Molecular Aspects of Biogenesis of Escherichia coli Dr Fimbriae: Characterization of DraB-DraE Complexes
http://www.100md.com
感染与免疫杂志 2005年第1期
Department of Microbiology, Gdask University of Technology, Gdask, Poland
Department of Obstetrics and Gynecology
Department of Microbiology and Immunology, The University of Texas Medical Branch, Galveston, Texas
ABSTRACT
The Dr hemagglutinin of uropathogenic Escherichia coli is a fimbrial homopolymer of DraE subunits encoded by the dra operon. The dra operon includes the draB and draC genes, whose products exhibit homology to chaperone-usher proteins involved in the biogenesis of surface-located polymeric structures. DraB is one of the periplasmic proteins belonging to the superfamily of PapD-like chaperones. It possesses two conserved cysteine residues characteristic of the FGL subfamily of Caf1M-like chaperones. In this study we obtained evidence that DraB cysteines form a disulfide bond in a mature chaperone and have the crucial function of forming the DraB-DraE binary complex. Expression experiments showed that the DraB protein is indispensable in the folding of the DraE subunit to a form capable of polymerization. Accumulation of DraB-DraEn oligomers, composed of head-to-tail subunits and the chaperone DraB, was observed in the periplasm of a recombinant E. coli strain which expressed DraB and DraE (but not DraC). To investigate the donor strand exchange mechanism during the formation of DraE oligomers, we constructed a series of DraE N-terminal deletion mutants. Deletion of the first three N-terminal residues of a potential donor strand resulted in a DraE protein lacking an oligomerization function. In vitro data showed that the DraE disulfide bond was not needed to form a binary complex with the DraB chaperone but was essential in the polymerization process. Our data suggest that assembly of Dr fimbriae requires a chaperone-usher pathway and the donor strand exchange mechanism.
INTRODUCTION
The recognition of host receptors and attachment to them are crucial steps in the process of bacterial pathogenesis. Gram-negative bacteria express surface polymeric adhesive organelles that are responsible for specific recognition and attachment to host tissues (32). In some cases (e.g., Escherichia coli type 1 fimbriae or Dr/Afa-III adhesins [10, 11, 18]) these structures are important in promoting the invasion of host cells by bacterial cells and have been implicated in recurrent or chronic infections. These polymeric structures may form fimbriae or afimbrial capsules surrounding bacterial cells (28, 32). Biogenesis of these polymers occurs in gram-negative bacteria by four main pathways, the chaperone-usher, alternative chaperone-usher, type IV secretion, and type IV pilus assembly pathways (32).
The family of more than 30 different fimbrial and nonfimbrial polymeric adhesins expressed by different gram-negative bacteria contains members assembled with a common chaperone-usher mechanism (29, 36). This family encompasses complex, heteropolymeric rigid fimbriae (e.g., P pili and type 1 pili of E. coli) (4, 12), much simpler homopolymeric fimbriae with a flexible structure (e.g., Dr fimbriae of uropathogenic E. coli) (25, 35), and nonfimbrial or amorphic structures (e.g., F1 capsular antigen of Yersinia pestis and Afa-III of E. coli) (8, 9). The process of biogenesis of polymeric adhesins in the chaperone-usher pathway depends on the action of two periplasmic proteins, the usher and chaperone proteins. The usher protein is a homooligomer that forms an outer membrane, donut-shaped channel by which the polymerization process occurs (1, 7, 31, 37). The chaperone is a periplasmic protein which has crucial roles in the assembly of polymers (20, 33). As shown first for P and type 1 pili, the subunits that are used to build these polymers after translocation to the periplasm are not able to fold spontaneously to form structures that are capable of polymerization (17, 30). In the periplasm, the chaperone protein interacts with fimbrial subunits, and based on steric information in the structure, the subunits fold to construct a native functional form (2, 23). Only when they are in a complex with the chaperone are subunits able to polymerize in contact with the usher protein to form surface-located adhesive polymers. Accumulation of free subunits in the periplasm induces the toxic effect that results in their degradation by proteases, mainly DegP (HtrP) (17). The chaperone bound to a subunit stabilizes it and caps its interactive regions, thereby protecting the periplasm from a potential toxic effect.
The periplasmic chaperones engaged in assembly of polymeric structures belong to a conserved superfamily whose members have high sequence homology and a common structure (29). Thus far, structures of the following four chaperones have been elucidated: PapD (P pili of E. coli) (13, 15), FimC (type 1 pili of E. coli) (6), SfaE (S pilus of E. coli) (19), and Caf1M (F1 antigen of Y. pestis) (39). All of these molecules are composed of two immunoglobulin-like domains with the overall shape of a boomerang. Comparison of the sequences of known chaperones with crystallographic structure data permits division of the superfamily into two groups, FGL (F1-G1 Long) and FGS (F1-G1 short) (5, 16). The base of this division is the number of residues in the loop connecting strands. Chaperones belonging to the FGS subfamily are engaged in the assembly of rod-like fimbriae (PapD and FimC), whereas FGL chaperones appear to assemble thin fimbrial and afimbrial structures (Caf1M) (16).
The crystallographic structures of the chaperone-subunit binary complexes PapD-PapK (27), PapD-PapE (26), FimC-FimH (6, 14), and Caf1M-Caf1 and the ternary complex Caf1M-Caf1'-Caf1" (39) permitted discovery of the molecular mechanism of biogenesis of organelles which are assembled by a chaperone-usher pathway. The crystallographic data showed that all subunits possess incomplete immunoglobulin structures with the seventh -strand missing, which results in exposure of a deep, hydrophobic cleft. In complex crystallographic structures, the chaperone completes this cleft with its G1 -strand in a donor strand complementation process. During polymerization, the chaperone is removed from the last subunit of the growing polymer by an incoming subunit, resulting in head-to-tail polymerization. This mechanism is called donor strand exchange, in which the G1 donor -strand of the chaperone is replaced by a similar interaction arising from the N-terminal strand of the subunit (29). Recently published crystallographic structures of Caf1M-Caf1 complexes permitted researchers to determine the differences in the activities of FGL and FGS chaperone-usher pathways (3, 39).
The adhesins of the Dr family in uropathogenic and diarrheal E. coli are encoded by operons that have similar gene organizations. Each operon contains genes encoding the proteins responsible for adhesin biogenesis and the group of genes encoding proteins that regulate the process of expression (21, 35). The amino acid sequences of accessory proteins are rather conserved in the entire family, but adhesin subunits have limited homology.
The Dr hemagglutinin is encoded by a dra operon in uropathogenic E. coli. DraE subunits form a fimbrial structure with an adhesive capacity. The dra gene cluster also encodes the DraC usher and DraB chaperone proteins. DraB-DraE has only limited homology to the previously characterized binary chaperone-subunit complexes PapD-PapK, PapD-PapE, FimC-FimH, and Caf1M-Caf1. It is, therefore, difficult to predict whether the DraB-DraE binary complex operates in a manner similar to these systems.
Here we constructed a recombinant in vitro and in vivo system for expression of DraB and DraE in the periplasm. We then constructed DraE deletion mutants and investigated the donor strand exchange mechanism in DraE oligomer formation. The results of our experiments are consistent with the interpretation that Dr fimbriae are assembled by the conserved chaperone-usher pathway (as proposed previously on the basis of sequence analyses of the dra operon).
MATERIALS AND METHODS
Strains, plasmids, and mutagenesis. E. coli Top10F' (Invitrogen, Paisley, United Kingdom) was used as the host strain for cloning. E. coli BL21(DE3) (Novagen, Darmstadt, Germany) was used as the expression host. The degP-deficient E. coli B178(DE3)htrA63 expression host was constructed by introducing a gene encoding T7 RNA polymerase into the degP strain E. coli B178htrA63(mini-Tn10) (22) by using a DE3 lysogenization kit (Novagen).
All expression vectors described in this study were derived from the pET30b(+) plasmid (Novagen). Plasmid pET30b-sygDraBE encoding the DraB and DraE proteins in a bicistronic system was constructed in two steps. The draB gene encoding the native DraB protein (with the signal sequence) was amplified by PCR by using forward primer DraB-syg (5'-TCTAGAAATAATTTTGTTTAACTTTAAGAAGGAGATATACATATGAAAATGCGGGCTGTGGCTGTGTTCACCGGC) and reverse primer DraB-stop (5'-TATGGATCCTCAACCCTTCAGCTCTGCCTCAAACTGCTTAC). The 5' overhanging sequence of the forward primer contains an XbaI site (underlined) and a ribosome binding site (RBS) (boldface type), and the 5' overhanging sequence of the reverse primer contains a BamHI site (underlined). The 876-bp PCR product encoding the DraB protein was digested with BamHI and ligated into pUC19 (New England Biolabs) digested with SmaI and BamHI, resulting in recombinant plasmid pUC19-sygDraB. The draB gene was excised from the pUC19-sygDraB plasmid with XbaI and BamHI and ligated into the XbaI and BamHI sites of expression vector pET30b(+) to obtain pET30b-sygDraB. The draE gene was amplified with the following primers: forward primer DraE-syg (5'-ATAGGATCCAACTTTAAGAAGGAGATATACATATGAAAAAATTAGCGATCATGGCCGCGGCCAG) and DraE-stop (5'-TATAAGCTTTCATTTTGCCCAGTAACCCCCGGTCAGGGTC). The forward DraE-syg primer contained an RBS (boldface type) and a BamHI site (underlined), and the DraE-stop primer contained a HindIII site (underlined). The 521-bp PCR product encoding the native DraE protein was digested with BamHI and HindIII and ligated into pET30b-sygDraB digested with the same endonucleases to obtain the expression plasmid pET30b-sygDraBE.
The plasmids encoding native DraB and N-terminal deletion mutants of the DraE protein were constructed by cutting out the 432-bp SacI-HindIII DNA fragment from pET30b-sygDraBE and ligating the PCR products encoding N-truncated DraE proteins into the same restriction sites. The SacI site is located at the end of the sequence coding for the Sec-dependent signal sequence in the DraE protein. The PCR products encoding the N-terminal mutants of DraE were obtained by using the reverse primer DraE-stop described above and the following forward primers (SacI sites are underlined): 1DraE, 5'-ATAGAGCTCCGCTCACGCGTTCACCCCGAGTGGCACCACCG; 2DraE, 5'-ATAGAGCTCCGCTCACGCGACCCCGAGTGGCACCACCGG; 3DraE, 5'-ATAGAGCTCCGCTCACGCGCCGAGTGGCACCACCGGCACC; 4DraE, 5'-ATAGAGCTCCGCTCACGCGAGTGGCACCACCGGCACCAC; 5DraE, 5'-ATAGAGCTCCGCTCACGCGGGCACCACCGGCACCACCAAAC; 8DraE, 5'-ATAGAGCTCCGCTCACGCGGGCACCACCAAACTCACAGTTACCGAAGAGTGCCAGGT; and 11DraE, 5'-ATAGAGCTCCGCTCACGCGAAACTCACAGTTACCGAAGAGTGCCA.
The pET30b-syg8DraE and pET30b-syg11DraE plasmids were constructed by cutting the 861-bp XbaI-BamHI fragment containing the draB gene out of the respective bicistronic plasmids (pET30b-sygDraB8E and pET30b-sygDraB11E). The overhangs were filled in by using the Pwo DNA polymerase, and the blunt-ended plasmids were then religated, resulting in plasmids that expressed the 8DraE and 11DraE proteins.
The pET30b-sygDraBC103A plasmid, encoding an alanine mutant at cysteine 103 of the DraB protein, was created from the pET30b-sygDraB plasmid by using a QuikChange site-directed mutagenesis kit according to the instructions of the manufacturer (Stratagene, La Jolla, Calif.).
The pET30b-DraE plasmid was constructed by cloning the 441-bp PCR product encoding the DraE protein without the signal sequence into the NdeI and HindIII sites of the pET30b(+) plasmid with forward primer 5'-ATAATACATATGGGGTTCACCCCGAGTGG (NdeI site underlined) and reverse primer 5'-TTGAAGCTTTCATTTTGCCCAGTAACCCCC (HindIII site underlined). The pET30b-DraEC21A plasmid, encoding an alanine mutant with a mutation at cysteine 21 of the DraE protein, was constructed from the pET30b-DraE plasmid by using a QuikChange site-directed mutagenesis kit (Stratagene).
In all PCRs the Pwo DNA polymerase (DNA-Gdask II s.c., Gdask, Poland) was used. After enzymatic reactions the DNA fragments were purified by using a DNA Clean Up kit (A&A Biotechnology, Gdynia, Poland). The nucleotide sequences of all recombinant plasmids were confirmed by sequencing (DNA-Gdask II s.c.).
Protein expression and purification. The E. coli BL21(DE3) and B178(DE3)htrA63 strains transformed with appropriate plasmids were typically cultivated to an A600 of 0.4 in Luria-Bertani medium supplemented with 20 μg of kanamycin per ml, and protein expression was induced by adding isopropyl--D-thiogalactopyranoside (IPTG) to a final concentration of 0.5 mM for 2 h. Then the cultures were harvested by centrifugation, and periplasmic fractions containing the target proteins were extracted by the osmotic shock procedure (41). To purify complexes of the DraB chaperone with DraE oligomers, the periplasmic fraction obtained from 1 liter of culture was dialyzed overnight against buffer A containing 10 mM NaCl and 20 mM Tris-HCl (pH 7.0). The dialyzed sample was applied to a Resource Q 6-ml anion-exchange column and eluted with a 10 to 400 mM NaCl gradient. At this chromatography step the oligomers were not well resolved. The eluted fractions, which in a dot blot analysis reacted with anti-DraE antibodies, were dialyzed against buffer A and, after concentration to an appropriate volume, were applied to a Mono Q HR 5/5 anion-exchange column. The complexes were eluted with a 10 to 400 mM NaCl gradient, and five well-resolved peaks containing DraB in complexes with DraE oligomers were obtained. DraB-DraE binary complexes were purified on a Resource S 6-ml cation-exchange column by using buffer B (50 mM CH3COONa [pH 5.5]) with a 0 to 250 mM NaCl gradient. Finally, binary complexes were resolved by fractionation on a Superdex G-200 HR 10/30 column with buffer B containing 20 mM NaCl.
The periplasmic extract containing the DraB protein was dialyzed against 20 mM NaCl-20 mM Tris-HCl (pH 7.5) and then applied to a Resource Q 6-ml column. The DraB protein was eluted with a 20 to 300 mM NaCl gradient in 20 mM Tris-HCl (pH 7.5). Finally, the DraB chaperone was resolved by fractionation on a Superdex G-200 HR 10/30 column in buffer containing 20 mM NaCl and 20 mM Tris-HCl (pH 7.5).
The E. coli BL21(DE3) strain transformed with pET30b-DraE (or derivatives of this plasmid) was grown to an A600 of 0.4 in Luria-Bertani medium containing 20 μg of kanamycin per ml. Protein expression was induced by adding IPTG to a final concentration of 1 mM for 4 h. Then the cells were harvested by centrifugation and disrupted by sonication. The centrifuged fraction containing inclusion bodies composed mainly of DraE protein was washed twice with buffer containing 1% Triton X-100, 0.5 mM dithiothreitol (DTT), and 20 mM Tris-HCl (pH 7.5) to remove contaminating proteins. After this, the inclusion bodies were centrifuged and solubilized under denaturing conditions (4 M guanidine-HCl [GuHCl], 0.5 mM DTT, 20 mM Tris-HCl; pH 7.5) for 2 h at 37°C with vigorous shaking. After solubilization the sample was centrifuged, and the supernatant containing denatured DraE protein (purity, 85%) was collected. The purified DraE protein was concentrated by using an Amicon Ultra concentrator (molecular mass cutoff, 10 kDa; Millipore, Eschborn, Germany) to obtain a final concentration of 40 mg/ml.
Renaturation of DraE. DraE protein (5 μl of a denatured sample [40 mg/ml]) was mixed with 800 μl of refolding buffer (50 mM CH3COONa, pH 5.5) containing DraB protein at a concentration of 5 mg/ml for 1 h at 25°C. After this, the sample was centrifuged and passed through a 0.45-μm-pore-size filter to remove potential precipitates. The supernatant obtained was analyzed by exclusion chromatography on a Superdex G-200 HR 10/30 column in buffer containing 20 mM NaCl and 50 mM CH3COONa (pH 5.5). The eluted fractions were analyzed by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and Western blotting with anti-DraE and anti-DraB antibodies. In the negative control experiment, we used refolding buffer with bovine serum albumin (BSA) (5 mg/ml). To determine the specificity of binary complex formation, we used refolding buffer containing the following proteins: DraB (5 mg/ml), BSA (3 mg/ml; molecular mass, 66 kDa), carbonic anhydrase (3 mg/ml; from bovine erythrocytes; molecular mass, 30 kDa), and -lactalbumin (4 mg/ml; from bovine milk; molecular mass, 14.4 kDa). To promote the formation of disulfide bonds in DraE protein complexed with DraB, the sample obtained from exclusion chromatography was dialyzed against oxidation buffer containing 1 mM reduced glutathione, 0.2 mM oxidized glutathione, and 20 mM Tris-HCl (pH 7.3) overnight at 4°C.
Detection of free thiols and disulfide bonds. To detect disulfide bonds, we used a nondirect method based on Ellman's reagent, 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB). DTNB is used directly to estimate the presence of free thiol groups; DTNB reacts with the thiol to give mixed disulfide and 2-nitro-5-thiobenzoic acid (TNB), which is quantified by absorbance of the anion (TNB2–) at 412 nm. In the first step, the sample was carboxymethylated with iodoacetamide, which resulted in modification of free thiols, while the disulfide bonds were left intact. Next, the sample with derivatized free thiols was denatured in 6 M GuHCl-0.1 M Tris-HCl (pH 8.0), and this was followed by reduction of disulfide bonds with 10 mM DTT for 2 h. The excess DTT was removed by dialysis against a detection buffer containing 0.1 M phosphate (pH 8.0). To detect free thiols corresponding to disulfide bonds, after dialysis the sample was treated with Ellman's solution containing 10 mM DTNB and 0.1 M phosphate buffer (pH 8.0), and the absorbance at 412 nm was measured. The concentration of thiols was calculated for the molar absorbance of the TNB anion (E412 = 1.415 x 104 cm–1 M–1). To detect the disulfide bonds, we used 2 nM mature DraB protein or the corresponding cysteine mutant in each reaction.
Partial proteolysis. Partial proteolysis was performed by using proteinase K (Merck, Darmstadt, Germany) at concentrations of 0.1, 1, 10, and 30 μg/ml and 0.1, 1, and 2 mg/ml (total volume, 17 μl). In the assays we used 10 μl of purified DraB complexes with the N-terminal deletion mutants of the DraE protein (1 mg/ml) or purified unresolved complexes of DraB with native DraE, 1DraE, or 2DraE protein at concentrations of 2 mg/ml (as determined by densitometric analysis). Digestion was performed in a buffer containing 2 mM CaCl2 and 20 mM Tris-HCl (pH 7.8) for 20 min at 25°C. Reactions were stopped by adding a proteinase inhibitor, phenylmethylsulfonyl fluoride, to a final concentration of 2 mM.
Two-dimensional (2D) electrophoresis. Isoelectric focusing IEF (first dimension) was performed by using an Ettan IPGphor II system (Amersham Bioscience, Uppsala, Sweden) and 13-cm-long IPG strips with an immobilized pH 3 to 10 gradient (Amersham Bioscience). Periplasm extract samples were loaded onto the IPG strips during rehydration. The rehydration was carried out for 16 h in 250 μl of buffer containing 0.5% (vol/vol) IPG buffer (pH 3 to 10), 0.002% (wt/vol) bromophenol blue, and 80 μl of periplasmic extract (10 mM MgCl2). To protect DraE oligomers from depolymerization, we did not use recommended denaturing agents, such as urea, in the rehydration buffer. The high-molecular-weight oligomers of DraE proteins were not loaded onto IPG strips under the rehydration reaction conditions used. The IEF was performed with a standard voltage gradient (200 to 5,000 V) for 12 h. The IEF strips were then equilibrated for 20 min with buffer containing 30% (vol/vol) glycerol, 2% (wt/vol) SDS, 0.002% (wt/vol) bromophenol blue, and 50 mM Tris-HCl (pH 8.8); this was followed by SDS-PAGE in a 12% polyacrylamide gel. The gels were stained with a ProteoSilver staining kit (Sigma, St. Louis, Mo.) or were analyzed by Western blotting.
Other techniques. SDS-PAGE was carried out by using samples that were heated in Laemmli buffer at 98°C for 5 min or were incubated with Laemmli buffer at 25°C for 10 min but not denatured thermally. Immunoblot detection of DraE and DraB proteins was performed with polyclonal rabbit monospecific anti-DraE and anti-DraB antibodies, respectively, and secondary anti-rabbit goat antibodies labeled with horseradish peroxidase and diaminobenzidine as a reaction substrate.
Dr adhesins were detected on the surface of E. coli strains by using rabbit anti-DraE antibodies and secondary anti-rabbit goat antibodies conjugated with fluorescein isothiocyanate, and stained fimbriae were viewed with an Olympus BX-60 fluorescence microscope system. Densitometric analyses of SDS-PAGE gels stained with Coomassie brilliant blue G-250 were performed with an SDS low-molecular-weight calibration kit (Amersham Bioscience) by using a VersaDoc system (Bio-Rad, Hertfordshire, United Kingdom) with Quantity One software (Bio-Rad). DNA sequencing and determination of the N-terminal sequence of the mature DraB protein were performed commercially (DNA-Gdask II s.c.).
RESULTS AND DISCUSSION
DraB protein belongs to the FGL subfamily of PapD-like chaperones. A draB gene encodes a 265-amino-acid protein with a calculated molecular mass of 28.6 kDa. During the expression process, the DraB protein is secreted into the periplasm. N-terminal sequencing of the mature protein showed that the signal peptidase cut DraB between Ala29 and Ala30; the periplasmic chaperone DraB (236 residues) has a molecular mass of 25.7 kDa and a measured pI of 9.0 (Fig. 1; also see Fig. 4).
Sequence analysis showed that the DraB protein belongs to the conserved family of PapD-like chaperones engaged in the biogenesis of surface-located polymeric adhesive structures in gram-negative bacteria. Alignment of the amino acid sequences of the DraB protein and PapD-like chaperones with known structures (PapD, FimC, and Caf1M) showed that the DraB protein belongs to the FGL subfamily (Fig. 1). Additionally, mature DraB protein contains two conserved cysteines, Cys103 and Cys139, that are homologous to Cys101 and Cys140 of the Caf1M chaperone in the FGL subfamily. The cysteines of Caf1M form a disulfide bond with a crucial function in biogenesis of the F1 capsular antigen (41). The disulfide bond of Caf1M is not required for folding of this protein but is necessary for maintenance of the Caf1M-Caf1 adhesin subunit interaction (41). The mature DraB protein purified from a periplasm extract did not react with Ellman's reagent, which demonstrated that the C103 and C139 residues form a disulfide bond. To investigate the role of these conserved cysteines in the folding of the chaperone and the capacity of DraB to form complexes with the DraE protein, we constructed a DraB mutant in which cysteine 103 was changed to alanine (the pET30b-sygDraBC103A expression plasmid). The C103A DraB mutant purified from the periplasm reacted with Ellman's reagent with a molar stoichiometry that corresponded to one cysteine residue per DraB molecule. We then tested the capacity of native and cysteine mutant DraB proteins to form binary complexes with denatured DraE protein during a renaturation reaction. The DraE protein was expressed in the cytoplasm of E. coli cells as a mature protein (without a signal sequence) from the pET30b-DraE plasmid. In the absence of the DraB chaperone and the reducing environment of the cytoplasm, we observed considerable overproduction of DraE as insoluble inclusion bodies. The inclusion bodies were washed with buffer containing 1% Triton X-100 to remove contaminating proteins and then solubilized under denaturing conditions in 4 M GuHCl-0.5 mM DTT. The final concentration of DraE was estimated to be 40 mg/ml (Fig. 2). This renaturation strategy was based on dilution of a small volume of the concentrated denatured DraE protein (40 mg/ml) by addition of a 160-fold volume of the DraB protein (5 mg/ml) to the refolding buffer. When a denatured DraE sample was mixed with DraB in refolding buffer, the GuHCl concentration decreased from 4 M to 25 mM. Thus, we estimated that DraB-DraE binary complexes isolated from the periplasmic space were stable in buffers containing 25 mM GuHCl. The final concentration of DraE in the mixture was 0.25 mg/ml; accordingly, the renaturation reaction and DraB-DraE complex formation occurred with a 20-fold excess of the chaperone DraB. The refolding mixture was incubated for 20 min at room temperature, and the presence of DraB-DraE complexes was determined by molecular filtration. A single peak on a chromatogram was observed for the C103A mutant of DraB (pET30b-DraBC103A plasmid). This fraction reacted with anti-DraB antibodies but not with anti-DraE antibodies. Inversely, when the native mature DraB protein (with a disulfide bond) was tested, two peaks appeared. The first peak had the same retention time as binary DraB-DraE complexes (which reacted with anti-DraB and anti-DraE antibodies), and the second peak corresponded to native DraB and a cysteine mutant of DraB (which reacted with anti-DraB antibodies). The negative control included a denatured DraE sample diluted in the refolding buffer complemented with albumin at a concentration of 5 mg/ml. In this control assay, the DraE protein precipitated during mixing and was removed during the ultrafiltration step, before chromatography.
To test whether DraB-DraE binary complex formation resulted from a specific reaction between the chaperone and the fimbrial subunit, we performed the same experiment with refolding buffer which contained, in addition to the DraB chaperone (5 mg/ml), BSA (3 mg/ml), carbonic anhydrase (3 mg/ml), and -lactalbumin (4 mg/ml). In this experiment, only the mature DraB protein was able to form soluble binary complexes with the DraE protein. These results agreed with the results of experiments in which the role of the homologous cysteines in the CaF1M chaperone was determined.
Overproduced DraE adhesin has a toxic effect on E. coli cells. Previous studies showed that E. coli cells harboring the pBJN417 plasmid (24) (containing the dra operon with a transposon mutation in a draC gene encoding the outer membrane usher) do not express surface-located Dr fimbriae. The mutant strain also lost the ability to bind to CHO cells expressing the DAF receptor. A similar phenotypic effect was obtained with an E. coli strain harboring the pBJN413 plasmid (containing the dra operon with a transposon mutation in a draB gene) (24). As described above, we showed that the cytoplasmic form of the DraE protein (without its signal sequence) expressed from pET30b-DraE accumulated as insoluble inclusion bodies. When the DraE protein was expressed with its signal sequence (plasmid pET30b-sygDraE), cell growth and the DraE production levels depended on the strength of induction of the cell culture. When the protein was expressed at 30°C with IPTG added to a final concentration of 1 mM (A600, 0.30), it had a toxic effect in connection with overproduction of the DraE protein from a strong T7 promoter. The E. coli cells produced DraE adhesin by 2 h after induction and reached a maximal optical density (A600) of 0.41. Incubation for three additional hours resulted in a reduced optical density (0.32). Under the same cultivation conditions, the control cells, harboring the pET30b vector, grew logarithmically to a maximal optical density of 1.2 (6 h after induction) and reached a stable stationary phase (results not shown). SDS-PAGE and Western blot analysis with anti-DraE antibodies of total cell lysates (taken from cultures during expression) revealed the presence of the DraE protein with and without a signal sequence. When the IPTG concentration was reduced to 0.1 mM, the bacterial cell growth in both cases was almost the same, and Western blotting of total cell lysates showed only the presence of the DraE protein with a signal sequence (cytoplasmic form). The absence of the DraE protein in the periplasm (without a signal sequence) suggests that it was probably efficiently degraded by periplasmic proteases, and no toxic effect was observed. When there was a high level of overproduction (1 mM IPTG), the proteolytic systems probably could not efficiently degrade the misfolded DraE protein, and thus toxicity occurred. These data suggest that fimbrial DraE subunits produced without the DraB chaperone had toxic effects on expressing cells, resulting in accumulation of misfolded DraE protein in the periplasmic space. These data agree with the reports describing expression of pilin subunits of other polymeric structures assembled by the usher-chaperone system (17).
DraB-DraEn complexes accumulate in the periplasm in the absence of the usher DraC. To investigate the mechanism of DraB-DraE complex formation and polymerization of DraE subunits, we constructed the pET30b-sygDraBE plasmid. In this bicistronic construct, the draB and draE genes (each with its own ribosome binding site) were under the control of a common T7 promoter (Tabor-Studier expression system [34]). The draB gene was located proximal to the promoter region, and the draE gene was located distal to the promoter region. One feature of E. coli polycistronic systems is that translation from the first RBS proximal to the promoter is stronger than translation from the RBS sites located distally. Therefore, the level of expression of the DraB chaperone was higher than that of the DraE adhesin from the bicistronic plasmid constructed (Fig. 3A). In accordance with the chaperone-usher pathway, the subunits of outer membrane polymers are able to fold into the native form only by interacting with a cognate chaperone protein to form a chaperone-subunit complex. As first shown for P pili of E. coli, coexpression of the PapD chaperone with pilin subunits eliminated the toxic effect resulting from expression of the subunit alone (17). Thus, in the bicistronic vector constructed, the higher level of expression of the DraB chaperone than of the DraE protein eliminated the previously described toxic effect connected with accumulation of misfolded DraE hemagglutinin. The SDS-PAGE and Western blot analyses of crude extracts of expressing cells harboring plasmid pET30b-sygDraBE revealed two bands that reacted with anti-DraE antibodies corresponding to the DraE protein with and without its signal sequence and also two other bands that reacted with anti-DraB antibodies corresponding to the DraB protein with and without its signal sequence. In isolates isolated from periplasmic fractions, two bands corresponding to the mature forms of the DraB and DraE proteins were detected (Fig. 3C). Densitometric analyses of SDS-PAGE gels containing periplasmic samples showed that the DraB chaperone accounted for almost 40% of the total protein content and DraE accounted for 25% of the total protein content (Fig. 3A).
Two types of interactions are possible in the expression systems that allow production of periplasmic chaperone and fimbrial subunits without an usher protein: (i) formation of a chaperone subunit binary complex in the donor strand complementation reaction and (ii) oligomerization of subunits by the donor strand exchange mechanism. We found that oligomers of DraE subunits (dimers, trimers, tetramers, etc.) formed in the periplasm during expression were stable in Laemmli buffer at the ambient temperature. This property of the oligomers formed in the periplasm allowed analysis by SDS-PAGE when a sample was not denatured thermally. A similar approach was used to characterize Caf1 oligomers of Y. pestis (40). Analysis of a periplasmic extract of E. coli BL21(DE3)/pET30b-sygDraBE by SDS-PAGE (with a sample not denatured thermally) resulted in bands corresponding to the DraB chaperone and a monomer of the DraE subunit, as well as a ladder of bands corresponding to oligomers of the DraE subunit. Western blotting with anti-DraE antibodies revealed the same ladder of oligomer bands and a monomer form of the DraE protein (Fig. 3B). The same sample in a native polyacrylamide gel also produced a ladder of bands that reacted with anti-DraE antibodies and reacted weakly with anti-DraB antibodies (data not shown). These data are consistent with the proposal that DraE oligomers are formed by a conserved chaperone-usher pathway. Because the oligomers are formed in accordance with a donor strand exchange reaction, each is built from head-to-tail-connected DraE subunits. In agreement with this mechanism, the last subunit incorporated into the growing oligomer is complexed with the chaperone DraB. Further analysis by 2D electrophoresis (pH 3 to 10 gradient; molecular weight analysis; SDS—12% polyacrylamide gel) showed that the oligomers had pIs in the range from 4.7 to 4.9 (Fig. 4). 2D electrophoresis permitted us to distinguish five well-resolved oligomers corresponding to DraEn (n = 2 to 6) and high-molecular-weight oligomers.
Using two-step ion-exchange chromatography and the gel filtration technique, we purified oligomers corresponding to DraEn (n = 2 to 5). The DraB-DraE binary complex was unstable at pH 7.0, the pH at which we purified oligomers, and therefore we used cation-exchange chromatography and pH 5.5 buffers to purify the oligomers. Analysis of purified oligomers by SDS-PAGE revealed the presence of the DraE and DraB proteins. To determine the amount of the DraE protein in the periplasmic fraction in an oligomeric form in relation to the total amount of DraE, we performed densitometry analyses of SDS-polyacrylamide gels stained with Coomassie brilliant blue G-250. The total amount of DraE was evaluated on gels containing resolved periplasmic samples that were denatured at 98°C for 5 min. The concentration of the DraE protein not in oligomers was determined by using the same periplasmic samples incubated at room temperature in Laemmli buffer. The densitometric data showed that almost 70% of the total DraE protein was in an oligomeric form (Fig. 3B). These data showed that in E. coli BL21(DE3)/pET30b-sygDraBE, in the absence of the DraC usher protein, DraB-DraE binary complexes and DraB-DraEn oligomers accumulated in the periplasm. In the presence of excess chaperone DraB, the DraE protein appeared mainly in an oligomeric form, which suggested that the DraE polymerization process is more favorable than the formation of DraB-DraE binary complexes or that the DraE oligomers are more stable thermodynamically than the DraB-DraE complexes. These data agree with the thermodynamic models of subunit polymerization proposed independently by Sauer et al. (26) and Zavialov et al. (39). According to these models, the energy driving the donor strand exchange reaction is stored in a high-energy, molten-globule-like subunit in a complex with the chaperone. During polymerization, subunits rearrange structurally to form a more rigid core with lower energy.
N-terminal strand of the DraE subunit is required for DraE oligomer formation. In surface-located polymeric systems in which biogenesis occurs by the usher-chaperone mechanism, the subunits possess two critical sequences responsible for polymerization: the N-terminal donor strand and a C-terminal strand homologous to -strand F of P and type 1 pilus subunits (40). During polymerization, the incoming subunit for the donor strand exchange reaction depletes the chaperone protein from the complex with the last subunit of a growing polymer. Crystallographic data showed that during this reaction the N-terminal strand for the incoming subunit exchange prevented the donor strand of the chaperone from interacting with -strand F of the last subunit of the polymer. Due to the limited sequence similarity between polymeric subunits for FGL chaperones, we could not precisely predict the minimal sequence of a donor strand in the DraE protein. Thus, mutation analysis of the N-terminal region comprising residues 1 to 11 of the mature DraE protein was performed. In this sequence we distinguished two fragments: two consecutive GTT sequences (residues 6 to 11) and a region comprising the first five N-terminal residues, which had no unique features. Using the pET30b-sygDraBE plasmid, we constructed N-terminal deletion mutants of the DraE protein, including 1DraE, 2DraE, 3DraE, 4DraE, 5DraE, 8DraE, and 11DraE (Table 1). Periplasmic samples isolated from all of the mutants incubated at the ambient temperature or at 98°C were analyzed by SDS-PAGE and Western blotting by using anti-DraE and anti-DraB antibodies. A periplasmic extract from E. coli Bl21(DE3)/pET30b-sygDraBE was the reference point in the analysis of the extracts from DraE deletion mutants. To determine the region comprising the signal for polymerization, we first prepared three long N-terminal deletion mutants, 11DraE, 8DraE, and 5DraE. These mutants did not all polymerize to oligomeric forms detectable by SDS-PAGE and Western blotting (samples were not denatured thermally) (Fig. 5A). Using cation-exchange and exclusion chromatographic techniques, we purified all of the DraE mutants as stable complexes with the DraB chaperone. Western blot analysis of eluted fractions with the anti-DraE and anti-DraB antibodies proved that the DraE and DraB proteins were present. The level of recovery of 5DraE was comparable to the level of recovery of the wild-type DraE protein. In the case of 8DraE and 11DraE the amount of mutated protein was much smaller than the amount of wild-type DraE protein.
To investigate the potential effect of periplasmic DegP protease on the low levels of 8DraE and 11DraE recovery, we performed experiments with E. coli B178(DE3)htrA63 (with a transposon mutation in the htrA [degP] gene). When the 8DraE and 11DraE proteins were expressed without the chaperone DraB (plasmids pET30b-syg8DraE and pET30b-syg11DraE, respectively) in degP+ cells, we did not obtain a signal for these proteins in SDS-PAGE and Western blot analyses. Previously described data showed that expression in a degP+ strain harboring plasmid pET30b-sygDraB8/11E resulted in the formation of binary complexes of DraB with a cognate deletion mutant of DraE. This suggested that DraB reacts with long deletion mutants of DraE and, to some degree, protects them from proteolysis. Expression from E. coli strain B178(DE3)htrA63 transformed with plasmid pET30b-sygDraB8/11E resulted in a level of recovery of deletion mutants of DraE that was higher than the level obtained for the degP+ strain (data not shown). These data suggest that the long deletion mutants of DraE probably have a more flexible structure in complexes with the DraB chaperone and are more susceptible to degradation by periplasmic proteases.
To further determine the N-terminal amino acids of DraE required for polymerization, we constructed the following mutants: 4DraE, 3DraE, 2DraE, and 1DraE. The 3DraE and 4DraE mutant proteins did not polymerize (oligomers were not detectable in Western blots of SDS-polyacrylamide gels; samples were not denatured thermally) (Fig. 5B). The properties of these mutants were very similar to those described above for the 5DraE protein. The periplasmic extract from cells expressing the 2DraE mutant assayed by SDS-PAGE (samples were not denatured thermally) and Western blotting with anti-DraE antibodies contained oligomers composed of at most five subunits of 2DraE, and high-molecular-weight oligomers were not detected (Fig. 5B). As was the case with the 11DraE, 8DraE, and 5DraE deletion mutants, we purified the 4DraE, 3DraE, 2DraE, and 1DraE proteins as stable binary complexes with the DraB protein. The oligomeric properties of the 1DraE mutant were almost the same as those of the native DraE (Fig. 5B).
These data showed that deletion of the first three residues from the N terminus of the mature DraE protein resulted in a mutant which was capable of forming only binary complexes with chaperone DraB and was deficient in formation of DraE oligomers.
To determine the properties of the constructed N-terminal deletion mutants of the DraE protein for the formation of Dr fimbriae in vivo, the pBJN17 plasmid (containing the dra operon with a transposon mutation in the draE gene) (24) was introduced into E. coli BL21(DE3) strains containing constructed plasmids encoding DraE N-terminal mutants. The transformants obtained were cultivated on Luria-Bertani agar plates complemented with IPTG (6 mM), an inducer of DraE mutant protein expression from constructed plasmids. The presence of Dr fimbriae on the surface of E. coli cells was analyzed by an immunofluorescence assay by using anti-DraE antibodies. Only the strain transformed with pET30b-sygDraB1E expressed surface-located Dr fimbriae. The amount of 1DraE was similar to the amount in cells transformed with control plasmid pET30b-sygDraBE (data not shown).
Oligomerization of DraE subunits increases their resistance to proteolysis. The data described above showed that deletion of more than two residues from the N terminus of the DraE protein diminished the capacity to form DraE oligomers but did not influence the formation of DraB-DraE binary complexes. The longer deletions (8DraE and 11DraE) resulted in DraE proteins that were more susceptible to degradation by periplasmic proteases. To investigate the effect of N-terminal deletions on the stability of the DraE protein, we examined the resistance of all mutants obtained to proteolysis by proteinase K (Table 1). Protein complexes of the DraB chaperone with mutant or native DraE, purified by ion-exchange chromatography, were treated with different concentrations of proteinase K (0.1 μg/ml to 2 mg/ml) for 20 min at 25°C. The 8DraE and 11DraE proteins were not detected in samples treated with proteinase K at a concentration of 0.5 μg/ml. The 3DraE, 4DraE, and 5DraE mutants were much more resistant to degradation by proteinase K; the 3DraE protein was detected with anti-DraE antibodies in samples incubated with 30 μg of proteinase K per ml at 25°C.
To measure the resistance of oligomeric DraE proteins in complexes with DraB, we used samples purified by ion-exchange chromatography. Following proteolysis, samples were analyzed by SDS-PAGE (without thermal denaturation) and by Western blotting (Fig. 6). Treatment with proteinase K (10 μg/ml) resulted in the disappearance of monomeric mature DraE (DraB-DraE binary complex) and the appearance of a new band at a molecular mass that was 1.5 kDa lower than that of mature DraE. Western blotting showed that this band reacted with anti-DraE antibody (Fig. 6). In addition, the conentration of low-molecular-weight oligomers decreased. Treatment with 2 mg of proteinase K per ml resulted in only the bands corresponding to high-molecular-weight oligomers (Fig. 6), and similar results were obtained when a proteinase K concentration of 10 mg/ml was used (data not shown). The properties of the 1DraE protein were the same as those of the native DraE protein. The resistance of the 2DraE protein to proteolysis was similar to the resistance of low-molecular-weight oligomers of native DraE protein.
The efficiency of proteolysis by proteinase K, the endoprotease, depends on the level of accessibility to the core of the degraded protein. The results of proteolysis of the 8DraE and 11DraE mutants confirmed that these proteins have a more accessible core (flexible structure) than the native DraE protein and, therefore, are more susceptible to degradation. The crystalline structure of the ternary complex of the Caf1M chaperone with a dimer of Caf1 subunits (39) and the models of polymerization proposed for P and type 1 pili (12) show that during the formation of oligomers the subunits rearrange in a more rigid form. During the biogenesis of the rigid pili (P and type 1 pili), there are second-order interactions in addition to head-to-tail interactions between subunits, which also stabilize the polymer. Our data for proteolytic resistance of oligomeric forms of the native DraE protein agree with the data obtained previously for other polymers.
Disulfide bond of DraE is necessary in oligomer formation. An experiment in which we examined the refolding of denatured DraE protein showed that the native DraB protein formed binary complexes with the DraE subunit. The denatured DraE protein used in this experiment was in a form with reduced disulfide bonds. To investigate if the disulfide bond in DraE is necessary for oligomerization, the DraB-DraE (Cys-reduced) binary complexes obtained were dialyzed against oxidizing buffer containing reduced and oxidized glutathione. To detect the appearance of oligomers after oxidation, the samples were analyzed by SDS-PAGE (samples were not denatured thermally) and stained with silver. An intense band corresponding to the DraE monomer and three additional weak bands, which migrated as DraEn oligomers (n = 2 to 4), were observed. Western blot analysis showed that these bands reacted with anti-DraE antibodies. To verify that the oligomers of DraE detected were not formed due to intersubunit disulfide bonds, before electrophoresis the samples were incubated for 10 min at 50°C in Laemmli buffer with double the normal concentration of DTT. The same refolding and oxidizing experiment was performed with denatured DraE C21A mutant protein (expression plasmid pET30b-DraEC21A). DraEnC21A oligomers were not detected on a silver-stained SDS-polyacrylamide gel. These assays showed that the disulfide bond in the DraE protein is not required to form a binary complex with the DraB chaperone but is important in oligomerization.
Disulfide bond of the DraE subunit is not necessary for maintenance of the polymeric structure of DraE oligomers. In the experiments described above, samples of DraE oligomers prepared by incubation in Laemmli buffer without thermal denaturation were easily detected by SDS-PAGE. Additionally, the oligomers purified from the periplasm and treated with DTT at the same concentration as the concentration in Laemmli buffer reacted with Ellman's reagent. These experiments suggested that disulfide bonds in DraE subunits were not necessary for maintaining the oligomeric structure. We determined the influence of the DraE disulfide bond on the stability of the DraE oligomers by incubating them in Laemmli buffer with and without DTT at temperatures from 50 to 90°C (Fig. 7). Western blot analysis showed that oligomers incubated with buffer containing DTT were depolymerized almost totally at 60°C, but the oligomers incubated with the same buffer without DTT were highly visible at 70°C. Treatment with proteinase K showed that resistance to proteolysis increased continuously from the low-molecular-weight DraE oligomers to the high-molecular-weight DraE oligomers. Analysis of oligomer resistance to depolymerization in the temperature gradient showed that all oligomers disappeared simultaneously at the same temperature. This suggests that the observed depolymerization was probably a secondary effect induced by thermal denaturation of the DraE subunit. At this stage the lower thermal resistance to destruction of oligomers treated with DTT may be accomplished simply with a decrease in the stability of the DraE protein caused by reduction of the disulfide bond.
In summary, data presented in this paper support the hypothesis that the chaperone-usher pathway for FGL polymers is conserved and that there are specific properties characteristic only of FGL subfamily polymers and not of FGS subfamily polymers. The DraB and Caf1M chaperones have a conserved disulfide bond, which is important in the formation of a binary complex. In the DraB-DraE and Caf1M-Caf1 expression systems (with the usher protein absent) oligomers with the same stoichiometry of subunits accumulated in the periplasm. DraE and Caf1 oligomers also have similar physicochemical properties, DraE and Caf1 oligomers are resistant to depolymerization during incubation in Laemmli buffer at ambient temperatures, and binary complex chaperone subunits of both systems are not stable at pH values above 7.5 (40). A comparison of data obtained from partial proteolysis indicated that DraE oligomers and also binary complexes of the DraB chaperone with DraE mutants were more resistant to degradation than Caf1 oligomers and mutant complexes. This difference in proteolysis may be due to increased internal stability of the DraE subunit (the presence of a disulfide bond). A more extensive intersubunit interaction in the DraE polymer may contribute to greater stability of the fimbrial structure of E. coli Dr hemagglutinin. Inversely, lower internal stability and decreased intersubunit interactions may contribute to the collapse of the Caf1 polymer structure and capsular appearance of the F1 antigen of Y. pestis.
ACKNOWLEDGMENTS
This work was supported by the Polish State Committee for Scientific Research (projects 3 PO4B 008 23, 3 PO5A 079 23, and 3 PO5A 05023 to J.K.) and in part by grant NIH/R01 HD41687-01 to B.N.
REFERENCES
1. Barnhart, M. M., F. G. Sauer, J. S. Pinkner, and S. J. Hultgren. 2003. Chaperone-subunit-usher interactions required for donor strand exchange during bacterial pilus assembly. J. Bacteriol. 185:2723-2730.
2. Barnhart, M. M., J. S. Pinkner, G. E. Soto, F. G. Sauer, S. Langermann, G. Waksman, C. Frieden, and S. J. Hultgren. 2000. PapD-like chaperones provide the missing information for folding of pilin proteins. Proc. Natl. Acad. Sci. USA 97:7709-7714.
3. Behrens, S. 2003. Periplasmic chaperones—preservers of subunit folding energy for organelle assembly. Cell 113:587-596.
4. Bullitt, E., and L. Makowski. 1995. Structural polymorphism of bacterial adhesion pili. Nature 373:164-167.
5. Chapman, D. A., A. V. Zavialov, T. V. Chernovskaya, A. V. Karlyshev, G. A. Zav'yalova, A. M. Vasiliev, I. V. Dudich, V. M. Abramov, V. P. Zav'yalov, and S. MacIntyre. 1999. Structural and functional significance of the FGL sequence of the periplasmic chaperone Caf1M of Yersinia pestis. J. Bacteriol. 181:2422-2429.
6. Choudhury, D., A. Thompson, V. Stojanoff, S. Langermann, J. Pinkner, S. J. Hultgren, and S. D. Knight. 1999. X-ray structure of the FimC-FimH chaperone-adhesin complex from uropathogenic Escherichia coli. Science 285:1061-1066.
7. Dodson, K. W., F. Jacob-Dubuisson, R. T. Striker, and S. J. Hultgren. 1993. Outer-membrane PapC molecular usher discriminately recognizes periplasmic chaperone-pilus subunit complexes. Proc. Natl. Acad. Sci. USA 90:3670-4367.
8. Galyov, E. E., A. V. Karlishev, T. V. Chernovskaya, D. A. Dolgikh, O. Smirnov, K. I. Volkovoy, V. M. Abramov, and V. P. Zav'yalov. 1991. Expression of the envelope antigen F1 of Yersinia pestis is mediated by the product of caf1M gene having homology with the chaperone protein PapD of Escherichia coli. FEBS Lett. 286:79-82.
9. Garcia, M. I., A. Labigne, and C. Le Bouguenec. 1994. Nucleotide sequence of the afimbrial-adhesin-encoding afa-3 gene cluster and its translocation via flanking IS1 insertion sequences. J. Bacteriol. 176:7601-7613.
10. Garcia, M. I., M. Jouve, J. P. Nataro, P. Gounon, and C. Le Bouguenec. 2000. Characterization of the AfaD-like family of invasins encoded by pathogenic Escherichia coli associated with intestinal and extra-intestinal infections. FEBS Lett. 479:111-117.
11. Goluszko, P., S. L. Moseley, L. D. Truong, A. Kaul, J. R. Williford, R. Selvarangan, S. Nowicki, and B. Nowicki. 1997. Development of experimental model of chronic pyelonephritis with Escherichia coli O75:K5:H-bearing Dr fimbriae: mutation in the dra region prevented tubulointerstitial nephritis. J. Clin. Investig. 99:1662-1672.
12. Hahn, E., P. Wild, U. Hermanns, P. Sebbel, R. Glockshuber, M. Haner, N. Taschner, P. Burkhard, U. Aebi, and S. A. Muller. 2002. Exploring the 3D molecular architecture of Escherichia coli type 1 pili. J. Mol. Biol. 323:845-857.
13. Holmgren, A., and C. I. Branden. 1989. Crystal structure of chaperone protein PapD reveals an immunoglobulin fold. Nature 342:248-251.
14. Hung, C. S., J. Bouckaert, D. Hung, J. Pinkner, C. Widberg, A. DeFusco, C. G. Auguste, R. Strouse, S. Langermann, G. Waksman, and S. J. Hultgren. 2002. Structural basis of tropism of Escherichia coli to the bladder during urinary tract infection. Mol. Microbiol. 44:903-915.
15. Hung, D. L., J. S. Pinkner, S. D. Knight, and S. J. Hultgren. 1999. Structural basis of chaperone self-capping in P pilus biogenesis. Proc. Natl. Acad. Sci. USA 96:8178-8183.
16. Hung, D. L., S. D. Knight, R. M. Woods, J. S. Pinkner, and S. J. Hultgren. 1996. Molecular basis of two subfamilies of immunoglobulin-like chaperones. EMBO J. 15:3792-3805.
17. Jones, C. H., P. N. Danese, J. S. Pinkner, T. J. Silhavy, and S. J. Hultgren. 1997. The chaperone-assisted membrane release and folding pathway is sensed by two signal transduction systems. EMBO J. 16:6394-6406.
18. Jouve, M., M. I. Garcia, P. Courcoux, A. Labigne, P. Gounon, and C. Le Bouguenec. 1997. Adhesion to and invasion of HeLa cells by pathogenic Escherichia coli carrying the afa-3 gene cluster are mediated by the AfaE and AfaD proteins, respectively. Infect. Immun. 65:4082-4089.
19. Knight, S. D., D. Choudhury, S. Hultgren, J. Pinkner, V. Stojanoff, and A. Thompson. 2002. Structure of the S pilus periplasmic chaperone SfaE at 2.2 A resolution. Acta Crystallogr. D 58:1016-1022.
20. Kuehn, M. J., D. J. Ogg, J. Kihlberg, L. N. Slonim, K. Flemmer, T. Bergfors, and S. J. Hultgren. 1993. Structural basis of pilus subunit recognition by the PapD chaperone. Science 262:1234-1241.
21. Le Bouguenec, C., M. I. Garcia, V. Ouin, J. M. Desperrier, P. Gounon, and A. Labigne. 1993. Characterization of plasmid-borne afa-3 gene clusters encoding afimbrial adhesins expressed by Escherichia coli strains associated with intestinal or urinary tract infections. Infect. Immun. 61:5106-5114.
22. Lipinska, B., O. Fayet, L. Baird, and C. Georgopoulos. 1989. Identification, characterization, and mapping of the Escherichia coli htrA gene, whose product is essential for bacterial growth only at elevated temperatures. J. Bacteriol. 171:1574-1584.
23. Normark, S. 2000. Anfinsen comes out of the cage during assembly of the bacterial pilus. Proc. Natl. Acad. Sci. USA 97:7670-7672.
24. Nowicki, B., J. Moulds, R. Hull, and S. Hull. 1988. A hemagglutinin of uropathogenic Escherichia coli recognizes the Dr blood group antigen. Infect. Immun. 56:1057-1060.
25. Pham, T., A. Kaul, A. Hart, P. Goluszko, J. Moulds, S. Nowicki, D. M. Lublin, and B. J. Nowicki. 1995. dra-related X adhesins of gestational pyelonephritis-associated Escherichia coli recognize SCR-3 and SCR-4 domains of recombinant decay-accelerating factor. Infect. Immun. 63:1663-1668.
26. Sauer, F. G., J. S. Pinkner, G. Waksman, and S. J. Hultgren. 2002. Chaperone priming of pilus subunits facilitates a topological transition that drives fiber formation. Cell 111:543-551.
27. Sauer, F. G., K. Futterer, J. S. Pinkner, K. W. Dodson, S. J. Hultgren, and G. Waksman. 1999. Structural basis of chaperone function and pilus biogenesis. Science 285:1058-1061.
28. Sauer, F. G., M. Barnhart, D. Choudhury, S. D. Knight, G. Waksman, and S. J. Hultgren. 2000. Chaperone-assisted pilus assembly and bacterial attachment. Curr. Opin. Struct. Biol. 10:548-556.
29. Sauer, F. G., S. D. Knight, G. J. Waksman, and S. J. Hultgren. 2000. PapD-like chaperones and pilus biogenesis. Semin. Cell Dev. Biol. 11:27-34.
30. Saulino, E. T., D. G. Thanassi, J. S. Pinkner, and S. J. Hultgren. 1998. Ramifications of kinetic partitioning on usher-mediated pilus biogenesis. EMBO J. 17:2177-2185.
31. Saulino, E. T., E. Bullitt, and S. J. Hultgren. 2000. Snapshots of usher-mediated protein secretion and ordered pilus assembly. Proc. Natl. Acad. Sci. USA 97:9240-9245.
32. Soto, G. E., and S. J. Hultgren. 1999. Bacterial adhesins: common themes and variations in architecture and assembly. J. Bacteriol. 181:1059-1071.
33. Soto, G. E., K. W. Dodson, D. Ogg, C. Liu, J. Heuser, S. Knight, J. Kihlberg, C. H. Jones, and S. J. Hultgren. 1998. Periplasmic chaperone recognition motif of subunits mediates quaternary interactions in the pilus. EMBO J. 17:6155-6167.
34. Studier, F. W., A. H. Rosenberg, J. J. Dunn, and J. W. Dubendorff. 1990. Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol. 185:60-89.
35. Swanson, T. N., S. S. Bilge, B. Nowicki, and S. L. Moseley. 1991. Molecular structure of the Dr adhesin: nucleotide sequence and mapping of receptor-binding domain by use of fusion constructs. Infect. Immun. 59:261-268.
36. Thanassi, D. G., and S. J. Hultgren. 2000. Assembly of complex organelles: pilus biogenesis in gram-negative bacteria as a model system. Methods 20:111-126.
37. Thanassi, D. G., E. T. Saulino, M. J. Lombardo, R. Roth J. Heuser, and S. J. Hultgren. 1998. The PapC usher forms an oligomeric channel: implications for pilus biogenesis across the outer membrane. Proc. Natl. Acad. Sci. USA 95:3146-3151.
38. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673-4680.
39. Zavialov, A. V., J. Berglund, A. F. Pudney, L. J. Fooks, T. M. Ibrahim, S. MacIntyre, and S. D. Knight. 2003. Structure and biogenesis of the capsular F1 antigen from Yersinia pestis: preserved folding energy drives fiber formation. Cell 30:587-596.
40. Zavialov, A. V., J. Kersley, T. Korpela, V. P. Zav'yalov, S. MacIntyre, and S. D. Knight. 2002. Donor strand complementation mechanism in the biogenesis of non-pilus systems. Mol. Microbiol. 45:983-995.
41. Zav'yalov, V. P., T. V. Chernovskaya, D. A. Chapman, A. V. Karlyshev, S. MacIntyre, A. V. Zavialov, A. M. Vasiliev, A. I. Denesyuk, G. A. Zav'yalova, I. V. Dudich, T. Korpela, and V. M. Abramov. 1997. Influence of the conserved disulphide bond, exposed to the putative binding pocket, on the structure and function of the immunoglobulin-like molecular chaperone Caf1M of Yersinia pestis. Biochem. J. 324:571-578.(Rafa Pitek, Beata Zalewsk)
Department of Obstetrics and Gynecology
Department of Microbiology and Immunology, The University of Texas Medical Branch, Galveston, Texas
ABSTRACT
The Dr hemagglutinin of uropathogenic Escherichia coli is a fimbrial homopolymer of DraE subunits encoded by the dra operon. The dra operon includes the draB and draC genes, whose products exhibit homology to chaperone-usher proteins involved in the biogenesis of surface-located polymeric structures. DraB is one of the periplasmic proteins belonging to the superfamily of PapD-like chaperones. It possesses two conserved cysteine residues characteristic of the FGL subfamily of Caf1M-like chaperones. In this study we obtained evidence that DraB cysteines form a disulfide bond in a mature chaperone and have the crucial function of forming the DraB-DraE binary complex. Expression experiments showed that the DraB protein is indispensable in the folding of the DraE subunit to a form capable of polymerization. Accumulation of DraB-DraEn oligomers, composed of head-to-tail subunits and the chaperone DraB, was observed in the periplasm of a recombinant E. coli strain which expressed DraB and DraE (but not DraC). To investigate the donor strand exchange mechanism during the formation of DraE oligomers, we constructed a series of DraE N-terminal deletion mutants. Deletion of the first three N-terminal residues of a potential donor strand resulted in a DraE protein lacking an oligomerization function. In vitro data showed that the DraE disulfide bond was not needed to form a binary complex with the DraB chaperone but was essential in the polymerization process. Our data suggest that assembly of Dr fimbriae requires a chaperone-usher pathway and the donor strand exchange mechanism.
INTRODUCTION
The recognition of host receptors and attachment to them are crucial steps in the process of bacterial pathogenesis. Gram-negative bacteria express surface polymeric adhesive organelles that are responsible for specific recognition and attachment to host tissues (32). In some cases (e.g., Escherichia coli type 1 fimbriae or Dr/Afa-III adhesins [10, 11, 18]) these structures are important in promoting the invasion of host cells by bacterial cells and have been implicated in recurrent or chronic infections. These polymeric structures may form fimbriae or afimbrial capsules surrounding bacterial cells (28, 32). Biogenesis of these polymers occurs in gram-negative bacteria by four main pathways, the chaperone-usher, alternative chaperone-usher, type IV secretion, and type IV pilus assembly pathways (32).
The family of more than 30 different fimbrial and nonfimbrial polymeric adhesins expressed by different gram-negative bacteria contains members assembled with a common chaperone-usher mechanism (29, 36). This family encompasses complex, heteropolymeric rigid fimbriae (e.g., P pili and type 1 pili of E. coli) (4, 12), much simpler homopolymeric fimbriae with a flexible structure (e.g., Dr fimbriae of uropathogenic E. coli) (25, 35), and nonfimbrial or amorphic structures (e.g., F1 capsular antigen of Yersinia pestis and Afa-III of E. coli) (8, 9). The process of biogenesis of polymeric adhesins in the chaperone-usher pathway depends on the action of two periplasmic proteins, the usher and chaperone proteins. The usher protein is a homooligomer that forms an outer membrane, donut-shaped channel by which the polymerization process occurs (1, 7, 31, 37). The chaperone is a periplasmic protein which has crucial roles in the assembly of polymers (20, 33). As shown first for P and type 1 pili, the subunits that are used to build these polymers after translocation to the periplasm are not able to fold spontaneously to form structures that are capable of polymerization (17, 30). In the periplasm, the chaperone protein interacts with fimbrial subunits, and based on steric information in the structure, the subunits fold to construct a native functional form (2, 23). Only when they are in a complex with the chaperone are subunits able to polymerize in contact with the usher protein to form surface-located adhesive polymers. Accumulation of free subunits in the periplasm induces the toxic effect that results in their degradation by proteases, mainly DegP (HtrP) (17). The chaperone bound to a subunit stabilizes it and caps its interactive regions, thereby protecting the periplasm from a potential toxic effect.
The periplasmic chaperones engaged in assembly of polymeric structures belong to a conserved superfamily whose members have high sequence homology and a common structure (29). Thus far, structures of the following four chaperones have been elucidated: PapD (P pili of E. coli) (13, 15), FimC (type 1 pili of E. coli) (6), SfaE (S pilus of E. coli) (19), and Caf1M (F1 antigen of Y. pestis) (39). All of these molecules are composed of two immunoglobulin-like domains with the overall shape of a boomerang. Comparison of the sequences of known chaperones with crystallographic structure data permits division of the superfamily into two groups, FGL (F1-G1 Long) and FGS (F1-G1 short) (5, 16). The base of this division is the number of residues in the loop connecting strands. Chaperones belonging to the FGS subfamily are engaged in the assembly of rod-like fimbriae (PapD and FimC), whereas FGL chaperones appear to assemble thin fimbrial and afimbrial structures (Caf1M) (16).
The crystallographic structures of the chaperone-subunit binary complexes PapD-PapK (27), PapD-PapE (26), FimC-FimH (6, 14), and Caf1M-Caf1 and the ternary complex Caf1M-Caf1'-Caf1" (39) permitted discovery of the molecular mechanism of biogenesis of organelles which are assembled by a chaperone-usher pathway. The crystallographic data showed that all subunits possess incomplete immunoglobulin structures with the seventh -strand missing, which results in exposure of a deep, hydrophobic cleft. In complex crystallographic structures, the chaperone completes this cleft with its G1 -strand in a donor strand complementation process. During polymerization, the chaperone is removed from the last subunit of the growing polymer by an incoming subunit, resulting in head-to-tail polymerization. This mechanism is called donor strand exchange, in which the G1 donor -strand of the chaperone is replaced by a similar interaction arising from the N-terminal strand of the subunit (29). Recently published crystallographic structures of Caf1M-Caf1 complexes permitted researchers to determine the differences in the activities of FGL and FGS chaperone-usher pathways (3, 39).
The adhesins of the Dr family in uropathogenic and diarrheal E. coli are encoded by operons that have similar gene organizations. Each operon contains genes encoding the proteins responsible for adhesin biogenesis and the group of genes encoding proteins that regulate the process of expression (21, 35). The amino acid sequences of accessory proteins are rather conserved in the entire family, but adhesin subunits have limited homology.
The Dr hemagglutinin is encoded by a dra operon in uropathogenic E. coli. DraE subunits form a fimbrial structure with an adhesive capacity. The dra gene cluster also encodes the DraC usher and DraB chaperone proteins. DraB-DraE has only limited homology to the previously characterized binary chaperone-subunit complexes PapD-PapK, PapD-PapE, FimC-FimH, and Caf1M-Caf1. It is, therefore, difficult to predict whether the DraB-DraE binary complex operates in a manner similar to these systems.
Here we constructed a recombinant in vitro and in vivo system for expression of DraB and DraE in the periplasm. We then constructed DraE deletion mutants and investigated the donor strand exchange mechanism in DraE oligomer formation. The results of our experiments are consistent with the interpretation that Dr fimbriae are assembled by the conserved chaperone-usher pathway (as proposed previously on the basis of sequence analyses of the dra operon).
MATERIALS AND METHODS
Strains, plasmids, and mutagenesis. E. coli Top10F' (Invitrogen, Paisley, United Kingdom) was used as the host strain for cloning. E. coli BL21(DE3) (Novagen, Darmstadt, Germany) was used as the expression host. The degP-deficient E. coli B178(DE3)htrA63 expression host was constructed by introducing a gene encoding T7 RNA polymerase into the degP strain E. coli B178htrA63(mini-Tn10) (22) by using a DE3 lysogenization kit (Novagen).
All expression vectors described in this study were derived from the pET30b(+) plasmid (Novagen). Plasmid pET30b-sygDraBE encoding the DraB and DraE proteins in a bicistronic system was constructed in two steps. The draB gene encoding the native DraB protein (with the signal sequence) was amplified by PCR by using forward primer DraB-syg (5'-TCTAGAAATAATTTTGTTTAACTTTAAGAAGGAGATATACATATGAAAATGCGGGCTGTGGCTGTGTTCACCGGC) and reverse primer DraB-stop (5'-TATGGATCCTCAACCCTTCAGCTCTGCCTCAAACTGCTTAC). The 5' overhanging sequence of the forward primer contains an XbaI site (underlined) and a ribosome binding site (RBS) (boldface type), and the 5' overhanging sequence of the reverse primer contains a BamHI site (underlined). The 876-bp PCR product encoding the DraB protein was digested with BamHI and ligated into pUC19 (New England Biolabs) digested with SmaI and BamHI, resulting in recombinant plasmid pUC19-sygDraB. The draB gene was excised from the pUC19-sygDraB plasmid with XbaI and BamHI and ligated into the XbaI and BamHI sites of expression vector pET30b(+) to obtain pET30b-sygDraB. The draE gene was amplified with the following primers: forward primer DraE-syg (5'-ATAGGATCCAACTTTAAGAAGGAGATATACATATGAAAAAATTAGCGATCATGGCCGCGGCCAG) and DraE-stop (5'-TATAAGCTTTCATTTTGCCCAGTAACCCCCGGTCAGGGTC). The forward DraE-syg primer contained an RBS (boldface type) and a BamHI site (underlined), and the DraE-stop primer contained a HindIII site (underlined). The 521-bp PCR product encoding the native DraE protein was digested with BamHI and HindIII and ligated into pET30b-sygDraB digested with the same endonucleases to obtain the expression plasmid pET30b-sygDraBE.
The plasmids encoding native DraB and N-terminal deletion mutants of the DraE protein were constructed by cutting out the 432-bp SacI-HindIII DNA fragment from pET30b-sygDraBE and ligating the PCR products encoding N-truncated DraE proteins into the same restriction sites. The SacI site is located at the end of the sequence coding for the Sec-dependent signal sequence in the DraE protein. The PCR products encoding the N-terminal mutants of DraE were obtained by using the reverse primer DraE-stop described above and the following forward primers (SacI sites are underlined): 1DraE, 5'-ATAGAGCTCCGCTCACGCGTTCACCCCGAGTGGCACCACCG; 2DraE, 5'-ATAGAGCTCCGCTCACGCGACCCCGAGTGGCACCACCGG; 3DraE, 5'-ATAGAGCTCCGCTCACGCGCCGAGTGGCACCACCGGCACC; 4DraE, 5'-ATAGAGCTCCGCTCACGCGAGTGGCACCACCGGCACCAC; 5DraE, 5'-ATAGAGCTCCGCTCACGCGGGCACCACCGGCACCACCAAAC; 8DraE, 5'-ATAGAGCTCCGCTCACGCGGGCACCACCAAACTCACAGTTACCGAAGAGTGCCAGGT; and 11DraE, 5'-ATAGAGCTCCGCTCACGCGAAACTCACAGTTACCGAAGAGTGCCA.
The pET30b-syg8DraE and pET30b-syg11DraE plasmids were constructed by cutting the 861-bp XbaI-BamHI fragment containing the draB gene out of the respective bicistronic plasmids (pET30b-sygDraB8E and pET30b-sygDraB11E). The overhangs were filled in by using the Pwo DNA polymerase, and the blunt-ended plasmids were then religated, resulting in plasmids that expressed the 8DraE and 11DraE proteins.
The pET30b-sygDraBC103A plasmid, encoding an alanine mutant at cysteine 103 of the DraB protein, was created from the pET30b-sygDraB plasmid by using a QuikChange site-directed mutagenesis kit according to the instructions of the manufacturer (Stratagene, La Jolla, Calif.).
The pET30b-DraE plasmid was constructed by cloning the 441-bp PCR product encoding the DraE protein without the signal sequence into the NdeI and HindIII sites of the pET30b(+) plasmid with forward primer 5'-ATAATACATATGGGGTTCACCCCGAGTGG (NdeI site underlined) and reverse primer 5'-TTGAAGCTTTCATTTTGCCCAGTAACCCCC (HindIII site underlined). The pET30b-DraEC21A plasmid, encoding an alanine mutant with a mutation at cysteine 21 of the DraE protein, was constructed from the pET30b-DraE plasmid by using a QuikChange site-directed mutagenesis kit (Stratagene).
In all PCRs the Pwo DNA polymerase (DNA-Gdask II s.c., Gdask, Poland) was used. After enzymatic reactions the DNA fragments were purified by using a DNA Clean Up kit (A&A Biotechnology, Gdynia, Poland). The nucleotide sequences of all recombinant plasmids were confirmed by sequencing (DNA-Gdask II s.c.).
Protein expression and purification. The E. coli BL21(DE3) and B178(DE3)htrA63 strains transformed with appropriate plasmids were typically cultivated to an A600 of 0.4 in Luria-Bertani medium supplemented with 20 μg of kanamycin per ml, and protein expression was induced by adding isopropyl--D-thiogalactopyranoside (IPTG) to a final concentration of 0.5 mM for 2 h. Then the cultures were harvested by centrifugation, and periplasmic fractions containing the target proteins were extracted by the osmotic shock procedure (41). To purify complexes of the DraB chaperone with DraE oligomers, the periplasmic fraction obtained from 1 liter of culture was dialyzed overnight against buffer A containing 10 mM NaCl and 20 mM Tris-HCl (pH 7.0). The dialyzed sample was applied to a Resource Q 6-ml anion-exchange column and eluted with a 10 to 400 mM NaCl gradient. At this chromatography step the oligomers were not well resolved. The eluted fractions, which in a dot blot analysis reacted with anti-DraE antibodies, were dialyzed against buffer A and, after concentration to an appropriate volume, were applied to a Mono Q HR 5/5 anion-exchange column. The complexes were eluted with a 10 to 400 mM NaCl gradient, and five well-resolved peaks containing DraB in complexes with DraE oligomers were obtained. DraB-DraE binary complexes were purified on a Resource S 6-ml cation-exchange column by using buffer B (50 mM CH3COONa [pH 5.5]) with a 0 to 250 mM NaCl gradient. Finally, binary complexes were resolved by fractionation on a Superdex G-200 HR 10/30 column with buffer B containing 20 mM NaCl.
The periplasmic extract containing the DraB protein was dialyzed against 20 mM NaCl-20 mM Tris-HCl (pH 7.5) and then applied to a Resource Q 6-ml column. The DraB protein was eluted with a 20 to 300 mM NaCl gradient in 20 mM Tris-HCl (pH 7.5). Finally, the DraB chaperone was resolved by fractionation on a Superdex G-200 HR 10/30 column in buffer containing 20 mM NaCl and 20 mM Tris-HCl (pH 7.5).
The E. coli BL21(DE3) strain transformed with pET30b-DraE (or derivatives of this plasmid) was grown to an A600 of 0.4 in Luria-Bertani medium containing 20 μg of kanamycin per ml. Protein expression was induced by adding IPTG to a final concentration of 1 mM for 4 h. Then the cells were harvested by centrifugation and disrupted by sonication. The centrifuged fraction containing inclusion bodies composed mainly of DraE protein was washed twice with buffer containing 1% Triton X-100, 0.5 mM dithiothreitol (DTT), and 20 mM Tris-HCl (pH 7.5) to remove contaminating proteins. After this, the inclusion bodies were centrifuged and solubilized under denaturing conditions (4 M guanidine-HCl [GuHCl], 0.5 mM DTT, 20 mM Tris-HCl; pH 7.5) for 2 h at 37°C with vigorous shaking. After solubilization the sample was centrifuged, and the supernatant containing denatured DraE protein (purity, 85%) was collected. The purified DraE protein was concentrated by using an Amicon Ultra concentrator (molecular mass cutoff, 10 kDa; Millipore, Eschborn, Germany) to obtain a final concentration of 40 mg/ml.
Renaturation of DraE. DraE protein (5 μl of a denatured sample [40 mg/ml]) was mixed with 800 μl of refolding buffer (50 mM CH3COONa, pH 5.5) containing DraB protein at a concentration of 5 mg/ml for 1 h at 25°C. After this, the sample was centrifuged and passed through a 0.45-μm-pore-size filter to remove potential precipitates. The supernatant obtained was analyzed by exclusion chromatography on a Superdex G-200 HR 10/30 column in buffer containing 20 mM NaCl and 50 mM CH3COONa (pH 5.5). The eluted fractions were analyzed by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and Western blotting with anti-DraE and anti-DraB antibodies. In the negative control experiment, we used refolding buffer with bovine serum albumin (BSA) (5 mg/ml). To determine the specificity of binary complex formation, we used refolding buffer containing the following proteins: DraB (5 mg/ml), BSA (3 mg/ml; molecular mass, 66 kDa), carbonic anhydrase (3 mg/ml; from bovine erythrocytes; molecular mass, 30 kDa), and -lactalbumin (4 mg/ml; from bovine milk; molecular mass, 14.4 kDa). To promote the formation of disulfide bonds in DraE protein complexed with DraB, the sample obtained from exclusion chromatography was dialyzed against oxidation buffer containing 1 mM reduced glutathione, 0.2 mM oxidized glutathione, and 20 mM Tris-HCl (pH 7.3) overnight at 4°C.
Detection of free thiols and disulfide bonds. To detect disulfide bonds, we used a nondirect method based on Ellman's reagent, 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB). DTNB is used directly to estimate the presence of free thiol groups; DTNB reacts with the thiol to give mixed disulfide and 2-nitro-5-thiobenzoic acid (TNB), which is quantified by absorbance of the anion (TNB2–) at 412 nm. In the first step, the sample was carboxymethylated with iodoacetamide, which resulted in modification of free thiols, while the disulfide bonds were left intact. Next, the sample with derivatized free thiols was denatured in 6 M GuHCl-0.1 M Tris-HCl (pH 8.0), and this was followed by reduction of disulfide bonds with 10 mM DTT for 2 h. The excess DTT was removed by dialysis against a detection buffer containing 0.1 M phosphate (pH 8.0). To detect free thiols corresponding to disulfide bonds, after dialysis the sample was treated with Ellman's solution containing 10 mM DTNB and 0.1 M phosphate buffer (pH 8.0), and the absorbance at 412 nm was measured. The concentration of thiols was calculated for the molar absorbance of the TNB anion (E412 = 1.415 x 104 cm–1 M–1). To detect the disulfide bonds, we used 2 nM mature DraB protein or the corresponding cysteine mutant in each reaction.
Partial proteolysis. Partial proteolysis was performed by using proteinase K (Merck, Darmstadt, Germany) at concentrations of 0.1, 1, 10, and 30 μg/ml and 0.1, 1, and 2 mg/ml (total volume, 17 μl). In the assays we used 10 μl of purified DraB complexes with the N-terminal deletion mutants of the DraE protein (1 mg/ml) or purified unresolved complexes of DraB with native DraE, 1DraE, or 2DraE protein at concentrations of 2 mg/ml (as determined by densitometric analysis). Digestion was performed in a buffer containing 2 mM CaCl2 and 20 mM Tris-HCl (pH 7.8) for 20 min at 25°C. Reactions were stopped by adding a proteinase inhibitor, phenylmethylsulfonyl fluoride, to a final concentration of 2 mM.
Two-dimensional (2D) electrophoresis. Isoelectric focusing IEF (first dimension) was performed by using an Ettan IPGphor II system (Amersham Bioscience, Uppsala, Sweden) and 13-cm-long IPG strips with an immobilized pH 3 to 10 gradient (Amersham Bioscience). Periplasm extract samples were loaded onto the IPG strips during rehydration. The rehydration was carried out for 16 h in 250 μl of buffer containing 0.5% (vol/vol) IPG buffer (pH 3 to 10), 0.002% (wt/vol) bromophenol blue, and 80 μl of periplasmic extract (10 mM MgCl2). To protect DraE oligomers from depolymerization, we did not use recommended denaturing agents, such as urea, in the rehydration buffer. The high-molecular-weight oligomers of DraE proteins were not loaded onto IPG strips under the rehydration reaction conditions used. The IEF was performed with a standard voltage gradient (200 to 5,000 V) for 12 h. The IEF strips were then equilibrated for 20 min with buffer containing 30% (vol/vol) glycerol, 2% (wt/vol) SDS, 0.002% (wt/vol) bromophenol blue, and 50 mM Tris-HCl (pH 8.8); this was followed by SDS-PAGE in a 12% polyacrylamide gel. The gels were stained with a ProteoSilver staining kit (Sigma, St. Louis, Mo.) or were analyzed by Western blotting.
Other techniques. SDS-PAGE was carried out by using samples that were heated in Laemmli buffer at 98°C for 5 min or were incubated with Laemmli buffer at 25°C for 10 min but not denatured thermally. Immunoblot detection of DraE and DraB proteins was performed with polyclonal rabbit monospecific anti-DraE and anti-DraB antibodies, respectively, and secondary anti-rabbit goat antibodies labeled with horseradish peroxidase and diaminobenzidine as a reaction substrate.
Dr adhesins were detected on the surface of E. coli strains by using rabbit anti-DraE antibodies and secondary anti-rabbit goat antibodies conjugated with fluorescein isothiocyanate, and stained fimbriae were viewed with an Olympus BX-60 fluorescence microscope system. Densitometric analyses of SDS-PAGE gels stained with Coomassie brilliant blue G-250 were performed with an SDS low-molecular-weight calibration kit (Amersham Bioscience) by using a VersaDoc system (Bio-Rad, Hertfordshire, United Kingdom) with Quantity One software (Bio-Rad). DNA sequencing and determination of the N-terminal sequence of the mature DraB protein were performed commercially (DNA-Gdask II s.c.).
RESULTS AND DISCUSSION
DraB protein belongs to the FGL subfamily of PapD-like chaperones. A draB gene encodes a 265-amino-acid protein with a calculated molecular mass of 28.6 kDa. During the expression process, the DraB protein is secreted into the periplasm. N-terminal sequencing of the mature protein showed that the signal peptidase cut DraB between Ala29 and Ala30; the periplasmic chaperone DraB (236 residues) has a molecular mass of 25.7 kDa and a measured pI of 9.0 (Fig. 1; also see Fig. 4).
Sequence analysis showed that the DraB protein belongs to the conserved family of PapD-like chaperones engaged in the biogenesis of surface-located polymeric adhesive structures in gram-negative bacteria. Alignment of the amino acid sequences of the DraB protein and PapD-like chaperones with known structures (PapD, FimC, and Caf1M) showed that the DraB protein belongs to the FGL subfamily (Fig. 1). Additionally, mature DraB protein contains two conserved cysteines, Cys103 and Cys139, that are homologous to Cys101 and Cys140 of the Caf1M chaperone in the FGL subfamily. The cysteines of Caf1M form a disulfide bond with a crucial function in biogenesis of the F1 capsular antigen (41). The disulfide bond of Caf1M is not required for folding of this protein but is necessary for maintenance of the Caf1M-Caf1 adhesin subunit interaction (41). The mature DraB protein purified from a periplasm extract did not react with Ellman's reagent, which demonstrated that the C103 and C139 residues form a disulfide bond. To investigate the role of these conserved cysteines in the folding of the chaperone and the capacity of DraB to form complexes with the DraE protein, we constructed a DraB mutant in which cysteine 103 was changed to alanine (the pET30b-sygDraBC103A expression plasmid). The C103A DraB mutant purified from the periplasm reacted with Ellman's reagent with a molar stoichiometry that corresponded to one cysteine residue per DraB molecule. We then tested the capacity of native and cysteine mutant DraB proteins to form binary complexes with denatured DraE protein during a renaturation reaction. The DraE protein was expressed in the cytoplasm of E. coli cells as a mature protein (without a signal sequence) from the pET30b-DraE plasmid. In the absence of the DraB chaperone and the reducing environment of the cytoplasm, we observed considerable overproduction of DraE as insoluble inclusion bodies. The inclusion bodies were washed with buffer containing 1% Triton X-100 to remove contaminating proteins and then solubilized under denaturing conditions in 4 M GuHCl-0.5 mM DTT. The final concentration of DraE was estimated to be 40 mg/ml (Fig. 2). This renaturation strategy was based on dilution of a small volume of the concentrated denatured DraE protein (40 mg/ml) by addition of a 160-fold volume of the DraB protein (5 mg/ml) to the refolding buffer. When a denatured DraE sample was mixed with DraB in refolding buffer, the GuHCl concentration decreased from 4 M to 25 mM. Thus, we estimated that DraB-DraE binary complexes isolated from the periplasmic space were stable in buffers containing 25 mM GuHCl. The final concentration of DraE in the mixture was 0.25 mg/ml; accordingly, the renaturation reaction and DraB-DraE complex formation occurred with a 20-fold excess of the chaperone DraB. The refolding mixture was incubated for 20 min at room temperature, and the presence of DraB-DraE complexes was determined by molecular filtration. A single peak on a chromatogram was observed for the C103A mutant of DraB (pET30b-DraBC103A plasmid). This fraction reacted with anti-DraB antibodies but not with anti-DraE antibodies. Inversely, when the native mature DraB protein (with a disulfide bond) was tested, two peaks appeared. The first peak had the same retention time as binary DraB-DraE complexes (which reacted with anti-DraB and anti-DraE antibodies), and the second peak corresponded to native DraB and a cysteine mutant of DraB (which reacted with anti-DraB antibodies). The negative control included a denatured DraE sample diluted in the refolding buffer complemented with albumin at a concentration of 5 mg/ml. In this control assay, the DraE protein precipitated during mixing and was removed during the ultrafiltration step, before chromatography.
To test whether DraB-DraE binary complex formation resulted from a specific reaction between the chaperone and the fimbrial subunit, we performed the same experiment with refolding buffer which contained, in addition to the DraB chaperone (5 mg/ml), BSA (3 mg/ml), carbonic anhydrase (3 mg/ml), and -lactalbumin (4 mg/ml). In this experiment, only the mature DraB protein was able to form soluble binary complexes with the DraE protein. These results agreed with the results of experiments in which the role of the homologous cysteines in the CaF1M chaperone was determined.
Overproduced DraE adhesin has a toxic effect on E. coli cells. Previous studies showed that E. coli cells harboring the pBJN417 plasmid (24) (containing the dra operon with a transposon mutation in a draC gene encoding the outer membrane usher) do not express surface-located Dr fimbriae. The mutant strain also lost the ability to bind to CHO cells expressing the DAF receptor. A similar phenotypic effect was obtained with an E. coli strain harboring the pBJN413 plasmid (containing the dra operon with a transposon mutation in a draB gene) (24). As described above, we showed that the cytoplasmic form of the DraE protein (without its signal sequence) expressed from pET30b-DraE accumulated as insoluble inclusion bodies. When the DraE protein was expressed with its signal sequence (plasmid pET30b-sygDraE), cell growth and the DraE production levels depended on the strength of induction of the cell culture. When the protein was expressed at 30°C with IPTG added to a final concentration of 1 mM (A600, 0.30), it had a toxic effect in connection with overproduction of the DraE protein from a strong T7 promoter. The E. coli cells produced DraE adhesin by 2 h after induction and reached a maximal optical density (A600) of 0.41. Incubation for three additional hours resulted in a reduced optical density (0.32). Under the same cultivation conditions, the control cells, harboring the pET30b vector, grew logarithmically to a maximal optical density of 1.2 (6 h after induction) and reached a stable stationary phase (results not shown). SDS-PAGE and Western blot analysis with anti-DraE antibodies of total cell lysates (taken from cultures during expression) revealed the presence of the DraE protein with and without a signal sequence. When the IPTG concentration was reduced to 0.1 mM, the bacterial cell growth in both cases was almost the same, and Western blotting of total cell lysates showed only the presence of the DraE protein with a signal sequence (cytoplasmic form). The absence of the DraE protein in the periplasm (without a signal sequence) suggests that it was probably efficiently degraded by periplasmic proteases, and no toxic effect was observed. When there was a high level of overproduction (1 mM IPTG), the proteolytic systems probably could not efficiently degrade the misfolded DraE protein, and thus toxicity occurred. These data suggest that fimbrial DraE subunits produced without the DraB chaperone had toxic effects on expressing cells, resulting in accumulation of misfolded DraE protein in the periplasmic space. These data agree with the reports describing expression of pilin subunits of other polymeric structures assembled by the usher-chaperone system (17).
DraB-DraEn complexes accumulate in the periplasm in the absence of the usher DraC. To investigate the mechanism of DraB-DraE complex formation and polymerization of DraE subunits, we constructed the pET30b-sygDraBE plasmid. In this bicistronic construct, the draB and draE genes (each with its own ribosome binding site) were under the control of a common T7 promoter (Tabor-Studier expression system [34]). The draB gene was located proximal to the promoter region, and the draE gene was located distal to the promoter region. One feature of E. coli polycistronic systems is that translation from the first RBS proximal to the promoter is stronger than translation from the RBS sites located distally. Therefore, the level of expression of the DraB chaperone was higher than that of the DraE adhesin from the bicistronic plasmid constructed (Fig. 3A). In accordance with the chaperone-usher pathway, the subunits of outer membrane polymers are able to fold into the native form only by interacting with a cognate chaperone protein to form a chaperone-subunit complex. As first shown for P pili of E. coli, coexpression of the PapD chaperone with pilin subunits eliminated the toxic effect resulting from expression of the subunit alone (17). Thus, in the bicistronic vector constructed, the higher level of expression of the DraB chaperone than of the DraE protein eliminated the previously described toxic effect connected with accumulation of misfolded DraE hemagglutinin. The SDS-PAGE and Western blot analyses of crude extracts of expressing cells harboring plasmid pET30b-sygDraBE revealed two bands that reacted with anti-DraE antibodies corresponding to the DraE protein with and without its signal sequence and also two other bands that reacted with anti-DraB antibodies corresponding to the DraB protein with and without its signal sequence. In isolates isolated from periplasmic fractions, two bands corresponding to the mature forms of the DraB and DraE proteins were detected (Fig. 3C). Densitometric analyses of SDS-PAGE gels containing periplasmic samples showed that the DraB chaperone accounted for almost 40% of the total protein content and DraE accounted for 25% of the total protein content (Fig. 3A).
Two types of interactions are possible in the expression systems that allow production of periplasmic chaperone and fimbrial subunits without an usher protein: (i) formation of a chaperone subunit binary complex in the donor strand complementation reaction and (ii) oligomerization of subunits by the donor strand exchange mechanism. We found that oligomers of DraE subunits (dimers, trimers, tetramers, etc.) formed in the periplasm during expression were stable in Laemmli buffer at the ambient temperature. This property of the oligomers formed in the periplasm allowed analysis by SDS-PAGE when a sample was not denatured thermally. A similar approach was used to characterize Caf1 oligomers of Y. pestis (40). Analysis of a periplasmic extract of E. coli BL21(DE3)/pET30b-sygDraBE by SDS-PAGE (with a sample not denatured thermally) resulted in bands corresponding to the DraB chaperone and a monomer of the DraE subunit, as well as a ladder of bands corresponding to oligomers of the DraE subunit. Western blotting with anti-DraE antibodies revealed the same ladder of oligomer bands and a monomer form of the DraE protein (Fig. 3B). The same sample in a native polyacrylamide gel also produced a ladder of bands that reacted with anti-DraE antibodies and reacted weakly with anti-DraB antibodies (data not shown). These data are consistent with the proposal that DraE oligomers are formed by a conserved chaperone-usher pathway. Because the oligomers are formed in accordance with a donor strand exchange reaction, each is built from head-to-tail-connected DraE subunits. In agreement with this mechanism, the last subunit incorporated into the growing oligomer is complexed with the chaperone DraB. Further analysis by 2D electrophoresis (pH 3 to 10 gradient; molecular weight analysis; SDS—12% polyacrylamide gel) showed that the oligomers had pIs in the range from 4.7 to 4.9 (Fig. 4). 2D electrophoresis permitted us to distinguish five well-resolved oligomers corresponding to DraEn (n = 2 to 6) and high-molecular-weight oligomers.
Using two-step ion-exchange chromatography and the gel filtration technique, we purified oligomers corresponding to DraEn (n = 2 to 5). The DraB-DraE binary complex was unstable at pH 7.0, the pH at which we purified oligomers, and therefore we used cation-exchange chromatography and pH 5.5 buffers to purify the oligomers. Analysis of purified oligomers by SDS-PAGE revealed the presence of the DraE and DraB proteins. To determine the amount of the DraE protein in the periplasmic fraction in an oligomeric form in relation to the total amount of DraE, we performed densitometry analyses of SDS-polyacrylamide gels stained with Coomassie brilliant blue G-250. The total amount of DraE was evaluated on gels containing resolved periplasmic samples that were denatured at 98°C for 5 min. The concentration of the DraE protein not in oligomers was determined by using the same periplasmic samples incubated at room temperature in Laemmli buffer. The densitometric data showed that almost 70% of the total DraE protein was in an oligomeric form (Fig. 3B). These data showed that in E. coli BL21(DE3)/pET30b-sygDraBE, in the absence of the DraC usher protein, DraB-DraE binary complexes and DraB-DraEn oligomers accumulated in the periplasm. In the presence of excess chaperone DraB, the DraE protein appeared mainly in an oligomeric form, which suggested that the DraE polymerization process is more favorable than the formation of DraB-DraE binary complexes or that the DraE oligomers are more stable thermodynamically than the DraB-DraE complexes. These data agree with the thermodynamic models of subunit polymerization proposed independently by Sauer et al. (26) and Zavialov et al. (39). According to these models, the energy driving the donor strand exchange reaction is stored in a high-energy, molten-globule-like subunit in a complex with the chaperone. During polymerization, subunits rearrange structurally to form a more rigid core with lower energy.
N-terminal strand of the DraE subunit is required for DraE oligomer formation. In surface-located polymeric systems in which biogenesis occurs by the usher-chaperone mechanism, the subunits possess two critical sequences responsible for polymerization: the N-terminal donor strand and a C-terminal strand homologous to -strand F of P and type 1 pilus subunits (40). During polymerization, the incoming subunit for the donor strand exchange reaction depletes the chaperone protein from the complex with the last subunit of a growing polymer. Crystallographic data showed that during this reaction the N-terminal strand for the incoming subunit exchange prevented the donor strand of the chaperone from interacting with -strand F of the last subunit of the polymer. Due to the limited sequence similarity between polymeric subunits for FGL chaperones, we could not precisely predict the minimal sequence of a donor strand in the DraE protein. Thus, mutation analysis of the N-terminal region comprising residues 1 to 11 of the mature DraE protein was performed. In this sequence we distinguished two fragments: two consecutive GTT sequences (residues 6 to 11) and a region comprising the first five N-terminal residues, which had no unique features. Using the pET30b-sygDraBE plasmid, we constructed N-terminal deletion mutants of the DraE protein, including 1DraE, 2DraE, 3DraE, 4DraE, 5DraE, 8DraE, and 11DraE (Table 1). Periplasmic samples isolated from all of the mutants incubated at the ambient temperature or at 98°C were analyzed by SDS-PAGE and Western blotting by using anti-DraE and anti-DraB antibodies. A periplasmic extract from E. coli Bl21(DE3)/pET30b-sygDraBE was the reference point in the analysis of the extracts from DraE deletion mutants. To determine the region comprising the signal for polymerization, we first prepared three long N-terminal deletion mutants, 11DraE, 8DraE, and 5DraE. These mutants did not all polymerize to oligomeric forms detectable by SDS-PAGE and Western blotting (samples were not denatured thermally) (Fig. 5A). Using cation-exchange and exclusion chromatographic techniques, we purified all of the DraE mutants as stable complexes with the DraB chaperone. Western blot analysis of eluted fractions with the anti-DraE and anti-DraB antibodies proved that the DraE and DraB proteins were present. The level of recovery of 5DraE was comparable to the level of recovery of the wild-type DraE protein. In the case of 8DraE and 11DraE the amount of mutated protein was much smaller than the amount of wild-type DraE protein.
To investigate the potential effect of periplasmic DegP protease on the low levels of 8DraE and 11DraE recovery, we performed experiments with E. coli B178(DE3)htrA63 (with a transposon mutation in the htrA [degP] gene). When the 8DraE and 11DraE proteins were expressed without the chaperone DraB (plasmids pET30b-syg8DraE and pET30b-syg11DraE, respectively) in degP+ cells, we did not obtain a signal for these proteins in SDS-PAGE and Western blot analyses. Previously described data showed that expression in a degP+ strain harboring plasmid pET30b-sygDraB8/11E resulted in the formation of binary complexes of DraB with a cognate deletion mutant of DraE. This suggested that DraB reacts with long deletion mutants of DraE and, to some degree, protects them from proteolysis. Expression from E. coli strain B178(DE3)htrA63 transformed with plasmid pET30b-sygDraB8/11E resulted in a level of recovery of deletion mutants of DraE that was higher than the level obtained for the degP+ strain (data not shown). These data suggest that the long deletion mutants of DraE probably have a more flexible structure in complexes with the DraB chaperone and are more susceptible to degradation by periplasmic proteases.
To further determine the N-terminal amino acids of DraE required for polymerization, we constructed the following mutants: 4DraE, 3DraE, 2DraE, and 1DraE. The 3DraE and 4DraE mutant proteins did not polymerize (oligomers were not detectable in Western blots of SDS-polyacrylamide gels; samples were not denatured thermally) (Fig. 5B). The properties of these mutants were very similar to those described above for the 5DraE protein. The periplasmic extract from cells expressing the 2DraE mutant assayed by SDS-PAGE (samples were not denatured thermally) and Western blotting with anti-DraE antibodies contained oligomers composed of at most five subunits of 2DraE, and high-molecular-weight oligomers were not detected (Fig. 5B). As was the case with the 11DraE, 8DraE, and 5DraE deletion mutants, we purified the 4DraE, 3DraE, 2DraE, and 1DraE proteins as stable binary complexes with the DraB protein. The oligomeric properties of the 1DraE mutant were almost the same as those of the native DraE (Fig. 5B).
These data showed that deletion of the first three residues from the N terminus of the mature DraE protein resulted in a mutant which was capable of forming only binary complexes with chaperone DraB and was deficient in formation of DraE oligomers.
To determine the properties of the constructed N-terminal deletion mutants of the DraE protein for the formation of Dr fimbriae in vivo, the pBJN17 plasmid (containing the dra operon with a transposon mutation in the draE gene) (24) was introduced into E. coli BL21(DE3) strains containing constructed plasmids encoding DraE N-terminal mutants. The transformants obtained were cultivated on Luria-Bertani agar plates complemented with IPTG (6 mM), an inducer of DraE mutant protein expression from constructed plasmids. The presence of Dr fimbriae on the surface of E. coli cells was analyzed by an immunofluorescence assay by using anti-DraE antibodies. Only the strain transformed with pET30b-sygDraB1E expressed surface-located Dr fimbriae. The amount of 1DraE was similar to the amount in cells transformed with control plasmid pET30b-sygDraBE (data not shown).
Oligomerization of DraE subunits increases their resistance to proteolysis. The data described above showed that deletion of more than two residues from the N terminus of the DraE protein diminished the capacity to form DraE oligomers but did not influence the formation of DraB-DraE binary complexes. The longer deletions (8DraE and 11DraE) resulted in DraE proteins that were more susceptible to degradation by periplasmic proteases. To investigate the effect of N-terminal deletions on the stability of the DraE protein, we examined the resistance of all mutants obtained to proteolysis by proteinase K (Table 1). Protein complexes of the DraB chaperone with mutant or native DraE, purified by ion-exchange chromatography, were treated with different concentrations of proteinase K (0.1 μg/ml to 2 mg/ml) for 20 min at 25°C. The 8DraE and 11DraE proteins were not detected in samples treated with proteinase K at a concentration of 0.5 μg/ml. The 3DraE, 4DraE, and 5DraE mutants were much more resistant to degradation by proteinase K; the 3DraE protein was detected with anti-DraE antibodies in samples incubated with 30 μg of proteinase K per ml at 25°C.
To measure the resistance of oligomeric DraE proteins in complexes with DraB, we used samples purified by ion-exchange chromatography. Following proteolysis, samples were analyzed by SDS-PAGE (without thermal denaturation) and by Western blotting (Fig. 6). Treatment with proteinase K (10 μg/ml) resulted in the disappearance of monomeric mature DraE (DraB-DraE binary complex) and the appearance of a new band at a molecular mass that was 1.5 kDa lower than that of mature DraE. Western blotting showed that this band reacted with anti-DraE antibody (Fig. 6). In addition, the conentration of low-molecular-weight oligomers decreased. Treatment with 2 mg of proteinase K per ml resulted in only the bands corresponding to high-molecular-weight oligomers (Fig. 6), and similar results were obtained when a proteinase K concentration of 10 mg/ml was used (data not shown). The properties of the 1DraE protein were the same as those of the native DraE protein. The resistance of the 2DraE protein to proteolysis was similar to the resistance of low-molecular-weight oligomers of native DraE protein.
The efficiency of proteolysis by proteinase K, the endoprotease, depends on the level of accessibility to the core of the degraded protein. The results of proteolysis of the 8DraE and 11DraE mutants confirmed that these proteins have a more accessible core (flexible structure) than the native DraE protein and, therefore, are more susceptible to degradation. The crystalline structure of the ternary complex of the Caf1M chaperone with a dimer of Caf1 subunits (39) and the models of polymerization proposed for P and type 1 pili (12) show that during the formation of oligomers the subunits rearrange in a more rigid form. During the biogenesis of the rigid pili (P and type 1 pili), there are second-order interactions in addition to head-to-tail interactions between subunits, which also stabilize the polymer. Our data for proteolytic resistance of oligomeric forms of the native DraE protein agree with the data obtained previously for other polymers.
Disulfide bond of DraE is necessary in oligomer formation. An experiment in which we examined the refolding of denatured DraE protein showed that the native DraB protein formed binary complexes with the DraE subunit. The denatured DraE protein used in this experiment was in a form with reduced disulfide bonds. To investigate if the disulfide bond in DraE is necessary for oligomerization, the DraB-DraE (Cys-reduced) binary complexes obtained were dialyzed against oxidizing buffer containing reduced and oxidized glutathione. To detect the appearance of oligomers after oxidation, the samples were analyzed by SDS-PAGE (samples were not denatured thermally) and stained with silver. An intense band corresponding to the DraE monomer and three additional weak bands, which migrated as DraEn oligomers (n = 2 to 4), were observed. Western blot analysis showed that these bands reacted with anti-DraE antibodies. To verify that the oligomers of DraE detected were not formed due to intersubunit disulfide bonds, before electrophoresis the samples were incubated for 10 min at 50°C in Laemmli buffer with double the normal concentration of DTT. The same refolding and oxidizing experiment was performed with denatured DraE C21A mutant protein (expression plasmid pET30b-DraEC21A). DraEnC21A oligomers were not detected on a silver-stained SDS-polyacrylamide gel. These assays showed that the disulfide bond in the DraE protein is not required to form a binary complex with the DraB chaperone but is important in oligomerization.
Disulfide bond of the DraE subunit is not necessary for maintenance of the polymeric structure of DraE oligomers. In the experiments described above, samples of DraE oligomers prepared by incubation in Laemmli buffer without thermal denaturation were easily detected by SDS-PAGE. Additionally, the oligomers purified from the periplasm and treated with DTT at the same concentration as the concentration in Laemmli buffer reacted with Ellman's reagent. These experiments suggested that disulfide bonds in DraE subunits were not necessary for maintaining the oligomeric structure. We determined the influence of the DraE disulfide bond on the stability of the DraE oligomers by incubating them in Laemmli buffer with and without DTT at temperatures from 50 to 90°C (Fig. 7). Western blot analysis showed that oligomers incubated with buffer containing DTT were depolymerized almost totally at 60°C, but the oligomers incubated with the same buffer without DTT were highly visible at 70°C. Treatment with proteinase K showed that resistance to proteolysis increased continuously from the low-molecular-weight DraE oligomers to the high-molecular-weight DraE oligomers. Analysis of oligomer resistance to depolymerization in the temperature gradient showed that all oligomers disappeared simultaneously at the same temperature. This suggests that the observed depolymerization was probably a secondary effect induced by thermal denaturation of the DraE subunit. At this stage the lower thermal resistance to destruction of oligomers treated with DTT may be accomplished simply with a decrease in the stability of the DraE protein caused by reduction of the disulfide bond.
In summary, data presented in this paper support the hypothesis that the chaperone-usher pathway for FGL polymers is conserved and that there are specific properties characteristic only of FGL subfamily polymers and not of FGS subfamily polymers. The DraB and Caf1M chaperones have a conserved disulfide bond, which is important in the formation of a binary complex. In the DraB-DraE and Caf1M-Caf1 expression systems (with the usher protein absent) oligomers with the same stoichiometry of subunits accumulated in the periplasm. DraE and Caf1 oligomers also have similar physicochemical properties, DraE and Caf1 oligomers are resistant to depolymerization during incubation in Laemmli buffer at ambient temperatures, and binary complex chaperone subunits of both systems are not stable at pH values above 7.5 (40). A comparison of data obtained from partial proteolysis indicated that DraE oligomers and also binary complexes of the DraB chaperone with DraE mutants were more resistant to degradation than Caf1 oligomers and mutant complexes. This difference in proteolysis may be due to increased internal stability of the DraE subunit (the presence of a disulfide bond). A more extensive intersubunit interaction in the DraE polymer may contribute to greater stability of the fimbrial structure of E. coli Dr hemagglutinin. Inversely, lower internal stability and decreased intersubunit interactions may contribute to the collapse of the Caf1 polymer structure and capsular appearance of the F1 antigen of Y. pestis.
ACKNOWLEDGMENTS
This work was supported by the Polish State Committee for Scientific Research (projects 3 PO4B 008 23, 3 PO5A 079 23, and 3 PO5A 05023 to J.K.) and in part by grant NIH/R01 HD41687-01 to B.N.
REFERENCES
1. Barnhart, M. M., F. G. Sauer, J. S. Pinkner, and S. J. Hultgren. 2003. Chaperone-subunit-usher interactions required for donor strand exchange during bacterial pilus assembly. J. Bacteriol. 185:2723-2730.
2. Barnhart, M. M., J. S. Pinkner, G. E. Soto, F. G. Sauer, S. Langermann, G. Waksman, C. Frieden, and S. J. Hultgren. 2000. PapD-like chaperones provide the missing information for folding of pilin proteins. Proc. Natl. Acad. Sci. USA 97:7709-7714.
3. Behrens, S. 2003. Periplasmic chaperones—preservers of subunit folding energy for organelle assembly. Cell 113:587-596.
4. Bullitt, E., and L. Makowski. 1995. Structural polymorphism of bacterial adhesion pili. Nature 373:164-167.
5. Chapman, D. A., A. V. Zavialov, T. V. Chernovskaya, A. V. Karlyshev, G. A. Zav'yalova, A. M. Vasiliev, I. V. Dudich, V. M. Abramov, V. P. Zav'yalov, and S. MacIntyre. 1999. Structural and functional significance of the FGL sequence of the periplasmic chaperone Caf1M of Yersinia pestis. J. Bacteriol. 181:2422-2429.
6. Choudhury, D., A. Thompson, V. Stojanoff, S. Langermann, J. Pinkner, S. J. Hultgren, and S. D. Knight. 1999. X-ray structure of the FimC-FimH chaperone-adhesin complex from uropathogenic Escherichia coli. Science 285:1061-1066.
7. Dodson, K. W., F. Jacob-Dubuisson, R. T. Striker, and S. J. Hultgren. 1993. Outer-membrane PapC molecular usher discriminately recognizes periplasmic chaperone-pilus subunit complexes. Proc. Natl. Acad. Sci. USA 90:3670-4367.
8. Galyov, E. E., A. V. Karlishev, T. V. Chernovskaya, D. A. Dolgikh, O. Smirnov, K. I. Volkovoy, V. M. Abramov, and V. P. Zav'yalov. 1991. Expression of the envelope antigen F1 of Yersinia pestis is mediated by the product of caf1M gene having homology with the chaperone protein PapD of Escherichia coli. FEBS Lett. 286:79-82.
9. Garcia, M. I., A. Labigne, and C. Le Bouguenec. 1994. Nucleotide sequence of the afimbrial-adhesin-encoding afa-3 gene cluster and its translocation via flanking IS1 insertion sequences. J. Bacteriol. 176:7601-7613.
10. Garcia, M. I., M. Jouve, J. P. Nataro, P. Gounon, and C. Le Bouguenec. 2000. Characterization of the AfaD-like family of invasins encoded by pathogenic Escherichia coli associated with intestinal and extra-intestinal infections. FEBS Lett. 479:111-117.
11. Goluszko, P., S. L. Moseley, L. D. Truong, A. Kaul, J. R. Williford, R. Selvarangan, S. Nowicki, and B. Nowicki. 1997. Development of experimental model of chronic pyelonephritis with Escherichia coli O75:K5:H-bearing Dr fimbriae: mutation in the dra region prevented tubulointerstitial nephritis. J. Clin. Investig. 99:1662-1672.
12. Hahn, E., P. Wild, U. Hermanns, P. Sebbel, R. Glockshuber, M. Haner, N. Taschner, P. Burkhard, U. Aebi, and S. A. Muller. 2002. Exploring the 3D molecular architecture of Escherichia coli type 1 pili. J. Mol. Biol. 323:845-857.
13. Holmgren, A., and C. I. Branden. 1989. Crystal structure of chaperone protein PapD reveals an immunoglobulin fold. Nature 342:248-251.
14. Hung, C. S., J. Bouckaert, D. Hung, J. Pinkner, C. Widberg, A. DeFusco, C. G. Auguste, R. Strouse, S. Langermann, G. Waksman, and S. J. Hultgren. 2002. Structural basis of tropism of Escherichia coli to the bladder during urinary tract infection. Mol. Microbiol. 44:903-915.
15. Hung, D. L., J. S. Pinkner, S. D. Knight, and S. J. Hultgren. 1999. Structural basis of chaperone self-capping in P pilus biogenesis. Proc. Natl. Acad. Sci. USA 96:8178-8183.
16. Hung, D. L., S. D. Knight, R. M. Woods, J. S. Pinkner, and S. J. Hultgren. 1996. Molecular basis of two subfamilies of immunoglobulin-like chaperones. EMBO J. 15:3792-3805.
17. Jones, C. H., P. N. Danese, J. S. Pinkner, T. J. Silhavy, and S. J. Hultgren. 1997. The chaperone-assisted membrane release and folding pathway is sensed by two signal transduction systems. EMBO J. 16:6394-6406.
18. Jouve, M., M. I. Garcia, P. Courcoux, A. Labigne, P. Gounon, and C. Le Bouguenec. 1997. Adhesion to and invasion of HeLa cells by pathogenic Escherichia coli carrying the afa-3 gene cluster are mediated by the AfaE and AfaD proteins, respectively. Infect. Immun. 65:4082-4089.
19. Knight, S. D., D. Choudhury, S. Hultgren, J. Pinkner, V. Stojanoff, and A. Thompson. 2002. Structure of the S pilus periplasmic chaperone SfaE at 2.2 A resolution. Acta Crystallogr. D 58:1016-1022.
20. Kuehn, M. J., D. J. Ogg, J. Kihlberg, L. N. Slonim, K. Flemmer, T. Bergfors, and S. J. Hultgren. 1993. Structural basis of pilus subunit recognition by the PapD chaperone. Science 262:1234-1241.
21. Le Bouguenec, C., M. I. Garcia, V. Ouin, J. M. Desperrier, P. Gounon, and A. Labigne. 1993. Characterization of plasmid-borne afa-3 gene clusters encoding afimbrial adhesins expressed by Escherichia coli strains associated with intestinal or urinary tract infections. Infect. Immun. 61:5106-5114.
22. Lipinska, B., O. Fayet, L. Baird, and C. Georgopoulos. 1989. Identification, characterization, and mapping of the Escherichia coli htrA gene, whose product is essential for bacterial growth only at elevated temperatures. J. Bacteriol. 171:1574-1584.
23. Normark, S. 2000. Anfinsen comes out of the cage during assembly of the bacterial pilus. Proc. Natl. Acad. Sci. USA 97:7670-7672.
24. Nowicki, B., J. Moulds, R. Hull, and S. Hull. 1988. A hemagglutinin of uropathogenic Escherichia coli recognizes the Dr blood group antigen. Infect. Immun. 56:1057-1060.
25. Pham, T., A. Kaul, A. Hart, P. Goluszko, J. Moulds, S. Nowicki, D. M. Lublin, and B. J. Nowicki. 1995. dra-related X adhesins of gestational pyelonephritis-associated Escherichia coli recognize SCR-3 and SCR-4 domains of recombinant decay-accelerating factor. Infect. Immun. 63:1663-1668.
26. Sauer, F. G., J. S. Pinkner, G. Waksman, and S. J. Hultgren. 2002. Chaperone priming of pilus subunits facilitates a topological transition that drives fiber formation. Cell 111:543-551.
27. Sauer, F. G., K. Futterer, J. S. Pinkner, K. W. Dodson, S. J. Hultgren, and G. Waksman. 1999. Structural basis of chaperone function and pilus biogenesis. Science 285:1058-1061.
28. Sauer, F. G., M. Barnhart, D. Choudhury, S. D. Knight, G. Waksman, and S. J. Hultgren. 2000. Chaperone-assisted pilus assembly and bacterial attachment. Curr. Opin. Struct. Biol. 10:548-556.
29. Sauer, F. G., S. D. Knight, G. J. Waksman, and S. J. Hultgren. 2000. PapD-like chaperones and pilus biogenesis. Semin. Cell Dev. Biol. 11:27-34.
30. Saulino, E. T., D. G. Thanassi, J. S. Pinkner, and S. J. Hultgren. 1998. Ramifications of kinetic partitioning on usher-mediated pilus biogenesis. EMBO J. 17:2177-2185.
31. Saulino, E. T., E. Bullitt, and S. J. Hultgren. 2000. Snapshots of usher-mediated protein secretion and ordered pilus assembly. Proc. Natl. Acad. Sci. USA 97:9240-9245.
32. Soto, G. E., and S. J. Hultgren. 1999. Bacterial adhesins: common themes and variations in architecture and assembly. J. Bacteriol. 181:1059-1071.
33. Soto, G. E., K. W. Dodson, D. Ogg, C. Liu, J. Heuser, S. Knight, J. Kihlberg, C. H. Jones, and S. J. Hultgren. 1998. Periplasmic chaperone recognition motif of subunits mediates quaternary interactions in the pilus. EMBO J. 17:6155-6167.
34. Studier, F. W., A. H. Rosenberg, J. J. Dunn, and J. W. Dubendorff. 1990. Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol. 185:60-89.
35. Swanson, T. N., S. S. Bilge, B. Nowicki, and S. L. Moseley. 1991. Molecular structure of the Dr adhesin: nucleotide sequence and mapping of receptor-binding domain by use of fusion constructs. Infect. Immun. 59:261-268.
36. Thanassi, D. G., and S. J. Hultgren. 2000. Assembly of complex organelles: pilus biogenesis in gram-negative bacteria as a model system. Methods 20:111-126.
37. Thanassi, D. G., E. T. Saulino, M. J. Lombardo, R. Roth J. Heuser, and S. J. Hultgren. 1998. The PapC usher forms an oligomeric channel: implications for pilus biogenesis across the outer membrane. Proc. Natl. Acad. Sci. USA 95:3146-3151.
38. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673-4680.
39. Zavialov, A. V., J. Berglund, A. F. Pudney, L. J. Fooks, T. M. Ibrahim, S. MacIntyre, and S. D. Knight. 2003. Structure and biogenesis of the capsular F1 antigen from Yersinia pestis: preserved folding energy drives fiber formation. Cell 30:587-596.
40. Zavialov, A. V., J. Kersley, T. Korpela, V. P. Zav'yalov, S. MacIntyre, and S. D. Knight. 2002. Donor strand complementation mechanism in the biogenesis of non-pilus systems. Mol. Microbiol. 45:983-995.
41. Zav'yalov, V. P., T. V. Chernovskaya, D. A. Chapman, A. V. Karlyshev, S. MacIntyre, A. V. Zavialov, A. M. Vasiliev, A. I. Denesyuk, G. A. Zav'yalova, I. V. Dudich, T. Korpela, and V. M. Abramov. 1997. Influence of the conserved disulphide bond, exposed to the putative binding pocket, on the structure and function of the immunoglobulin-like molecular chaperone Caf1M of Yersinia pestis. Biochem. J. 324:571-578.(Rafa Pitek, Beata Zalewsk)