The Antibiotic Efflux Protein TolC Is a Highly Evolvable Target under Colicin E1 or TLS Phage Selection

Bacteriophages and bacterial toxins are promising antibacterial agents to treat infections caused by multidrug-resistant (MDR) bacteria. In fact, bacteriophages have recently been successfully used to treat life-threatening infections caused by MDR bacteria (Schooley RT, Biswas B, Gill JJ, Hernandez-Morales A, Lancaster J, Lessor L, Barr JJ, Reed SL, Rohwer F, Benler S, et al. 2017. Development and use of personalized bacteriophage-based therapeutic cocktails to treat a patient with a disseminated resistant Acinetobacter baumannii infection. Antimicrob Agents Chemother. 61(10); Chan BK, Turner PE, Kim S, Mojibian HR, Elefteriades JA, Narayan D. 2018. Phage treatment of an aortic graft infected with Pseudomonas aeruginosa. Evol Med Public Health. 2018(1):60–66; Petrovic Fabijan A, Lin RCY, Ho J, Maddocks S, Ben Zakour NL, Iredell JR, Westmead Bacteriophage Therapy Team. 2020. Safety of bacteriophage therapy in severe Staphylococcus aureus infection. Nat Microbiol. 5(3):465–472). One potential problem with using these antibacterial agents is the evolution of resistance against them in the long term. Here, we studied the fitness landscape of the Escherichia coli TolC protein, an outer membrane efflux protein that is exploited by a pore forming toxin called colicin E1 and by TLS phage (Pagie L, Hogeweg P. 1999. Colicin diversity: a result of eco-evolutionary dynamics. J Theor Biol. 196(2):251–261; Andersen C, Hughes C, Koronakis V. 2000. Chunnel vision. Export and efflux through bacterial channel-tunnels. EMBO Rep. 1(4):313–318; Koronakis V, Andersen C, Hughes C. 2001. Channel-tunnels. Curr Opin Struct Biol. 11(4):403–407; Czaran TL, Hoekstra RF, Pagie L. 2002. Chemical warfare between microbes promotes biodiversity. Proc Natl Acad Sci U S A. 99(2):786–790; Cascales E, Buchanan SK, Duché D, Kleanthous C, Lloubès R, Postle K, Riley M, Slatin S, Cavard D. 2007. Colicin biology. Microbiol Mol Biol Rev. 71(1):158–229). By systematically assessing the distribution of fitness effects of ∼9,000 single amino acid replacements in TolC using either positive (antibiotics and bile salts) or negative (colicin E1 and TLS phage) selection pressures, we quantified evolvability of the TolC. We demonstrated that the TolC is highly optimized for the efflux of antibiotics and bile salts. In contrast, under colicin E1 and TLS phage selection, TolC sequence is very sensitive to mutations. Finally, we have identified a large set of mutations in TolC that increase resistance of E. coli against colicin E1 or TLS phage without changing antibiotic susceptibility of bacterial cells. Our findings suggest that TolC is a highly evolvable target under negative selection which may limit the potential clinical use of bacteriophages and bacterial toxins if evolutionary aspects are not taken into account.     

Initial steps of colicin E1 import across the outer membrane of Escherichia coli

Data suggest a two-receptor model for colicin E1 (ColE1) translocation across the outer membrane of Escherichia coli. ColE1 initially binds to the vitamin B(12) receptor BtuB and then translocates through the TolC channel-tunnel, presumably in a mostly unfolded state. Here, we studied the early events in the import of ColE1. Using in vivo approaches, we show that ColE1 is cleaved when added to whole cells. This cleavage requires the presence of the receptor BtuB and the protease OmpT, but not that of TolC. Strains expressing OmpT cleaved ColE1 at K84 and K95 in the N-terminal translocation domain, leading to the removal of the TolQA box, which is essential for ColE1's cytotoxicity. Supported by additional in vivo data, this suggests that a function of OmpT is to degrade colicin at the cell surface and thus protect sensitive E. coli cells from infection by E colicins. A genetic strategy for isolating tolC mutations that confer resistance to ColE1, without affecting other TolC functions, is also described. We provide further in vivo evidence of the multistep interaction between TolC and ColE1 by using cross-linking followed by copurification via histidine-tagged TolC. First, secondary binding of ColE1 to TolC is dependent on primary binding to BtuB. Second, alterations to a residue in the TolC channel interfere with the translocation of ColE1 across the TolC pore rather than with the binding of ColE1 to TolC. In contrast, a substitution at a residue exposed on the cell surface abolishes both binding and translocation of ColE1.     

Biosynthesis of the outer membrane receptor for vitamin B12, E colicins, and bacteriophage BF23 by Escherichia coli: kinetics of phenotypic expression after the introduction of bfe+ and bfe alleles.

The bfe locus codes for the cell surface receptor for vitamin B12, the E colicins, and bacteriophage BF23 in the Escherichia coli outer membrane. When the bfe+ allele, which is closely linked to the argH locus, was introduced into an argH bfe recipient by conjugation, arg+ recombinant cells rapidly and simultaneously acquired sensitivity to colicin E3 and phage BF23. In the reciprocal experiment introducing bfe into an argH bfe+ recipient, it was found that colicin E3-resistant, arg+ cells began to appear shortly after the arg+ recombinant population began to divide. This was far earlier than would have been predicted on the basis of 220 receptors per haploid cell. Moreover, there was a lag between the appearance of colicin resistance and the appearance of resistance to killing by phage BF23, and hence a period of time during which some arg+ recombinant cells were sensitive to the phage but resistant to the colicin. Colicin E3 added to cells during this period of time protected against phage killing, indicating that the colicin-resistant cells still had receptors capable of binding colicin on their surface. The modification of the phenotypic expression of colicin and phage resistance by inhibitors of deoxyribonucleic acid, ribonucleic acid, and protein synthesis was also investigated. The results obtained indicate that the receptor protein coded for by the bfe locus can exist on the cell surface in several different functional states.     

Replication of colicin E1 plasmid DNA in vivo requires no plasmid-encoded proteins.

A derivative of bacteriophage lambda containing a colicin E1 plasmid replicon was constructed by recombinant DNA techniques. This phage, lambdacol100, has two functional modes of DNA replication; it can replicate via either plasmid or phage replication systems. lambdacol100 has been used to introduce the colicin E1 plasmid replicon into Escherichia coli previously treated with chloramphenicol to block protein synthesis. Under these conditions, lambdacol100 DNA is replicated normally as a colicin E1 plasmid. This suggests that colicin E1 plasmid replication in vivo does not require any plasmid-encoded proteins.     

A hybrid toxin from bacteriophage f1 attachment protein and colicin E3 has altered cell receptor specificity.

A hybrid protein was constructed in vitro which consists of the first 372 amino acids of the attachment (gene III) protein of filamentous bacteriophage f1 fused, in frame, to the carboxy-terminal catalytic domain of colicin E3. The hybrid toxin killed cells that had the F-pilus receptor for phage f1 but not F- cells. The activity of the hybrid protein was not dependent upon the presence of the colicin E3 receptor, BtuB protein. The killing activity was colicin E3 specific, since F+ cells expressing the colicin E3 immunity gene were not killed. Entry of the hybrid toxin was also shown to depend on the products of tolA, tolQ, and tolR which are required both for phage f1 infection and for entry of E colicins. TolB protein, which is required for killing by colicin E3, but not for infection by phage f1, was also found to be necessary for the killing activity of the hybrid toxin. The gene III protein-colicin E3 hybrid was released from producing cells into the culture medium, although the colicin E3 lysis protein was not present in those cells. The secretion was shown to depend on the 18-amino-acid-long gene III protein signal sequence. Deletion of amino acids 3 to 18 of the gene III moiety of the hybrid protein resulted in active toxin, which remained inside producing cells unless it was mechanically released.     

Lysis gene t of T-even bacteriophages: evidence that colicins and bacteriophage genes have common ancestors.

The lysis gene t of the T-even-like bacteriophage K3 has been cloned and sequenced. The gene codes for a protein with a predicted molecular weight of 25,200. Expression of the complete lysis protein was impossible, but peptides complementing T4 amber mutants in t are described. No known lysis protein of other phages is homologous to protein T. Also, the Escherichia coli phospholipase A is different from protein T. CelB, the lysis protein of the colicin E2 operon, shows a similarity to protein T. Sequences of colicins A, E1, and E2 are related to gene 38 sequences, the gene preceding t and coding for the phage adhesin. A common origin for colicin genes and phage genes is discussed, and a protein region in colicins that is responsible for receptor recognition is predicted.     

Genetic analysis of an MDR-like export system: the secretion of colicin V.

The extracellular secretion of the antibacterial toxin colicin V is mediated via a signal sequence independent process which requires the products of two linked genes: cvaA and cvaB. The nucleotide sequence of cvaB reveals that its product is a member of a subfamily of proteins, involved in the export of diverse molecules, found in both eukaryotes and prokaryotes. This group of proteins, here referred to as the 'MDR-like' subfamily, is characterized by the presence of a hydrophobic region followed by a highly conserved ATP binding fold. By constructing fusions between the structural gene for colicin V, cvaC, and a gene for alkaline phosphatase, phoA, lacking its signal sequence, it was determined that 39 codons in the N-terminus of cvaC contained the structural information to allow CvaC-PhoA fusion proteins to be efficiently translocated across the plasma membrane of Escherichia coli in a CvaA/CvaB dependent fashion. This result is consistent with the location of point mutations in the cvaC gene which yielded export deficient colicin V. The presence of the export signal at the N-terminus of CvaC contrasts with the observed C-terminal location of the export signal for hemolysin, which also utilizes an MDR-like protein for its secretion. It was also found that the CvaA component of the colicin V export system shows amino acid sequence similarities with another component involved in hemolysin export, HlyD. The role of the second component in these systems and the possibility that other members of the MDR-like subfamily will also have corresponding second components are discussed. A third component used in both colicin V and hemolysin extracellular secretion is the E. coli host outer membrane protein, TolC.     

A common receptor protein for phage T5 and colicin M in the outer membrane of Escherichia coli B

The receptor protein for phage T5 was isolated from the outer membrane of Escherichia coli B and found to be also a receptor for colicin M. The receptor protein from a phage-resistant mutant inactivates neither the phage nor the colicin. Binding of colicin M to the receptor prevents binding of phage T5. It is concluded that phage T5 and colicin M bind to the same active area of this receptor protein. The receptor protein seems to consist of one polypeptide chain with a molecular weight of 85000.     

Antibiotic-sensitive TolC mutants and their suppressors

The TolC protein of Escherichia coli, through its interaction with AcrA and AcrB, is thought to form a continuous protein channel that expels inhibitors from the cell. Consequently, tolC null mutations display a hypersensitive phenotype. Here we report the isolation and characterization of tolC missense mutations that direct the synthesis of mutant TolC proteins partially disabled in their efflux role. All alterations, consisting of single amino acid substitutions, were localized within the periplasmic alpha-helical domain. In two mutants carrying an I106N or S350F substitution, the hypersensitivity phenotype may be in part due to aberrant TolC assembly. However, two other alterations, R367H and R390C, disrupted efflux function by affecting interactions among the helices surrounding TolC's periplasmic tunnel. Curiously, these two TolC mutants were sensitive to a large antibiotic, vancomycin, and exhibited a Dex(+) phenotype. These novel phenotypes of TolC(R367H) and TolC(R390C) were likely the result of a general influx of molecules through a constitutively open tunnel aperture, which normally widens only when TolC interacts with other proteins during substrate translocation. An intragenic suppressor alteration (T140A) was isolated from antibiotic-resistant revertants of the hypersensitive TolC(R367H) mutant. T140A also reversed, either fully (R390C) or partially (I106N and S350F), the hypersensitivity phenotype of other TolC mutants. Our data suggest that this global suppressor phenotype of T140A is the result of impeded antibiotic influx caused by tapering of the tunnel passage rather than by correcting individual mutational defects. Two extragenic suppressors of TolC(R367H), mapping in the regulatory region of acrAB, uncoupled the AcrR-mediated repression of the acrAB genes. The resulting overexpression of AcrAB reduced the hypersensitivity phenotype of all the TolC mutants. Similar results were obtained when the chromosomal acrR gene was deleted or the acrAB genes were expressed from a plasmid. Unlike the case for the intragenic suppressor T140A, the overexpression of AcrAB diminished hypersensitivity towards only erythromycin and novobiocin, which are substrates of the TolC-AcrAB efflux pump, but not towards vancomycin, which is not a substrate of this pump. This showed that the two types of suppressors produced their effects by fundamentally different means, as the intragenic suppressor decreased the general influx while extragenic suppressors increased the efflux of TolC-AcrAB pump-specific antibiotics.     

In vivo and in vitro characterization of overproduced colicin E9 immunity protein.

We report the overproduction of the immunity protein for the DNase colicin E9 and its characterization both in vivo and in vitro. The genes for colicin immunity proteins are normally co-expressed from Col plasmids with their corresponding colicins. In the context of the enzymatic colicins, the two proteins form a complex, thereby protecting the host bacterium from the antibiotic activity of the colicin. This complex is then released into the medium, whereupon the colicin alone translocates (through the appropriate receptor) into sensitive bacterial strains, resulting in bacterial cell death. The immunity protein for colicin E9 (Im9) has been overproduced in a bacterial host in the absence of its colicin, to enable sufficient material to be isolated for structural studies. As a prelude to such studies, the in-vivo and in-vitro properties of overproduced Im9 were analysed. Electrospray mass spectrometry verified the molecular mass of the purified protein and analytical ultracentrifugation indicated that the native protein approximates a symmetric monomer. Fluorescence-enhancement and gel-filtration experiments show that purified Im9 binds to colicin E9 in a 1:1 molar ratio and that this binding neutralizes the DNase activity of the colicin. These results lay the foundations for a full biophysical and structural characterization of the colicin E9 DNase inhibitor protein, Im9.     

Receptor specificity of the Escherichia coli T-even type phage Ox2. Mutational alterations in host range mutants

The T-even type Escherichia coli phage Ox2 uses the outer membrane protein OmpA as a receptor. The protein is recognized with the ends of the virion's long tail fibers. The 266 residue protein 38 is located at this site and acts as an adhesin. Host-range mutants had previously been isolated from Ox2. Mutant Ox2h5 is able to infect cells possessing an altered OmpA protein, which renders the cell resistant to Ox2. Ox2h10 was selected from Ox2h5. This phage recognizes the OmpC protein in addition to the OmpA protein. Ox2h12, which stems from Ox2h10, binds to OmpC with high affinity, but has lost efficient binding to OmpA. The mutational alterations caused in genes 38 are: Asp231----Asn(h5) and His170----Arg(h10). The triple mutant Ox2h12 possesses an insertion of a Gly residue next to Gly121. The three mutants have additionally acquired mutations affecting their base plate, making them "trigger-happy". When protein 38 was compared with the same protein derived from other E. coli phages, it was found to contain two constant and one variable domains, the latter harboring four hypervariable regions flanked by a largely conserved glycine-rich sequence. The h5 and h10 mutations occurred within two hypervariable areas, while the additional Gly residue was present in one of the flanking conserved sequences. On the basis of these results, as well as those obtained from host-range mutants analyzed previously, a model for such adhesins is proposed. Receptor recognition is most likely performed via the hypervariable regions, which may form loops held together in close proximity by the oligoglycine sequences. The latter may achieve this by being part of highly compact omega loops.     

The role of the TolC family in protein transport and multidrug efflux. From stereochemical certainty to mechanistic hypothesis.

Gram-negative bacteria are enveloped by a system of two membranes, and they use specialized multicomponent, energy-driven pumps to transport molecules directly across this double-layered partition from the cell interior to the extra-cellular environment. One component of these pumps is embedded in the outer-membrane, and the paradigm for its structure and function is the TolC protein from Escherichia coli. A common component of a wide variety of efflux pumps, TolC and its homologues are involved in the export of chemically diverse molecules ranging from large protein toxins, such as alpha-hemolysin, to small toxic compounds, such as antibiotics. TolC family members thus play important roles in conferring pathogenic bacteria with both virulence and multidrug resistance. These pumps assemble reversibly in a transient process that brings together TolC or its homologue, an inner-membrane-associated periplasmic component, an integral inner-membrane translocase and the substrate itself. TolC can associate in this fashion with a variety of different partners to participate in the transport of diverse substrates. We review here the structure and function of TolC and the other components of the efflux/transport pump.     

Conjugation-deficient mutants of Escherichia coli distinguish classes of functions of the outer membrane OmpA protein

Summary Sixty-two E. coli mutants, selected as being deficient as recipients in F factor conjugation, are altered either in the amount or function of the outer membrane OmpA protein or in lipopolysaccharide structure. These two components may function together in conjugation, since the residual conjugation activity of a mutant lacking OmpA protein was unaffected by the additional presence of a lipopolysaccharide defect. Sixty of the strains carried mutations mapping to ompA, and these could be divided into classes depending on the amount of OmpA protein in their membranes. Representatives of these classes of mutant alleles failed to complement in diploids, indicating that they all affect the ompA structural gene and nearby sequences needed for its expression. The properties of these classes distinguish three groups of OmpA protein functions: 1) the structural function in the outer membrane in providing resistance to chelating agents and the hydrophobic antibiotic novobiocin, 2) the receptor functions in phage Tull^* and K3 infection, and 3) the functions of binding cells together during conjugation, facilitating the uptake of receptorbound colicin K or L, and allowing phage Ox2 to infect. Different cellular amounts or sites in OmpA protein are thus required for these three groups of functions.     

Conversion of the coprogen transport protein FhuE and the ferrioxamine B transport protein FoxA into ferrichrome transport proteins

The FhuA protein of Escherichia coli K-12 transports ferrichrome and the structurally related antibiotic albomycin across the outer membrane and serves as a receptor for the phages T1, T5, and φ80 and for colicin M. In this paper, we show that chimeric proteins consisting of the central part of FhuA and the N- and C-terminal parts of FhuE (coprogen receptor) or the N- and/or C-terminal parts of FoxA (ferrioxamine B receptor), function as ferrichrome transport proteins. Although the hybrid proteins contained the previously identified gating loop of FhuA, which is the principal binding site of the phages T5, T1, and φ80, only the hybrid protein consisting of the N-terminal third of FoxA and the C-terminal two thirds of FhuA conferred weak phage sensitivity to cells. Apparently, the gating loop is essential, but not sufficient for wild-type levels of ferrichrome transport and for phage sensitivity. The properties of FhuA-FoxA hybrids suggest different regions of the two receptors for ferric siderophore uptake.     

Colicin occlusion of OmpF and TolC channels: outer membrane translocons for colicin import

The interaction of colicins with target cells is a paradigm for protein import. To enter cells, bactericidal colicins parasitize Escherichia coli outer membrane receptors whose physiological purpose is the import of essential metabolites. Colicins E1 and E3 initially bind to the BtuB receptor, whose beta-barrel pore is occluded by an N-terminal globular "plug". The x-ray structure of a complex of BtuB with the coiled-coil BtuB-binding domain of colicin E3 did not reveal displacement of the BtuB plug that would allow passage of the colicin (Kurisu, G., S. D. Zakharov, M. V. Zhalnina, S. Bano, V. Y. Eroukova, T. I. Rokitskaya, Y. N. Antonenko, M. C. Wiener, and W. A. Cramer. 2003. Nat. Struct. Biol. 10:948-954). This correlates with the inability of BtuB to form ion channels in planar bilayers, shown in this work, suggesting that an additional outer membrane protein(s) is required for colicin import across the outer membrane. The identity and interaction properties of this OMP were analyzed in planar bilayer experiments.OmpF and TolC channels in planar bilayers were occluded by colicins E3 and E1, respectively, from the trans-side of the membrane. Occlusion was dependent upon a cis-negative transmembrane potential. A positive potential reversibly opened OmpF and TolC channels. Colicin N, which uses only OmpF for entry, occludes OmpF in planar bilayers with the same orientation constraints as colicins E1 and E3. The OmpF recognition sites of colicins E3 and N, and the TolC recognition site of colicin E1, were found to reside in the N-terminal translocation domains. These data are considered in the context of a two-receptor translocon model for colicin entry into cells.     

Identification of receptor binding sites by competitive peptide mapping: phages T1, T5, and phi 80 and colicin M bind to the gating loop of FhuA.

Previously we proposed a transmembrane model of the FhuA receptor protein in the outer membrane of Escherichia coli. Removal of the largest loop at the cell surface converted the FhuA transport protein into an open channel and rendered cells resistant to the FhuA-specific phages T1, T5, and phi 80 and to colicin M. In the present study we employed acetylated hexapeptide amides covering the entire surface loop to investigate binding of the phages and of colicin M. Competitive peptide mapping proved to be a powerful technique to uncover three ligand binding sites within a region of 34 amino acid residues. Hexapeptides derived from three specific regions of the surface loop inhibited infection of cells by the phages and killing by colicin M. Two of these regions were common among all four FhuA ligands. Electron microscopy of phage T5 revealed that one inhibitory peptide triggered a strong conformational change leading to the release of DNA from the phage head. These results suggest that the FhuA gating loop is the target for specific binding of phages T1, T5, and phi 80 and colicin M.     

An aspartate deletion mutation defines a binding site of the multifunctional FhuA outer membrane receptor of Escherichia coli K-12.

The FhuA protein of the outer membrane serves as a receptor for phages T5, T1, and phi 80, for colicin M, for the antibiotic albomycin, and for ferrichrome and related siderophores. To identify protein regions important for the multiple FhuA activities, fhuA genes of spontaneous chromosomal mutants which expressed wild-type amounts of the FhuA protein were sequenced. A mutant which was partially T5 sensitive but impaired in all other functions was missing aspartate residue 348 of the mature protein as a result of a three-base deletion. This aspartate residue is part of the hydrophilic sequence Asp-Asp-Glu-Lys. Replacement by site-specific mutagenesis of each of the Asp residues by Tyr, of Glu by Val, and of Lys by Met reduced FhuA activity but less than the Asp deletion did. Ferrichrome inhibited binding of phage phi 80 and of colicin M to these mutants in an allele-specific manner. A completely resistant derivative of the Asp deletion mutant contained, in addition, a leucine-to-proline substitution at position 106 and eight changed bases, converting at positions 576 to 578 an Arg-Pro-Leu sequence to Ala-Arg-Cys. The latter mutations and the Leu-to-Pro replacement alone did not alter sensitivity to the phages but reduced sensitivity to colicin M and albomycin 10- to 1,000-fold. The proline replacements probably disturb FhuA conformation and, in concert with the Asp deletion, inactivate FhuA completely. It is concluded that the Asp deletion site defines a region of FhuA which directly participates in binding of all FhuA ligands. Growth promotion studies on iron-limited media revealed that certain siderophores of the hydroxamate type, such as butylferrichrome, ferrichrysin, and ferrirubin, are taken up not only via FhuA but also via the FhuE outer membrane receptor protein.     

Inactivation of FhuA at the cell surface of Escherichia coli K-12 by a phage T5 lipoprotein at the periplasmic face of the outer membrane.

Inactivation of phage T5 by lysed cells after phage multiplication is prevented by a phage-encoded lipoprotein (Llp) that inactivates the FhuA outer membrane receptor protein (K. Decker, V. Krauel, A. Meesmann, and K. Heller, Mol. Microbiol. 12:321-332, 1994). Using FhuA derivatives carrying insertions of 4 and 16 amino acid residues and point mutations, we determined whether FhuA inactivation is caused by binding of Llp to FhuA and which regions of FhuA are important for inactivation by Llp. Cells expressing Llp were resistant not only to phage T5 but to all FhuA ligands tested, such as phage phi 80, colicin M, and albomycin, and they were strongly reduced in the uptake of ferrichrome. Most of the FhuA derivatives which were not affected by Llp were, according to a previously published FhuA transmembrane topology model, located in periplasmic turns and in the TonB box close to the periplasm. Since the ligands bind to the cell surface, interaction of FhuA with Llp in the periplasm may induce a FhuA conformation which impairs binding of the ligands. This conclusion was supported by the increase rather than decrease of colicin M sensitivity of two mutants in the presence of Llp. The only Llp-resistant FhuA derivatives with mutations at the cell surface contained insertions of 16 residues in the loop that determines the permeability of the FhuA channel and serves as the principal binding site for all FhuA ligands. This region may be inactivated by steric hindrance in that a portion of Llp penetrates into the channel. Outer membranes prepared with 0.25% Triton X-100 from cells expressing Llp contained inactivated FhuA, suggesting Llp to be an outer membrane protein whose interaction with FhuA was not abolished by Triton X-100. Llp solubilized in 1.1% octylglucoside prevented T5 inactivation by FhuA dissolved in octylglucoside.     

Functional implications of an intermeshing cogwheel-like interaction between TolC and MacA in the action of macrolide-specific efflux pump MacAB-TolC

Macrolide-specific efflux pump MacAB-TolC has been identified in diverse gram-negative bacteria including Escherichia coli. The inner membrane transporter MacB requires the outer membrane factor TolC and the periplasmic adaptor protein MacA to form a functional tripartite complex. In this study, we used a chimeric protein containing the tip region of the TolC α-barrel to investigate the role of the TolC α-barrel tip region with regard to its interaction with MacA. The chimeric protein formed a stable complex with MacA, and the complex formation was abolished by substitution at the functionally essential residues located at the MacA α-helical tip region. Electron microscopic study delineated that this complex was made by tip-to-tip interaction between the tip regions of the α-barrels of TolC and MacA, which correlated well with the TolC and MacA complex calculated by molecular dynamics. Taken together, our results demonstrate that the MacA hexamer interacts with TolC in a tip-to-tip manner, and implies the manner by which MacA induces opening of the TolC channel.     

In vivo and in vitro studies on interactions between the components of the hemolysin (HlyA) secretion machinery of Escherichia coli

The glycopeptide antibiotic vancomycin blocks cell wall synthesis in Escherichia coli only when it can reach its target site in the periplasm. In vivo, sensitivity to vancomycin is enhanced in the presence of the hemolysin (hly) determinant of E. coli or its translocator portion hlyBD. Two different mutations in hlyD alter the cell's susceptibility to vancomycin: mutations in the tolC-homologous region of hlyD increase vancomycin resistance, whereas mutations at the 3′-terminus of hlyD lead to hypersensitivity to vancomycin and to the accumulation of large periplasmic and cytoplasmic pools of this antibiotic in E. coli. These effects are only observed in the presence of functional HlyB and TolC, the two other components of the hemolysin secretion machinery. A defect in TolC causes hyperresistance to vancomycin, even when present together with a mutant HlyD protein which in the presence of TolC renders E. coli hypersensitive to vancomycin. Lipid bilayer experiments in vitro revealed specific interactions between TolC and vancomycin or HlyD protein. Second-site suppressor mutations in hlyD and hlyB were obtained, which abolish the hypersensitive phenotype caused by the 3′-terminal mutations in hlyD. Our results are compatible with the idea that (a) TolC, together with the TolC-homologous part of HlyD, forms a pore in the outer membrane through which hemolysin is released and vancomycin taken up; and (b) the C-terminal sequence of HlyD interacts with periplasmic loop(s) of HlyB to form a closed channel spanning the periplasm.