Colin Hughes

Crystal structure of the bacterial membrane protein TolC central to multidrug efflux and protein export.

Diverse molecules, from small antibacterial drugs to large protein toxins, are exported directly across both cell membranes of Gram-negative bacteria. This export is brought about by the reversible interaction of substrate-specific inner-membrane proteins with an outer-membrane protein of the TolC family, thus bypassing the intervening periplasm. Here we report the 2.1-Å crystal structure of TolC from Escherichia coli, revealing a distinctive and previously unknown fold. Three TolC protomers assemble to form a continuous, solvent-accessible conduit—a ‘channel-tunnel’ over 140 Å long that spans both the outer membrane and periplasmic space. The periplasmic or proximal end of the tunnel is sealed by sets of coiled helices. We suggest these could be untwisted by an allosteric mechanism, mediated by protein–protein interactions, to open the tunnel. The structure provides an explanation of how the cell cytosol is connected to the external environment during export, and suggests a general mechanism for the action of bacterial efflux pumps.

Substrate-induced assembly of a contiguous channel for protein export from E.coli: reversible bridging of an inner-membrane translocase to an outer membrane exit pore.

The toxin HlyA is exported from Escherichia coli, without a periplasmic intermediate, by a type I system comprising an energized inner‐membrane (IM) translocase of two proteins, HlyD and the traffic ATPase HlyB, and the outer‐membrane (OM) porin‐like TolC. These and the toxin substrate were expressed separately to reconstitute export and, via affinity tags on the IM proteins, cross‐linked in vivo complexes were isolated before and after substrate engagement. HlyD and HlyB assembled a stable IM complex in the absence of TolC and substrate. Both engaged HlyA, inducing the IM complex to contact TolC, concomitant with conformational change in all three exporter components. The IM–OM bridge was formed primarily by HlyD, which assembled to stable IM trimers, corresponding to the OM trimers of TolC. The bridge was transient, components reverting to IM and OM states after translocation. Mutant HlyB that bound, but did not hydrolyse ATP, supported IM complex assembly, substrate recruitment and bridging, but HlyA stalled in the channel. A similar picture was evident when the HlyD C‐terminus was masked. Export thus occurs via a contiguous channel which is formed, without traffic ATPase ATP hydrolysis, by substrate‐induced, reversible bridging of the IM translocase to the OM export pore.

Interactions underlying assembly of the Escherichia coli AcrAB–TolC multidrug efflux system

The major Escherichia coli multidrug efflux pump AcrAB–TolC expels a wide range of antibacterial agents. Using in vivo cross‐linking, we show for the first time that the antiporter AcrB and the adaptor AcrA, which form a translocase in the inner membrane, interact with the outer membrane TolC exit duct to form a contiguous proteinaceous complex spanning the bacterial cell envelope. Assembly of the pump appeared to be constitutive, occurring in the presence and absence of drug efflux substrate. This contrasts with substrate‐induced assembly of the closely related TolC‐dependent protein export machinery, possibly reflecting different assembly dynamics and degrees of substrate responsiveness in the two systems. TolC could be cross‐linked independently to AcrB, showing that their large periplasmic domains are in close proximity. However, isothermal titration calorimetry detected no interaction between the purified AcrB and TolC proteins, suggesting that the adaptor protein is required for their stable association in vivo. Confirming this view, AcrA could be cross‐linked independently to AcrB and TolC in vivo, and calorimetry demonstrated energetically favourable interactions of AcrA with both AcrB and TolC proteins. AcrB was bound by a polypeptide spanning the C‐terminal half of AcrA, but binding to TolC required interaction of N‐ and C‐terminal polypeptides spanning the lipoyl‐like domains predicted to present the intervening coiled‐coil to the periplasmic coils of TolC. These in vivo and in vitro analyses establish the central role of the AcrA adaptor in drug‐independent assembly of the tripartite drug efflux pump, specifically in coupling the inner membrane transporter and the outer membrane exit duct.

Isolation and analysis of the C-terminal signal directing export of Escherichia coli hemolysin protein across both bacterial membranes.

We have studied the C‐terminal signal which directs the complete export of the 1024‐amino‐acid hemolysin protein (HlyA) of Escherichia coli across both bacterial membranes into the surrounding medium. Isolation and sequencing of homologous hlyA genes from the related bacteria Proteus vulgaris and Morganella morganii revealed high primary sequence divergence in the three HlyA C‐termini and highlighted within the extreme terminal 53 amino acids the conservation of three contiguous sequences, a potential 18‐amino‐acid amphiphilic alpha‐helix, a cluster of charged residues, and a weakly hydrophobic terminal sequence rich in hydroxylated residues. Fusion of the C‐terminal 53 amino acid sequence to non‐exported truncated Hly A directed wild‐type export but export was radically reduced following independent disruption or progressive truncation of the three C‐terminal features by in‐frame deletion and the introduction of translation stop codons within the 3′ hlyA sequence. The data indicate that the HlyA C‐terminal export signal comprises multiple components and suggest possible analogies with the mitochondrial import signal. Hemolysis assays and immunoblotting confirmed the intracellular accumulation of non‐exported HlyA proteins and supported the view that export proceeds without a periplasmic intermediate. Comparison of cytoplasmic and extracellular forms of an independently exported extreme C‐terminal 194 residue peptide showed that the signal was not removed during export.

Co-ordinate expression of virulence genes during swarm-cell differentiation and population migration of Proteus mirabilis

The uropathogenic Gram‐negative bacterium Proteus mirabilis exhibits a form of multicellular behaviour termed swarming, which involves cyclical differentiation of typical vegetative cells into filamentous, multi‐nucleate, hyperflagellate swarm cells capable of rapid and co‐ordinated population migration across surfaces. We observed that differentiation into swarm cells was accompanied by substantial increases in the activities of intracellular urease and extracellular haemolysjn and metalloprotease, which are believed to be central to the pathogenicity of P. mirabills. In addition, the ability of P. mirabilis to invade human urothelial cells in vitro was primarily a characteristic of differentiated swarm cells, not vegetative cells. These virulence factor activities fell back as the cells underwent cyclical reversion to the vegetative form (consolidation), in parallel with the diagnostic modulation of flagellin levels on the cell surface. Control cellular alkaline phosphatase activities did not increase during differentiation or consolidation. Non‐flagellated, non‐motile transposon insertion mutants were unable to invade urothelial cells and they generated only low‐level activities of haemolysin, urease and protease (0–10% of wild type). Motile mutants unable to differentiate into swarm cells were comparably reduced in their haemolytic, ureolytic and invasive phenotypes and generated threefold less protease activity. Mutants that were able to form swarm cells but exhibited various aberrant patterns of swarming migration produced wild‐type activities of haemotysin, urease and protease, but their ability to enter urothelial cells was three‐ to 10‐fold lower. Analysis of haemolysin (hpmA) transcripts showed that during swarm cell differentiation there was a c. 50‐fold increase in the level of the predicted major 5.2 kb and minor 6.9 kb mRNAs transcribed from the hpm operon, and assay of mRNA complimentary to urease (ureC) and flagellin (fliC) gene sequences confirmed that modulation of virulence factor activity during the swarming cycle resulted from differential expression of virulence genes in parallel with fliC gene expression. Hybridization of stage‐specific mRNA with 30 random, non‐overlapping chromosomal gene probes provided no evidence for universal changes in the expression of the P. mirabiis genome, suggesting that differential virulence gene expression may be a specific strategy.

Substrate-specific binding of hook-associated proteins by FlgN and FliT, putative chaperones for flagellum assembly

During flagellum assembly by motile enterobacteria, flagellar axial proteins destined for polymerization into the cell surface structure are thought to be exported through the 25–30 Å flagellum central channel as partially unfolded monomers. How are premature folding and oligomerization in the cytosol prevented? We have shown previously using hyperflagellated Proteus mirabilis and a motile but non‐swarming flgN transposon mutant that the apparently cytosolic 16.5 kDa flagellar protein FlgN facilitates efficient flagellum filament assembly. Here, we investigate further whether FlgN, predicted to contain a C‐terminal amphipathic helix typical of type III export chaperones, acts as a chaperone for axial proteins. Incubation of soluble radiolabelled FlgN from Salmonella typhimurium with nitrocellulose‐immobilized cell lysates of wild‐type S. typhimurium and a non‐flagellate class 1 flhDC mutant indicated that FlgN binds to flagellar proteins. Identical affinity blot analysis of culture supernatants from the wild‐type and flhDC, flgI, flgK, flgL, fliC or fliD flagellar mutants showed that FlgN binds to the flagellar hook‐associated proteins (HAPs) FlgK and FlgL. This was confirmed by blotting artificially expressed individual HAPs in Escherichia coli. Analysis of axial proteins secreted into the culture medium by the original P. mirabilis flgN mutant demonstrated that export of FlgK and FlgL was specifically reduced, with concomitant increased release of unpolymerized flagellin (FliC), the immediately distal component of the flagellum. These data suggest that FlgN functions as an export chaperone for FlgK and FlgL. Parallel experiments showed that FliT, a similarly small (14 kDa), potentially helical flagellar protein, binds specifically to the flagellar filament cap protein, FliD (HAP2), indicating that it too might be an export chaperone. Flagellar axial proteins all contain amphipathic helices at their termini. Removal of the HAP C‐terminal helical domains abolished binding by FlgN and FliT in each case, and polypeptides comprising each of the HAP C‐termini were specifically bound by FlgN and FliT. We suggest that FlgN and FliT are substrate‐specific flagellar chaperones that prevent oligomerization of the HAPs by binding to their helical domains before export.

RfaH and the ops element, components of a novel system controlling bacterial transcription elongation

The RfaH protein controls the transcription of a specialized group of Escherichia coli and Salmonella operons that direct the synthesis, assembly and export of the lipopolysaccharide core, exopolysaccharide, F conjugation pilus and haemolysin toxin. RfaH is a specific regulator of transcript elongation; its loss increases transcription polarity in these operons without affecting initiation from the operon promoters. The operons of the RfaH‐dependent regulon contain a short conserved 5′ sequence, the ops element, deletion of which increases operon polarity to an extent similar to that caused by loss of RfaH. The ops element is also present upstream of polysaccharide gene clusters of Shigella flexneri, Yersinia enterocolitica, Vibrio cholerae and Klebsiella pneumoniae and the RP4 fertility operon of Pseudomonas aeruginosa, suggesting that this is a widely spread control system. The mechanistic coupling of RfaH and the ops element has been demonstrated in vitro and in vivo, and we suggest that the ops element recruits RfaH and potentially other factors to the RNA polymerase complex, modifying the complex to increase its processivity and allowing transcription to proceed over long distances.

Transition to the open state of the TolC periplasmic tunnel entrance

The TolC channel-tunnel spans the bacterial outer membrane and periplasm, providing a large exit duct for protein export and multidrug efflux when recruited by substrate-engaged inner membrane complexes. The sole constriction in the single pore of the homotrimeric TolC is the periplasmic tunnel entrance, which in its resting configuration is closed by dense packing of the 12 tunnel-forming α-helices. Recruitment of TolC must trigger opening for substrate transit to occur, but the mechanism underlying transition from the closed to the open state is not known. The high resolution structure of TolC indicates that the tunnel helices are constrained at the entrance by a circular network of intra- and intermonomer hydrogen bonds and salt bridges. To assess how opening is achieved, we disrupted these connections and monitored changes in the aperture size by measuring the single channel conductance of TolC derivatives in black lipid bilayers. Elimination of individual connections caused incremental weakening of the circular network, accompanied by gradual relaxation from the closed state and increased flexibility of the entrance. Simultaneous abolition of the key links caused a substantial increase in conductance, generating an aperture that corresponds to the modeled open state, with the capacity to allow access and passage of diverse substrates. The results support a model in which transition to the open state of TolC is achieved by an iris-like realignment of the tunnel entrance helices.

From flagellum assembly to virulence: the extended family of type III export chaperones

acteria have evolved severalexport systems to handleproteins destined for the surrounding medium or assemblyinto cell surface macromolecularstructures. The type III secretionmechanism exports many viru-lence effector proteins involved inthe subversion of eukaryotic cellfunction, and has been describedin a variety of animal and plantpathogens, including species of

A CELL-SURFACE POLYSACCHARIDE THAT FACILITATES RAPID POPULATION MIGRATION BY DIFFERENTIATED SWARM CELLS OF PROTEUS MIRABILIS

Swarming by Proteus mirabilis is characterized by cycles of rapid population migration across surfaces, following differentiation of typical vegetative rods into long, hyperflagellated, virulent swarm cells. A swarm‐defective TnphoA insertion mutant was isolated that was not defective in cell motility, differentiation or control of the migration cycle, but was specifically impaired in the ability to undergo surface translocation as a multicellular mass. The mutation, previously shown to compromise urinary tract virulence, was located within a 1112 bp gene that restored normal swarming of the mutant when expressed in trans. The gene encoded a 40.6 kDa protein that is related to putative sugar transferases required for lipopolysaccharide (LPS) core modification in Shigella and Salmonella. The immediately distal open reading frame encoded a protein that is related to dehydrogenases involved in the synthesis of LPS O‐side‐chains, enterobacterial common antigen and extracellular polysaccharide (PS). Gel electrophoresis and electron microscopy showed that the mutant still made LPS but it had lost the ability to assemble a surface (capsular) PS, which gas‐liquid chromatography and mass spectrometry indicated to be an acidic type II molecule rich in galacturonic acid and galactosamine. We suggest that this surface PS facilitates translocation of differentiated cell populations by reducing surface friction.