface exposure of IcmQ could be due to susceptibility of L. pneumophila cells to release of cellular contents under the fixation conditions used. Moreover, it is difficult to interpret the significance of characterization of IcmQ in the absence of IcmR. IcmR is not present in the closely related Dot/Icm T4SS found in Coxiella burnetii, although this system does encode an IcmQ homolog. However, the C. burnetii IcmQ protein has an amino-terminal extension that may eliminate the need for IcmR in this pathogen (Zamboni et al., 2003). Thus, examination of the role of IcmQ in the absence of IcmR may not be physiologically relevant, as these proteins appear to be covalently linked in C. burnetii and may never exist separately in L. pneumophila. While it is clear that IcmQ is required for the function of the Dot/Icm complex, its specific role remains unclear. read more ...

IcmT

IcmT is a predicted inner membrane protein that is required for growth in macrophages and amoebae (Segal and Shuman, 1999). IcmT has a transmembrane domain at its amino terminus and is thought to be a component of the secretion apparatus. Mutations that result in carboxy-terminal truncations of IcmT have been identified that affect the ability of L. pneumophila to form pores and escape from host cells (Molmeret et al., 2002a; Molmeret et al., 2002b; Bitar et al., 2005). These data demonstrate that IcmT is required for both secretion of substrates involved in formation of the L. pneumophila replicative phagosome and for lysis of host cells.

Remaining Dot/Icm proteins

Little is known about the remaining Dot/Icm proteins. DotK, like the core subcomplex proteins DotC and DotD, contains a lipobox motif that is essential for its function (Yerushalmi et al., 2005). DotK is the only outer membrane component that is partially required for intracellular growth and its role in the T4SS is unclear (Segal and Shuman, 1999). Only one Dot/Icm protein localizes exclusively to the periplasm, IcmX (Matthews and Roy, 2000). Although this protein has been shown to be required for the function of the type IV secretion system, its specific role is also unknown.

Similarly, the roles of the remaining inner membrane proteins, DotE, DotI, DotJ, DotP, DotV and IcmV, remain uncharacterized. Three of these proteins, DotE, DotP, and DotV, exhibit a significant degree of homology, although they are not functionally redundant as each is required for intracellular growth (Segal et al., 2005). Based on its size and predicted secondary structure, DotP is a candidate to form the Dot/Icm pilus, although this hypothesis has not been tested. According to this model, DotE and DotV may function akin to pseudopilins, but this also has not been tested. Interestingly, DotJ and DotV (like the cytoplasmic chaperone IcmR) do not have homologues in the C. burnetii Dot/Icm T4SS, indicating that their roles may be unique to L. pneumophila (Vogel, 2004).

The Dot/Icm complex consists of at least two critical subassemblies. The first is composed of the inner membrane proteins DotF and DotG and the outer membrane proteins DotC, DotD, and DotH. This subcomplex forms the 'core' of the T4SS, bridging the inner and outer membranes, and it was designated as the core-transmembrane complex (Vincent et al., 2006b). The second subassembly includes the type IV coupling protein DotL, the inner membrane proteins DotM and DotN, and the cytoplasmic adaptor proteins IcmS and IcmW. This subcomplex is likely to be involved in the recognition of substrates by the secretion apparatus.

DotD/DotC/DotF/DotG/DotH core-transmembrane subcomplex

The core subcomplex of the Dot/Icm secretion apparatus is composed of the inner membrane proteins DotF and DotG and the outer membrane proteins DotC, DotD, and DotH (Fig. 9.1) (Vincent et al., 2006b). One of the inner membrane components, DotF, was previously shown via a bacterial two-hybrid screen to interact with several Dot/Icm substrates in the cytoplasm (Luo and Isberg, 2004). Our findings show that DotF is required for the interactions between the inner membrane and outer membrane components of the core subcomplex. Together, these finding suggest that DotF plays an important role in the secretion apparatus. It is somewhat surprising, then, that a ΔdotF mutant is only partially attenuated for intracellular growth, demonstrating that DotF is not a strictly required component of the Dot/Icm T4SS (Purcell and Shuman, 1998). In light of this, it appears likely that DotF functions with DotL (discussed below) either as a co-receptor for secreted substrates or as the protein to which DotL 'hands off' substrates in the process of translocation.

The other inner membrane component of the core subcomplex, DotG, exhibits limited homology over its carboxy-terminus to the A. tumefaciens protein VirB10. VirB10, a component of the A. tumefaciens core subcomplex, has been shown to function as an energy transducer that adopts different conformations in response to ATP hydrolysis by the VirB11 and VirD4 proteins (Cascales and Christie, 2004). In the presence of ATP, VirB10 interacts with the outer membrane component VirB9, but this interaction is not detected when cellular ATP levels are depleted. According to the current model, VirB10 responds to the energetic status of the complex by forming or disassembling a bridge between the inner and outer membrane components of the secretion system. It is likely that DotG performs the same role in the Dot/Icm T4SS. Supporting this notion, we have shown that DotG is an inner membrane protein with an identical topology to VirB10. In addition, we have shown that DotG interacts with the outer membrane components DotC, DotD, and DotH, similar to the interactions between VirB10 and the outer membrane proteins VirB7 and VirB9. Whether DotG adopts different conformations depending on ATP utilization by DotB and DotL remains to be experimentally determined.

The outer membrane components of the core subcomplex are DotC, DotD, and DotH. Two of these proteins, DotC and DotD, contain lipobox motifs and thus are likely to be covalently attached to a lipid that anchors them to the outer membrane (Yerushalmi et al., 2005). DotC and DotD are required for insertion of DotH in the outer membrane, and in the absence of either of these lipoproteins, DotH accumulates in the periplasm. The dependence on a lipoprotein for the outer membrane insertion of DotH is reminiscent of secretins, which also require outer membrane lipoproteins called pilot proteins for insertion and oligomerization in the outer membrane (Bayan et al., 2006). Secretins are found in a variety of secretion systems, including type II and type III secretion systems and type IV pilus biogenesis systems, although they have not been described in type IV secretion systems. In each of these diverse systems, secretins are though to perform a similar function as the outer membrane channel through which substrates are secreted. Although most secretins require a single pilot protein, the Shigella T3SS secretin MxiD is known to require two pilot proteins for outer membrane insertion (Schuch and Maurelli, 2001). Even though DotH does not exhibit detectable homology to the secretin family, we predict that is performs a similar role as an oligomeric channel for the secretion of substrates by the Dot/Icm T4SS. read more ...

DotL/DotM/DotN/IcmS/IcmW T4CP subcomplex

A second putative Dot/Icm subcomplex contains the inner membrane proteins DotL, DotM, and DotN and the cytoplasmic proteins IcmS and IcmW (C.D. Vincent and J.P. Vogel, unpublished). This subcomplex is of particular interest because of the key role proposed for DotL a member of the type IV coupling protein family (Buscher et al., 2005). T4CPs function as inner membrane receptors for secreted substrates, linking the cytoplasmic substrates to the secretion machinery, and provide energy to the complex through ATP hydrolysis.

In support of DotL's predicted role in providing energy to the complex through ATP hydrolysis, we have observed that the DotL Walker box motif is essential for its function. In addition, expression of a DotL Walker box mutant inhibits the intracellular growth of L. pneumophila (C.D. Vincent and J.P. Vogel, unpublished) consistent with previous observations that other T4CPs form hexameric rings and display a dominant negative phenotype when their Walker box motif is mutated. In support of DotL's role as the inner membrane receptor for secreted substrates, DotL has been shown to interact with several components of the Dot/Icm secretion apparatus (C.D. Vincent and J.P. Vogel, unpublished). DotL was previously proposed to interact with DotM and DotN based on a shared phenotype of the dotL, dotM, and dotN mutants; all are essential genes but the lethality caused by their deletion can be suppressed by mutations in other dot/icm genes (Buscher et al., 2005).

Although direct interactions between DotL and secreted substrates have not been reported, DotL has been found to interact with the secretion adaptor proteins IcmS and IcmW (C.D. Vincent and J.P. Vogel, unpublished). IcmS and IcmW are small, acidic, cytoplasmic proteins that are only partially required for intracellular growth (Segal and Shuman, 1997; Zuckman et al., 1999; Coers et al., 2000). Early evidence that IcmS and IcmW represented a unique class of Dot/Icm proteins included the observation that, unlike the majority of dot/icm mutants, icmS and icmW mutants still exhibit wild-type levels of contact-dependent cytotoxicity (Zuckman et al., 1999; Coers et al., 2000). In addition, icmS and icmW mutations do not suppress ΔdotL lethality (Buscher et al., 2005). Because contact-dependent cytotoxicity and ΔdotL lethality require the presence of a functional secretion apparatus, these findings indicate that the Dot/Icm T4SS is assembled and functional in the icmS and icmW mutants. In spite of this, the mutants are severely attenuated in intracellular growth, suggesting that IcmS and IcmW may play a critical role in substrate secretion.

It was proposed that the role of IcmS and IcmW might be similar to type III secretion chaperones, proteins that bind to secreted substrates and facilitate their secretion (Page and Parsot, 2002; Parsot et al., 2003). Consistent with this proposal, several reports have demonstrated interactions of IcmS and IcmW with substrates and a requirement for substrate export (Coers et al., 2000; Bardill et al., 2005; Ninio et al., 2005). IcmS was shown to interact with the SidE family of secreted substrates, IcmW was found to interact with the substrates WipB and SidH, and both IcmS and IcmW were shown to bind to the substrates WipA and SidG (Bardill et al., 2005; Ninio et al., 2005). Notably, although IcmS and IcmW are required for the stable expression of each other, neither IcmS nor IcmW is required for the stable expression of their cognate secreted substrates. Therefore it is thought that the primary role of these proteins is to mediate interactions between substrates and the secretion apparatus. For these reasons, they have been termed 'adaptor proteins.'

Although several substrates require the adaptor proteins for secretion, at least one, RalF, does not require either IcmS or IcmW (Bardill et al., 2005; Ninio et al., 2005). This indicates that either some substrates are secreted by an adaptor-independent mechanism or that additional Dot/Icm adaptor proteins remain undiscovered. In fact, the LvgA protein appears to function as a third adaptor protein for the Dot/Icm T4SS (Vincent and Vogel, 2006). LvgA was originally identified as a virulence factor in an animal model of L. pneumophila infection, but its specific function was not determined (Edelstein et al., 1999; Edelstein et al., 2003). Subsequently, LvgA was shown to interact directly with IcmS and an ΔlvgA mutant was shown to phenotypically resemble a ΔicmS and ΔicmW mutant. In addition, LvgA was demonstrated to not be a component of the secretion apparatus nor be secreted by the Dot/Icm T4SS. As LvgA is more strictly required for growth in mouse bone marrow-derived macrophages than in human macrophages or in amoebae, LvgA appears to confer host specificity to L. pneumophila. Although a specific substrate(s) that requires LvgA for export has not yet been revealed, lvgA was identified in a screen for L. pneumophila mutants that were defective in the recruitment of Rab1 to the LCV (Murata et al., 2006). The same report identified two secreted substrates, SidM and LidA, which are also required for recruitment of Rab1 to the LCV. It is tempting to speculate that the Rab1 recruitment defect of the lvgA mutant is due to a failure to secrete SidM and LidA, and that LvgA functions as an adaptor protein for one or both of these substrates. read more ...

Assembly factors for the Dot/Icm secretion apparatus

Considering that the Dot/Icm secretion apparatus is an extremely complicated multi-protein structure, it is not surprising that its assembly is not well understood. For example, the Dot/Icm complex likely requires factors to assist in the proper assembly of the secretion apparatus. In fact, two L. pneumophila proteins that are not components of the secretion apparatus, DjlA and LdsA, were recently shown to be required for the assembly of the Dot/Icm complex (Vincent et al., 2006a). Although the L. pneumophila djlA and ldsA mutants still synthesize the Dot/Icm proteins, the mutants are unable to grow in macrophages and are defective for the phenotypes associated with an assembled, functional Dot/Icm secretion system (Vincent et al., 2006a).

In E. coli, DjlA is an inner membrane protein that has homology to the heat shock protein DnaJ (Clarke et al., 1996), is involved in the sensing of envelope stress and has been proposed to function as a chaperone for membrane proteins (Kelley and Georgopoulos, 1997). Like DnaJ, DjlA interacts with the chaperone DnaK (HSP70) and stimulates its ATPase activity (Wall et al., 1994). DjlA was shown to be a virulence factor in the strain L. dumoffii, although its role in virulence is not understood (Ohnishi et al., 2004). In L. pneumophila, DjlA is required for the assembly or activation of the Dot/Icm complex (Vincent et al., 2006a). Based on the involvement of DjlA in assembly of the Dot/Icm complex, it is likely that the chaperone DnaK is also involved in folding and assembly of the Dot/Icm proteins.

In contrast to DjlA, the LdsA protein does not contain any clear functional motifs, thus obfuscating predictions of its function. LdsA appears to be a polytopic inner membrane protein, and although it is possible that this protein is part of the Dot/Icm secretion apparatus, it is unlikely because (1) it is not homologous to any known type IV secretion system proteins and (2) the ldsA gene is encoded distantly from the dot/icm pathogenicity islands. Instead LdsA may function in an accessory role perhaps as a specific chaperone for the assembly and/or activation of the Dot/Icm apparatus. Further characterization of the roles of DjlA and LdsA will likely provide insight into how the Dot/Icm system is assembled and activated. read more ...

Conclusions

A significant amount of progress has been made in understanding the Legionella Dot/Icm type IVB secretion system since the initial identification of DotA and IcmV/IcmW/IcmX in 1993 (Berger and Isberg, 1993; Sadosky et al., 1993). The recent identification of two critical subcomplexes of the Dot/Icm complex exemplifies the progress in understanding how this secretion apparatus assembles and functions (Vincent et al., 2006b). The first subcomplex is composed of five proteins (DotC, DotD, DotF, DotG, and DotH), functions as a molecular bridge between the inner and outer membranes, and has been designated as the 'core-transmembrane' subcomplex. In this subcomplex, DotH localization to the outer membrane requires DotC and DotD and DotH likely functions similarly to a secretin, forming a channel in the outer membrane. The inner membrane proteins DotF and DotG interact with each other and with the DotC/DotD/DotH proteins in the outer membrane, likely transducing energy from ATP hydrolysis to the outer membrane. The existence of a Legionella core-transmembrane complex demonstrates a functionally conserved core assembly in the evolutionarily diverse type IVA and type IVB secretion systems. The second Dot/Icm subcomplex contains DotL, DotM, DotN, IcmS, and IcmW. This subcomplex likely functions as the membrane receptor for the Dot/Icm substrates as DotL is related to type IV coupling proteins and IcmS and IcmW have been shown to interact with a number of the Dot/Icm substrates. In addition to these two subcomplexes, other subcomplexes likely consist of the remaining sixteen Dot/Icm proteins.

Finally, analysis of the Legionella Dot/Icm apparatus revealed a surprising difference between type IVA and type IVB secretion systems. Of the 26 Legionella Dot/Icm proteins, five are in the cytoplasm, one in the periplasm, and four in the outer membrane but sixteen are found associated with the inner membrane (Vincent et al., 2006b). In contrast, type IVA systems consist of one cytoplasmic protein, one periplasmic protein, three outer membrane proteins, two pilus proteins and only five inner membrane proteins (Schroder and Lanka, 2005). Thus, it appears that type IVB systems have undergone a proliferation of non-conserved inner membrane proteins. Why the type IVB systems would require so many inner membrane components remains unclear. However, these 'extra' proteins do not appear to be accessory factors, because with the exception of DotU and IcmF, all are strictly required for the function of the Dot/Icm secretion apparatus. Moreover, the majority of these proteins are conserved in other type IVB systems, including the C. burnetii Dot/Icm secretion system and the plasmid transfer apparatus of the self-transmissible plasmids R64 and ColIb-P9, further supporting the idea that these proteins are essential for the function of the type IVB systems. Elucidation of the functions of these components will likely reveal important differences between the type IVA and type IVB secretion systems.

from Legionella: Molecular Microbiology
See also: Bacterial Secreted Proteins: Secretory Mechanisms and Role in Pathogenesis
See also: Bacterial Polysaccharides: Current Innovations and Future Trends

Biology Blogs

Conferences

Journals

Open Access Biology Home | PCR | Microbiology Books | Molecular Biology Books | Molecular Biology Gateway | PCR Protocols | Real-Time PCR |