Knowledge of the structure and dynamics of cellular protein complexes is essential for understanding their functionally relevant interactions. In Gram-negative bacteria, the complex machinery associated with the type II secretion system (T2SS) polymerises inner membrane pseudopilin proteins into thin filaments, to export substrates such as toxins, hydrolases and cytochromes. Here, computational simulations were used to study proteins from the Klebsiella oxytoca T2SS, focusing on the substrate pullulanase PulA, the major pseudopilin PulG, and the putative chaperone PulM. Chapter 3 contains an in silico study of both post-translationally acylated PulA (lipoPulA) and non-acylated PulA (PulANA) in association with a lipid bilayer, representing an approximation of the biological state prior to secretion; this study examined PulA dynamics and the possible role of the acyl tail in protein-membrane interactions before secretion. Novel insights into the interactions of a key residue necessary for Type 2 secretion were gained via simulations performed on a PulANA D2S variant, extending prior in vitro results. In Chapter 4, PulA was simulated in conditions closer to the physiological environment, using counter-ions to investigate the possible effect of the high periplasmic calcium concentration on protein conformation and lipid interactions prior to secretion. In Chapter 5, variants of the major pseudopilin PulG containing one transmembrane helix were simulated, demonstrating N-terminal interactions made possible by wild-type methylation of residue Phe1. Simulations of several monomeric PulG variants provided insight into the roles of the essential residue Glu5 and Phe1 methylation, previously identified by experimental work to be important. Simulations of the PulG dimer demonstrated the dynamic nature of the membrane-embedded dimer interface, and showed how computational analysis can predict in vivo contacts. Finally, Chapter 6 extended the T2SS studies to coarse-grained methods, sampling possible conformations and predicting the PulG-PulM interface within the membrane, prior to PulG presentation to the remaining secretion apparatus.

The Type Three Secretion Systems (T3SSs) of Gram-negative bacteria, including Shigella, Salmonella, and Enteropathogenic/Enterohaemorrhagic Escherichia coli (EPEC/EHEC), pass virulence factors directly into the host to mediate invasion. Prior to secretion down the narrow T3SS channel, effector proteins associate with chaperone proteins. The binding enables the T3SS to keep effectors soluble and partially unfolded for secretion. In the first part of this thesis, the association of one promiscuous chaperone, Spa15 of Shigella flexneri, with three of its cognate effectors has been studied. In addition to the role this plays in secretion, the binding of one particular substrate leads to Spa15 being involved in the regulation of the T3SS. The oligomerisation and impact of substrate binding upon Spa15 has been determined by crystallography and EPR. Once secreted, T3SS effectors subvert the host cytoskeleton for the benefit of the bacteria. Soluble homologues of Spa15 effectors from EHEC and Salmonella have been purified, and their interactions with host GTPases which lead to stress fibre phenotypes observed. The Twin Arginine Translocation (Tat) pathway provides a contrasting view of bacterial secretion. Instead of preventing folding in the cytoplasm, it is a criterion of transport that the protein be folded. One of the reasons for internal folding is the necessity to insert cofactors which could not be incorporated externally. In the second part of this thesis, a protein which exemplifies this necessity is studied. This is PhoD, the model protein for Tat export from Bacillus subtilis. PhoD is an alkaline phosphodiesterase expressed to scavenge phosphate in times of phosphate deficiency. The structure of PhoD has been solved, and the protein is shown to be able to cleave a component of its own cell wall. It uses an unusual catalytic site more reminiscent of the eukaryotic purple acid phosphatases than of other currently known alkaline phosphatases. Furthermore this site appears to require metal binding before export from the bacterial cytoplasm.

Das Typ III Sekretionssystem (T3SS) ist ein Proteinkomplex den Gramnegative Bakterien nutzen um in einem Schritt Effektorproteine (Effektoren) aus dem Zytosol über die Doppelmembran zu sekretieren. Für viele Bakterien ist das T3SS ein essenzieller Virulenzfaktor, der es ihnen erlaubt mit ihrem Wirt zu interagieren und diesen zu manipulieren. Charakteristisch für das T3SS ist die strukturelle Komponente, der Nadelkomplex. Dieser ähnelt strukturell einer Spritze, deren Basalkörper die bakteriellen Membranen und das Periplasma durchspannt und einer Nadel, die vom Basalkörper aus dem Bakterium ragt. Basierend auf dem Modell einer Spritze wird angenommen, dass Effektoren entfaltet und anschließend durch Basalkörper und Nadelkanal sekretiert werden. Trotz der kontinuierlichen Forschung an T3SS entbehrt dieses Modell einer experimentellen Grundlage und der Mechanismus ist nicht vollständig erklärt. Ziel der Arbeit war es, eine experimentelle Basis für den Sekretionsmechanismus des T3SS zu schaffen. Um zu verstehen, wie das T3SS Effektoren sekretiert, wurden zunächst Fusionsproteine konstruiert, welche aus einem Effektor und einem stabil gefalteten Knotenprotein bestehen. Aufgrund des Knotens in der Fusion ist davon auszugehen, dass dieser während der Sekretion nicht entfalten kann. Die Effektordomäne wird sekretiert während der Knoten im Kanal verbleibt und diesen verstopft. Nach unseremWissen ist diese Arbeit die erste Visualisierung von Effektorfusionen an isolierten Nadelkomplexen. Die Effektorfusion wird N-terminal voran durch den Kanal sekretiert, wobei der Kanal das Substrat umschließt und gegen Proteasen und chemische Modifikationen abschirmt. Die Ergebnisse dieser Arbeit untermauern eine Grundidee der Funktionsweise des T3SS und liefern eine vielversprechende Strategie für in situ-Strukturanalysen. Dieser Ansatz lässt sich auch auf andere Proteinsekretionssysteme übertragen, bei welchen Substrate vor dem Transport entfaltet werden müssen. / The Type III Secretion System (T3SS) is a complex used by Gram-negative bacteria to secrete effector proteins from the cytoplasm across the bacterial envelope in a single step. For many pathogens, the T3SS is an essential virulence factor that enables the bacteria to interact with and manipulate their respective host. A characteristic structural feature of the T3SS is the needle complex (NC). The NC resembles a syringe with a basal body spanning both bacterial membranes and a long needle-like structure that protrudes from the bacterium. Based on the paradigm of a syringe-like mechanism, it is generally assumed that effectors are unfolded and secreted from the bacterial cytoplasm through the basal body and needle channel. Despite extensive research on T3SS, this hypothesis lacks experimental evidence and the mechanism of secretion is not fully understood. This work aimed to provide an experimental basis for the model of the T3SS mechanism. In order to elucidate details of the effector secretion mechanism, fusion proteins consisting of an effector and a bulky protein containing a knotted motif were generated. It is assumed that the knot cannot be unfolded during secretion of the chimera. Consequently, these fusions are accepted as T3SS substrates but remain inside the NC channel and obstruct the T3SS. This is, to our best knowledge, the first time effector fusions have been visualized together with isolated NCs and it demonstrates that effector proteins are secreted directly through the channel with their N-terminus first. The channel encloses the substrate and shields it from a protease and chemical modifications. These results corroborate an elementary understanding of how the T3SS works and provide a powerful tool for in situ-structural investigations. This approach might also be applicable to other protein secretion systems that require unfolding of their substrates prior to secretion.