Characterization of the secretins, large outer membrane channels of gram-negative bacteria : a thesis presented in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biochemistry at Massey University, Palmerston North, New Zealand

Abstract

Secretins, a family of large outer membrane channels, mediate secretion and/or assembly of virulence factors and/or complex proteinaceous structures, such as rods, type IV pili and filamentous bacteriophage. Secretins form large radially-symmetric channels composed of 12 to 14 identical subunits, with internal diameters of up to 10 nm, whose lumen is interrupted by a septum/plug structure that very likely represents a gate or a valve. The identity of septum in the primary sequence of secretins has not been determined as yet, however the cryo-EM and SPA analyses point to the C-terminal domain forming these structures, whereas mutagenesis specifically identified two regions in this domain, named GATE1 and GATE2, as having an important role in gating of the filamentous phage secretion system secretin pIV. However, it is not known whether these regions are also involved in gating of the secretins from type II and type III secretion systems. In this work, twelve “leaky-gate” mutants in the secretin PulD from the type II secretion system (T2SS), selected from a random mutant library based on ability to utilise 829 Da oligosaccharide maltopentaose in the absence of maltoporin, were analysed in detail. Most of PulD leaky-gate mutants clustered in the GATE1 and GATE2 regions. All point mutants were positive for secretion of the cognate PulD substrate, enzyme Pullulanase (PulA), whereas a 5-residue in-frame deletion (?477-481) was negative. Two severely leaky GATE1 region mutants, G458S and ?477-481,sensitised E. coli to all tested antibiotics whose molecular weight is too high to pass through porins: rifamycin SV (720 Da), bacitracin (1423 Da) vancomycin (1449 Da) and daptomycin (1621 Da). The GATE1 of this secretin is therefore a potential drug target, for design of molecules that can sensitise secretin-containing pathogenic Gram-negative bacteria to > 600 Da antibiotics and/or block the secretion of substrates, including virulence factors. Engineered chimeras between PulD and pIV were used to probe functional compartmentalisation among the secretin domains and the segments involved in gating. This analysis showed that the N-terminal domains, GATE1 region and channel-forming secretin homology domain are interdependent with respect to function in secretion/assembly of the substrates, and to different extends for folding and multimerisation. This work further analysed the gating properties of the type III secretion system (T3SS) secretins EscC and InvG. When expressed in E. coli K12, these secretins were naturally “leaky” and mistargeted to the inner membrane, resulting in growth retardation. The survival of E. coli expressing these secretins depended on the PspF, positive regulator of the key inner membrane stress response Psp. Therefore, in the T3SS secretin-expressing or toxin-secreting cells, PspF is a potential target for design of molecules that could kill T3SS-containing toxin-secreting Gram-negative bacteria.