Integral β-barrel proteins are found in the outer membranes of Gram-negative bacteria, mitochondria, and chloroplasts. The assembly of these proteins requires a proteinaceous apparatus of which Omp85 is an evolutionary conserved central component. To study its molecular mechanism, we have produced Omp85 from Escherichia coli in inclusion bodies and refolded it in vitro. The interaction of Omp85 with its substrate proteins was studied in lipid-bilayer experiments, where it formed channels. The properties of these channels were affected upon addition of unfolded outer-membrane proteins (OMPs) or synthetic peptides corresponding to their C-terminal signature sequences. The interaction exhibited species specificity, explaining the inefficient assembly of OMPs from Neisseria in E. coli. Accordingly, the in vivo assembly of the neisserial porin PorA into the E. coli outer membrane was accomplished after adapting its signature sequence. These results demonstrate that the Omp85 assembly machinery recognizes OMPs by virtue of their C-terminal signature sequence.

SummaryAfter attaching to human intestinal epithelial cells, enteropathogenic Escherichia coli (EPEC) induces the formation of an actin-rich pedestal-like structure. The signalling pathway leading to pedestal formation is initiated by the bacterial protein Tir, which is inserted into the host cell plasma membrane. The domain exposed on the cell surface binds to another bacterial protein, intimin, while one of the cytoplasmic domains binds the adaptor protein Nck. This leads to recruitment of other cytoskeletal proteins including neural Wiskott-Aldrich syndrome protein (N-WASP) and Arp2/3, resulting in focused actin polymerization at the site of bacterial attachment. In this study we investigated the role of the large GTPase dynamin 2 (Dyn2) in pedestal formation. We found that in HeLa cells, both endogenous and overexpressed Dyn2 were recruited to sites of EPEC attachment. Recruitment of endogenous Dyn2 required the presence of Tir, Nck and N-WASP but was independent of cortactin and Arp2/3. Knockdown of Dyn2 expression by RNA interference reduced actin polymerization and pedestal formation. Overexpression of dominant-negative mutants of Dyn2 also reduced pedestal formation and prevented recruitment of N-WASP, Arp3 and cortactin, but not Nck. Together, our results indicate that Dyn2 is an integral component of the signalling cascade leading to actin polymerization in EPEC pedestals.

The subcellular localization of the major type II secretion system of Pseudomonas aeruginosa, the Xcp system, was studied microscopically using a biarsenical ligand that becomes fluorescent upon binding to a tetracysteine motif (Lumio tag), which was fused to several Xcp components. Fusion of the Lumio tag to the C termini of the XcpR and XcpS proteins did not affect the functionality of these proteins. Fluorescence microscopy showed that they were predominantly localized to the poles of P. aeruginosa cells, when produced at levels comparable to chromosomally encoded XcpR and XcpS. In most labelled cells, the proteins were found at one of the poles, although bipolar localization was also observed. When produced in the absence of other Xcp components, labelled XcpS was still found to locate at the poles, whereas XcpR was evenly distributed in the cell. These data suggest that XcpS, but not XcpR, contains information required for polar localization. The polar location of the Xcp machinery was further confirmed by the visualization of protease secretion with an intramolecularly quenched casein conjugate.