An inner membrane protein N-terminal signal sequence is able to promote efficient localisation of an outer membrane protein in Escherichia coli

To test the importance of N-terminal pre-sequences in translocation of different classes of membrane proteins, we exchanged the normal signal sequence of an Escherichia coli outer membrane protein, OmpF, for the pre-sequence of the inner membrane protein, DacA. The DacA-OmpF hybrid was efficiently assembled into the outer membrane in a functionally active form. Thus the pre-sequence of DacA, despite its relatively low hydrophobicity compared with that of OmpF, contains all the essential information necessary to initiate the translocation of OmpF to the outer membrane. Since processing of DacA was also shown to be dependent upon SecA we conclude that the initiation of translocation of this inner membrane polypeptide across the envelope occurs by the same mechanism as outer membrane and periplasmic proteins. The N-terminal 11 amino acids of mature OmpF, which in the hybrid are replaced by the N-terminal nine amino acids of DacA, carry no essential assembly signals since the hybrid protein is apparently assembled with equal efficiency to OmpF.     

Role of outer membrane barrier in efflux-mediated tetracycline resistance of Escherichia coli.

Accumulation of tetracycline in Escherichia coli was studied to determine its permeation pathway and to provide a basis for understanding efflux-mediated resistance. Passage of tetracycline across the outer membrane appeared to occur preferentially via the porin OmpF, with tetracycline in its magnesium-bound form. Rapid efflux of magnesium-chelated tetracycline from the periplasm was observed. In E. coli cells that do not contain exogenous tetracycline resistance genes, the steady-state level of tetracycline accumulation was decreased when porins were absent or when the fraction of Mg(2+)-chelated tetracycline was small. This is best explained by assuming the presence of a low-level endogenous active efflux system that bypasses the outer membrane barrier. When influx of tetracycline is slowed, this efflux is able to reduce the accumulation of tetracycline in the cytoplasm. In contrast, we found no evidence of a special outer membrane bypass mechanism for high-level efflux via the Tet protein, which is an inner membrane efflux pump coded for by exogenous tetA genes. Fractionation and equilibrium density gradient centrifugation experiments showed that the Tet protein is not localized to regions of inner and outer membrane adhesion. Furthermore, a high concentration of tetracycline was found in the compartment that rapidly equilibrated with the medium, most probably the periplasm, of Tet-containing E. coli cells, and the level of tetracycline accumulation in Tet-containing cells was not diminished by the mutational loss of the OmpF porin. These results suggest that the Tet protein, in contrast to the endogenous efflux system(s), pumps magnesium-chelated tetracycline into the periplasm. A quantitative model of tetracycline fluxes in E. coli cells of various types is presented.     

Specific localization of the lysis protein of bacteriophage MS2 in membrane adhesion sites of Escherichia coli.

Specific localization of the lysis (L) protein of bacteriophage MS2 in the cell wall of Escherichia coli was determined by immunoelectron microscopy. After induction of the cloned lysis gene, the cells were plasmolyzed, fixed, and embedded in either Epon or Lowicryl K4M. Polyclonal L-protein-specific antiserum was purified by preabsorption to membranes from cells harboring a control plasmid. Protein A-gold was used to label the protein-antibody complexes. Between 42.8% (Lowicryl) and 33.8% (Epon) of the label was found in inner and outer membranes, but 30.3% (Lowicryl) and 32.8% (Epon) was present mostly in clusters in the adhesion sites visible after plasmolysis. The remaining label (26.9 and 33.4%, respectively) appeared to be present in the periplasmic space but may also have been part of membrane junctions not visible because of poor contrast of the specimen. In contrast, a quite different distribution of the L protein was found in cells grown under conditions of penicillin tolerance, i.e., at pH 5, a condition that had previously been shown to protect cells from L-protein-induced lysis. At tolerant conditions, only 21.0% of the L protein was in the adhesion sites; most of the protein (68.2%) was found in inner and outer membranes. It is concluded that lysis of the host, E. coli, was a result of the formation of specific L-protein-mediated membrane adhesion sites.     

Characterization of the precursor of Serratia marcescens serine protease and COOH-terminal processing of the precursor during its excretion through the outer membrane of Escherichia coli.

The Serratia marcescens serine protease, which is directed by the gene encoding a precursor composed of a typical NH2-terminal signal sequence, a mature enzyme domain, and a large COOH-terminal domain, was excreted through the outer membrane of Escherichia coli. The precursor, with the expected molecular size (110 kilodaltons), was detected in an insoluble form in the periplasmic space of E. coli cells after induction with isopropyl-beta-D-thiogalactopyranoside of the expression of the gene under the control of the tac promoter. Upon membrane fractionation of the disrupted cells by sucrose density gradient centrifugation, the precursor was recovered from a fraction slightly heavier than the outer membrane fraction but not from the inner membrane fraction. Conversion of the precursor into the mature form, which was accompanied by its excretion into the medium, was observed even in the absence of de novo protein synthesis caused by the addition of chloramphenicol. The mutated gene product lacking all of the COOH-terminal domain was localized in the periplasmic space only and was not excreted into the medium. Additional mutant genes were generated by site-directed mutagenesis to test the role of some amino acids in the excretion of this protease in E. coli. The mutant protein with no protease activity because of the change of the catalytic residue Ser-341 to Thr was still excreted into the medium but with abnormal processing. Both self-processing and host-dependent processing of the precursor seem to be involved in the excretion of the mature enzyme. Replacement of the four Cys residues, two in the mature enzyme and two in the COOH-terminal domain, with Ser in different combinations caused a distinct or complete loss of excretion, suggesting that a certain conformation possibly formed via disulfide bonding was important for the excretion of the S. marcescens protease.     

Secretion into the culture medium of a foreign gene product from Escherichia coli: use of the ompF gene for secretion of human beta-endorphin.

The ompF gene codes for a major outer membrane protein of Escherichia coli. A plasmid was constructed in which the structural gene for human beta-endorphin is preceded by the upstream region of the ompF gene consisting of the promoter region and the coding regions for the signal peptide and the N terminus of the OmpF protein. When the plasmid was introduced into E. coli N99, and OmpF-beta-endorphin fused peptide was synthesized and secreted into the culture medium through both the cytoplasmic and outer membranes. The OmpF signal peptide was cleaved correctly during the secretion, indicating that the export of the fused protein across the cytoplasmic membrane was dependent on the signal peptide. The secretion into the culture medium was apparently selective. Neither beta-lactamase nor alkaline phosphatase (both are periplasmic proteins) appeared in the culture medium in significant amounts. The mode of passage of the fused peptide across the outer membrane is discussed.     

Regulation of envelope protein composition during adaptation to osmotic stress in Escherichia coli.

Adaptation to osmotic stress alters the amounts of several specific proteins in the Escherichia coli K-12 envelope. The most striking feature of the response to elevated osmolarity was the strong induction of a periplasmic protein with an Mr of 31,000. This protein was absent in mutants with lambda plac Mu insertions in an osmotically inducible locus mapping near 58 min. The insertions are likely to be in proU, a locus encoding a transport activity for the osmoprotectants glycine betaine and proline. Factors affecting the extent of proU induction were identified by direct examination of periplasmic proteins on sodium dodecyl sulfate gels and by measuring beta-galactosidase activity from proU-lac fusions. Expression was stimulated by increasing additions of salt or sucrose to minimal medium, up to a maximum at 0.5 M NaCl. Exogenous glycine betaine acted as an osmoregulatory signal; its addition to the high-osmolarity medium substantially repressed the expression of the 31,000-dalton periplasmic protein and the proU-lac+ fusions. Elevated osmolarity also caused the appearance of a second periplasmic protein (Mr = 16,000), and severe reduction in the amounts of two others. In the outer membrane, the well-characterized repression of OmpF by high osmolarity was observed and was reversed by glycine betaine. Additional changes in membrane composition were also responsive to glycine betaine regulation.     

A genetic engineering approach to study the mode of assembly of the OmpF porin in the envelope of E coli

Inducible hybrid genes encoding two large domains, a periplasmic domain consisting of the PhoS sequence and an outer membrane domain corresponding to various lengths of the OmpF mature sequence were constructed. The synthesized hybrid polypeptides are correctly processed during the early times of induction, their precursor forms being accumulated at later times. These hybrids restore sensitivity toward colicin A to ompF E coli B strain which suggests an outer membrane location. At least 2 of them are indeed localized in the outer membrane after immunogold labelling on ultrathin cryosections. Insertion of a hydrophobic sequence between PhoS and OmpF improves the trimerization and the assembly of the OmpF part. Only the hybrids presenting the last C-terminal 29 residues of OmpF are able to promote the colicin N killing action and to exhibit a trimeric conformation which is recognized by specific antibodies. Moreover, the deletion of the C-terminal region impairs the functional insertion of the OmpF domain; this indicates that the last membrane-spanning region of OmpF is necessary for the correct folding and orientation of the protein in the outer membrane.     

A morphologic study of filamentous phage infection of Escherichia coli using biotinylated phages

Using biotinylated phage (BIO-phages), we observed the infection of filamentous phages into Escherichia coli JM109 morphologically. BIO-phages and BIO-phage-derived proteins, mainly pVIII, were detected in E. coli by using the avidin-biotin-peroxidase complex method with electron microscopy. Infected cells revealed positive staining on the outer and inner membranes and in the periplasmic space. Some cells showed specific or predominant staining of the outer membrane, whereas others showed predominant staining of the inner membrane or equivalent staining of the outer and inner membranes. The periplasmic spaces in some infected cells were expanded and filled with reaction products. Some cells showed wavy lines of positive staining in the periplasmic space. BIO-phages were detected as thick filaments or clusters covered with reaction products. The ends of the infecting phages were located on the surface of cells, in the periplasmic space, or on the inner membrane. These findings suggest that phage major coat proteins are integrated into the outer membrane and that phages cause periplasmic expansion during infection.     

Release of outer membrane fragments from wild-type Escherichia coli and from several E. coli lipopolysaccharide mutants by EDTA and heat shock treatments.

EDTA-induced outer membrane losses from whole cells of wild-type Escherichia coli (O111:B4) and several lipopolysaccharide (LPS) mutants derived from E. coli K-12 D21 were analyzed. EDTA treatment induced losses of LPS (up to 40%), outer membrane proteins OmpA, OmpF/C, and lipoprotein, periplasmic proteins, and phosphatidylethanolamine. The extent of these releases was strain specific. Successively more EDTA was necessary to induce these losses from strains containing LPS with increasing polysaccharide chain length. An additional heat shock immediately following the EDTA treatment had no effect on LPS release, but it decreased the release of outer membrane proteins and reduced the leakage of periplasmic proteins, suggesting that the temporary increase in outer membrane "permeability" caused by Ca2+-EDTA treatment was rapidly reversed by the redistribution of outer membrane components, a process which is favored by a mild heat shock. The fact that the material released from E. coli C600 showed a constant ratio of lipoprotein, OmpA, and phosphatidylethanolamine at all EDTA concentrations tested suggests that the material is lost as specific outer membrane patches. The envelope alterations caused by EDTA did not result in cell lysis.     

Localization of the dicarboxylate binding protein in the cell envelope of Escherichia coli K12

Examination of the localization of the dicarboxylate binding protein (DBP) in the cell envelope of Escherichia coli K12 reveals that this protein is present on the cell surface, and also in the inner and outer regions of the periplasmic space. The cell surface DBP is release by treating the cells with EDTA. This protein can be surface labeled by lactoperoxidase radioiodination, and by diazo[125I]iodosulfanilic acid in whole cells. It also binds tightly, but not covalently, to lipopolysaccharide. The DBP located in the outer region of the periplasmic space is released when the outer membrane is dissociated by EDTA-osmotic shock treatment. The DBP located in the inner region of the periplasmic space is released only when the EDTA-osmotic shocked cells are subjected to lysozyme treatment. At the moment, it is not certain whether this protein is bound to or trapped by the peptidoglycan network. This protein cannot be surface labeled in whole cells or in EDTA-osmotic shock treated cells; and it is not associated with lipopolysaccharide. Analysis of transport mutants indicates that these DBP are coded by the same gene.     

A novel labeling approach supports the five-transmembrane model of subunit a of the Escherichia coli ATP synthase.

Cysteine mutagenesis and surface labeling has been used to define more precisely the transmembrane spans of subunit a of the Escherichia coli ATP synthase. Regions of subunit a that are exposed to the periplasmic space have been identified by a new procedure, in which cells are incubated with polymyxin B nonapeptide (PMBN), an antibiotic derivative that partially permeabilizes the outer membrane of E. coli, along with a sulfhydryl reagent, 3-(N-maleimidylpropionyl) biocytin (MPB). This procedure permits reaction of sulfhydryl groups in the periplasmic space with MPB, but residues in the cytoplasm are not labeled. Using this procedure, residues 8, 27, 37, 127, 131, 230, 231, and 232 were labeled and so are thought to be exposed in the periplasm. Using inside-out membrane vesicles, residues near the end of transmembrane spans 1, 64, 67, 68, 69, and 70 and residues near the end of transmembrane spans 5, 260, 263, and 265 were labeled. Residues 62 and 257 were not labeled. None of these residues were labeled in PMBN-permeabilized cells. These results provide a more detailed view of the transmembrane spans of subunit a and also provide a simple and reliable technique for detection of periplasmic regions of inner membrane proteins in E. coli.     

Thermosensitive omsA mutation of Escherichia coli that causes thermoregulated release of periplasmic proteins.

A mutant of Escherichia coli with a thermosensitive defect, possibly in the outer membrane (omsA mutant), was isolated from E. coli K-12 by mutagenization and selection for thermosensitivity and beta-lactam supersensitivity of growth. The mutant also showed very high sensitivity to other antibiotics, such as macarbomycin, midecamycin, rifampin, and bacitracin. The mutation was recessive to the wild type and was mapped at about 4 min on the E. coli chromosome between fhuA and metD. The mutation caused rapid release into the medium of periplasmic enzymes such as RTEM penicillinase but practically no cytoplasmic enzyme when cells grown at 30 degrees C were transferred to 37 or 42 degrees C. Electron microscopic observations showed many large double-layered vesicles attached to the surface of cells incubated at 42 degrees C. We conclude that the mutant had a mutation that caused a temperature-dependent defect in the outer membrane structure or its assembly (named an oms mutation). The omsA mutant may be useful for production of periplasmic proteins, which it releases into the culture medium on shift up of temperature.     

SurA assists the folding of Escherichia coli outer membrane proteins.

Many proteins require enzymatic assistance in order to achieve a functional conformation. One rate-limiting step in protein folding is the cis-trans isomerization of prolyl residues, a reaction catalyzed by prolyl isomerases. SurA, a periplasmic protein of Escherichia coli, has sequence similarity with the prolyl isomerase parvulin. We tested whether SurA was involved in folding periplasmic and outer membrane proteins by using trypsin sensitivity as an assay for protein conformation. We determined that the efficient folding of three outer membrane proteins (OmpA, OmpF, and LamB) requires SurA in vivo, while the folding of four periplasmic proteins was independent of SurA. We conclude that SurA assists in the folding of certain secreted proteins.     

Heterologous expression of Escherichia coli porin genes in Salmonella typhi Ty2: regulation by medium osmolarity, temperature and oxygen availability

Abstract Electrophoretic analysis of outer membrane proteins showed that Salmonella typhi OmpC expression is not reciprocally regulated relative to OmpF as described for Escherichia coli and S. typhimurium . When bacteria were grown in minimal media, both OmpC and OmpF were repressed as the osmolarity increased. However, in Luria broth, expression of OmpC was slightly induced by osmolarity up to 0.3 osmM. Plasmids bearing E. coli ompC-lacZ or ompF-lacZ gene fusions were studied for their expression in S. typhi and E. coli . Under anaerobic growth conditions, expression of ompC-lacZ in S. typhi was maximal at 0.16 osmM, while in E. coli expression was maximal at 0.7 osmM. ompF-lacZ expression was similarly repressed by medium osmolarity and anaerobiosis in both species. In contrast, a drastic difference in the regulation of OmpF by temperature was observed; at 37 °C ompF-lacZ expression was repressed in E. coli . while in S. typhi it was induced.     

Release of a chimeric protein into the medium from Escherichia coli using the C-terminal secretion signal of haemolysin.

Recently, we have identified a novel topogenic sequence at the C terminus of Escherichia coli haemolysin (HlyA) which is essential for its efficient secretion into the medium. This discovery has introduced the possibility of using this secretion system for the release of chimeric proteins from E. coli directly into the medium. We have now successfully fused this C-terminal signal to a hybrid protein containing a few residues of beta-galactosidase and the majority of the E. coli outer membrane porin OmpF lacking its own N-terminal signal sequence. We find that this chimeric protein is specifically translocated across the inner and outer membranes and is released into the medium. In addition, we have further localized the HlyA secretion signal to the final 113 amino acids of the C terminus. In fact, a specific secretion signal appears to reside at least in part within the last 27 amino acids of HlyA.     

Permeability of the boundary layers of Bdellovibrio bacteriovorus 109J and its bdelloplasts to small hydrophilic molecules.

Measurements of the sucrose-permeable and -impermeable volumes during Bdellovibrio bacteriovorus attack on Escherichia coli or Pseudomonas putida showed that the volume of the bdelloplast increased over that of the substrate cell. Although the pattern of the increase differed with the two organisms, the volumes reached maximum at about 60 min into the bdellovibrio growth cycle. By this time, the cytoplasmic membranes of the attacked cells were completely permeable to sucrose. The kinetics of increase in sucrosepermeable volumes were similar to the kinetics of attachment and penetration (Varon and Shilo, J. Bacteriol. 95:744-753, 1968). These data show that the original cytoplasmic and periplasmic compartmentalization of the substrate cell ceases to exist with respect to small hydrophilic molecules during bdellovibrio attack. In contrast, the effective pore size of the outer membrane of the substrate cell to small oligosaccharides remains unaltered during bdelloplast formation as was shown by direct measurements of its exclusion limits. The major porin protein of E. coli, OmpF, was recoverable from the bdelloplast outer membrane fraction until the onset of lysis. The Braun lipoprotein was removed from the bdelloplast wall early, and OmpA was lost in the terminal part of the bdellovibrio growth cycle.     

Unimolecular study of the interaction between the outer membrane protein OmpF from E. coli and an analogue of the HP(2–20) antimicrobial peptide

Little is known on antimicrobial peptide permeation through outer membrane channels in Gram-negative bacteria. Herein, we probed at a single-molecule level the interaction of two different peptides, magainin 2 and HPA3P with OmpF from E. coli. HPA3P is an analogue of the antimicrobial peptide HP(2–20) isolated from the N-terminal region of the Helicobacter pylori ribosomal protein. Our data show that the shorter and more charged HPA3P peptide is more accessible to the inner volume of the OmpF than magainin 2. We demonstrate the ability of HPA3P peptides to interact with OmpF in a voltage- and concentration-dependent manner, which does not rule out a novel mechanism by which such peptides could reach the periplasmic space of Gram-negative bacteria. Unexpectedly, we found that increasing the applied voltage led to an increase of the residence time of HPA3P peptide inside the pore, possibly reflecting electric field-induced changes in pore and peptide geometry.     

TolA central domain interacts with Escherichia coli porins.

TolA is an inner membrane protein with three domains: a transmembrane N-terminus and periplasmic central and C-terminal domains. The interaction of TolA with outer membrane porins of Escherichia coli was investigated. Western blot analyses of cell extracts with anti-TolA antibodies indicated that TolA forms high molecular weight complexes specifically with trimeric OmpF, OmpC, PhoE and LamB, but not with OmpA. The interaction of purified TolA domains with purified porins was also studied. TolA interacted with OmpF, PhoE and LamB porins via its central domain, but not with either their denatured monomeric forms or OmpA. Moreover, the presence or absence of lipopolysaccharides associated with trimeric porins did not modify the interactions. These results suggest that the specific interaction of TolA with outer membrane porins might be relevant to the function of Tol proteins.