The FliN-FliH interaction mediates localization of flagellar export ATPase FliI to the C ring complex

FliH regulates the flagellar export ATPase FliI, preventing nonproductive ATP hydrolysis. FliH has been shown to stably associate with the C ring protein FliN. Analysis of this complex reveals that FliH is required for FliI localization to the C ring, and thus FliH not only inhibits FliI ATPase activity but also may act to target FliI to the basal body. Quantitative binding studies revealed a KD of 110 nM for FliH binding to FliN. The KD for FliH binding of a FliN variant from a temperature-sensitive nonflagellate fliN point mutant was determined to be 270 nM, suggesting a molecular explanation for its phenotype. Another variant FliN from a temperature-sensitive mutant with a different phenotype displayed binding with an intermediate affinity. Weak export activity in a fliN null mutant was greatly increased by overproduction of FliI, mimicking a previously observed FliH bypass effect and supporting the conclusion that FliN-FliH binding is important for localization of FliI to the C ring and thus the membrane-embedded export apparatus beyond. A model incorporating the present findings is presented.     

Protein export through the bacterial flagellar type III export pathway

For construction of the bacterial flagellum, which is responsible for bacterial motility, the flagellar type III export apparatus utilizes both ATP and proton motive force across the cytoplasmic membrane and exports flagellar proteins from the cytoplasm to the distal end of the nascent structure. The export apparatus consists of a membrane-embedded export gate made of FlhA, FlhB, FliO, FliP, FliQ, and FliR and a water-soluble ATPase ring complex consisting of FliH, FliI, and FliJ. FlgN, FliS, and FliT act as substrate-specific chaperones that do not only protect their cognate substrates from degradation and aggregation in the cytoplasm but also efficiently transfer the substrates to the export apparatus. The ATPase ring complex facilitates the initial entry of the substrates into the narrow pore of the export gate. The export gate by itself is a proton-protein antiporter that uses the two components of proton motive force, the electric potential difference and the proton concentration difference, for different steps of the export process. A specific interaction of FlhA with FliJ located in the center of the ATPase ring complex allows the export gate to efficiently use proton motive force to drive protein export. The ATPase ring complex couples ATP binding and hydrolysis to its assembly–disassembly cycle for rapid and efficient protein export cycle. This article is part of a Special Issue entitled: Protein trafficking and secretion in bacteria. Guest Editors: Anastassios Economou and Ross Dalbey.     

Oligomerization and activation of the FliI ATPase central to bacterial flagellum assembly

FliI is the peripheral membrane ATPase pivotal to the type III protein export mechanism underlying the assembly of the bacterial flagellum. Gel filtration and multiangle light scattering showed that purified soluble native FliI protein was in a monomeric state but, in the presence of ATP, FliI showed a propensity to oligomerize. Electron microscopy revealed that FliI assembles to a ring structure, the yield of which was increased by the presence of a non-hydrolysable ATP analogue. Single particle analysis of the resulting electron micrograph images, to which no symmetry was applied, showed that the FliI ring structure has sixfold symmetry and an external diameter of approximately 10 nm. The oligomeric ring has a central cavity of 2.5-3.0 nm, which is comparable to the known diameter of the flagellar export channel into which export substrates feed. Enzymatic activity of the FliI ATPase showed positive co-operativity, establishing that oligomerization and enzyme activity are coupled. Escherichia coli phospholipids increased enzyme co-operativity, and in vitro cross-linking demonstrated that they promoted FliI multimerization. The data reveal central facets of the structure and action of the flagellar assembly ATPase and, by extension, the homologous ATPases of virulence-related type III export systems.     

Interaction between FliJ and FlhA,Components of the Bacterial Flagellar Type III Export Apparatus

A soluble protein, FliJ, along with a membrane protein, FlhA, plays a role in the energy coupling mechanism for bacterial flagellar protein export. The water-soluble FliH_X-FliI₆ ATPase ring complex allows FliJ to efficiently interact with FlhA. However, the FlhA binding site of FliJ remains unknown. Here, we carried out genetic analysis of a region formed by well-conserved residues—Gln38, Leu42, Tyr45, Tyr49, Phe72, Leu76, Ala79, and His83—of FliJ. A structural model of the FliI₆-FliJ ring complex suggests that they extend out of the FliI₆ ring. Glutathione S-transferase (GST)-FliJ inhibited the motility of and flagellar protein export by both wild-type cells and a fliH-fliI flhB(P28T) bypass mutant. Pulldown assays revealed that the reduced export activity of the export apparatus results from the binding of GST-FliJ to FlhA. The F72A and L76A mutations of FliJ significantly reduced the binding affinity of FliJ for FlhA, thereby suppressing the inhibitory effect of GST-FliJ on the protein export. The F72A and L76A mutations were tolerated in the presence of FliH and FliI but considerably reduced motility in their absence. These two mutations affected neither the interaction with FliI nor the FliI ATPase activity. These results suggest that FliJ(F72A) and FliJ(L76A) require the support of FliH and FliI to exert their export function. Therefore, we propose that the well-conserved surface of FliJ is involved in the interaction with FlhA.     

Intrinsic membrane targeting of the flagellar export ATPase FliI: interaction with acidic phospholipids and FliH

The specialised ATPase FliI is central to export of flagellar axial protein subunits during flagellum assembly. We establish the normal cellular location of FliI and its regulatory accessory protein FliH in motile Salmonella typhimurium, and ascertain the regions involved in FliH(2)/FliI heterotrimerisation. Both FliI and FliH localised to the cytoplasmic membrane in the presence and in the absence of proteins making up the flagellar export machinery and basal body. Membrane association was tight, and FliI and FliH interacted with Escherichia coli phospholipids in vitro, both separately and as the preformed FliH(2)/FliI complex, in the presence or in the absence of ATP. Yeast two-hybrid analysis and pull-down assays revealed that the C-terminal half of FliH (H105-235) directs FliH homodimerisation, and interacts with the N-terminal region of FliI (I1-155), which in turn has an intra-molecular interaction with the remainder of the protein (I156-456) containing the ATPase domain. The FliH105-235 interaction with FliI was sufficient to exert the FliH-mediated down-regulation of ATPase activity. The basal ATPase activity of isolated FliI was stimulated tenfold by bacterial (acidic) phospholipids, such that activity was 100-fold higher than when bound by FliH in the absence of phospholipids. The results indicate similarities between FliI and the well-characterised SecA ATPase that energises general protein secretion. They suggest that FliI and FliH are intrinsically targeted to the inner membrane before contacting the flagellar secretion machinery, with both FliH105-235 and membrane phospholipids interacting with FliI to couple ATP hydrolysis to flagellum assembly.     

Interaction between FliI ATPase and a flagellar chaperone FliT during bacterial flagellar protein export

FliT is a flagellar type III export chaperone specific for the filament-capping protein FliD. The FliT/FliD complex binds to the FliI ATPase of the flagellar export apparatus. The C-terminal α4 helix of FliT controls its interaction with FliI but it remains unknown how it does so. Here, we analysed the FliI-FliT interaction by pull-down assays using GST affinity chromatography. FliT94, missing the C-terminal α4 helix, bound to the extreme N-terminal region of FliI (FliI(EN)) with high affinity and to the C-terminal ATPase domain (FliI(CAT)) with low affinity. The C-terminal α4 helix of FliT suppressed the interaction with FliI(EN). FliH and FliT94 bound to a common binding site on FliI(EN) and hence FliH induced the release of FliI from FliT94 in an ATP-independent manner. FliD increased the binding affinity of FliI(CAT) for FliT. These results raise a possible hypothesis that the FliH/FliI complex binds to the FliT/FliD complex through FliI(CAT) to escort it from the cytoplasm to the export gate made up of six integral membrane proteins and that, upon dissociation of FliD from FliT, FliT94 may bind to FliI(EN) and then FliI may transfer from FliT94 to FliH by the direct competition of FliT94 and FliH for FliI(EN).     

Interaction of the Extreme N-Terminal Region of FliH with FlhA Is Required for Efficient Bacterial Flagellar Protein Export

The flagellar type III protein export apparatus plays an essential role in the formation of the bacterial flagellum. FliH forms a complex along with FliI ATPase and is postulated to provide a link between FliI ring formation and flagellar protein export. Two tryptophan residues of FliH, Trp7 and Trp10, are required for the effective docking of the FliH-FliI complex to the export gate made of six membrane proteins. However, it remains unknown which export gate component interacts with these two tryptophan residues. Here, we performed targeted photo-cross-linking of the extreme N-terminal region of FliH (FliH(EN)) with its binding partners. We replaced Trp7 and Trp10 of FliH with p-benzoyl-phenylalanine (pBPA), a photo-cross-linkable unnatural amino acid, to produce FliH(W7pBPA) and FliH(W10pBPA). They were both functional and were photo-cross-linked with one of the export gate proteins, FlhA, but not with the other gate proteins, indicating that these two tryptophan residues are in close proximity to FlhA. Mutant FlhA proteins that are functional in the presence of FliH and FliI but not in their absence showed a significantly reduced function also by N-terminal FliH mutations even in the presence of FliI. We suggest that the interaction of FliH(EN) with FlhA is required for anchoring the FliI hexamer ring to the export gate for efficient flagellar protein export.     

Roles of the extreme N‐terminal region of FliH for efficient localization of the FliH–FliI complex to the bacterial flagellar type III export apparatus

Most bacterial flagellar proteins are exported by the flagellar type III protein export apparatus for their self‐assembly. FliI ATPase forms a complex with its regulator FliH and facilitates initial entry of export substrates to the export gate composed of six integral membrane proteins. The FliH–FliI complex also binds to the C ring of the basal body through a FliH–FliN interaction for efficient export. However, it remains unclear how these reactions proceed within the cell. Here, we analysed subcellular localization of FliI–YFP by fluorescence microscopy. FliI–YFP was localized to the flagellar base, and its localization required both FliH and the C ring. The ATPase activity of FliI was not required for its localization. FliI–YFP formed a complex with FliHΔ1 (missing residues 2–10) but the complex did not show any localization. FliHΔ1 did not interact with FliN, and alanine‐scanning mutagenesis revealed that only Trp‐7 and Trp‐10 of FliH are essential for the interaction with FliN. Overproduction of the FliH–FliI complex improved the export activity of the fliN mutant whereas neither of the FliH(W7A)‐FliI nor FliH(W10A)‐FliI complexes did, suggesting that Trp‐7 and Trp‐10 of FliH are also required for efficient localization of the FliH–FliI complex to the export gate.     

Interaction of FliI, a component of the flagellar export apparatus, with flagellin and hook protein.

FliI is a key component of the flagellar export apparatus in Salmonella typhimurium. It catalyzes the hydrolysis of ATP which is necessary for flagellar assembly. Affinity blotting experiments showed that purified flagellin and hook protein, two flagellar axial proteins, interact specifically with FliI. The interaction of either of the two proteins with FliI, increases the intrinsic ATPase activity. The presence of either flagellin or hook protein stimulates ATPase activity in a specific and reversible manner. A Vmax of 0.12 nmol Pi min-1 microgram-1 and a Km for MgATP of 0.35 mM was determined for the unstimulated FliI; the presence of flagellin increased the Vmax to 0.35 nmol Pi min-1 microgram-1 and the Km for MgATP to 1.1 mM. The stimulation induced by the axial proteins was fully reversible suggesting a direct link between the catalytic activity of FliI and the export process.     

Oligomerization of the bacterial flagellar ATPase FliI is controlled by its extreme N-terminal region

Salmonella FliI is the flagellar ATPase which converts the energy of ATP hydrolysis into the export of flagellar proteins. It forms a ring-shaped oligomer in the presence of ATP, its analogs, or phospholipids. The extreme N-terminal region of FliI has an unstable conformation and is responsible for the interaction with other components of the export apparatus and for regulation of the catalytic mechanism. To understand the role of this N-terminal region in more detail, we used multi-angle light-scattering, analytical ultracentrifugation, far-UV CD and biochemical methods to characterize a partially functional variant of FliI, missing its first seven amino acid residues (His-FliI(Delta1-7)), whose ATPase activity is about ten times lower than that of wild-type FliI. His-FliI(Delta1-7) is monomeric in solution. The deletion increased the content of alpha-helix, suggesting that the deletion stabilizes the unstable N-terminal region into an alpha-helical conformation. The deletion did not influence the K(m) value for ATP. However, unlike the wild-type, ATP and acidic phospholipids did not induce oligomerization of His-FliI(Delta1-7) or increase its ATPase activity. These results suggest that the deletion suppresses the oligomerization of FliI, and that a conformational change in the unstable N-terminal region is required for FliI oligomerization to effectively couple the energy of ATP hydrolysis to the translocation of flagellar proteins.     

Genetic and biochemical analysis of Salmonella typhimurium FliI, a flagellar protein related to the catalytic subunit of the F0F1 ATPase and to virulence proteins of mammalian and plant pathogens.

FliI is a Salmonella typhimurium protein that is needed for flagellar assembly and may be involved in a specialized protein export pathway that proceeds without signal peptide cleavage. FliI shows extensive sequence similarity to the catalytic beta subunit of the F0F1 ATPase (A. P. Volger, M. Homma, V. M. Irikura, and R. M. Macnab, J. Bacteriol. 173:3564-3572, 1991). It is even more similar to the Spa47 protein of Shigella flexneri (M. M. Venkatesan, J. M. Buysse, and E. V. Oaks, J. Bacteriol. 174:1990-2001, 1992) and the HrpB6 protein of Xanthomonas campestris (S. Fenselau, I. Balbo, and U. Bonas, Mol. Plant-Microbe Interact. 5:390-396, 1992), which are believed to play a role in the export of virulence proteins. Site-directed mutagenesis of residues in FliI that correspond to catalytically important residues in the F1 beta subunit resulted in loss of flagellation, supporting the hypothesis that FliI is an ATPase. FliI was overproduced and purified almost to homogeneity. It demonstrated ATP binding but not hydrolysis. An antibody raised against FliI permitted detection of the protein in wild-type cells and an estimate of about 1,500 subunits per cell. An antibody directed against the F1 beta subunit of Escherichia coli cross-reacted with FliI, confirming that the proteins are structurally related. The relationship between three proteins involved in flagellar assembly (FliI, FlhA, and FliP) and homologs in a variety of virulence systems is discussed.     

Role of the N-terminal domain of FliI ATPase in bacterial flagellar protein export

FliI, the ATPase involved in bacterial flagellar protein export, forms a complex with its regulator FliH in the cytoplasm and hexamerizes upon docking to the export gate composed of integral membrane proteins. The extreme N-terminal region of FliI is involved not only in its interaction with FliH but also in its oligomerization, but the regulatory mechanism of oligomerization remains unclear. Using in-frame 10-residue deletions within the 100 residues of the N-terminal domain, we demonstrate that the first 20 residues are required for FliH binding and that the conformation of the N-terminal domain is sensitive to the export function, even though the oligomerization and FliH-binding ability are retained and the ATPase activity is maintained in most of the deletion variants.     

Interactions between C ring proteins and export apparatus components: a possible mechanism for facilitating type III protein export

The flagellar switch proteins of Salmonella, FliG, FliM and FliN, participate in the switching of motor rotation, torque generation and flagellar assembly/export. FliN has been implicated in the flagellar export process. To address this possibility, we constructed 10-amino-acid scanning deletions and larger truncations over the C-terminal domain of FliN. Except for the last deletion variant, all other variants were unable to complement a fliN null strain or to restore the export of flagellar proteins. Most of the deletions showed strong negative dominance effects on wild-type cells. FliN was found to associate with FliH, a flagellar export component that regulates the ATPase activity of FliI. The binding of FliM to FliN does not interfere with this FliN-FliH interaction. Furthermore, a five-protein complex consisting of FliG, His-tagged FliM, FliN, FliH and FliI was purified by nickel-affinity chromatography. FliJ, a putative general chaperone, is bound to FliM even in the absence of FliH. The importance of the C ring as a possible docking site for export substrates, chaperones and FliI through FliH for their efficient delivery to membrane components of the export apparatus is discussed.     

The ATPase FliI can interact with the type III flagellar protein export apparatus in the absence of its regulator,FliH

Salmonella FliI is the ATPase that drives flagellar protein export. It normally exists as a complex together with the regulatory protein FliH. A fliH null mutant was slightly motile, with overproduction of FliI resulting in substantial improvement of its motility. Mutations in the cytoplasmic domains of FlhA and FlhB, which are integral membrane components of the type III flagellar export apparatus, also resulted in substantially improved motility, even at normal FliI levels. Thus, FliH, though undoubtedly important, is not essential.     

Molecular dissection of Salmonella FliH,a regulator of the ATPase FliI and the type III flagellar protein export pathway

FliH is a soluble component of the flagellar export apparatus that binds to the ATPase FliI, and negatively regulates its activity. The 235-amino-acid FliH dimerizes and interacts with FliI to form a hetero-trimeric (FliH)2FliI complex. In the present work, the importance of different regions of FliH was examined. A set of 24 scanning deletions of 10 amino acids was constructed over the entire FliH sequence, along with several combined deletions of 40 amino acids and truncations of both N- and C-termini. The mutant proteins were examined with respect to (i) complementation; (ii) dominance and multicopy effects; (iii) interaction with wild-type FliH; (iv) interaction with FliI; (v) inhibition of the ATPase activity of FliI; and (vi) interaction with the putative general chaperone FliJ. Analysis of the deletion mutants revealed a clear functional demarcation between the FliH N- and C-terminal regions. The 10-amino-acid deletions throughout most of the N-terminal half of the sequence complemented and were not dominant, whereas those throughout most of the C-terminal half did not complement and were dominant. FliI binding was disrupted by C-terminal deletions from residue 101 onwards, indicating that the C-terminal domain of FliH is essential for interaction with FliI. FliH dimerization was abolished by deletion of residues 101-140 in the centre of the sequence, as were complementation, dominance and interaction with FliI and FliJ. The importance of this region was confirmed by the fact that fragment FliHC2 (residues 99-235) interacted with FliH and FliI, whereas fragment FliHC1 (residues 119-235) did not. FliHC2 formed a relatively unstable complex with FliI and showed biphasic regulation of ATPase activity, suggesting that the FliH N-terminus stabilizes the (FliH)2FliI complex. Several of the N-terminal deletions tested permitted close to normal ATPase activity of FliI. Deletion of the last five residues of FliH caused a fivefold activation of ATPase activity, suggesting that this region of FliH governs a switch between repression and activation of FliI. Deletion of the first 10 residues of FliH abolished complementation, severely reduced its interaction with FliJ and uncoupled its role as a FliI repressor from its other export functions. Based on these data, a model is presented for the domain construction and function of FliH in complex with FliI and FliJ.     

EscO,a Functional and Structural Analog of the Flagellar FliJ Protein,Is a Positive Regulator of EscN ATPase Activity of the Enteropathogenic Escherichia coli Injectisome

Type III secretion systems (T3SSs) are multiprotein molecular devices used by many Gram-negative bacterial pathogens to translocate effector proteins into eukaryotic cells. A T3SS is also used for protein export in flagellar assembly, which promotes bacterial motility. The two systems are evolutionarily related, possessing highly conserved components in their export apparatuses. Enteropathogenic Escherichia coli (EPEC) employs a T3SS, encoded by genes in the locus of enterocyte effacement (LEE) pathogenicity island, to colonize the human intestine and cause diarrheal disease. In the present work, we investigated the role of the LEE-encoded EscO protein (previously Orf15 or EscA) in T3SS biogenesis. We show that EscO shares similar properties with the flagellar FliJ and the Yersinia YscO protein families. Our findings demonstrate that EscO is essential for secretion of all categories of T3SS substrates. Consistent with its central role in protein secretion, it was found to interact with the ATPase EscN and its negative regulator, EscL, of the export apparatus. Moreover, we show that EscO stimulates EscN enzymatic activity; however, it is unable to upregulate ATP hydrolysis in the presence of EscL. Remarkably, EscO partially restored the swimming defect of a Salmonella flagellar fliJ mutant and was able to stimulate the ATPase activity of FliI. Overall, our data indicate that EscO is the virulence counterpart of the flagellar FliJ protein.     

Interaction of a bacterial flagellar chaperone FlgN with FlhA is required for efficient export of its cognate substrates

FlgN chaperone acts as a bodyguard to protect its cognate substrates, FlgK and FlgL, from proteolysis in the cytoplasm. Docking of the FlgN-FlgK complex with the FliI ATPase of the flagellar type III export apparatus is key to the protein export process. However, a ΔfliH-fliI flhB(P28T) mutant forms some flagella even in the absence of FliH and FliI, raising the question of how FlgN promotes the export of its cognate substrates. Here, we report that the interaction of FlgN with an integral membrane export protein, FlhA, is directly involved in efficient protein export. A ΔfliH-fliI flhB(P28T) ΔflgN mutant caused extragenic suppressor mutations in the C-terminal domain of FlhA (FlhA(C) ). Pull-down assays using GST affinity chromatography showed an interaction between FlgN and FlhA(C) . The FlgN-FlgK complex bound to FlhA(C) and FliJ to form the FlgN-FlgK-FliJ-FlhA(C) complex. The FlgN-FlhA(C) interaction was enhanced by FlgK but not by FliJ. FlgN120 missing the last 20 residues still bound to FlgK and FliJ but not to FlhA(C) . A highly conserved Tyr-122 residue was required for the interaction with FlhA(C) . These results suggest that FlgN efficiently transfers FlgK/L subunits to FlhA(C) to promote their export.     

Interactions among components of the Salmonella flagellar export apparatus and its substrates

We have examined the cytoplasmic components (FliH, FliI and FliJ) of the type III flagellar protein export apparatus, plus the cytoplasmic domains (FlhAC and FlhBC) of two of its six membrane components. FliH, FlhAC and FliJ, when overproduced, caused inhibition of motility of wild-type cells and inhibition of the export of substrates such as the hook protein FlgE. Co-overproduction of FliH and FliI substantially relieved the inhibition caused by FliH, suggesting that it is excess free FliH that is inhibitory and that FliH and FliI form a complex. We purified His-FLAG-tagged versions of: (i) export components FliH, FliI, FliJ, FlhAC and FlhBC; (ii) rod/hook-type export substrates FlgB (rod protein), FlgE (hook protein), FlgD (hook capping protein) and FliE (basal body protein); and (iii) filament-type export substrates FlgK and FlgL (hook-filament junction proteins) and FliC (flagellin). We tested for protein-protein interactions by affinity blotting. In many cases, a given protein interacted with more than one other component, indicating that there are likely to be multiple dynamic interactions or interactions that involve more than two components. Interactions of FlhBC with rod/hook-type substrates were strong, whereas those with filament-type substrates were very weak; this may reflect the role of FlhB in substrate specificity switching. We propose a model for the flagellar export apparatus in which FlhA and FlhB and the other four integral membrane proteins of the apparatus form a complex at the base of the flagellar motor. A soluble complex of at least three proteins (FliH, FliI and FliJ) bind the protein to be exported and then interact with the complex at the motor to deliver the protein, which is then exported in an ATP-dependent process mediated by FliI.     

Self-assembly and type III protein export of the bacterial flagellum

The bacterial flagellum is a supramolecular structure consisting of a basal body, a hook and a filament. Most of the flagellar components are translocated across the cytoplasmic membrane by the flagellar type III protein export apparatus in the vicinity of the flagellar base, diffuse down the narrow channel through the nascent structure and self-assemble at its distal end with the help of a cap structure. Flagellar proteins synthesized in the cytoplasm are targeted to the export apparatus with the help of flagellum-specific chaperones and pushed into the channel by an ATPase, whose activity is controlled by its regulator to enable the energy of ATP hydrolysis to be efficiently coupled to the translocation reaction. The export apparatus switches its substrate specificity by monitoring the state of flagellar assembly in the cell exterior, allowing this huge and complex macromolecular assembly to be built efficiently by a highly ordered and well-regulated assembly process.     

Weak Interactions between Salmonella enterica FlhB and Other Flagellar Export Apparatus Proteins Govern Type III Secretion Dynamics

The bacterial flagellum contains its own type III secretion apparatus that coordinates protein export with assembly at the distal end. While many interactions among export apparatus proteins have been reported, few have been examined with respect to the differential affinities and dynamic relationships that must govern the mechanism of export. FlhB, an integral membrane protein, plays critical roles in both export and the substrate specificity switching that occurs upon hook completion. Reported herein is the quantitative characterization of interactions between the cytoplasmic domain of FlhB (FlhB_C) and other export apparatus proteins including FliK, FlhA_C and FliI. FliK and FlhA_C bound with micromolar affinity. K_D for FliI binding in the absence of ATP was 84 nM. ATP-induced oligomerization of FliI induced kinetic changes, stimulating fast-on, fast-off binding and lowering affinity. Full length FlhB purified under solubilizing, nondenaturing conditions formed a stable dimer via its transmembrane domain and stably bound FliH. Together, the present results support the previously hypothesized central role of FlhB and elucidate the dynamics of protein-protein interactions in type III secretion.