The outer membrane of bacteria is a complex and important structure representing the first (and most ’fortified’) line of defense against insults from the extracellular medium as well as a platform for adhesion, recognition, and nutrient acquisition. As such, the outer membrane is a prime target for rational design of new drugs - both in terms of designing new classes of bacteriocidal agents, and also to sensitize bacteria to improve the efficacy of existing antibiotics. Our current knowledge suggests that all essential routes to the assembly of the outer membrane itself (both the lipids and proteins that form its structure) pass through the β-barrel assembly machinery complex (BAM complex), either directly or indirectly. This nanoscale machine is conserved across all bacteria containing an outer membrane and although the exact constituent parts vary, there is a common architecture scaffolded around one completely conserved protein, BamA. The BAM complex is responsible for the ATP- and protonmotive-independent assembly of integral transmembrane β-barrel proteins commonly refered to as outer membrane proteins (OMPs). Despite its essential role, the availability of high resolution structures, and over 15 years of biochemical studies, many questions about its mechanism of action remain unanswered. In this thesis new methods for studying OMP biogenesis through the use of crosslinking mass spectrometry and cryogenic super-resolution microscopy have been developed and applied to study the interaction with a model OMP, OmpA, and its chaperones Skp and SurA, as well as mapping its interaction with the BAM complex during folding, giving new insights into the mechanism of SurA chaperoning and suggesting a possible mechanism and route for the transit of an OMP from SurA and through BAM. Cryogenic super-resolution microscopy is used to provide preliminary insights into the nanoscale organisation of the BAM complex and OmpA, as well as two-colour co-localisation of these proteins. Kinetic assays are used alongside fluorescent probes of lipid order and single-molecule FRET to study the role of the lipid environment on BAM-catalysed, BamA-catalysed, and uncatalysed folding of tOmpA, both with and without SurA.
In Chapter 3, a new crosslinking method was developed and validated on the Skp-OmpA chaperone-substrate pair. This was then applied to the other major OMP chaperone, SurA, where it could be shown that the binding activity of SurA resides almost exclusively in the core N- and C-terminal domains. Finally, this approach was used to try and capture a folding intermediate of OmpA as it was passed from SurA through the BAM complex and then analyse the interactions from OmpA to these partners during folding. A position at the bottom of the first (N-terminal) β-strand of OmpA makes crosslinks with the POTRA1, 4, and 5 domains of BamA as well as with the N-terminus and P2 of SurA, suggesting a greater recruitment of SurA P2 during OMP folding. This pattern of crosslinks from OmpA also implies a possible route from POTRA1, via BamD near the interface with BamA-POTRA5, onto β1 of BamA, with the crosslinks at POTRA4 formed last as the final β-strand is appended to the nascent barrel.
In Chapter 4, the hypothesis that BAM functions by disordering lipids in the membrane was tested by using a number of techniques. DMPC was used as a model bilayer to be able to control the phase of lipids by conducting experiments at, below, or above, the transition temperature of 24 °C. The kinetics of tOmpA folding into DMPC liposomes showed that the full BAM complex is a much better catalyst for OMP folding (as measured by t50) than BamA alone when below or at the transition temperature of DMPC, and slightly better when above. While the BAM complex could accelerate the formation of folded tOmpA almost 16X over uncatalyzed folding at the transition temperature (24 °C), folding via BamA was only marginally faster (at 1.5 fold the uncatalyzed t50) which prompted further studies into the ability of these proteins to affect lipid order. The packing of lipids was assessed directly using the lipid order probe, laurdan, and the dynamics and conformational ensemble of the BAM complex was measured at a single-molecule level using FRET (smFRET). Laurdan experiments found that the presence of the BAM complex causes a broadening of the phase transition region as well as a 2 °C fall in the transition temperature implying a stabilisation of the liquid phase by BAM. smFRET studies showed that two populations of the BAM complex exist in solution, corresponding to the predicted FRET efficiencies of the open and closed states and these do not appear to interconvert on a 100s of μs to 100s of ms timescale.
In Chapter 5, the organisation of the BAM complex is probed by using a novel method in super-resolution microscopy, cryoSTORM. By plunge-freezing samples of E. coli expressing BamA and/or OmpA fused with fluorescent proteins or the self-labelling HaloTag protein, these proteins could be visualised in their assembled state on the surface of a bacteria trapped in a frozen-hydrated state with sub 5 nm localisation precision. This showed the arrangement of molecules of BAM into discrete ‘islands’ spotted throughout the cell surface as well as smaller islands formed by OmpA showing that OMPs are prone to cluster together in small islands. Initial two-colour studies of OmpA and BamA suggest a relatively low degree of co-localisation for these proteins.
