Periplasmic Face Research Articles

α-Helical conformation in the C-terminal anchoring domains of E. coli penicillin-binding proteins 4, 5 and 6

The E. coli low molecular mass penicillin-binding proteins (PBP's) are penicillin sensitive, enzymes involved in the terminal stages of peptidoglycan biosynthesesis. These PBP's are believed to anchor to the periplasmic face of the inner membrane via C-terminal amphiphilic α-helices but to date the only support for this hypothesis has been obtained from theoretical analysis. In this paper, the conformational behaviour of synthetic peptides corresponding to these C-terminal anchoring domains was studied as a function of solvent, pH, sodium dodecyl sulphate micelles and phospholipid (DOPC, DOPG) vesicles using circular dichroism (CD) spectroscopy. The CD data showed that in 2,2,2-trifluoroethanol or sodium dodecylsulphate, all three peptides have the capacity to form an α-helical conformation but in aqueous solution or in the presence of phospholipid vesicles only those peptides corresponding to the PBP5 and PBP6 C-termini were observed to do so. A pH dependent loss of α-helical conformation in the peptide corresponding to the PBP5 C-terminus was found to correlate with the susceptibility of PBP5 to membrane extraction. This correlation would agree with the hypothesis that an α-helical conformation is required for membrane interaction of the PBP5 C-terminal region.

Tromp1, a putative rare outer membrane protein, is anchored by an uncleaved signal sequence to the Treponema pallidum cytoplasmic membrane

Treponema pallidum rare outer membrane protein 1 (Tromp1) has extensive sequence homology with substrate-binding proteins of ATP-binding cassette transporters. Because such proteins typically are periplasmic or cytoplasmic membrane associated, experiments were conducted to clarify Tromp1's physicochemical properties and cellular location in T. pallidum. Comparison of the sodium dodecyl sulfate-polyacrylamide gel electrophoresis mobilities of (i) native Tromp1 and Tromp1 synthesized by coupled in vitro transcription-translation and (ii) native Tromp1 and recombinant Tromp1 lacking the N-terminal signal sequence revealed that the native protein is not processed. Other studies demonstrated that recombinant Tromp1 lacks three basic porin-like properties: (i) the ability to form aqueous channels in liposomes which permit the influx of small hydrophilic solutes, (ii) an extensive beta-sheet secondary structure, and (iii) amphiphilicity. Subsurface localization of native Tromp1 was demonstrated by immunofluorescence analysis of treponemes encapsulated in gel microdroplets, while opsonization assays failed to detect surface-exposed Tromp1. Incubation of motile treponemes with 3-(trifluoromethyl)-3-(m-[125I]iodophenyl)-diazarine, a photoactivatable, lipophilic probe, also did not result in the detection of Tromp1 within the outer membranes of intact treponemes but, instead, resulted in the labeling of a basic 30.5-kDa presumptive outer membrane protein. Finally, analysis of fractionated treponemes revealed that native Tromp1 is associated predominantly with cell cylinders. These findings comprise a body of evidence that Tromp1 actually is anchored by an uncleaved signal sequence to the periplasmic face of the T. pallidum cytoplasmic membrane, where it likely subserves a transport-related function.

An investigation into the ability of C-terminal homologues of Escherichia coli low molecular mass penicillin-binding proteins 4, 5 and 6 to undergo membrane interaction

The Escherichia coli low molecular mass penicillin-binding proteins (PBP4, PBP5 and PBP6) are a group of penicillin-sensitive enzymes involved in the final stages of cell wall assembly. It has been suggested that these proteins may interact with the periplasmic face of the inner membrane via C-terminal amphiphilic α-helices. Theoretical analysis has predicted that these C-terminal helical regions may be membrane interactive. We have tested this hypothesis by assaying PBP C-terminal homologues (P4, P5 and P6) for haemolytic activity. Our results show that the PBP5 and PBP6 C-terminal homologues readily lyse sheep erythrocytes in a pH-dependent manner with LD 50's of 3.5 × 10 −6 M and 6.8 × 10 −7 M respectively at pH 7. These results appear to support the present model for the membrane anchoring of PBP5 and PBP6. The PBP4 C-terminal homologue shows no evidence of haemolytic activity which could imply a different means of membrane association for PBP4.

Histidine 225, a Residue of the NhaA-Na+/H+ Antiporter of Escherichia coli Is Exposed and Faces the Cell Exterior

Cysteine residues were found nonessential in the mechanism of the NhaA antiporter activity of Escherichia coli. The functional C-less NhaA has provided the groundwork to study further histidine 225 of NhaA which has previously been suggested to play an important role in the activation of NhaA at alkaline pH (Rimon, A., Gerchman, Y., Olami, Y., Schuldiner, S. and Padan, E. (1995) J. Biol. Chem. 270, 26813-26817). C-less H225C was constructed and shown to possess an antiporter activity 60% of that of C-less antiporter and a pH profile similar to that of both the C-less or wild-type antiporters. Remarkably, whereas neither the wild-type nor the C-less antiporters were affected by N-ethylmaleimide, C-less H225C was inhibited by this reagent. To determine the degree of alkylation of the antiporter protein by N-ethylmaleimide, antiporter derivatives tagged at their C termini with six histidines residues were constructed. Alkylation of C-less H225C was measured by labeling of everted membrane vesicles with [14C]N-ethylmaleimide, affinity purification of the His-tagged antiporter, and determination of the radioactivity of the purified protein. This assay showed that H225C is alkylated to a much higher level than any of the native cysteinyl residues of NhaA reaching saturation at alkyl/NhaA stoichiometry of 1. The wild-type derivative showed at least 10-fold less alkylation even at higher concentrations, suggesting that H225C resides in a domain that is much more exposed to N-ethylmaleimide than the native cysteinyl residues of NhaA. Since H225C residues both in right-side out and inside-out membrane vesicles were quantitatively alkylated by N-ethylmaleimide, this assay was used to determine the accessibility of H225C to other SH reagents by titrating the H225C left free to react with N-ethylmaleimide, following exposure of the membranes to the reagents. Furthermore, since membrane-impermeant probes can react with residues in membrane-embedded protein only if accessible to the medium containing the reagent, the assay was used to determine the membrane topology of H225C. As expected for a membrane-permeant probe, p-chloromercuribenzoate reacted with H225C as efficiently as N-ethylmaleimide in both membrane orientations. Similar results were obtained with methanethiosulfonate ethylammonium supporting the recent observations that this probe is membrane-permeant. On the other hand, both membrane-impermeant reagents p-chloromercuribenzosulfonate and methanethiosulfonate ethyl-trimethyl ammonium bromide reacted with H225C 10-fold more in right-side out than in inside-out vesicles, and p-chloromercuribenzosulfonate also blocked completely the H225C in intact cells. These results strongly suggest that H225C is exposed at the periplasmic face of the membrane.

SecA is an intrinsic subunit of the Escherichia coli preprotein translocase and exposes its carboxyl terminus to the periplasm.

SecA is the dissociable ATPase subunit of the Escherichia coli preprotein translocase, and cycles in a nucleotide-modulated manner between the cytosol and the membrane. Overproduction of the integral subunits of the translocase, the SecY, SecE and SecG polypeptides, results in an increased level of membrane-bound SecA. This fraction of SecA is firmly associated with the membrane as it is resistant to extraction with the chaotropic agent urea, and appears to be anchored by SecYEG rather than by lipids. Topology analysis of this membrane-associated form of SecA indicates that it exposes a carboxy-terminal domain to the periplasmic face of the membrane.

Extragenic suppression of motA missense mutations of Escherichia coli.

The MotA and MotB proteins are thought to comprise elements of the stator component of the flagellar motor of Escherichia coli. In an effort to understand interactions among proteins within the motor, we attempted to identify extragenic suppressors of 31 dominant, plasmid-borne alleles of motA. Strains containing these mutations were either nonmotile or had severely impaired motility. Four of the mutants yielded extragenic suppressors mapping to the FlaII or FlaIIIB regions of the chromosome. Two types of suppression were observed. Suppression of one type (class I) probably results from increased expression of the chromosomal motB gene due to relief of polarity. Class I suppressors were partial deletions of Mu insertion sequences in the disrupted chromosomal motA gene. Class I suppression was mimicked by expressing the wild-type MotB protein from a second, compatible plasmid. Suppression of the other type (class II) was weaker, and it was not mimicked by overproduction of wild-type MotB protein. Class II suppressors were point mutations in the chromosomal motB or fliG genes. Among 14 independent class II suppressors characterized by DNA sequencing, we identified six different amino acid substitutions in MotB and one substitution in FliG. A number of the strongest class II suppressors had alterations of residues 136 to 138 of MotB. This particular region within the large, C-terminal periplasmic domain of MotB has previously not been associated with a specific function. We suggest that residues 136 to 138 of MotB may interact directly with the periplasmic face of MotA or help position the N-terminal membrane-spanning helix of MotB properly to interact with the membrane-spanning helices of the MotA proton channel.

C-terminal half of Salmonella enterica WbaP (RfbP) is the galactosyl-1-phosphate transferase domain catalyzing the first step of O-antigen synthesis.

We previously showed that the product of the wbaP gene of Salmonella enterica serovar Typhimurium has two functions: it is involved in the first step of O-antigen synthesis (the galactosyltransferase [GT] function) and in a later step (the T function), first thought to be the flipping of the O-antigen subunit on undecaprenyl pyrophosphate from the cytoplasmic face to the periplasmic face of the cytoplasmic membrane. We now locate two wbaP(T) mutations within the first half of the wbaP gene by sequencing. Both mutants retain GT activity, although one was a frameshift mutation resulting in a stop codon 10 codons after the frameshift to give an open reading frame containing only 138 of the 476 codons in WbaP. We also show that there is a secondary translation starting within the wbaP gene resulting in the synthesis of a polypeptide with GT activity. These results indicate that the N- and C-terminal halves of WbaP are the T and GT functional domains, respectively. We now propose that the T block operates prior to the flippase function, probably at the release of undecaprenyl pyrophosphate-linked galactose from WbaP.

Display of β-lactamase on the Escherichia coli surface: outer membrane phenotypes conferred by Lpp′–OmpA′–β-lactamase fusions

Bacterial cell-surface exposure of foreign peptides and soluble proteins has been achieved recently by employing a fusion protein methodology. An Lpp'-OmpA(46-159)-Bla fusion protein has been shown previously to display the normally periplasmic enzyme beta-lactamase (Bla) on the cell surface of the Gram-negative bacterium Escherichia coli. Here, we have investigated the role of the OmpA domain of the tripartite fusion protein in the surface display of the passenger domain (Bla) and have characterized the effects of the fusion proteins on the integrity and permeability of the outer membrane. We show that in addition to OmpA(46-159), a second OmpA segment, consisting of amino acids 46-66, can also mediate the display of Bla on the cell surface. Other OmpA domains of various lengths (amino acids 46-84, 46-109, 46-128, 46-141 and 46-145) either anchored the Bla domain on the periplasmic face of the outer membrane or caused a major disruption of the outer membrane, allowing the penetration of antibodies into the cell. Detergent and antibiotic sensitivity and periplasmic leakage assays showed that changes in the permeability of the outer membrane are an unavoidable consequence of displaying a large periplasmic protein on the surface of E. coli. This is the first systematic report on the effects that cell surface engineering may have on the integrity and permeability properties of bacterial outer membranes.

Depletion of anionic phospholipids has no observable effect on the anchoring of penicillin binding protein 5 to the inner membrane of Escherichia coli

Escherichia coli penicillin-binding protein 5 (PBP5) is anchored to the periplasmic face of the inner membrane via a C-terminal amphiphilic α-helix. The results of washing experiments have suggested an electrostatic contribution to the anchoring mechanism which may involve the cationic region of the C-terminal α-helix. Similarities between this anchor domain and some surface active agents, such as melittin, suggest that the cationic region of the PBP5 anchor may require the presence of anionic phospholipids for membrane interaction. Washing experiments performed on membranes of HDL11, an E. coli mutant in which the expression of the major anionic phospholipids is under lac control, found no such requirement. The results are discussed in relation to the hypothesis that the cationic region may interact with other sources of negative charge, possibly arising from a PBP complex.

Depletion of anionic phospholipids has no observable effect on the anchoring of penicillin binding protein 5 to the inner membrane of Escherichia coli.

Escherichia coli penicillin-binding protein 5 (PBP5) is anchored to the periplasmic face of the inner membrane via a C-terminal amphiphilic alpha-helix. The results of washing experiments have suggested an electrostatic contribution to the anchoring mechanism which may involve the cationic region of the C-terminal alpha-helix. Similarities between this anchor domain and some surface active agents, such as melittin, suggest that the cationic region of the PBP5 anchor may require the presence of anionic phospholipids for membrane interaction. Washing experiments performed on membranes of HDL11, an E. coli mutant in which the expression of the major anionic phospholipids is under lac control, found no such requirement. The results are discussed in relation to the hypothesis that the cationic region may interact with other sources of negative charge, possibly arising from a PBP complex.

Yersinia spp. HMWP2, a cytosolic protein with a cryptic internal signal sequence which can promote alkaline phosphatase export.

The iron starvation-induced, 2,042-amino-acid protein HMWP2 of Yersinia enterocolitica has two internal hydrophobic segments which might promote its export and association with the cytoplasmic membrane. To determine whether part of HMWP2 could be exported beyond the periplasmic face of the cytoplasmic membrane, we used TnphoA mutagenesis to construct 10 hybrid proteins in which periplasmic alkaline phosphatase (PhoA) was fused to the end of C-terminally truncated HMWP1 (at amino acid positions 1751 and 1753 two independent isolates]) had high alkaline phosphate activity (close to that of the native enzyme), both in Escherichia coli and in Y. pseudotuberculosis, indicating that the PhoA segment of the hybrid reached the periplasm. Deletion studies showed that the export signal resides in the second hydrophobic segment of HMWP2. This result would be compatible with the topology of the protein in the cytoplasmic membrane predicted from the distribution of charged amino acids at either end of the two hydrophobic segments. However, two hybrids in which the junction was even further toward the C terminus of HMMWP2 (at positions 1793 and 1999) had only weak alkaline phosphatase activity, suggesting that the predicted topology is incorrect. The location of HMWP2 was therefore determined by subcellular fractionation. The results indicate that HMPW2 is mainly cytoplasmic, consistent with its presumed role in the ATP-dependent, nonribosomal synthesis of an unknown peptide. We propose that the high alkaline phosphatase activity associated with some of the HMWP-2-PhoA hybrids results from the unmasking of the cryptic export signal activity in the second hydrophobic segment of HMPW2.

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.

Membrane topology of multidrug resistance protein expressed in Escherichia coli. N-terminal domain.

Expression of eukaryotic polytopic membrane proteins in Escherichia coli could provide an invaluable system for structure-function studies. Recently, the functional expression of a mouse multidrug resistance protein (Mdr1) in E. coli was described (Bibi, E., Gros, P., and Kaback, H. R. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 9209-9213). In the present study, the phoA gene fusion approach has been utilized to determine the membrane topology of the N-terminal domain of Mdr. The results support the idea that the N-terminal half of Mdr contains six transmembrane helices (TMs). However, our observations suggest that the previously proposed TM4 (amino acid residues Thr214-Ala232) is located at the periplasmic face of the membrane. Alternatively, we propose that another stretch of amino acid residues (Leu243 (out) to Ile260 (in)) traverses the membrane. This assignment is indicated also by the following observations: 1) deletion of a segment containing the originally predicted TM4 (delta T214-K241) does not reverse the orientation of the Mdr-alkaline phosphatase hybrid that is located in the following cytoplasmic loop; 2) a "sandwich" chimera, in which alkaline phosphatase is inserted in-frame between amino acid residues Leu226 and Ser227, exhibits high alkaline phosphatase activity. The existence of an externally exposed hydrophobic domain in Mdr may have special structural and functional implications, and these may also be relevant to other members of the ABC superfamily.

Involvement of the galactosyl-1-phosphate transferase encoded by the Salmonella enterica rfbP gene in O-antigen subunit processing.

rfbT of Salmonella enterica LT2 was previously thought, together with rfaL, to be involved in the ligation of polymerized O antigen to core-lipid A, and three mutants were known. We report the mapping of the mutations to rfbP, the galactosyl-1-phosphate transferase gene, which is now shown to encode a bifunctional protein. The mutations which have the former rfbT phenotype are referred to as rfbP(T). We also show that rfbP(T) mutants are not blocked in the ligation step as previously believed but in an earlier step, possibly in flipping the O-antigen subunit on undecaprenyl pyrophosphate from the cytoplasmic to periplasmic face of the cytoplasmic membrane.

Analysis of the membrane-anchoring properties of the putative amphiphilic -helical anchor at the C-terminus of Escherichia coli PBP 6

Penicillin-binding protein (PBP) 6 is anchored to the periplasmic face of the Escherichia coli inner membrane. Analysis of the C-terminal 20 amino acids of PBP 6 implies the presence of a C-terminal amphiphilic alpha-helical anchor comparable to that of PBP 5. A C-terminal deletion of PBP 6 was constructed; it resulted in the release of the protein from the inner membrane into the periplasm, thus confirming that this region is essential for anchoring. Treatment of E. coli K12 membrane vesicles with various reagents was used to probe the membrane-binding characteristics of both PBP 5 and PBP 6. The results indicate that, although the strength of membrane anchoring of PBP 6 is weaker than that of PBP 5, both modes of anchoring involve a large hydrophobic element and have similar membrane-binding characteristics. This is in agreement with the hypothesis that both proteins exhibit the same novel method of anchoring.

Membrane interaction of Escherichia coli penicillin binding protein 5 is modulated by the ectomembranous domain

E. coli penicillin binding protein (PBP) 5 is anchored to the periplasmic face of the inner membrane by a C-terminal domain which is predicted to form an amphipihilic α-helix. Here we show that the presence of a substrate analogue, benzyl penicillin, causes the protein to be converted from a membrane bound urea inaccessible form to a urea extractable form. If the anchor region is fused to the periplasmic protein, β-lactamase, the fusion protein becomes membrane bound but is unable to exhibit the changes in urea extractability which are observed with PBP5. We therefore conclude that although the C-terminus of PBP5 is sufficient to anchor the protein to the membrane surface the ectomembranous domain can affect the state of the anchor and in vivo changes in the state of anchoring may be related to enzyme activity.

Evidence for energy-dependent transposition of core lipopolysaccharide across the inner membrane of Salmonella typhimurium

The uncoupler 2,4-dinitrophenol blocks the final step of lipopolysaccharide assembly--transfer of O antigen from undecaprenyl pyrophosphate to core lipopolysaccharide--in intact Salmonella typhimurium but not in isolated membrane fractions. The O-antigen ligase enzyme is not inhibited by dinitrophenol in vitro, and core lipopolysaccharide synthesized in the presence of uncoupler in vivo is functional as acceptor of O antigen in vitro. The evidence strongly suggests that maintenance of proton motive force is required for transmembrane transposition of core lipopolysaccharide to the active site of O-antigen ligase at the periplasmic face of the inner membrane.

Binding of the immunity protein inactivates colicin M

Colicin M (Cma) displays a unique mode of action in that it inhibits peptidoglycan and lipopolysaccharide biosynthesis through interference with bactoprenyl phosphate recycling. Protection of Cma-producing cells by the immunity protein (Cmi) was studied. The amount of Cmi determined the degree of inhibition of in vitro peptidoglycan synthesis by Cma. In cells, immunity breakdown could be achieved by overexpression of the Cma uptake system. Full immunity was restored after raising the cmi gene copy number. In sphaeroplasts, Cmi was degraded by trypsin, but this could be prevented by the addition of Cma. The N-terminal end includes the only hydrophobic amino acid sequence of Cmi, suggesting a function in anchoring of Cmi in the cytoplasmic membrane. It is proposed that Cmi does not act catalytically but binds Cma at the periplasmic face of the cytoplasmic membrane, thereby resulting in Cma inactivation. Two other possible modes of colicin M immunity, interference of Cmi with the uptake of Cma, and interaction of Cmi with the target of Cma, were ruled out by the data.

Localization of the terminal steps of O-antigen synthesis in Salmonella typhimurium

Previous immunoelectron microscopic studies have shown that both the final intermediate in O-antigen synthesis, undecaprenol-linked O polymer, and newly synthesized O-antigenic lipopolysaccharide are localized to the periplasmic face of the inner membrane (C. A. Mulford and M. J. Osborn, Proc. Natl. Acad. Sci. USA 80:1159-1163, 1983). In vivo pulse-chase experiments now provide further evidence that attachment of O antigen to core lipopolysaccharide, as well as polymerization of O-specific polysaccharide chains, takes place at the periplasmic face of the membrane. Mutants doubly conditional in lipopolysaccharide synthesis [kdsA(Ts) pmi] were constructed in which synthesis of core lipopolysaccharide and O antigen are temperature sensitive and mannose dependent, respectively. Periplasmic orientation of O antigen:core lipopolysaccharide ligase was established by experiments showing rapid chase of undecaprenol-linked O polymer, previously accumulated at 42 degrees C in the absence of core synthesis, into lipopolysaccharide following resumption of core formation at 30 degrees C. In addition, chase of the monomeric O-specific tetrasaccharide unit into lipopolysaccharide was found in similar experiments in an O-polymerase-negative [rfc kdsA(Ts) pmi] mutant, suggesting that polymerization of O chains also occurs at the external face of the inner membrane.

Minimal deletion of amino acids from the carboxyl terminus of the B subunit of heat-labile enterotoxin causes defects in its assembly and release from the cytoplasmic membrane of Escherichia coli.

Minimal alterations at the carboxyl terminus of the B subunit (EtxB) of heat-labile enterotoxin from Escherichia coli were found to have a marked effect on the assembly and release of this polypeptide into the periplasm. Nine mutant EtxB polypeptides were obtained by genetic manipulation of the 3'-end of the etxB gene using Bal31 nuclease digestion and codon substitution. A correlation was observed between the magnitude of the changes introduced at the carboxyl terminus and the extent to which the mutant polypeptides were defective in assembly and release. Some of the mutant B subunits, exemplified by those in which the last 2 amino acids had been deleted or in which the last 4 residues had been replaced by three different ones, were found to be only partially defective, with a proportion being associated with the periplasmic face of the cytoplasmic membrane and the remainder being exported to the periplasm. The portion associated with membranes was detected as monomers on sodium dodecyl sulfate-polyacrylamide gels, whereas the portion exported to the periplasm were detected as assembled oligomers. We conclude that the last few amino acids at the carboxyl terminus of EtxB exert a profound influence on the assembly and release of the B subunit from the cytoplasmic membrane during export in E. coli.