We have been using an autotransporter produced by E. coli O157:H7 called EspP as a model protein to study autotransporter biogenesis. In one major line of investigation we have been examining the mechanism by which the EspP passenger domain is translocated across the outer membrane (OM). It was originally proposed that the passenger domain is secreted through a channel formed by the covalently linked beta barrel domain (whence the name autotransporter), but results that we obtained from both biochemical and structural studies are inconsistent with this hypothesis. We found that the insertion of a small linker into the EspP passenger domain effectively creates a translocation intermediate by transiently stalling the translocation reaction (which is normally extremely rapid) near the site of the insertion. By using a site-specific photocrosslinking approach we found that residues adjacent to the stall point interact with BamA, a component of a heterooligomeric barrel assembly machinery complex (Bam complex) that catalyzes OM protein assembly by an unknown mechanism. These results support a model in which the Bam complex plays a major role in facilitating both the integration of the beta barrel domain into the OM and the translocation of the passenger domain across the OM. In a recent study, we found that the mutation of an unusual lipid-exposed lysine residue in the EspP beta barrel domain impairs a previously unidentified late folding step that follows both the membrane insertion of the beta barrel domain and the secretion of the passenger domain but that precedes proteolytic maturation. Our results demonstrate that beta barrel assembly can be completed at a post-insertion stage and raise the possibility that interactions with membrane lipids can promote folding in vivo. Furthermore, by showing that the passenger domain is secreted before the beta barrel domain is fully assembled, our results also provide evidence against the hypothesis that autotransporters are autonomous protein secretion systems. In our early experiments on EspP we obtained evidence that the beta barrel domain begins to fold in the periplasm before it interacts with the Bam complex. To obtain additional insight into the folding of bacterial OM proteins, we analyzed the biogenesis of UpaG, a trimeric autotransporter adhesin (TAA) produced by uropathogenic E. coli. TAAs are a poorly characterized class of bacterial virulence factors composed of three identical subunits that together form a single beta barrel domain and a coiled-coil passenger domain. Using a site-specific photocrosslinking approach to obtain spatial and temporal insights into UpaG assembly, we found that UpaG beta barrel segments fold into a trimeric structure in the periplasm that persists until the termination of passenger domain translocation. In addition to obtaining further evidence that at least some beta barrel proteins begin to fold before they interact with the Bam complex, we identified several discrete steps in the assembly of TAAs. In addition to analyzing the assembly of EspP in vivo, we have also reconstituted the assembly of the protein in vitro using purified components. We found that the Bam complex and a periplasmic chaperone (SurA) are both necessary and sufficient to promote the integration of the EspP beta barrel domain into proteoliposomes and the translocation of the passenger domain into the lumen of the vesicles. The reaction does not require ATP or other exogenous energy source. Previous studies that examine the spontaneous insertion of simple OM proteins into pure lipid vesicles under non-physiological conditions (e.g., high pH) have suggested that the efficiency of insertion is strongly influenced by lipid head groups. To determine the role of lipids in the assembly of OM proteins under more physiological conditions, we recently examined the assembly of EspP and another model OM protein, OmpA, into vesicles that contain both the Bam complex and a variety of different synthetic lipids. We found that both proteins folded efficiently regardless of the lipid composition. Interestingly, both proteins folded into membranes composed of a gel phase lipid that mimics the rigid bacterial OM. We also found that EspP, OmpA and another model protein (OmpG) folded at significantly different rates and that the alpha helix that is normally embedded inside the EspP beta barrel accelerates folding. Our results show that the Bam complex overcomes any effects that lipids exert on OM protein assembly and suggest that specific interactions between the Bam complex and an OM protein influence its rate of folding.
Showing the most recent 10 out of 20 publications