HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 39, 30202-30210, September 29, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/39/30202    most recent M005515200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Guédin, S. Articles by Jacob-Dubuisson, F. Search for Related Content PubMed PubMed Citation Articles by Guédin, S. Articles by Jacob-Dubuisson, F. Social Bookmarking                      What's this?

Novel Topological Features of FhaC, the Outer Membrane Transporter Involved in the Secretion of the Bordetella pertussis Filamentous Hemagglutinin^*

Sandrine Guédin, Eve Willery, Jan Tommassen^§, Emmanuelle Fort, Hervé Drobecq^¶, Camille Locht, and Françoise Jacob-Dubuisson

From INSERM U447, IBL, Institut Pasteur de Lille, 1 rue Calmette, 59019 Lille Cedex, France, the ^§ Department of Molecular Cell Biology, University of Utrecht, Padualaan 8, 3584 CH Utrecht, The Netherlands, and ^¶ CNRS UMR 8525, IBL, Institut Pasteur de Lille, 1 rue Calmette, 59019 Lille Cedex, France

Received for publication, June 23, 2000, and in revised form, July 19, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Many pathogenic Gram-negative bacteria secrete virulence factors across the cell envelope into the extracellular milieu. The secretion of filamentous hemagglutinin (FHA) by Bordetella pertussis depends on the pore-forming outer membrane protein FhaC, which belongs to a growing family of protein transporters. Protein alignment and secondary structure predictions indicated that FhaC is likely to be a -barrel protein with an odd number of transmembrane -strands connected by large surface loops and short periplasmic turns. The membrane topology of FhaC was investigated by random insertion of the c-Myc epitope and the tobacco etch virus protease-specific cleavage sequence. FhaC was fairly permissive to short linker insertions. Furthermore, FhaC appeared to undergo conformational changes upon FHA secretion. Surface detection of the inserted sequences indicated that several predicted loops in the C-terminal moiety as well as the N terminus of the protein are exposed. However, a large surface-predicted region in the N-terminal moiety of FhaC was inaccessible from the surface. In addition, the activity and the stability of the protein were affected by insertions in that region, indicating that it may have important structural and/or functional roles. The surface exposure of the N terminus and the presence of an odd number of -strands are novel features for -barrel outer membrane proteins.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Protein secretion into the extracellular milieu is a complex process, since the exoproteins must be transported across biological membranes. The cell envelope of Gram-negative bacteria adds further complexity to protein secretion, because of the presence of an outer membrane in addition to the cytoplasmic membrane. Various protein secretion pathways have been developed by these organisms. Newly synthesized exoproteins devoid of a cleavable N-terminal signal peptide take Sec-independent pathways, mostly the ABC transporter-based type I system (1-3), or the host contact-dependent type III system (4-6). In contrast, exoproteins synthesized with a cleavable N-terminal signal peptide cross the cell envelope in two steps (7). They are first translocated across the cytoplasmic membrane via the ubiquitous Sec machinery and reach the periplasm following proteolytic removal of their signal peptides. They are then translocated across the outer membrane by one of various systems called the terminal branches of the general export pathway. A widely used system of this kind is the type II secretory pathway, a multiprotein cell envelope machinery (7).

Bordetella pertussis, the whooping cough agent, secretes several toxins, enzymes, and adhesins essential to its pathogenicity (8). In particular, its main adhesin, filamentous hemagglutinin (FHA), is secreted in large quantities into the extracellular milieu (9). The highly efficient secretion pathway of FHA requires a single specific accessory protein, FhaC, located in the outer membrane (10, 11). FHA has a cleavable N-terminal signal peptide and is likely to cross the cytoplasmic membrane via the Bordetella Sec machinery (12). It then transits through the periplasm and interacts specifically with FhaC to reach the cell surface. FHA most likely crosses the outer membrane in an extended conformation and acquires its tertiary structure at the cell surface, following an extensive C-terminal proteolytic maturation (13-15).

FhaC belongs to a family of accessory outer membrane proteins involved in the secretion of hemolysins or adhesins in various Gram-negative pathogens (10, 16-19). FHA shares with these hemolysins and adhesins N-proximal motifs involved in secretion and hypothesized to interact with FhaC (20-22). Thus, FhaC and homologs form a family of outer membrane transporters, which probably represents a distinct terminal branch of the general export pathway. Despite their homology, FhaC and its hemolysin-secreting relatives are not exchangeable, indicating the molecular specificity of these systems (22). Both FhaC and its Serratia marcescens homolog ShlB have recently been shown to form sugar- or ion-permeable pores in artificial membranes (23, 24). Therefore, our current model on the mode of action of FhaC is that it forms specific channels in the outer membrane for the outward translocation of FHA.

According to circular dichroism analyses, FhaC contains approximately 40% -structure (24). Furthermore, its sequence is rich in amphipathic stretches that could form a -barrel in the outer membrane. In this study, a working model for the topology of FhaC was constructed. It was tested by assessing the accessibility of inserted recognition sites for an antibody and a protease. Our results indicate that FhaC probably forms a transmembrane -barrel with an unprecedented uneven number of strands such that the N terminus of the protein is extracellular, whereas its C terminus is periplasmic.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Plasmid Construction-- The unique BamHI restriction site of pEC24 (13) was eliminated by treatment with Klenow enzyme following linearization of the plasmid. The resulting pEC24BamHI was digested using the following restriction enzymes that generate blunt ends and cut within fhaC: Eco47III, NruI, EcoRV, NaeI, PvuII, XmnI, HaeIII, BstUI, RsaI, MspAI, ScaI, EheI, BsaAI, or HincII. Partial restrictions were performed in the presence of ethidium bromide (25-75 µg/ml final concentration) (25). A nonphosphorylated palindromic 6-mer BamHI linker, 5'-GGATCC-3', was ligated with the linearized plasmid. The positions of BamHI in fhaC were determined by DNA sequencing with several primers internal to fhaC, using the ABI Prism 377 DNA Sequencer from Perkin-Elmer. The resulting plasmids were called pFccx (where x represents the position of the codon with a BamHI insert). In addition, BamHI sites were introduced in positions 125 and 503 by site-directed mutagenesis using the Altered Sites® II kit from Promega (Charbonnières, France) with the oligonucleotides 5'-CGTGGTGGATGGCGGATCCGTGCTCAAGCTGAAG-3' and 5'-CATCCCGACGCGGGATCCCGCACGATACGCATG-3', respectively (the BamHI sites are underlined). The sequence of fhaC was verified by DNA sequencing. In position 532, a BamHI site was inserted by a double polymerase chain reaction using pEC24 as a template and the oligonucleotides 5'-GTGGGCGCGGATGTCGGAGCGCTCA-3' and 5'-CGCGCCGGATCCCACGGGCTTCGAGTACGTGAACGA-3', and 5'-TAAGCTTGGCTGCAGGTCGACCCGG-3' and 5'-CCCGTGGGATCCGGCGCGCAGCCCGGCGGCGCACCA-3' as primers (the BamHI sites are underlined). Following a polymerase chain reaction with a mixture of the two amplicons as a template and the two outermost oligonucleotides as primers, the new amplicon was restricted by Eco47III and NotI and cloned into pEC24BamHI to replace the corresponding fhaC fragment. The inserted fragment was sequenced entirely. For the construction of the pMycx series, three pairs of nonphosphorylated oligonucleotides coding for the c-Myc epitope, EQKLISEEDL, in the three reading frames were used: 5'-GATCCGAACAGAAGCTGATCTCGGAAGAGGATCTGGGCCCCG-3' and 5'-GATCCGGGGCCCAGATCCTCTTCCGAGATCAGCTTCTGTTCG-3'; 5'-GATCCGGAACAGAAGCTGATCTCGGAAGAGGATCTGGGCCCG-3' and 5'-GATCCGGGCCCAGATCCTCTTCCGAGATCAGCTTCTGTTCCG-3'; and 5'-GATCGGAGAACAGAAGCTGATCTCGGAAGAGGATCTGGGCCCACG-3' and 5'-GATCCGTGGGCCCAGATCCTCTTCCGAGATCAGCTTCTGTTCTCC-3' (the ApaI sites used to screen the recombinant clones are underlined). The appropriate double-stranded linkers were ligated with various BamHI-restricted pFccx, to generate the corresponding in-frame c-Myc insertions. Similarly, for the construction of the pTEVx series, three pairs of oligonucleotides coding for the cleavage site of the tobacco etch virus (TEV) protease, ENLYFQG, were used: 5'-GATCCAGGAAAACCTGTACTTCCAGGGGGCCCGG-3' and 5'-GATCCCGGGCCCCCTGGAAGTACAGGTTTTCCGT-3'; 5'-GATCTGGAAAACCTGTACTTCCAGGGGGCCCGG-3' and 5'-GATCCCGGGCCCCCTGGAAGTACAGGTTTTCCA-3'; and 5'-GATCCGAAAACCTGTACTTCCAGGGGGCCCGTG-3' and 5'-GATCCACGGGCCCCCTGGAAGTACAGGTTTTCG-3'. The sequence and orientation of each insert were determined by DNA sequencing. For expression of the fhaC derivatives in B. pertussis, selected pFccx, pMycx, and pTEVx were restricted by PvuI and SacI. The resulting fragments corresponding to fhaC derivatives were cloned into pFJD16 (24) to replace the wild type fhaC fragment, resulting in pFcbx, pMycbx, or pTEVbx plasmids. Deletions in fhaC were obtained by combining the 5' portion of the gene from a given pFccx with the 3' portion of fhaC from a given pFccy with BamHI in the same reading frame further down the gene, resulting in internal truncations of fhaC from codons x to y. Strains and plasmids used are listed in Table I.

Determination of the Activity and Heat Modifiability of FhaC Derivatives-- The activities of the FhaC derivatives in Escherichia coli were determined as described (11, 13). Culture supernatants were analyzed by immunoblotting using anti-Fha44 chicken IgY (12). To determine the activities of FhaC derivatives in B. pertussis, the fhaC-deficient BPEC strain (13) was transformed by the pFcbx, pTEVbx, or pMycbx series, and complementation was assessed by the restoration of FHA secretion. Heat modifiability of the mutant proteins was detected as described (24).

Immunoblotting-- E. coli UT5600 producing c-Myc-FhaC or TEV-FhaC derivatives were grown to an optical density at 600 nm (A₆₀₀) of 1, resuspended in volume of 20 mM Hepes (pH 7.4), 10 µg/ml DNase, 0.5 mg/ml 4-(2-aminoethyl)benzenesulfonyl fluoride, and lysed by a passage in a French pressure cell. Total membranes were pelleted by a centrifugation for 1 h at 50,000 × g at 12 °C. The preparations were analyzed by SDS-PAGE and immunoblotting. For dot immunoblotting, 30 µl of total membranes were spotted onto a nitrocellulose membrane, which was incubated with the anti-c-Myc 9E10 monoclonal antibody (mAb) (Roche Molecular Biochemicals). The rat antiserum against FhaC was prepared by Eurogentec (Liège, Belgium) and purified by chromatography on CNBr-activated Sepharose saturated with an E. coli lysate. The anti-porin antiserum was raised in a rabbit. The immunoblots were developed using alkaline phosphatase-conjugated secondary antibodies followed by the addition of nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate if the primary antibodies were polyclonal and horseradish peroxidase-conjugated secondary antibodies and the Chemiluminescence Reagent Plus kit from NEN Life Science Products if the primary antibody was monoclonal, unless otherwise indicated.

Detection of the c-Myc Epitope on Intact Cells by Flow Cytometry-- 1-ml culture aliquots of E. coli UT5600 producing each of the c-Myc-FhaC derivatives grown to an A₆₀₀ of 1 were centrifuged to harvest the cells. The cells were resuspended in phosphate-buffered saline supplemented with 100 mM EDTA (26) and incubated overnight at 4 °C with the anti-c-Myc mAb. The addition of EDTA to the cells appeared to increase the percentage of labeled cells, while it did not expose epitopes that were shielded in its absence (not shown). After three washes with phosphate-buffered saline, the cells were incubated with anti-mouse fluorescein isocyanate-conjugated antibodies for 1 h at room temperature. Then three washes with phosphate-buffered saline were performed before the detection of fluorescent cells in a flow cytometer (FACSort, Becton Dickinson, San Jose, CA) at an excitation wavelength of 488 nm and an emission wavelength of 530 nm.

TEV Protease Digestions-- For digestions on intact cells, E. coli UT5600 producing the TEV-FhaC derivatives with or without Fha44, were grown to an A₆₀₀ of 1 at 600 nm. When required, 1 mM isopropyl--D thiogalactopyranoside was added to the cultures to induce Fha44 synthesis for 1 h. Cells from 1-ml aliquots were harvested by centrifugation and resuspended into 1 ml of 50 mM Tris-HCl (pH 8.0), 0.5 mM EDTA. For proteolysis, 10 mM dithiothreitol and 100 units of TEV protease (Life Technologies, Inc.) were added to 80 µl of cells, and the digestion was performed at 30 °C for 4 h (27). The reaction was stopped by the addition of 12% trichloroacetic acid, and the samples were analyzed by SDS-PAGE followed by immunoblotting using the anti-FhaC antibodies.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Topology Model of FhaC-- To construct a working model of the topology of FhaC, the following approach was used (28). First, the amino acid sequence of FhaC was aligned with those of six homologs, i.e. HpmB, ShlB, HhdB, EthB, HMW1B, and HxuB (data not shown). This alignment was used to identify hypervariable regions, which usually correspond to the cell surface-exposed domains of outer membrane proteins. All sequences were used to identify potential -turns, using the criteria described by Paul and Rosenbusch (29), as adapted for -sheet proteins. These turns are predicted to be surface-exposed or periplasmically exposed. The sequences were then scanned for putative -strands consisting of stretches of approximately 10 residues containing nonpolar amino acids (according to the normalized consensus scale of Eisenberg (30)) at every second position, which could form a -sheet configuration. Putative -strands with flanking aromatic residues were favored wherever possible because rings of aromatic residues surround the -barrels on both sides of the membrane in all outer membrane proteins of which the structures have been solved. The vast majority of bacterial outer membrane proteins contain a consensus sequence at the C terminus, consisting of a C-terminal phenylalanine preceded at every second position by four nonpolar amino acids (31). This consensus sequence is also present in FhaC and its homologs. In all outer membrane proteins of which the structures have been solved, this segment of the protein forms a membrane-spanning -strand with an N_out-C_in topology. Hence, we assume that the same is true for the C-terminal segment of FhaC. Model building then proceeded from the C terminus toward the N terminus reconciling the sequence features described above with features indicating potential short turns at the periplasmic side, long loops on the cell surface, and transmembrane strands of 6-12 amino acids. Potential conservation of these features in the homologs of FhaC was demanded. The resulting topology model of FhaC is depicted in Fig. 1.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Predicted topology of FhaC in the outer membrane. The prediction was established based on sequence analyses and alignment of FhaC and six homologs. FhaC was predicted to have 19 amphipathic -strands and to form a -barrel in the outer membrane. The putative  strands (b1 to b19) are connected by large loops at the cell surface (L1 to L10) and short turns on the periplasmic side. The N terminus is predicted to be surface-exposed. Amino acid residues are numbered according to their position in FhaC. The residues corresponding to the codons in which a BamHI linker was inserted are numbered and shown in italic type. The oligonucleotides coding for two types of peptides, the TEV cleavage site (represented by black squares) or the c-Myc epitope (black circles), were inserted into the BamHI sites of several mutants.

Construction of a Library of Insertion Mutants-- The topology of FhaC was studied by insertion of epitopes within the protein. First, a collection of mutants was generated by random BamHI linker insertions into fhaC (25). Each of the 48 clones contains a single linker (Lnk) insertion in fhaC, the position of which is given by the number of the codon within or after which the unique BamHI site was inserted. The numbering corresponds to mature FhaC (Fig. 1). Since FhaC is functional in E. coli for the secretion of an 80-kDa FHA derivative named Fha44 (11), we used this expression system for most of the following experiments. The 48 mutant alleles were expressed from pEC24, an fhaC-bearing plasmid for the production of FhaC in E. coli UT5600 at a level similar to that in B. pertussis (13). All 48 Lnk-FhaC derivatives were detected in the membranes of E. coli, although some of them at a somewhat lower level than wild type FhaC (Fig. 2). Their activities were analyzed by assessing the secretion of Fha44. Some Lnk-FhaC derivatives did not secrete Fha44 into the culture supernatant, suggesting that their function may be affected by the 2-amino acid residue insert. A few other mutant proteins showed a reduced secretion activity (Fig. 2). In agreement with the topology model, several inactive Lnk-FhaC derivatives (positions 111, 243, 288, 314, 430, 486, and 527) had an insert within or very close to a predicted transmembrane segment, which could disrupt the structure of the protein. However, Lnk125-, Lnk232-, Lnk513-, Lnk550- and Lnk551-FhaC were active, although their inserts were also within predicted transmembrane segments. To detect very low levels of activity, the more sensitive assay of fhaC-deficient B. pertussis BPEC complementation was used for some mutants. Lnk33-, Lnk243-, and Link434-FhaC derivatives proved to be slightly active for FHA secretion in B. pertussis, whereas Lnk104-, Lnk134-, and Lnk462-FhaC were not (Fig. 2).

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Production of FhaC derivatives in E. coli. Total membranes of E. coli UT5600 cells producing each of the linker-FhaC derivatives were analyzed by SDS-PAGE followed by immunoblotting using anti-FhaC antibodies. Similar amounts of total protein were loaded in all lanes, as estimated by the amounts of OmpF/C (not shown). The activities of the FhaC derivatives were assessed in E. coli UT5600 coproducing Fha44 and each FhaC derivative. The relative levels of Fha44 secretion in the culture supernatants are given as follows. +, levels of secretion at least 10% of that mediated by wild type FhaC; +, low levels of secretion (<10% of wild type level); , absence of secretion. Black and open triangles represent detectable and no detectable FHA secretion activities in B. pertussis, respectively, while black and open circles represent detectable and no detectable heat-modifiable forms, respectively.

The lack of activity of mutant proteins with insertions outside predicted transmembrane segments may point to regions that are important for the structure or the function of FhaC. To determine whether their structure was affected, some of these inactive mutant proteins were tested for heat modifiability. This property is typical of -rich integral outer membrane proteins, including FhaC (24). Lnk307-FhaC, Lnk342-FhaC, and Lnk462-FhaC were heat-modifiable (Fig. 2), arguing against major conformational alterations. In contrast, Lnk104-FhaC, Lnk134-FhaC, Lnk314-FhaC, and Lnk527-FhaC were not heat-modifiable. Together, these results indicate that FhaC is fairly tolerant to short insertions.

Topology Analysis of FhaC Using c-Myc Epitope Insertion-- The Lnk-FhaC derivatives that retained full or partial activity or heat modifiability were used for epitope insertion. The c-Myc epitope, EQKLISEEDL, was inserted into various positions using the BamHI sites generated above. Envelopes of E. coli UT5600 producing the c-Myc-FhaC derivatives were analyzed by SDS-PAGE and immunoblotting with anti-FhaC antibodies (Fig. 3A). All of the c-Myc-FhaC derivatives were produced, except for c-Myc88-FhaC and c-Myc395-FhaC. However, insertion of another peptide at these positions did not interfere with the production of FhaC (see below), suggesting that the very sequence of the inserted epitope can affect the stability of the protein. The envelopes of E. coli UT5600 cells producing each c-Myc-FhaC were also analyzed by SDS-PAGE and immunoblotting with an anti-c-Myc mAb. Except for c-Myc88- and c-Myc395-FhaC, the epitope was detected in all FhaC derivatives (Fig. 3B). However, the relative levels of c-Myc reactivity of the mutant proteins did not correlate well with their levels of production. In addition, the electrophoretic mobilities of the FhaC derivatives were somewhat heterogeneous. These observations suggest that some proteins may not have been fully denatured or may have locally renatured during SDS-PAGE and immunoblotting.

View larger version (33K): [in this window] [in a new window]   Fig. 3.   Production of c-Myc-FhaC derivatives in E. coli A. Envelopes of E. coli UT5600 producing the c-Myc-FhaC derivatives were analyzed by SDS-PAGE followed by immunoblotting using an anti-porin antiserum (upper panel) or anti-FhaC antibodies (lower panel). The upper panel shows that similar amounts of proteins were loaded in all lanes. B, the same samples were analyzed by SDS-PAGE followed by immunoblotting using the anti-c-Myc mAb. The positions of the molecular mass standard proteins are shown at the left.

The activities of the FhaC derivatives were assessed by the secretion of Fha44 into E. coli culture supernatants. In addition, heat modifiability of the mutant proteins was analyzed to determine whether their structure was affected by the insertions. Fha44 secretion and heat modifiability correlated well in the sense that both were abolished in several cases. This suggested that c-Myc insertions in various parts of the protein, especially in the second part of the loop L1 and in the loop L2 probably disrupted its structure (Table II).

View this table: [in this window] [in a new window]   Table II Characterization of the c-Myc-FhaC derivatives E. coli UT5600 producing the c-Myc-FhaC derivatives were analyzed for the secretion activity of the mutant proteins (column 2), their heat modifiability (column 3), and the accessibility of the c-Myc epitope at the cell surface (column 4) and on cell envelopes (column 5). The activities were assessed by the levels of Fha44 secreted in culture supernatants. In column 2, + and + indicate levels of Fha44 secretion between 10 and 100% and <10% of that mediated by wild type FhaC, respectively, and  indicates the absence of Fha44 secretion. In column 3, + indicates that a proportion of the mutant protein was heat-modifiable, and  indicates the absence of a detectable heat-modifiable form. In column 4, numbers in boldface type indicate values significantly higher than the negative control, which was taken as the background fluorescence of UT5600(pEC24). In column 5, + and ± indicate high and low levels of immunoreactivity in dot blot, respectively, while  indicates the absence of reactivity.

The accessibility of the c-Myc epitope from the cell surface was tested by immunolabeling of intact cells and flow cytometry (Table II). The three inserts in the very N-terminal part of FhaC (first portion of L1, Fig. 1) were easily detected on intact cells. The surface exposure of the N-terminal region of FhaC is in good agreement with the predicted model and represents an unprecedented feature among -barrel outer membrane proteins.

The c-Myc insert was also detected in loops L8 and L9 of the C-terminal part of FhaC, suggesting that the corresponding regions are surface-exposed as predicted in the model. In addition, the epitope was detected when inserted within the predicted last transmembrane strand, but in this case the mutant protein was neither active nor heat-modifiable. This suggests that the epitope disturbed the structure of FhaC. In all of the other positions, the percentages of labeled cells were at background levels, which suggests that the epitope is not accessible to the antibody from the surface.

To determine whether some inserts were more accessible on isolated membranes, dot immunoblotting with the anti-c-Myc mAb was performed with envelopes of E. coli UT5600 producing the c-Myc-FhaC derivatives (Fig. 4). As expected, the epitope was readily accessible on envelopes containing the proteins with surface-accessible c-Myc, i.e. those with insertions in the N-terminal region or in several surface-predicted loops of the C-terminal moiety of FhaC. The epitope was also detected at high (positions 73, 150, 171, 193, 383, and 551) and medium (positions 93, 104, 221, and 462) levels, respectively, for several inactive mutant proteins. However, except for Myc383-FhaC, it is likely that these mutant proteins were not properly inserted in the outer membrane, since they were also not heat-modifiable (Table II). The epitope was also detected in envelopes containing the active Myc125-FhaC, Myc206-FhaC, and Myc260-FhaC derivatives, although they were not detected on intact cells. This suggests that c-Myc may be oriented toward the periplasm in these three proteins. In the other sites (positions 79, 307, 322, 367, 416, and 532), there was virtually no detection of c-Myc. In these cases, the epitope might be somewhat buried in the protein or structurally constrained. Three insertion sites (positions 322, 367, and 416) are predicted to correspond to short periplasmic turns.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Detection of the c-Myc epitope on E. coli envelopes. Envelopes of E. coli UT5600 producing the Myc-FhaC derivatives were spotted onto a nitrocellulose membrane. Accessibility of the c-Myc epitope was determined by immunoblotting using the anti-c-Myc mAb. Envelopes of E. coli UT5600(pEC24) producing wild type FhaC (WT) were used as a negative control.

Altogether, insertion of c-Myc allowed to determine that three regions of FhaC are surface-exposed: the N-terminal portion of L1 and the L8 and L9 loops in the C-terminal part of the protein, consistent with the predicted model.

Topology Analysis of FhaC Using TEV Protease Cleavage Site Insertion-- To confirm and extend the above results on the topology of FhaC, a second tagging and detection system was used. The cleavage site for the TEV protease, ENLYFQG, was inserted into FhaC in order to allow for site-specific cleavage of the protein (27). The production, heat modifiability, and secretion activity of the TEV-FhaC derivatives were determined (Table III). Most of the mutant proteins were stably produced, although some of them in somewhat lower amounts than wild type FhaC (Table III). Interestingly, both TEV88-FhaC and TEV395-FhaC were produced, unlike the corresponding c-Myc derivatives. In contrast, TEV171-FhaC and TEV221-FhaC were hardly detected (not shown). The activity and heat modifiability of FhaC were in general less affected by the TEV inserts than by the c-Myc inserts (compare Tables II and III), possibly because the high negative charge of the c-Myc epitope may be more disruptive.

View this table: [in this window] [in a new window]   Table III Analysis of the TEV-FhaC derivatives E. coli UT5600 producing the TEV-FhaC derivatives were tested for the production (column 2), the secretion activity (column 3), and the heat modifiability (column 4) of each of the TEV-FhaCs. In column 2, + indicates amounts of FhaC similar to that of wild type FhaC, and + indicates low but detectable amounts of FhaC. In column 3, + and + indicate levels of Fha44 secretion between 10 and 100% and <10% of that mediated by wild type FhaC, respectively, and  indicates the absence of Fha44 secretion. In column 4, + indicates that a proportion of the mutant protein is heat-modifiable, and  indicates the absence of a detectable heat-modifiable form; ND, not determined. Proteolytic digestion with the TEV protease on intact cells was performed (column 5). + indicates TEV-FhaC cleavage by exogenous TEV protease, and  indicates the absence of detectable cleavage products. The last column shows the calculated masses of the fragments, with the boldface numbers indicating those that were observed.

To test the accessibility of the TEV sequence on intact cells, E. coli UT5600 producing the TEV-FhaC derivatives were digested by exogenous TEV protease (Table III, Fig. 5). Proteins and proteolytic fragments were analyzed by immunoblotting using the anti-FhaC antibodies. Partial or total digestions were observed for 8 TEV-FhaC derivatives, in which the inserts were predicted to be surface-exposed (Fig. 5). For TEV26-, TEV47-, TEV383-, TEV395-, TEV434-, TEV462-, TEV503-, and TEV532-FhaC, immunoreactive bands were detected at the molecular masses expected for the larger fragments (Table III, Fig. 5). The smaller proteolytic fragments were not detected, probably because they were too small or were not recognized by the anti-FhaC antibodies. Alternatively, some proteolytic products could have undergone further proteolysis by other proteases (32).

View larger version (13K): [in this window] [in a new window]   Fig. 5.   Proteolytic digestion of the TEV-FhaC derivatives on intacts cells. Intact E. coli UT5600 cells producing the TEV-FhaC derivatives were treated with exogenously added TEV protease. The proteins were analyzed by SDS-PAGE followed by immunoblotting using anti-FhaC antibodies. Only those TEV-FhaC derivatives for which a proteolytic product was observed are shown. In the case of the insertions at positions 462 and 503, cleavage by TEV protease was only observed when Fha44 was co-produced. The positions of the molecular mass standard proteins are shown at the left.

Interestingly, TEV462-FhaC and TEV503-FhaC were only digested when Fha44 was co-produced. This is an indication that additional portions of FhaC might become surface-exposed upon secretion of FHA, probably because of conformational changes in the protein. That loop L8 may have an important functional role is corroborated by the observation that even short insertions therein drastically affected the activity of FhaC but not its structure, since the protein remained heat-modifiable.

Proteolytic digestion by exogenously added TEV protease was not detected in large portions of FhaC. While some of these sites (206, 322, 367, 416) were predicted to be oriented toward the periplasm, many others were predicted to be part of surface loops (Fig. 1). We performed proteolysis on cell extracts containing some of the mutant proteins (TEV47-, TEV73-, TEV88-, TEV206-, TEV260-, and TEV416-FhaC) using the TEV enzyme. Only TEV47-FhaC was digested using this procedure (not shown), similar to the results of proteolysis on intact cells.

In summary, the results afforded by the TEV method confirm and extend those obtained above. The first portion of the L1 region, as well as the loops L7 to L10 are most likely surface-exposed, in keeping with the predicted model, while the location of the central portion of FhaC, in particular the large L2 region, remains uncertain.

In-frame Deletions in fhaC-- Since large surface-predicted regions in the N-terminal moiety of FhaC were refractory to detection from the cell surface, we attempted to address the possibility that their structure blocked the interactions between the probe molecules and their target sites. We reasoned that small deletions might help to locally disrupt the structure and consequently modify the accessibility of such regions. Several in-frame deletions were obtained in the N-terminal moiety of FhaC (Fig. 6). Seven of them were positioned in predicted loops, while two included putative transmembrane strands, which could be severely disruptive. To determine whether the FhaC derivatives were stably produced, cellular proteins of E. coli UT5600 expressing the mutant genes were analyzed by SDS-PAGE and immunoblotting using the anti-FhaC antibodies (Fig. 6). An immunoreactive band corresponding to FhaC was detected in all cell extracts, albeit in lower amounts than the wild type protein. The large (150-221) deletion affected the production of FhaC as well as two shorter ones (73-104 and 3-33). This latter observation indicates that such regions could be important for the stability of FhaC.

View larger version (21K): [in this window] [in a new window]   Fig. 6.   Effect of deletions on FhaC production and activity. A, the N-terminal portion of the protein sequence is represented schematically as a stippled line. The black arrows represent predicted transmembrane -strands, and the deletions are represented as lines below the sequence. The numbers correspond to amino acid residue positions in the protein. B, proteins of E. coli UT5600 expressing the nine mutant genes were analyzed by SDS-PAGE followed by immunoblotting using anti-FhaC antibodies. To improve sensitivity, the immunoblot was developed with horseradish peroxidase-conjugated antibodies and chemiluminescence detection. The activities of the FhaC derivatives were estimated using Fha44 secretion. + and  indicate low levels of secretion (<10% of that mediated by wild type FhaC) and the absence of secretion, respectively. The positions of the molecular mass standard proteins are shown at the left.

The Fha44 secretion activity of the deletion derivatives was also assessed (Fig. 6B). Only (221-228)-FhaC and (3-26)-FhaC were active to some extent, suggesting that the 221-228 and 3-26 regions are dispensable for activity. In contrast, deletions further into the N-terminal L1 region led to loss of Fha44 secretion, and similarly, the second predicted loop (L2) appeared important for activity, stability, and/or folding. Epitopes were inserted into the stably produced deletion mutant derivatives in an attempt to unmask additional surface-exposed regions of FhaC. However, the inserts severely destabilized the proteins (not shown). Thus, no additional information was obtained.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Exoprotein secretion in Gram-negative bacteria involves one of several families of transporter proteins in the outer membrane. FhaC is the prototype of one such family. We built a topology model of FhaC based on the alignment of seven homologous proteins. Using prediction methods similar to those used for porins, FhaC appears to consist of 19 amphipathic transmembrane -strands connected by short periplasmic turns and longer surface loops, forming a -barrel in the outer membrane. The high -sheet contents of FhaC assessed by circular dichroism (24) are consistent with this proposal. This model predicts several novel features for outer membrane proteins. Unlike porins (33-36), only the C-terminal moiety of FhaC appears to be rich in -strands, and an odd number of strands is predicted, placing the N terminus of the protein at the cell surface. In addition, two predicted extracellular regions of FhaC, with 108 (L1) and 65 (L2) residues, respectively, are larger than the surface loops found in porins or other outer membrane proteins, the structures of which have been solved (33-36).

In the absence of crystallographic data, the insertion of small epitopes is the method of choice to experimentally address the topology of outer membrane proteins. Such insertions are minimally disruptive of the protein structure and folding compared with gene fusions (37-41). Our two-step strategy consisted first in the random insertion of two amino acid residues along the protein, followed by the addition of larger, 11-15-residue inserts for detection purposes. FhaC proved to be fairly tolerant to the short insertions, since 35 derivatives out of 48 retained secretion activity. For the other derivatives, the absence of activity could result from a negative effect of the insert on the structure of FhaC. Alternatively, the insert could disrupt a region important for the recognition or the secretion of FHA. To distinguish between these possibilities, the criterion of heat modifiability was used to find additional "permissive" insertion sites (37) outside predicted transmembrane segments. The second step of the strategy then consisted in the insertion of the c-Myc epitope or the TEV protease cleavage sequence into permissive sites. In some sites, the protein may impose constraints on the conformation of the insert, which can affect recognition by the cognate antibody or protease (38). The use of two different inserts can minimize such events. In addition, the two modes of detection, using an antibody or a protease, are complementary. On the one hand, the antibody allows for detection of both surface-exposed epitopes on intact cells and periplasm-exposed epitopes on isolated cell envelopes. On the other hand, the significantly smaller, 27-kDa TEV protease can be advantageous for the detection of less accessible extracellular sites. In addition, treatment of cell envelopes with the TEV protease may sometimes also lead to the identification of periplasmically exposed loops. Conceivably, periplasmic production of the enzyme might lead to in vivo proteolysis of periplasmic loops of the target protein. In our hands, however, attempts to produce active periplasmic TEV protease have been unsuccessful.

The sets of results obtained by both methods are in good agreement. The experimental results appeared to confirm the topology prediction for the C-terminal moiety of FhaC, indicating that it is composed of transmembrane -strands connected by short periplasmic turns and larger extracellular loops. In all of the outer membrane proteins of known structure, the last transmembrane strand is oriented with the C terminus toward the periplasm. Our data are compatible with predicted loops L7 to L10 being accessible from the surface. It is thus plausible that the C terminus of FhaC is also oriented toward the periplasm. This is corroborated by the presence of the characteristic signature for outer membrane proteins in the C-terminal strand (31).

Interestingly, both techniques confirm that the N-terminal 50 residues of FhaC are surface-exposed and support the prediction that the -barrel consists of an odd number of strands, which is unusual for outer membrane proteins. The very N-terminal region of the protein appeared permissive to insertions and up to 20-residue deletions. Thus, it is not essential for the structure or the activity of FhaC. In contrast, inserts further into L1 and in L2 were not detected on the cell surface, and they affected the stability and secretion activity of the protein. This latter observation may indicate that these parts of the protein have structural and possibly also functional roles.

Our experimental approaches could not confirm that the predicted loop 2 is surface-exposed. Two different inserts in several positions of the same loop failed to be surface-accessible. Similar results were reported for other outer membrane proteins, and several hypotheses have been put forth to explain such results (39, 42). An epitope could remain undetected even if inserted into an extracellular loop, because it is masked from antibody or protease recognition. Major causes for steric hindrance could be the shielding by surface structures such as lipopolysaccharides (43) or the conformation of the loop itself. For instance, in the case of the extracellular loop 3 of PhoE, which folds back into the channel lumen, as many as three copies of the epitope were necessary to force surface exposure (44). Such a strategy would probably be too disruptive for FhaC, given the detrimental effect of single insertions in the problematic regions.

An alternative possibility is that these inserts lie within a periplasmic region. A periplasmic location for L2 implies that some of the predicted transmembrane -strands would actually not be in the membrane. In particular, b2 (Fig. 1) is perhaps not a transmembrane segment, because neither the c-Myc nor the TEV inserts in position 125 affected the activity of FhaC. In addition, b9 is not predicted with certainty, because it has an Arg on the hydrophobic, lipid-facing side. Such modifications to the predicted FhaC topology would set it further from porins but closer to those of the FhuA and FepA siderophore transporters, whose three-dimensional structures have been solved recently (45, 46). These proteins were initially thought to contain 32 and 29 transmembrane strands, respectively, based on predictions and epitope insertion studies (43, 47). Both are actually composed of two domains, a C-terminal -barrel domain consisting of 22 transmembrane anti-parallel -strands, and a 150-residue N-terminal globular domain, called plug or cork, that folds into the barrel lumen and interacts with its inner surface (45, 46). The lack of cleavage by the TEV protease in the FhaC loop L2 is consistent with this region forming such a highly structured domain possibly folded back into the channel. It is likely that protein-specific outer membrane transporters such as FhaC have some distinctive structural features. Notably, these proteins, unlike porins, should contact their cognate exoprotein from the periplasmic side.

The S. marcescens hemolysin-secreting ShlB is one of the FhaC homologs used in our alignment. FhaC and ShlB share 18% identity, but they are not functionally exchangeable (22). The topology of ShlB has recently been studied by epitope insertion (23). The FhaC and ShlB models are in reasonable agreement. However, Könninger et al. (23) have predicted an additional N-proximal -strand in ShlB, thus placing the 50 N-terminal residues in the periplasm. They have further suggested that this region might be involved in specific interactions with incoming ShlA (23). Here, we show that it is clearly not the case for FhaC. Both models and experimental results are in good agreement for the C-terminal half of the proteins. In addition, for ShlB epitopes inserted in L1 (corresponding to the C-terminal portion of the FhaC loop L1) or L2 (roughly corresponding to FhaC loop L2) could be detected from the surface by dot blotting of intact cells. This technique proved to be poorly reproducible in the case of FhaC. Furthermore, the structure and activity of FhaC appeared much more readily disrupted by insertions than those of ShlB. Despite these differences and only 18% identity between the two proteins, however, it is likely that FhaC and ShlB have similar topologies. FepA and FhuA also share fewer than 20% identical residues, yet they have similar structural arrangements (45, 46).

Finally, this study provides the first clue that conformational changes may take place in FhaC upon FHA secretion. In two surface-exposed loops of FhaC, site-specific proteolysis by the TEV protease became possible by the concomitant production of Fha44. We have also observed that in B. pertussis, immunodetection of FhaC from the cell surface is increased upon co-production with FHA.² One of two loops, in particular L8, may play an important functional role. Insertions in this loop greatly affected the activity of the protein but not its structure. It will be interesting to characterize the specific molecular interactions between FHA and FhaC, perhaps particularly loop 8, upon secretion.

	ACKNOWLEDGEMENTS

We thank M. Ehrmann for advice in the early stages of this work. J. Dubuisson is acknowledged for critical reading of the manuscript.

	FOOTNOTES

^* This work was supported in part by INSERM, the Institut Pasteur de Lille, and the Région Nord-Pas de Calais.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Supported by the Ministère de l'Education Nationale, de la Recherche et de la Technologie.

Researcher of the CNRS. To whom correspondence should be addressed. Tel.: 33 3 20 87 11 55; Fax: 33 3 20 87 11 58; E-mail francoise.jacob@pasteur-lille.fr.

Published, JBC Papers in Press, July 20, 2000, DOI 10.1074/jbc.M005515200

² S. Guédin and F. Jacob-Dubuisson, unpublished observations.

	ABBREVIATIONS

The abbreviations used are: FHA, filamentous hemagglutinin; PAGE, polyacrylamide gel electrophoresis; mAb, monoclonal antibody.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				J. Nesper, A. Brosig, P. Ringler, G. J. Patel, S. A. Muller, J. H. Kleinschmidt, W. Boos, K. Diederichs, and W. Welte Omp85Tt from Thermus thermophilus HB27: an Ancestral Type of the Omp85 Protein Family J. Bacteriol., July 1, 2008; 190(13): 4568 - 4575. [Abstract] [Full Text] [PDF]

				H. Li, S. Grass, T. Wang, T. Liu, and J. W. St. Geme III Structure of the Haemophilus influenzae HMW1B Translocator Protein: Evidence for a Twin Pore J. Bacteriol., October 15, 2007; 189(20): 7497 - 7502. [Abstract] [Full Text] [PDF]

				B. Clantin, A.-S. Delattre, P. Rucktooa, N. Saint, A. C. Meli, C. Locht, F. Jacob-Dubuisson, and V. Villeret Structure of the Membrane Protein FhaC: A Member of the Omp85-TpsB Transporter Superfamily Science, August 17, 2007; 317(5840): 957 - 961. [Abstract] [Full Text] [PDF]

				R. Balder, J. Hassel, S. Lipski, and E. R. Lafontaine Moraxella catarrhalis Strain O35E Expresses Two Filamentous Hemagglutinin-Like Proteins That Mediate Adherence to Human Epithelial Cells Infect. Immun., June 1, 2007; 75(6): 2765 - 2775. [Abstract] [Full Text] [PDF]

				L. Ran, F. Huang, M. Ekman, J. Klint, and B. Bergman Proteomic analyses of the photoauto- and diazotrophically grown cyanobacterium Nostoc sp. PCC 73102 Microbiology, February 1, 2007; 153(2): 608 - 618. [Abstract] [Full Text] [PDF]

				J. F. Stegmeier and C. Andersen Characterization of Pores Formed by YaeT (Omp85) from Escherichia coli J. Biochem., August 1, 2006; 140(2): 275 - 283. [Abstract] [Full Text] [PDF]

				N. K. Surana, A. Z. Buscher, G. G. Hardy, S. Grass, T. Kehl-Fie, and J. W. St. Geme III Translocator Proteins in the Two-partner Secretion Family Have Multiple Domains J. Biol. Chem., June 30, 2006; 281(26): 18051 - 18058. [Abstract] [Full Text] [PDF]