This Article Abstract Full Text (PDF) Supplemental material Other Versions of this Article: JB.01480-06v1 189/5/1496    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Cescau, S. Articles by Wandersman, C. Search for Related Content PubMed PubMed Citation Articles by Cescau, S. Articles by Wandersman, C.

Probing the In Vivo Dynamics of Type I Protein Secretion Complex Association through Sensitivity to Detergents ^,

Unité des Membranes Bactériennes CNRS URA 2172,¹ Unité de Biologie Moléculaire du Gène chez les Extrêmophiles, Département de Microbiologie, Institut Pasteur, 75724 Paris Cedex 15, France²

Received 19 September 2006/ Accepted 27 November 2006

	ABSTRACT

Top Abstract Introduction Materials and Methods Results Discussion References

 
The Serratia marcescens hemophore is secreted by a type I secretion system consisting of three proteins: a membrane ABC protein, an adaptor protein, and the TolC-like outer membrane protein. Assembly of these proteins is induced by substrate binding to the ABC protein. Here we show that a hemophore mutant lacking the last 14 C-terminal amino acids is not secreted but rather interacts with the ABC protein and promotes a stable multiprotein complex. Strains expressing the transporter and the mutant protein are sensitive to detergents (sodium dodecyl sulfate [SDS]). TolC is trapped in the transporter jammed by the truncated substrate and therefore is not present at sufficient concentrations to allow the efflux pumps to expel detergents. Using an SDS sensitivity assay, we showed that the hemophore interacts with the ABC protein via two nonoverlapping sites. We also demonstrated that the C-terminal peptide, which functions as an intramolecular signal sequence in the complete substrate, may also have intermolecular activity and triggers complex dissociation in vivo when it is provided as a distinct peptide. The SDS sensitivity test on plates enables workers to study type I secretion protein association and dissociation independent of the secretion process itself.

	INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

 
Gram-negative bacteria possess several distinct pathways for secreting proteins across the two membranes which constitute the cell envelope. The type I secretion system transports proteins across both membranes without a periplasmic intermediate, via a mechanism which is independent of a cleavable N-terminal signal peptide.

The sequences, sizes, and functions of the proteins secreted by pathway I are very diverse, and these proteins include toxins, hydrolytic enzymes, such as proteases and lipases, hemophores, which scavenge heme to feed cells with iron, surface-attached S-layer proteins, etc. The proteins lack an N-terminal cleavable signal peptide but have a noncleavable secretion signal located in the last 50 to 60 amino acids. The secretion signals do not exhibit strong primary sequence identity, and in most cases they drive secretion through dedicated transporters. Nevertheless, type I secretion apparatuses have conserved features (6). They are all made up of three envelope proteins: an inner membrane ATP-driven transporter belonging to the ubiquitous ABC family, a membrane fusion protein (MFP) or adaptor also located in the inner membrane, and an outer membrane protein belonging to the TolC family (22). TolC homologues are characterized by a small ß-barrel inserted in the outer membrane and a large protruding domain forming a so-called periplasmic channel tunnel (12). While the genes encoding the secreted protein and inner membrane components of the translocator are generally clustered, genes encoding the outer membrane components are often unlinked and play a role in other cell functions, such as drug efflux (8, 25). Many systems, including toxin, protease, lipase, and hemophore exporters, are functional when they are reconstituted in Escherichia coli K-12. In E. coli K-12 the uropathogenic hemolysin of E. coli is secreted by a translocator consisting of the dedicated proteins HlyB (ABC protein) and HlyD (MFP) and the host protein TolC (17, 22). The Serratia marcescens hemophore (HasA) is secreted in E. coli K-12 both by its genuine apparatus consisting of HasD (ABC protein), HasE (MFP), and HasF (S. marcescens TolC analogue) and by a hybrid apparatus in which E. coli TolC replaces HasF (3). Genetic and biochemical studies performed mainly with hemolysin and hemophore secretion systems have demonstrated that each substrate interacts with its cognate ABC protein and that this interaction is essential for assembling the complete secretion apparatus (13, 20). Hemolysin binding to the HlyB-HlyD complex triggers TolC association with inner membrane components. Hemophore binding to HasD induces an ordered association of HasE and TolC. Experimental data have indicated that the interaction between the substrate and the ABC protein involves at least the C-terminal secretion signal, but the possibility that other interaction domains are involved has not been excluded (6). Since the secretion signal is synthesized last, it must remain accessible on the complete protein. In the case of hemophore secretion, the general chaperone SecB binds to HasA and maintains it in a secretion-competent state prior to transport (7). SecB slows folding and facilitates secretion of hemophores during the course of translation. When synthesis and secretion are uncoupled, the hemophore folds in the cytoplasm and becomes incompetent for secretion, a status which is not reversed by SecB overexpression (5).

In previous work, we observed that cytoplasmic accumulation of folded HasA inhibits secretion of newly synthesized HasA molecules (5). This strongly suggests that folded molecules still interact with the transporter in vivo, as demonstrated in vitro by hemin affinity copurification of HasA with the ABC protein HasD, which by itself does not bind to hemin (13).

The present work was undertaken in order to elucidate the molecular basis of inhibition of secretion by cytoplasmic folded polypeptides.

We demonstrate here that (i) the folded polypeptide interacts with the ABC protein, (ii) this interaction is mediated by regions other than the extreme C-terminal secretion signal, (iii) this interaction in the absence of the C terminus triggers stable in vivo TolC recruitment into the transporter, and (iv) TolC recruitment can be reversed by the C terminus alone when it is provided as a distinct polypeptide.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

 
Bacterial strains and media. Strains and plasmids used in this study are listed in Table 1. Strains were grown at 30°C or 37°C in rich medium (LB) or in minimal medium (M9) supplemented with vitamin B₁ and glycerol as the carbon source. When necessary, ampicillin was added at a concentration of 100 µg/ml, spectinomycin was added at a concentration of 50 µg/ml, tetracycline was added at a concentration of 10 µg/ml, and chloramphenicol was added at a concentration of 50 µg/ml. Isopropyl-ß-D-thiogalactopyranoside (IPTG) and arabinose were used to induce expression of genes under control of the P_lac and P_BAD promoters, respectively (Table 1). In assays for sensitivity, vancomycin (Sigma) was used at various concentrations (0 to 600 µg/ml), and novobiocin (Sigma) was used at concentrations ranging from 0 to 10 µg/ml.

pTrc99A-Cter-hasA was constructed in the same way as pTrc99A-hasA_1-174 using primers 3 and 4 for the PCR, and pTrc99A-Cm-Cter-hasA was constructed in the same way as pTrc99A-Cm-hasA_1-174.

pAM-hasA_1-174 was constructed by cloning the AatII-DraI fragment of pTrc99A-hasA1_-174 into pAM-hasA cut by the same enzymes.

pACYC184-Cm was constructed by deletion of the DraI-DraI fragment (339 bp) and religation.

pACYC184-Cter-hasA was constructed by cloning a PCR product containing the Cter-hasA coding sequence cut by the NcoI and BspEI enzymes into pACYC184 cut by the same enzymes. This PCR product was obtained with primers 7 and 8 using pBR322-Cter-hasA as the template (24).

pTrc99A-Cm-hasA_(133-174) was constructed by directed mutagenesis using primers 5 and 6 and pTrc99A-Cm-hasA as the template.

All PCR products were sequenced to check for the absence of mutation. The following primers were used for construction of the plasmids: primer 1, 5'-TTCACCATGGCATTTTCAGTCAATTATGAC-3'; primer 2, 5'-CGACTCTAGATCACGCCGTCGCCGCCGCCA-3'; primer 3, 5'-CATGCCATGGCCATGATTACGAATTTGCACCAGGT-3'; primer 4, 5'-CTAGTCTAGATCAGGCCGCCAGCAGTTCCG-3'; primer 5, 5'-AGGGTCATGATGGCGTGGTGGTGGGCGTGCAGCACGCCGAC-3'; primer 6, 5'-TCGGCGTGCTGCACGCCCACCACCACGCCATCATGACCCTG-3'; primer 7, 5'-CATGCCATGGTCAGGCCGCCAGCAGTT-3'; and primer 8, 5'-TCCGGATGAAGGAGGAATTCACCATGACCATGA-3'.

Extraction, manipulation of plasmids, and in vitro cloning. Isolation of plasmids, transformation of E. coli, ligation with T4 DNA ligase, agarose gel electrophoresis of DNA, and DNA fragment purification were performed as described previously (19).

Protein preparation. HasA produced in MC4100(pSYC-K150) was purified as previously described (11).

HasA_1-174 was prepared from the pellet of a 500-ml culture grown to an optical density at 600 nm (OD₆₀₀) of 1.5. After precipitation with ammonium sulfate (65% saturation concentration) and centrifugation at 10,000 rpm at 4°C for 45 min, the protein pellet was resuspended in 3 to 5 ml of 50 mM Tris-HCl and dialyzed at 4°C against the same buffer. The proteins were then purified in a two-step process, first on a Q-Sepharose column (samples were eluted with buffer A containing 50 mM Tris-HCl and 1 M NaCl) and then on a Sephacryl 200 column (samples were eluted with buffer B containing 20 mM phosphate buffer [pH 7]).

HasA_(133-174) produced in LMD439 was prepared from a culture supernatant of a 500-ml LB medium culture induced with 0.2% arabinose and 100 µM IPTG grown to an OD₆₀₀ of 1.5. After precipitation with ammonium sulfate (65% saturation concentration) and centrifugation at 10,000 rpm at 4°C for 45 min, the protein pellet was resuspended in 3 to 5 ml of 50 mM Tris-HCl and dialyzed at 4°C against the same buffer. The proteins were then purified on two successive Source-15Q columns. Samples were eluted with buffer A.

HasA spectroscopy. HasA spectroscopy was performed as described previously (5).

Isolation of total membranes and membrane protein solubilization. Five-hundred-milliliter cultures of E. coli bearing the appropriate plasmids and grown at 30°C in LB medium were induced with 0.2% arabinose if necessary and were harvested at an OD₆₀₀ of 1. The samples were centrifuged for 15 min at 5,000 x g at 4°C, and the cell pellets were washed once with 10 ml of 60 mM Tris-HCl (pH 8) and resuspended in 20 ml of the same buffer. Cells were broken with a French press operating at 10,000 lb/in². A crude membrane pellet was collected by centrifugation (45 min, 21,000 x g, 4°C). The pellet was resuspended in 60 mM Tris-HCl (pH 8) to which EDTA-free protease inhibitor (Roche cocktail) was added as recommended by the supplier. Membrane proteins were solubilized by incubation for 1 h at 4°C in buffer C (60 mM Tris-HCl [pH 7.5], 20% glycerol, 5 mM MgCl₂, 100 mM NaCl, 20 mM imidazole, 0.7% lauryl maltoside). Insoluble material was removed by centrifugation (45 min, 21,000 x g, 4°C). The supernatant containing the solubilized membrane proteins was kept frozen at –80°C until it was used.

Ni-NTA chromatography. For His₆::HasD protein purification, solubilized membranes prepared as described above were loaded onto a 1-ml Ni-nitrilotriacetic acid (NTA) column with buffer D (buffer C containing only 0.03% lauryl maltoside) at a flow rate of 0.2 ml/min. The column was washed with 15 ml of the same buffer, and then His₆::HasD proteins were eluted with the equilibrium buffer supplemented with 500 mM imidazole. Fractions were collected, analyzed by gel electrophoresis, and kept frozen at –80°C until they were used.

Complex copurification with Ni-NTA beads. Membrane proteins of strain MC4100(pBAD24-His₆::hasD-hasE) expressing pAM238, pAM-hasA, or pAM-hasA_1-174 were solubilized and purified as described below in the absence or presence of purified HasA (at micromolar concentrations) in all buffers.

Nickel agarose beads were washed three times in buffer E (HEPES-20 mM NaOH [pH 7.8], 20% glycerol, 5 mM MgCl₂, 100 mM NaCl, 20 mM imidazole, 0.03% lauryl maltoside). Mixtures containing 1 ml of a solubilized membrane preparation and 100 µl of Ni-NTA beads were incubated for 1 h at 4°C on a wheel and then centrifuged for 5 min at 5,000 x g. The supernatants corresponding to the unbound fraction (fraction U) were collected, and the proteins were resolved by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE). The pellets were washed four times in 1 ml of the same buffer. Then proteins were eluted (fraction E1) in buffer F (HEPES-20 mM NaOH [pH 7.8], 20% glycerol, 5 mM MgCl₂, 100 mM NaCl, 500 mM imidazole, 0.03% lauryl maltoside). The beads were then resuspended in 50 µl of SDS sample buffer and boiled for 5 min. The agarose beads were pelleted by centrifugation at 5,000 x g for 5 min, and the remaining eluted proteins were resolved by SDS-PAGE (fraction E2).

Overlay experiments. Aliquots of purified His₆::HasD protein were adsorbed on 0.25-cm² nitrocellulose membranes. The membranes were incubated in 150 µl of solution containing purified HasA or HasA_1-174 at a concentration of at 700 µM in Tween buffer containing sodium Tricine. The nitrocellulose membranes were probed with polyclonal anti-HasA used at a dilution of 1/5,000.

Electrophoresis and immunoblotting techniques. Protein samples from cell pellets were resuspended at 4°C in 10 mM Tris-HCl-1 mM EDTA (pH 8) and then precipitated by adding trichloroacetic acid at a final concentration of 10%; supernatants were precipitated directly with trichloroacetic acid. After centrifugation and washing with 1 volume of 100% acetone, dry samples were resuspended in sample buffer and loaded onto SDS-polyacrylamide gels. When necessary, gels were transferred to nitrocellulose membranes using a semidry apparatus. Membranes were probed with polyclonal anti-HasA (1/5,000), anti-TolC (1/2,000), anti-HasD (1/5,000), or anti-HasE (1/5,000) antibodies.

Vancomycin and novobiocin sensitivity assays. The antibacterial activities of different agents were determined on LB agar plates containing IPTG (0.25 mM) and arabinose (0.002%). To determine the MICs, bacteria were grown on LB agar plates at 37°C overnight; cells were diluted in the same medium and tested using a final inoculum size of 10¹, 10³, or 10⁵ CFU following incubation at 37°C for 20 h. The MIC was the lowest concentration of compound that inhibited cell growth.

SDS sensitivity assays. Cell growth in the presence of SDS was examined in two ways. The first assay was performed with solid medium, and single colonies were streaked onto LB agar plates supplemented with antibiotics, IPTG (0, 100, 250, or 500 µM), 0.002% arabinose, and SDS at various concentrations (0, 0.1%, 1%, and 2%); growth of colonies was recorded after overnight incubation at 37°C. In the second assay, growth was tested using liquid cultures. Cells were grown until the OD₆₀₀ was 1, and then the cultures were diluted with LB medium supplemented with arabinose (0.002%), IPTG (0.5 mM), and SDS at various concentrations to obtain concentrations of about 10⁶ to 10⁷ cells/ml. The cultures were grown for 17 h before the OD₆₀₀ was determined. Each culture was diluted 1/10,000, and 100-µl portions were spread on LB agar plates supplemented with antibiotics. The plates were incubated overnight at 37°C. The colonies were then counted.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

 
HasA_1-174 is not secreted and folds in vivo. A folded, cytoplasmic hemophore that is not competent for secretion inhibits the secretion of newly synthesized hemophore polypeptides (5). We tried to determine whether the secretion signal was involved in this inhibition.

HasA, like most proteins secreted by ABC transporters, has a C-terminal signal (16). A 56-amino-acid C-terminal peptide (Cter-HasA) is autonomously secreted by its cognate secretion apparatus (HasD, HasE, TolC) (14), showing that the signal is located in this sequence, but more precise localiztion of the signal was not available. For several proteins secreted by the ABC pathway, the secretion signal was at the end of the C terminus and had to be exposed (9). In HasA, the last 14 residues were flexible and not visible on the crystal structure (1). Thus, we hypothesized that deletion of this segment might have no impact on HasA folding but might affect secretion. A HasA deletion mutant encoding HasA_1-174 and lacking the last 14 residues was cloned in pTrc99A under control of the P_trc promoter and was repressed by the plasmid-encoded LacI protein. Strain MC4100 was transformed with either pTRC99A-hasA or pTRC99A-hasA_1-174, and the truncated protein folding in the cytoplasm was monitored on the basis of its ability to bind heme in vivo. The absorption spectra of MC4100(pTRC99A-hasA) and MC4100(pTRC99A-hasA_1-174) whole cells had a peak centered around 407 nm (the Soret band), indicating that cytoplasmic HasA_1-174 was loaded with heme, as previously shown for complete HasA (5) (Fig. 1A). A comparison of the intracellular HasA_1-174 amounts (estimated by immunoblotting) and the amplitude of the Soret band indicated that >90% of the protein was loaded with heme, as was the case for wild-type HasA (data not shown). Thus, the truncated form also folded in the cytoplasm. pTRC99A-Cm-hasA_1-174 was introduced into strain LMD439, in which the hasDE genes were integrated into the chromosome at the att site and expressed under control of the arabinose promoter. The truncated polypeptide was not detected in the culture supernatant of strain LMD439(pTRC99A-Cm-hasA_1-174). Instead, it accumulated in the cytoplasm (Fig. 1B), showing that the last 14 residues were necessary for secretion. The SDS gels were also analyzed by immunoblotting with anti-HasA polyclonal serum to identify the Coomassie blue-stained bands as HasA polypeptides (Fig. 1C).

View larger version (16K): [in this window] [in a new window]   FIG. 1. Folding and secretion of truncated HasA proteins. (A) Folding of holo-HasA mutant in the cytoplasm of E. coli. The visible spectra of MC4100 cells expressing plasmids pAM-hasA(H32A-Y75A-H83A), pAM-hasA, and pAM-hasA1-174 were recorded, as described in Materials and Methods. The spectra show the 385- to 425-nm region after normalization by subtraction of the absorption of cells carrying pAM238. (B) Coomassie blue staining of HasA mutant proteins. E. coli LMD439 carrying pTrc99A-Cm-hasA(133-174), pTrc99A-Cm-hasA1-174, or pTrc99A-Cm-hasA was grown at 30°C in LB medium supplemented with 100 µM IPTG and 0.02% arabinose. An SDS-14% PAGE analysis and Coomassie blue staining of the cellular pellet (lanes P) and culture supernatant (lanes S) were performed. An amount of culture supernatant or cellular pellet equivalent to an OD600 of 0.2 was loaded in each lane. Lanes 1, HasA; lanes 2, HasA1-174; lanes 3, HasA(133-174). (C) Immunodetection of HasA mutant proteins. The same aliquots of culture supernatants (lanes S) and cellular pellets (lanes P) that were used for panel B were separated by SDS-14% PAGE and transferred to a nitrocellulose membrane. Immunodetection was performed with anti-HasA antibodies.

(133-174)

View larger version (9K): [in this window] [in a new window]   FIG. 5. SDS sensitivity of strains expressing various HasA proteins. SDS sensitivity assays were performed on LB agar plates supplemented with 1% SDS with or without IPTG and 0.002% arabinose. WT, wild type; S, no colonies; R, normal-size colonies; +, secretion; –, no secretion. MTMITNL corresponds to the first seven residues of ß-galactosidase fused to the 56 residues of HasA, present in Cter-HasA.

In the absence of IPTG or at a low IPTG concentration, MC4100(pAM238-hasISRADE, pTRC99A-hasA_1-174) secreted HasA as well as strain MC4100(pAM238-hasISRADE) (Fig. 2, lanes 2 and 4). At higher IPTG concentrations, the extracellular HasA concentration decreased (Fig. 2, lanes 6 and 8), whereas intracytoplasmic HasA_1-174 accumulated (Fig. 2, lanes 5 and 7). The amount of residual HasA secreted in the presence of a high IPTG concentration was at the level of cellular lysis (less than 5% of the total amount of HasA). Cytoplasmic wild-type HasA, which is encoded by a low-copy-number plasmid, was not visible on Coomassie blue-stained gels but was detected in the cytoplasmic fraction by Western blotting. The amounts of cytoplasmic wild-type HasA, as evaluated by Western blotting, were larger at high IPTG concentrations (data not shown). Accumulation of HasA_1-174, a protein lacking the secretion signal, in the cytoplasm inhibited secretion of a wild-type copy of HasA. This suggested that HasA and the transporter had at least one additional interaction region that was different from the C-terminal extremity.

View larger version (26K): [in this window] [in a new window]   FIG. 2. Folded HasA1-174 inhibits secretion of a wild-type copy of HasA. MC4100 strains carrying plasmid pAM-hasISRADE (lanes 1 and 2) or plasmids pAM-hasISRADE and pTrc99A-hasA1-174 (lanes 3 to 8) were grown at 30°C in LB medium with 0.2 mM 2,2'-dipyridyl. When the OD600 of the MC4100(pTrc99A-hasA1-174, pAM-hasISRADE) culture reached 0.1, the culture was aliquoted, transferred into flasks containing various amounts of IPTG, and grown for 3 h. Proteins present in supernatants (lanes 2, 4, 6, and 8) and cell pellets (lanes 1, 3, 5, and 7) were visualized after separation on 14% polyacrylamide gels and stained with Coomassie blue. An amount equivalent to an OD600 of 0.15 was loaded in each lane.

View larger version (19K): [in this window] [in a new window]   FIG. 3. Interaction between wild-type HasA or HasA1-174 and the ABC protein of the His6::HasD transporter. Purified His6::HasD proteins were adsorbed on nitrocellulose and incubated with purified wild-type HasA or HasA1-174. The membranes were probed with anti-HasA antibodies. Spot 1, His6::HasD plus HasA; spot 2, buffer plus HasA; spot 3, His6::HasD plus HasA1-174; spot 4, buffer plus HasA1-174; spot 5, His6::HasD plus buffer; spot 6, buffer plus buffer; spot 7, histidine-rich hexapeptide plus HasA.

Both HasA and HasA_1-174 induce TolC recruitment in a multiprotein complex in vitro. Substrate-induced (HasD-HasE-TolC) transporter protein association was observed previously after protein copurification with HasA on hemin agarose (13). We first examined whether HasA, HasE, and TolC could coelute with HasD in Ni-NTA chromatography using His₆::HasD. Membrane proteins were prepared from JP313 cells harboring plasmid pBAD24-His₆::hasD-hasE expressing His₆::HasD and HasE under control of an AraC-regulated promoter and, as a second plasmid, either control plasmid pAM238, pAM238-hasA encoding wild-type HasA, or pAM238-hasA_1-174 encoding the truncated protein. Protein complexes were purified on an Ni-NTA column, washed, and eluted with various buffers. Pure HasA, which could have a stabilizing effect on the HasA-HasD interaction, was added to some elution buffers as indicated in Fig. 4. Eluted proteins were separated by SDS-PAGE and immunodetected with the anti-HasD, anti-HasA, anti-HasE, and anti-TolC antibodies (note that TolC migrates as a doublet band; both bands are missing in a tolC null mutant). In all experiments, HasA or HasA_1-174 copurified with His₆::HasD on Ni-NTA agarose (data not shown). Neither TolC nor HasE bound directly to Ni-NTA agarose columns (data not shown). Figure 4 shows only immunodetection of TolC, but immunodetection with anti-HasE and anti-TolC antibodies revealed the presence of the two proteins together in the secretion complex (data not shown). With membrane proteins prepared from JP313(pBAD24-His₆::hasD-hasE, pAM238), TolC was not bound (Fig. 4). Thus, as previously reported based on hemin agarose affinity chromatography, there was no transporter protein association when the proteins were prepared from cells that did not coexpress the transporter substrate. Exogenously added substrate did not promote association.

View larger version (7K): [in this window] [in a new window]   FIG. 4. Secretion protein complex copurification with nickel affinity resin. The solubilized membranes were obtained from strains JP313(pBAD24-His6::hasD-hasE pAM238), JP313(pBAD24-His6::hasD-hasE pAM-hasA), and JP313(pBAD24-His6::hasD-hasE pAM-hasA1-174). Eluted polypeptides were separated by 12% SDS-PAGE and immunodetected with anti-TolC antibodies. U, unbound fraction; E1 and E2, two elution fractions.

In vivo TolC trapping in the secretion protein complex is induced by nonsecreted substrates. Nonsecreted HasA polypeptides resulting either from premature intracellular folding or from deletion of the secretion signal both inhibited HasA secretion and induced HasD-HasE-TolC protein complex formation. Such a complex, if stable, might have titrated TolC molecules which otherwise would have been available to associate with other protein partners, such as AcrA and AcrB, for drug and detergent efflux (25). As mutants lacking either AcrA or AcrB or mutants lacking TolC are hypersensitive to SDS (8), we tested the availability of TolC for drug efflux by measuring the differential effects of SDS on various strains expressing HasD and HasE in the presence or absence of HasA or HasA_1-174. LMD439 was transformed either with the empty vector pTRC99A-Cm or with pTRC99A-Cm-hasA or pTRC99A-Cm-hasA_1-174. In the resulting strains, the uncoupling of hasDE expression and hasA expression led to accumulation of folded HasA when increasing concentrations of IPTG were added. SDS sensitivity was monitored on LB agar plates containing 1% SDS and 0.002% arabinose to induce expression of hasDE. IPTG was added at a concentration of 0.5 mM to induce HasA and its variants (inducing conditions). LMD439(pTRC99A-Cm) formed colonies that were the normal size on LB medium plates containing 1% SDS under inducing or noninducing conditions. This was also the case for LMD439(pTRC99A-Cm-hasA) and LMD439(pTRC99A-Cm-hasA_1-174) under noninducing conditions. On the other hand, neither LMD439(pTRC99A-Cm-hasA) nor LMD439(pTRC99A-Cm-hasA_1-174) formed visible colonies in the presence of SDS at a concentration of 1% under inducing conditions (Fig. 5). There was no SDS sensitivity when the hasD and hasE genes were not induced at arabinose concentrations lower than 0.0002%, indicating that accumulation of HasA and HasA_1-174 was not responsible for SDS sensitivity per se. Thus, the strains were SDS sensitive when wild-type HasA or the truncated proteins were overexpressed.

The amounts of TolC and cellular localization in strains LMD439(pTRC99A-Cm), LMD439(pTRC99A-Cm-hasA), and LMD439(pTRC99A-Cm-hasA_1-174) grown under inducing conditions (LB medium containing 0.002% arabinose and 0.5 mM IPTG) were compared after cellular fractionation and TolC protein immunodetection with anti-TolC antibodies. All of the strains tested had similar amounts of TolC proteins in the outer membrane, while only traces were detected in the inner membrane and soluble fractions. All cell extracts contained similar amounts of TolC in aggregates and cell debris (about 5% of the total proteins) (data not shown). Thus, the SDS sensitivity of strains expressing HasA or HasA_1-174 was not the result of a lower level of TolC production or TolC mislocalization or misfolding. Hence, it is likely that TolC was trapped in the secretion protein complex and was not present at concentrations sufficient to associate with the efflux pumps.

SDS sensitivity is reversed by overproduction of TolC or HasF. To determine whether the SDS sensitivity observed was due to titration of TolC, we overproduced TolC. pAM-tolC, a plasmid expressing tolC under the P_lac promoter, was introduced into LMD439(pTRC99A-Cm-hasA_1-174). Western blot analysis of protein extracts of strains LMD439(pTRC99A-Cm-hasA_1-174, pAM238) and LMD439(pTRC99A-Cm-hasA_1-174, pAM-tolC) with anti-TolC antibodies showed that there was a threefold overall increase in the amount of TolC in strains harboring pAM238-tolC (data not shown). The SDS sensitivities of strains LMD439(pTRC99A-Cm, pAM238), LMD439(pTRC99A-Cm-hasA_1-174, pAM238), and LMD439(pTRC99A-Cm-hasA_1-174, pAM-tolC) were determined by using liquid cultures growing in LB media supplemented with SDS at various concentrations. In liquid cultures, the highest SDS concentration which allowed efficient growth was 0.05%. Viable cell counts were evaluated by cell plating after 17 h of growth at 30°C. Figure 6 shows that LMD439(pTRC99A-Cm-hasA_1-174, pAM-tolC) grew well in LB medium under inducing conditions in the presence of 0.05% SDS, whereas LMD439(pTRC99A-CmhasA_1-174, pAM238) exhibited sharply reduced growth. Thus, overexpression of TolC restored SDS resistance. Similar results were obtained with the S. marcescens TolC homologue HasF (data not shown). Other drugs, including substrates of the AcrA-AcrB-TolC efflux pump, were also tested. The sensitivities to vancomycin (a drug which is not expelled by AcrA-AcrB-TolC) were similar (MIC, 300 µg/ml) for strains LMD439(pTRC99A-Cm-hasA_1-174), LMD439(pTRC99A-Cm-hasA), and LMD439 (pTRC99A-Cm) under inducing conditions (0.002% arabinose, 0.5 mM IPTG). In contrast, LMD439(pTRC99A-Cm-hasA_1-174) and LMD439(pTRC99A-Cm-hasA) exhibited 20-fold-increased sensitivity to novobiocin, a substrate of the AcrA-AcrB-TolC pump (the novobiocin MIC changed from up to 10 µg/ml to 0.6 µg/ml), under inducing conditions, strengthening the interpretation that drug sensitivity is the result of TolC titration by the Has transporter.

View larger version (28K): [in this window] [in a new window]   FIG. 6. SDS sensitivity is reversed by TolC overproduction or by Cter-HasA coexpression. A total of 106 cells of strains LMD439(pTRC99A-Cm, pAM238) (WT), LMD439(pTRC99A-Cm-hasA1-174, pAM238) (NC), LMD439(pTRC99A-Cm-hasA1-174, pAM-tolC) (TolC), and LMD439 (pTRC99A-Cm-hasA1-174, pACYC184-Cter-hasA) (Cter-HasA) from exponentially LB medium growing cultures were inoculated into LB medium supplemented with 0.05% SDS, 0.002% arabinose, and 0.5 mM IPTG, and the viable cell counts were evaluated by cell plating after 17 h of growth at 30°C in liquid LB medium. CFU were normalized by using CFU without SDS.

(140-188)

Cter-HasA reversed SDS sensitivity induced by HasA_1-174. The secretion signal does not promote SDS sensitivity. Thus, we examined whether it could reverse this phenotype induced by the truncated HasA_1-174 protein when it was expressed as a distinct polypeptide. Strain LMD439(pTRC99A-Cm-hasA_1-174, pACYC184-Cter-hasA) grew as well as strain LMD439(pTRC99A-Cm, pACYC184-Cm) with 0.002% arabinose and in the absence or presence of IPTG at an SDS concentration of 0.5%, indicating that Cter-HasA could reach the transporter and reverse SDS sensitivity even in the presence of the inhibitor molecule HasA_1-174 (Fig. 6).

Thus, we tested whether such an interaction also led to Cter-HasA secretion. Strain LMD439(pTRC99-Cm-hasA_1-174, pACYC184-Cter-hasA) secreted Cter-HasA only in the absence of IPTG when HasA_1-174 synthesis was not induced (data not shown). This indicated that Cter-HasA was not secreted in the presence of the truncated substrate.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

 
Uncoupling hemophore synthesis from secretion leads to intracytoplasmic accumulation of folded HasA molecules, incompetent for further secretion, which are not unfolded by SecB overexpression and which block secretion of newly synthesized HasA molecules (5).

To determine whether the C-terminal secretion signal is involved in this inhibition, a HasA mutant lacking the last 14 C-terminal amino acids was constructed. The truncated protein was not secreted; it was folded in the cytoplasm in an active heme-binding form and inhibited wild-type HasA molecule secretion. We show here that the truncated protein interacted in vitro with the ABC protein. As the last 14 HasA residues are necessary and sufficient for directing HasA secretion, this indicated that HasA interacts with the ABC protein via two distinct regions, the secretion signal and another region.

Secretion protein association. We showed previously that the three components of the hemophore and the protease transporters are not permanently associated. Coexpression of the substrate with its exporter was necessary to associate the secretion proteins. Multiprotein complex formation was observed with functional pairs of substrates and transporters and with the C-terminus-derived secretion peptides, but also with abortive pairs, such as HasA and the protease transporter (13). This raised the question of whether the truncated HasA protein could also assemble the transporter. We show here that the truncated substrate lacking the C terminus is also able to trigger multiprotein complex association. While secretion protein complex copurification with wild-type HasA required addition of exogenous HasA during the purification steps, this was not the case with the truncated HasA protein. Since it is likely that the addition of exogenous HasA shifted the HasA-HasD equilibrium toward association, this suggests that HasA_1-174 has a more stable interaction with His₆::HasD than with the wild-type protein. Direct interaction between His₆::HasD and HasA, measured by quantification in an overlay experiment, also showed that the affinity of the truncated HasA for His₆::HasD was threefold higher than the affinity of the wild-type substrate (data not shown). The presence of the last 14 amino acids drives complex dissociation, most likely by returning the ABC protein to its unloaded state, as suggested for hemolysin secretion (20).

TolC trapping in the secretion protein complex. The E. coli TolC protein also interacts with the inner membrane proton antiporter pump AcrB and the membrane fusion protein AcrA, allowing efflux of toxic compounds, such as antibiotics, metals, and detergents (25). TolC, AcrA, and AcrB mutants are hypersensitive to these compounds (8). Several cross-linking experiments have shown that, unlike hemolysin and hemophore transporter assembly, the AcrA-AcrB-TolC efflux pump proteins are permanently associated independent of addition of a drug (21). Thus, in E. coli strains actively secreting either the hemophore or the hemolysin, TolC has to be shared by the type I secretion system and the efflux pump. Yet such cells do not exhibit higher detergent sensitivity (S. Cescau, unpublished), either because there is enough TolC for both transporters or because TolC association with type I secretion systems is transient and TolC dissociates from the type I secretion complex at each substrate molecule exit.

We hypothesized that substrates that are not secreted, either due to premature folding or to the absence of the secretion signal, might jam the Has transporter and trap TolC in vivo, making it no longer available to pump out detergents. We found that both cells exhibiting uncoupled HasA synthesis and secretion that accumulated folded HasA and strains expressing the transporter and the protein lacking the secretion signal were sensitive to 1% SDS. Sensitivity to vancomycin, a compound which is not an AcrA-AcrB-TolC pump substrate (2), was not enhanced in strains expressing the Has transporter and the wild-type or truncated HasA proteins. This indicates that the HasD-HasE-TolC protein complex does not nonspecifically enhance outer membrane permeability, as is the case for the HlyB-HlyD-TolC transporter (4, 23).

SDS sensitivity was reversed by plasmid overexpression of TolC (or of the S. marcescens TolC orthologue HasF), demonstrating that SDS sensitivity resulted from a lack of available TolC molecules. A decrease in the amount of TolC was not due to increased TolC instability in cells expressing HasA_1-174, since Western blot analysis showed that the amounts of TolC were similar in cells harboring the various plasmids (data not shown).

The SDS sensitivity phenotype of strains fully expressing HasA_1-174 and HasD-HasE was not as strong as that of a tolC null mutant which did not grow at an SDS concentration of 0.1%, most likely indicating that TolC sequestration was not complete, which left few molecules available for an interaction with AcrAB, resulting in a residual SDS efflux.

C-terminal secretion peptide can promote multicomplex dissociation as an independent polypeptide. The 60-amino-acid C-terminal peptide carrying the secretion signal expressed as a single molecule also reversed the SDS sensitivity of strains producing HasD-HasE and HasA_1-174. This demonstrates that the C-terminal peptide, which functions as an intramolecular signal sequence with the entire substrate, can also function in an intermolecular process and probably triggers the complex dissociation in vivo. For this function, the C terminus must still access the proper site on the ABC protein despite the presence of the truncated polypeptide on the ABC protein. This result appears to be paradoxical, since HasA_1-174 is able to totally inhibit HasA protein secretion. This might result from a level of expression of Cter-HasA (from pACYC184) that is higher than the level of expression of the whole polypeptide (from pAM238) (data not shown). Alternatively, isolated Cter-HasA could have a higher affinity for the transporter. Although Cter-HasA reaches the transporter, it is not secreted in the presence of the truncated polypeptide, which most likely induces steric hindrance.

Two modes of interaction induced by two substrate sites. The two HasA sites are distinct and do not overlap, since the secretion signal is located in the last 14 residues, whereas the second site is located upstream of the last 60 Cter-HasA residues. Preliminary results showed that the second site is located in the first 132 amino acids. At present, we do not know whether the second site is a linear or folded HasA region. It also remains to be determined whether the two HasA binding sites recognize the same or different sites on the ABC protein.

Our work has revealed two modes of interaction between the hemophore and its ABC protein. When the C terminus is absent or when the folded HasA protein accumulates in the cytoplasm, there is a stable interaction between the ABC protein and the substrate which permanently traps TolC and makes it unavailable for association with the AcrA-AcrB efflux pump. Engagement of the extreme C terminus dissociates the complex. Thanabalu et al. (20) showed that TolC was not cross-linked to the HlyB-HlyD inner membrane complex when the HlyA intracellular pool was cleared out. TolC dissociation corresponded to the return to an unloaded state. ATP hydrolysis has been shown to be required for hemolysin exporter disassembly (20).

Here we present evidence that dissociation is an active process triggered by the interaction of the C terminus with the ABC protein.

The second site does not appear to be essential for secretion, since the 60-amino-acid C-terminal peptide (which does not carry the second site) is efficiently secreted. The second site could act as an anchor enabling preassociation of the complex during HasA synthesis prior to completion and engagement of the C terminus in the transporter, leading to dissociation. Nevertheless, it is likely that the second sites are important for optimizing secretion, particularly of long polypeptides. We do not know yet whether second sites are present on other proteins secreted by pathway I.

Finally, we developed a simple Petri dish test for evaluating in vivo trapping of TolC in a stable type I secretion complex. This test enables workers to study transporter association and dissociation independent of the secretion process itself. Screening of SDS-resistant revertants should enable identification of the hemophore second site, the hemophore binding site(s) on the ABC protein, and the role of ATP binding and hydrolysis during the association/dissociation cycle.

	ACKNOWLEDGMENTS

 
We thank Nicolas Wolff, Philippe Delepelaire, and Sylvie Létoffé for providing plasmids and antibodies and for constant interest and help with this work. We also thank Elie Dassa for critical reading of the manuscript and Francis Biville for helpful suggestions.

Sandra Cescau received a Ph.D. grant from the Ministère de la recherche, Paris, France. This work was supported by the Fondation pour la Recherche Médicale.

	FOOTNOTES

 
* Corresponding author. Mailing address: Unité des Membranes Bactériennes, Département de Microbiologie Fondamentale et Médicale, Institut Pasteur, 25-28, rue du Dr. Roux, 75724 Paris Cedex 15, France. Phone: 33 1 40 61 32 75. Fax: 33 1 45 68 87 90. E-mail: cwander{at}pasteur.fr.

Published ahead of print on 8 December 2006.

Supplemental material for this article may be found at http://jb.asm.org/.

	REFERENCES

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Arnoux, P., R. Haser, N. Izadi, A. Lecroisey, M. Delepierre, C. Wandersman, and M. Czjzek. 1999. The crystal structure of HasA, a hemophore secreted by Serratia marcescens. Nat. Struct. Biol. 6:516-520.[CrossRef][Medline]
2. Augustus, A., T. Celaya, F. Husain, M. Humbard, and R. Misra. 2004. Antibiotic-sensitive TolC mutants and their suppressors. J. Bacteriol. 186:1851-1860.[Abstract/Free Full Text]
3. Binet, R., and C. Wandersman. 1996. Cloning of the Serratia marcescens hasF gene encoding the Has ABC exporter outer membrane component: a TolC analogue. Mol. Microbiol. 22:265-273.[CrossRef][Medline]
4. Blight, M., A. Pimenta, J. Lazzaroni, C. Dando, L. Kotelevets, S. Seror, and I. Holland. 1994. Identification and preliminary characterization of temperature-sensitive mutations affecting HlyB, the translocator required for the secretion of haemolysin (HlyA) from Escherichia coli. Mol. Gen. Genet. 245:431-440.[CrossRef][Medline]
5. Debarbieux, L., and C. Wandersman. 2001. Folded HasA inhibits its own secretion through its ABC exporter. EMBO J. 20:4657-4663.[CrossRef][Medline]
6. Delepelaire, P. 2004. Type I secretion in gram-negative bacteria. Biochim. Biophys. Acta 1694:149-161.[Medline]
7. Delepelaire, P., and C. Wandersman. 1998. The SecB chaperone is involved in the secretion of the Serratia marcescens HasA protein through an ABC transporter. EMBO J. 17:936-944.[CrossRef][Medline]
8. Fralick, J. A., and L. L. Burns-Kehliher. 1994. Additive effect of tolC and rfa mutations on the hydrophobic barrier of the outer membrane of Escherichia coli K-12. J. Bacteriol. 176:6404-6412.[Abstract/Free Full Text]
9. Ghigo, J. M., and C. Wandersman. 1994. A carboxyl-terminal four-amino acid motif is required for secretion of the metalloprotease PrtG through the Erwinia chrysanthemi protease secretion pathway. J. Biol. Chem. 269:8979-8985.[Abstract/Free Full Text]
10. Ghigo, J.-M., S. Létoffé, and C. Wandersman. 1997. A new type of hemophore-dependent heme acquisition system of Serratia marcescens reconstituted in Escherichia coli. J. Bacteriol. 179:3572-3579.[Abstract/Free Full Text]
11. Izadi, N., Y. Henri, J. Haladjan, M. E. Goldberg, C. Wandersman, M. Delepierre, and A. Lecroisey. 1997. Purification and characterization of an extra cellular heme binding protein, HasA, involved in heme iron acquisition. Biochemistry 36:7050-7057.[CrossRef][Medline]
12. Koronakis, V., A. Sharff, E. Koronakis, B. Luisi, and C. Hughes. 2000. Crystal structure of the bacterial membrane protein TolC central to multidrug efflux and protein export. Nature 405:914-919.[CrossRef][Medline]
13. Létoffé, S., P. Delepelaire, and C. Wandersman. 1996. Protein secretion in Gram negative bacteria: assembly of the three components of ABC protein mediated exporters is ordered and promoted by substrate binding. EMBO J. 15:5804-5811.[Medline]
14. Létoffé, S., J. M. Ghigo, and C. Wandersman. 1994. Secretion of the Serratia marcescens HasA protein by an ABC transporter. J. Bacteriol. 176:5372-5377.[Abstract/Free Full Text]
15. Létoffé, S., C. Deniau, N. Wolff, E. Dassa, P. Delepelaire, A. Lecroisey, and C. Wandersman. 2001. Haemophore-mediated bacterial haem transport: evidence for a common or overlapping site for haem-free and haem-loaded haemophore on its specific outer membrane receptor. Mol. Microbiol. 41:439-450.[CrossRef][Medline]
16. Mackman, N., K. Baker, L. Gray, and I. B. Holland. 1987. Release of a chimeric protein into the medium from Escherichia coli using the C-terminal secretion signal of haemolysin. EMBO J. 6:2835-2841.[Medline]
17. Mackman, N., J. M. Nicaud, V. Gray, and I. B. Holland. 1986. Secretion of hemolysin by Escherichia coli. Curr. Top. Microbiol. Immunol. 125:159-181.[Medline]
18. Pogliano, J., A. S. Lynch, D. Belin, E. C. Lin, and J. Beckwith. 1997. Regulation of Escherichia coli cell envelope proteins involved in protein folding and degradation by the Cpx two-component system. Genes Dev. 11:1169-1182.[Abstract/Free Full Text]
19. Sambrook, J., E. F. Fritsch, and T. E. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
20. Thanabalu, T., E. Koronakis, C. Hughes, and V. Koronakis. 1998. Substrate-induced assembly of a contiguous channel for protein export from E. coli: reversible bridging of an inner-membrane translocase to an outer membrane exit pore. EMBO J. 17:6487-6496.[CrossRef][Medline]
21. Touze, T., J. Eswaran, E. Bokma, E. Koronakis, C. Hughes, and V. Koronakis. 2004. Interactions underlying assembly of the Escherichia coli AcrAB-TolC multidrug efflux system. Mol. Microbiol. 53:697-706.[CrossRef][Medline]
22. Wandersman, C., and P. Delepelaire. 1990. TolC, an Escherichia coli outer membrane protein required for hemolysin secretion. Proc. Natl. Acad. Sci. USA 87:4776-4780.[Abstract/Free Full Text]
23. Wandersman, C., and S. Letoffe. 1993. Involvement of lipopolysaccharide in the secretion of Escherichia coli alpha-haemolysin and Erwinia chrysanthemi proteases. Mol. Microbiol. 7:141-150.[CrossRef][Medline]
24. Wolff, N., P. Delepelaire, J. M. Ghigo, and M. Delepierre. 1997. Spectroscopic studies of the C-terminal secretion signal of the Serratia marcescens haem acquisition protein (HasA) in various membrane-mimetic environments. Eur. J. Biochem. 243:400-407.[Medline]
25. Zgurskaya, H. I., and H. Nikaido. 1999. Bypassing the periplasm: reconstitution of the AcrAB multidrug efflux pump of Escherichia coli. Proc. Natl. Acad. Sci. USA 96:7190-7195.[Abstract/Free Full Text]

This Article Abstract Full Text (PDF) Supplemental material Other Versions of this Article: JB.01480-06v1 189/5/1496    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Cescau, S. Articles by Wandersman, C. Search for Related Content PubMed PubMed Citation Articles by Cescau, S. Articles by Wandersman, C.