| dc.description.abstract | Shigella flexneri is a Gram negative bacterium that invades human intestinal epithelial cells and causes bacillary dysentery (shigellosis). An alarming feature of Shigella species is that as few as 10 organisms is a large enough dose to cause disease, and the transmission between humans, especially children, occurs easily by the fecal to oral route or by consumption of contaminated water. Moreover, the CDC has reported increased rates of antibiotic resistance among Shigella species and this has complicated treatment strategies. Thus, developing anti-infective strategies that target important virulence factors may be an attractive new treatment strategy. One target for such agents is the type III secretion apparatus (T3SA), so-call injectisome, of Shigella. Shigella uses its T3SA to contact and directly translocate effector proteins into eukaryotic host cells. These effector proteins then alter host cell functions so that Shigella is able to enter and survive within the cytoplasm of intestinal epithelial cells. Because of its central role in Shigella virulence, the T3SA is a reasonable target for developing new anti-infective therapeutics, however, to generate such agents it will be important to better understand the architecture and molecular functions of the various moving parts of the T3SA, which is still not completely understood. Recently, cryo-electron tomography (cryo-ET) has revealed the overall architecture of the in situ Shigella T3SA, which is now known to be composed of four distinct parts: 1) a cytoplasmic sorting platform (SP) located within the bacterial cytoplasm that participates in the selection of effector proteins and provides the energy needed for translocation; 2) a basal body that spans the bacterium’s two membranes and cell wall and upon which the entire T3SA is built; 3) an extracellular needle that is an elongated polymer that extends outward from the basal body; and 4) a needle tip complex that makes contact with the host cell and is needed for insertion of a translocon pore within the host membrane. A prominent feature of the SP are pod-like structures that are present with an evenly spaced six-fold symmetry. Each pod is connected to a centrally located ATPase called Spa47 via radial spokes consisting of a protein called MxiN. Directly above the hexameric Spa47 ATPase is MxiA, which forms the export gate for the effector protein secretion. The Spa47 and MxiA complexes are associated with a connecting Spa13 stalk. The pod structures associate with the cytoplasmic portion of the basal body via the cytoplasmic domain of MxiG (here called MxiGCyt). Along with MxiJ, MxiG forms the inner membrane ring (IR) of the basal body. The periplasmic domain of MxiG (MxiGPeri) has a densely packed 24-fold symmetry with the smaller MxiGCyt having a less densely packed but evenly spaced 24-fold symmetry. It is anticipated that the symmetry transition between MxiGCyt and the pods has an important role in the dynamic mechanisms responsible for secretion, however, to dissect this process, we need to generate a deeper understanding of the protein components of the pod structures and the structural features that guide their interactions. The positions of Spa33 and MxiN within the SP have been determined using cryo-ET, however, the protein MxiK is known to be part of the SP, but its location within this complex has not yet been determined. We hypothesize that MxiK occupies a density near the IR where it acts as an adaptor between MxiGCyt (basal body) and the Spa33 density of the SP pods. To localize the position of MxiK, we generated a library of bacteriophage T4 lysozyme (T4L) insertion mutants of MxiK for functional and structural analysis. Many of these insertion mutants continue to be active in restoring type III secretion to a Shigella mxiK null mutant and one of these (MxiK-T4L-C) was used for cryo-ET studies to observe where the additional electron density from T4L appeared within the SP. Identification of the T4L density helped us to confirm the position of MxiK as being within the pod component that forms the interface with MxiGCyt. Moreover, the MxiK electron density suggested it is either lobed or possibly has two domains. To confirm the position of MxiK within the SP (between Spa33 and MxiGCyt ), the SP interactome was determined using a bacterial adenylate cyclase two-hybrid (BACTH) system with the identified interactions confirmed, when possible, using Biolayer Interferometry (BLI). Both approaches indicated that MxiK’s binding partners are MxiGCyt and Spa33, with the latter being the dominant component of the pod structures. Spa33, in turn, interacts with MxiN. Here I will combine all of the available in vivo, in vitro, and in situ data to provide a definitive description of the architecture of the Shigella T3SA sorting platform. The pod structures that make up the Shigella T3SA sorting platform are now proposed to be composed of a single copy of MxiK and a heterotrimeric Spa33 complex with both MxiK and Spa33 being essential for the assembly and stability of the overall SP structure. Consistent with this, cryo-ET images of the T3SA formed by a mxiK null S. flexneri strain completely lacks the entire SP and all of the extracellular components of the T3SA. These data indicate that MxiK is key component of the pods and is required for pod assembly as it forms the IR-SP interface. Thus, we propose that disruption of MxiK interactions with MxiGCyt or Spa33 are potential targets for new T3SA-targeting anti-infective agents. In this investigation, we identify the molecular basis for interactions between Spa33 and MxiK, as well as between MxiK and MxiGCyt, using both insertional and site-directed mutagenesis. From the MxiK-T4L insertion and Ala substitutional mutant libraries, two regions of MxiK completely disrupted the type III secretion system (T3SS) activities and these disruptions were not due to protein misfolding based on circular dichroism (CD) spectroscopy data. Subsequent BACTH analysis for these mutations revealed that both regions of MxiK are essential for the MxiK interactions in the SP. In parallel experiments, we investigated the structural bases for MxiK binding to MxiGCyt. Based on the MxiGCyt structure, this protein is a β-rich forkhead-associated (FHA) domain. Canonical FHA domains recognize phosphorylated threonine residues to promote specific protein-protein interactions, however, a previous study and work presented here show that while MxiGCyt is involved in protein-protein interactions, this does not occur via phosphothreonine recognition. In fact, when we introduced a continuous string of Ala residues into the MxiGCyt 61-66 loop together with Ala residues in the 81-85 loop region, we found that MxiGCyt interactions with MxiK were eliminated and Shigella hemolytic activity was lost. As importantly, cryo-ET analysis showed that while this mutation allowed for T3SA basal body formation, it did not allow for SP assembly or formation of the extracellular needle. These data suggest that MxiG interacts with MxiK via loop regions on its MxiGCyt domain and the loss of this interaction gives rise to a mxiG null mutation phenotype. In this investigation, we demonstrate the position of all major SP components using cryo-ET with confirmatory biochemical analyses. Our next goal will be to determine high-resolution protein structures for these components such as MxiK and MxiN. Such structures are not available for most of the SP components. This information will be needed to rationally design new anti-infective agents. | |
