A Clostridium difficile gene encoding flagellin

Albert Tasteyre1, Marie-Claude Barc1, Tuomo Karjalainen1, Paul Dodson2, Susan Hyde2, Pierre Bourlioux1 and Peter Borriello2,3

Université de Paris-Sud, Faculté de Pharmacie, Département de Microbiologie, 5 rue JB Clément, 92296 Châtenay-Malabry cedex, France1
Institute of Infection and Immunity, Queen’s Medical Centre, University of Nottingham, University Park, Nottingham NG7 2RD, UK2
PHLS Central Public Health Laboratory, 61 Colindale Ave, London NW9 5HT, UK3

Author for correspondence: Marie-Claude Barc. Tel: +33 1 46 83 55 49. Fax: +33 1 46 83 58 83. e-mail: marie-claude.barc{at}cep.u-psud.fr


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Six strains of Clostridium difficile examined by electron microscopy were found to carry flagella. The flagella of these strains were extracted and the N-terminal sequences of the flagellin proteins were determined. Four of the strains carried the N-terminal sequence MRVNTNVSAL exhibiting up to 90% identity to numerous flagellins. Using degenerate primers based on the N-terminal sequence and the conserved C-terminal sequence of several flagellins, the gene encoding the flagellum subunit (fliC) was isolated and sequenced from two virulent strains. The two gene sequences exhibited 91% inter-strain identity. The gene consists of 870 nt encoding a protein of 290 amino acids with an estimated molecular mass of 31 kDa, while the extracted flagellin has an apparent molecular mass of 39 kDa on SDS-PAGE. The FliC protein displays a high degree of identity in the N- and C-terminal amino acids whereas the central region is variable. A second ORF is present downstream of fliC displaying homology to glycosyltransferases. The fliC gene was expressed in fusion with glutathione S-transferase, purified and a polyclonal monospecific antiserum was obtained. Flagella of C. difficile do not play a role in adherence, since the antiserum raised against the purified protein did not inhibit adherence to cultured cells. PCR-RFLP analysis of amplified flagellin gene products and Southern analysis revealed inter-strain heterogeneity; this could be useful for epidemiological and phylogenetic studies of this organism.

Keywords: Clostridium difficile, flagella, flagellin, PCR, cloning

Abbreviations: GST, glutathione S-transferase

The GenBank accession numbers for the sequences reported in this paper are AF065259 (strain 79-685) and AF077341 (strain VPI 10463).


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Clostridium difficile is recognized as the major aetiological agent of nosocomial infections such as pseudomembranous colitis and antibiotic-associated colitis and diarrhoea (Bartlett et al., 1978 ; Larson et al., 1978 ; George, 1984 ). Several confirmed and putative virulence factors that could play a role in C. difficile pathogenesis have been identified. Virulence is mainly due to the production of two protein exotoxins, toxins A and B, although other toxic factors have been described (Lyerly et al., 1988 ; Perelle et al., 1997 ). There is considerable evidence that some strains are more virulent than others; a number of factors have been proposed to explain inter-strain diversity: (i) capsule, an antiphagocytic factor (Davies & Borriello, 1990 ); (ii) proteolytic enzymes, which may play a role in releasing suitable substrates from available protein sources for metabolism and which could be involved in mucus penetration (Seddon & Borriello, 1992 ; Poilane et al., 1998 ); (iii) adhesins which are involved in mucus and cell association (Borriello et al., 1988b ; Eveillard et al., 1993 ; Karjalainen et al., 1994 ); (iv) fimbriae, the role of which is obscure (Borriello et al., 1988a ); and (v) flagella, the potential role of which in colonization is under study. Delmée et al. (1990) have developed a serogrouping method for C. difficile and flagella appear to be responsible for cross-reactions between strains. In other bacteria, flagella have been implicated in the adherence to mucus and cells and colonization by Pseudomonas aeruginosa (Ritchings et al., 1995 ; Arora et al., 1996 ), Vibrio cholerae (Richardson, 1991 ; Gardel & Mekalanos, 1996 ), Vibrio anguillarum (McGee et al., 1996 ; Milton et al., 1996 ), Helicobacter pylori (Eaton et al., 1996 ) and Burkholderia pseudomallei (Brett et al., 1994 ). Furthermore, they contribute to the invasiveness of Campylobacter jejuni (Morooka et al., 1985 ; Grant et al., 1993 ; Szymanski et al., 1995 ), Salmonella typhi (Liu et al., 1988 ) and Proteus mirabilis (Mobley et al., 1996 ).

One aspect of C. difficile virulence that has been studied by us is its interaction with target cells (Eveillard et al., 1993 ; Karjalainen et al., 1994 ). Adhesion and colonization of animal tissue by bacteria is an important step in establishing infection. It is probable that without attachment, C. difficile cannot colonize and will be quickly removed by non-specific host defence mechanisms.

We are interested in finding out whether flagella play a role in C. difficile colonization. In this study we undertook the isolation of C. difficile flagella from clinical strains and characterization and expression of the flagellin subunit gene. In addition, Southern analysis and PCR amplification of flagellin genes coupled with RFLP analysis were used in a preliminary attempt to differentiate between clinical isolates.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains and media.
Six C. difficile isolates were investigated and are presented in Table 1. The VPI 10463 strain was obtained from Dr Wilkins (Virginia Polytechnic Institute, Blacksburg, USA). Strains of Clostridium perfringens, Bacillus subtilis and Clostridium sordellii (Institut Pasteur, Paris, France) were used as negative controls. The strains were grown under anaerobic conditions on agar plates (Oxoid) supplemented with 7% horse blood (bioMérieux) or in TGY (tryptone/glucose/yeast infusion broth; Difco) for 48 h.


View this table:
[in this window]
[in a new window]
 
Table 1. C. difficile strains studied

 
Escherichia coli strain XL-1 Blue MRF' (Stratagene) was used as a host in general cloning procedures and E. coli BL21 (Pharmacia) was used for expression and purification of GST–FliC fusion protein.

Flagellin protein of C. difficile: isolation, characterization and N-terminal sequencing.
Flagellin proteins were isolated by the procedure described by Delmée et al. (1990) . The strains were grown on blood agar plates under anaerobic conditions for 48 h. Bacteria were harvested in 5 ml distilled water; the suspensions were strongly shaken for 1 min and centrifuged at 5000 g for 30 min at 4 °C. The supernatants were centrifuged at 25000 g for 1 h at 4 °C and the pellets were suspended in 100 µl PBS (pH 7·4).

SDS-PAGE was carried out as described by Laemmli (1970) using an SDS-PAGE gel (7·5%, w/v, acrylamide). Gels were stained with Coomassie blue or used for electric transfer onto nitrocellulose membrane for immunoblotting. The nitrocellulose membrane was incubated for 30 min at room temperature in blocking buffer (0·2% Tween, 3% skim milk in PBS) and then overnight in a rabbit polyclonal antiflagellin serum to serogroup A (strain W1194), kindly provided by M. Delmée (UCL, Brussels, Belgium) (1:2000 dilution). The membranes were washed in 10 mM Tris/HCl, 150 mM NaCl, 0·05% Tween 20 buffer (TNT) and bound antibodies were detected with goat anti-rabbit IgG alkaline phosphatase conjugate (1:2500 dilution; Sigma) with nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate as substrates.

For the N-terminal sequencing of the flagellin, proteins were electroblotted onto a Hybond-N membrane (Amersham) with CAPS buffer (Sigma) and stained with amido black. The bands of interest were excised and N-terminal analysis was performed (Polymer synthesis and Analysis unit, Department of Biochemistry, University of Nottingham). The method of Edman (1950) was used with an Applied Biosystems 473A amino acid sequencer on a PTH-RP-PHLC C18 column. Identification was performed using 610A analysis software (Applied Biosystems).

Electron and immunoelectron microscopy of flagella and strains.
Flagellum preparations or 24 h cultures of C. difficile strains as described above were resuspended in distilled water. They were negatively stained with a phosphotungstic acid solution for 5 min and adsorbed on carbon-stabilized nitrocellulose film copper specimen grids (2 min) by placing each grid on a drop of the bacterium-staining solution. After drying, the grids were observed with an EM 301 Philips transmission electron microscope.

For immunoelectron microscopy, after adsorption of bacteria or flagellum preparation to the grids for 5 min, the grids were floated on PBS with 1% BSA for 30 min, then incubated with antiflagellin antibodies diluted 1:100 for 1 h and washed three times in PBS. They were incubated with a 1:20 dilution of protein A labelled with 15-nm-diameter colloidal gold particles (British Biocen International) for 1 h, washed three times with PBS and subsequently fixed with 3% glutaraldehyde. After three washings the grids were stained with phosphotungstic acid before observation by transmission electron microscopy. A strain of B. subtilis was used as a negative control.

Motility assays.
Motility assays were conducted in BHI broth with 0·175% agar. The medium was placed in tubes which were inoculated from a colony by stabbing the agar with a toothpick. The tubes were incubated at 37 °C for 2 d under anaerobic conditions.

DNA techniques.
Plasmid isolations were performed by the alkaline lysis procedure using a kit from Qiagen. Ligations and restriction endonuclease digestions were done according to Sambrook et al. (1989) and protocols provided by the vendors. The TSB (Transformation Storage Buffer) method was used for transformation of E. coli (Chung et al., 1989 ).

PCR amplification and nucleotide sequencing of the C. difficile flagellin gene.
The N-terminal sequence of the C. difficile 79-685 flagellin and conserved motifs in the C-terminal sequences of flagellins from various bacteria (V. cholerae, Vibrio parahaemolyticus, B. subtilis, Clostridium tyrobutyricum) were used to design degenerate primers synthesized by Life Technologies. The N-terminal primer was ATGMGAGTWAATGTWTCWGCGCTY; the C-terminal primer was TTGWCGAAYTGTTGGTTWGCAGCWAGCAG. Genomic DNA was extracted from C. difficile strains as described by Karjalainen et al. (1994) or bacteria were harvested from blood agar plates, resuspended in sterile distilled water and boiled for 5 min and used directly in a standard amplification mixture. Amplifications were carried out in 100 µl volumes containing 1 µg genomic DNA, 2 µl MgCl2, 1 U Taq polymerase (Promega), 200 pmol of each deoxynucleotide triphosphate, 1 µM of each primer and 10 µl 10xpolymerase buffer for 34 cycles consisting of denaturation at 95 °C (1 min), annealing at 55 °C (1 min) and extension at 72 °C (2 min) (Perkin Elmer Thermocycler 480). At the end of the amplification, 20 µl samples were subjected to electrophoresis on a standard 1% agarose gel alongside a PCR size marker (100 bp ladder; Pharmacia) to confirm the presence of an amplified product. Amplified products were purified by a gel extraction kit (Wizard Gel Extraction kit; Promega).

The nucleotide sequences of both strands of amplified products of the C. difficile flagellin of strains 79-685 and VPI 10403 were obtained by using the Taq Dye Deoxy Terminator and Dye Primer cycle sequencing protocols developed by Applied Biosystems (Perkin Elmer) using fluorescence-labelled dideoxynucleotides and primers, respectively. The labelled extension products were analysed with an ABI PRISM 310 Genetic Analyzer (Perkin Elmer). More nucleotide sequence upstream and downstream of the first amplified region of the flagellin gene was obtained by amplification by PCR from the {lambda}ZapII genomic library (Karjalainen et al., 1994 ), converted into a plasmid library by excision en masse, using a gene-specific primer (ACGAACCTTCTGCTGTTTGTAC) and a primer (GGAAACAGCTATGACCATG) that hybridized with the pBluescript polylinker (M13rev). For the upstream region a PCR product of 500 bp was obtained. For the downstream region a PCR product of 1·1 kb was obtained with the gene-specific primer CTTTAGAGAATGTTACAGCAGC and a vector-specific primer GTAAAACGACGGCCAGT (M13 ‘-20’). Secondary structure of the flagellin protein was predicted by using the Chou and Fasman algorithm (Chou & Fasman, 1978 ). Additional sequencing of the flagellin gene was performed with internal primers.

Southern blotting and RFLP analysis.
DNA of C. difficile strains was prepared as described by Karjalainen et al. (1994) . DNA was digested with HindIII and electrically transferred to a nylon membrane (Boehringer Mannheim). The 850 bp amplified PCR product was used as a flagellin gene specific probe. The DNA probe was labelled and detected by using the ECL direct nucleic acid labelling and detection system from Amersham. Washing of membranes was performed under low stringency (0·5xSSC at 42 °C). For RFLP amplified DNA was obtained by PCR from colonies grown on blood agar plates as described above. The DNA was digested with four restriction enzymes, HindIII, EcoRV, DraI and SacI, under the conditions recommended by the supplier (New England Biolabs). These digests were then subjected to electrophoresis on a 1% agarose gel alongside a PCR size marker (100 bp ladder; Pharmacia).

Cloning, expression, purification and identification of the fusion protein.
For the cloning of the C. difficile 79-685 flagellin gene into an expression vector, two oligonucleotide primers, CCCCTGGGATCCATGAGAGTTAATACAAATGTAAGTGC and CCGGGAATTCCTATCCTAATAATTGTAAAACTCC, incorporating the BamHI and EcoRI restriction site, respectively, were synthesized and used to amplify by PCR the full-length coding region of the fliC gene of strain 79-685 (Taq polymerase, Promega; 1 U per 100 µl reaction volume). The resulting 895 bp DNA product was digested with BamHI and EcoRI and cloned in-frame into the corresponding sites of pGEX-6P-1 (Pharmacia). The nucleotide sequence of the junction between vector and insert was confirmed by sequencing analysis to be correct. The plasmid was transformed into E. coli BL21.

For the expression and purification of the fusion protein, an overnight culture of E. coli BL21 containing pGEX-6P-1-fliC was diluted 1:100 into 4 l 2xYT medium (Life Technologies) containing ampicillin and the culture was grown to OD600 0·6 at 37 °C. The expression of the fusion protein was induced by adding IPTG at 1 mM for 2 h. Bacteria were collected by centrifugation and resuspended in 200 ml ice-cold PBS. The bacteria were lysed by sonication (intervals of 5 s for 1 h at 80% power; Bioblock Scientific 72442 Vibra Cell). Insoluble material was removed by centrifugation at 8000 g for 10 min, and the fusion protein was purified from the supernatant by a single-step affinity chromatography using glutathione–Sepharose-4B and protocols from Pharmacia. A 2 ml bed volume was used for each 200 ml sonicate; the column was washed three times with 20 ml PBS, followed by cleavage of the FliC moiety bound to glutathione–Sepharose with 80 units Prescission protease per 1 ml bed volume. Identification of the fusion protein was carried out by SDS-PAGE as described above and by immunoblot (serum dilution 1:2000).

A polyclonal antiflagellin serum was raised against the purified FliC recombinant protein. The gel band located at 39 kDa corresponding to the purified flagellin was cut out and injected into a rabbit. The polyclonal, monospecific antiserum was obtained according to a protocol described previously (Karjalainen et al., 1994 ) and used at a 1:2000 dilution in Western blots.

Computer analyses.
Nucleotide and protein sequence alignments were performed with the CLUSTALX program (Thompson et al., 1997 ). Homology searches were conducted with FASTA3 (EBI) or BLAST (National Institute for Biotechnology Information, Washington).

Adherence inhibition assays.
Cell culture and adherence inhibition assays were performed as previously described (Karjalainen et al., 1994 ). For adherence inhibition with antibodies, bacteria were incubated with preimmune serum (dilution 1:2) or immune serum (dilution 1:2) for 1 h prior to contact of 1 h with cultured Vero cells. Non-adherent bacteria were eliminated by five washings in PBS (10 mM phosphate buffer, 150 mM NaCl) (pH 7·0) and the cells were fixed and stained with May–Grünwald–Giemsa (Sigma). The adhesion index is given as the mean number of adhering bacteria per cell (counted at a magnification of x1000) from at least three different assays.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Flagella of C. difficile: isolation, N-terminal analysis and functionality
An SDS-PAGE profile of flagella isolated from the six C. difficile strains is shown in Fig. 1(a). Each extract carries a band of 39 kDa corresponding to the flagellin as shown previously by Delmée et al. (1990) . The rabbit polyclonal flagellin antiserum raised by Delmée to crude flagellin preparations from strain W1194 reacted with a 39 kDa protein in all six strains on Western blots (Fig. 1b). In four strains, the N-terminal sequence of the flagellin was MRVNTNVSAL (Table 1).



View larger version (58K):
[in this window]
[in a new window]
 
Fig. 1. (a) SDS-PAGE of surface proteins including flagellin from six strains of C. difficile. Isolated proteins were denatured in sample buffer and electrophoresed on a 5% stacking gel and 12·5% separating polyacrylamide gel. Proteins were isolated from strains: 79-685 (lane A); 51187 (lane B); W1194 (lane C); B-1 (lane D); 56026 (lane E); Kohn (lane F). Low-molecular-mass standards of 104, 80, 47, 33, 28 and 20 kDa (Bio-Rad) are in lane M. The arrow indicates the band corresponding to the 39 kDa flagellin. (b) Isolated surface proteins reacted with a 1:2000 dilution of polyclonal antiserum raised against flagella of strain W1194 in a Western blot. Order of strains as in Fig. 1(a).

 
Because the flagellin proteins of some other bacteria are glycosylated (Ge et al., 1988 ; Guerry, 1997 ), the possible glycosylation of the C. difficile flagellin was examined by the DIG Glycan detection kit (Roche), which detects glycoconjugates on proteins by immunoassay. The results revealed that flagella of C. difficile strain 79-685 were not glycosylated (data not shown). In addition, the functional activity of the flagella of the six strains of C. difficile studied was determined by investigating cell motility. All the strains were motile (data not shown).

Detection of flagella by electron microscopy
Electron microscopy was used to observe the flagella of C. difficile strains, either on intact bacteria or in isolated flagellum preparations (Fig. 2). Some strains, such as Kohn and W1194, had numerous flagella, whereas other strains carried fewer flagella. It was evident by immunoelectron microscopy that the flagella were labelled by the Delmée flagellin antiserum A. B. subtilis and a C. sordellii strain, used as negative controls, showed no labelled flagella.



View larger version (143K):
[in this window]
[in a new window]
 
Fig. 2. Electron micrographs showing flagellated C. difficile strains Kohn (a) and 56026 (b) and isolated flagella from strain Kohn (c), 51187 (d) and W1194 (e). The flagella were labelled with antiflagellum serum to strain W1194 and visualized with protein A coupled to 15 nm colloidal gold particles. Bars, 1 µm.

 
DNA sequence analysis of the flagellin filament gene
PCR amplification with primers described in Methods allowed us to amplify an 850 bp product from the C. difficile genome. Sequences upstream and downstream were obtained by amplifying DNA from a genomic library constructed in {lambda}ZapII. The nucleotide sequence of a 1·6 kb fragment of two strains, 79-685 and VPI 10463, was determined; analysis of the DNA sequence revealed the presence of two ORFs: ORF1 composed of 870 nt corresponding to 290 amino acids, and a partial ORF 90 bases downstream. Based on the comparison of the deduced amino acid sequence with flagellin sequences from both Gram-positive and Gram-negative bacteria, ORF1 was identified as the flagellin gene of C. difficile, which we named fliC. FliC of C. difficile has a calculated molecular mass of 30·9 kDa: thus it differs from the estimated molecular mass of 39 kDa determined by SDS-PAGE (Fig. 1a). The FliC protein (Fig. 3) displays highest homology to the corresponding protein of C. tyrobutyricum (61% identity, 76% similarity) and exhibits features found in other flagellins: variable central part; predominantly {alpha}-helical conformation with frequent alanine and few proline residues; no signal sequence, the sequence LIAN resembling the consensus sequence N(I/L)AN that serves as an export signal for flagellar subunits (Heinzerling et al., 1997 ). The gene has a G+C value of 33·5 mol%, a value higher than that of the C. difficile genome (28 mol%) (Sneath, 1986 ). Analysis of the codon usage reveals a marked preference for A or T for the third position of the triplets and a total of 29 codons are not used. A motif similar to the {sigma}28 consensus binding site (TAAAGTN12GCCGATAA) is found in the promoter: TAAAGTN13TCCGATAA. The translational termination codon is followed by an imperfect inverted repeat that can form a stem–loop structure with a {Delta}G (25 °C) of -63·9 kJ; it could constitute a {rho}-independent transcriptional terminator.



View larger version (38K):
[in this window]
[in a new window]
 
Fig. 3. Sequence alignment of the deduced amino acid sequence of FliC of C. difficile strains 79-685 and VPI 10463. Identical residues are indicated with an asterisk; functionally identical residues are indicated with a colon. The alignment was performed with the CLUSTAL W program. The proteins show 90·6% identity (90·3% at the nucleotide level).

 
Conservation and expression of the fliC gene in C. difficile strains
To investigate the conservation of the fliC gene region in different strains, we amplified by PCR the flagellin gene from the six C. difficile strains. All strains showed the presence of an 850 bp band corresponding to the fliC gene (data not shown). The amplified DNA was used as a probe in Southern hybridization of chromosomal DNA of these strains. Hybridization under low stringency conditions showed that DNAs of all isolates studied hybridized with the fliC-specific probe. Only one copy of the gene was present in each strain (Fig. 4). Some strains carry a HindIII site and therefore show the presence of two bands.



View larger version (66K):
[in this window]
[in a new window]
 
Fig. 4. Southern blot of chromosomal DNA isolated from C. difficile strains hybridized under low stringency to the fliC probe. DNA was digested with HindIII, electrophoresed and transferred to nylon membrane. Lanes: 1, 850 bp PCR product used as probe (positive control); 2, DNA from C. sordellii (negative control); 3, DNA from strain 79-685; 4, DNA from 51187; 5, DNA from W1194; 6, DNA from B-1; 7, DNA from 56026; 8, DNA from Kohn. The two bands in lanes 4, 5, 6 and 8 are produced by the HindIII site near base 937 of the flagellin gene. The molecular mass standards (1 kb ladder; Gibco-BRL) are shown on the left.

 
The amplified DNA was digested with HindIII, EcoRV, DraI and SacI restriction enzymes; the results are shown in Fig. 5. The six strains were classified into two groups according to the RFLP of the fliC gene.



View larger version (34K):
[in this window]
[in a new window]
 
Fig. 5. RFLP patterns of PCR-amplified flagellin genes. Lanes: 1, 100 bp DNA ladder (PCR low ladder; Sigma); 2, strain 79-685; 3, strain 51187; 4, strain W1194; 5, strain B-1; 6, strain 56026; 7, strain Kohn; 8, 850 bp PCR product from the fliC gene of strain 79-685 (positive control). Amplified DNA was digested with HindIII (a), EcoRV (b), DraI (c) and SacI (d).

 
Expression, purification and immunological detection of flagellin
Screening of the C. difficile genomic library with the Delmée antiserum revealed that it recognizes a second protein of 39 kDa (GenBank accession no. AF065260) which is not the flagellin (data not shown). Therefore we decided to purify the flagellin protein by recombinant technology in order to produce a polyclonal, monospecific antiserum. The coding region of fliC was amplified by PCR and inserted into the E. coli expression vector pGEX-6P-1. The GST–FliC fusion protein was purified by affinity chromatography and was cleaved with Prescission protease. As shown in Fig. 6(a), a major 39 kDa band was observed in SDS-PAGE, free of contaminating GST. This band reacted in Western blot with the Delmée antiflagellar antiserum raised against the W1194 strain (not shown). The band was excised from the gel and injected into a rabbit in order to obtain polyclonal antibodies. The monospecific antiserum thus obtained reacted with a 39 kDa protein in the flagellar preparation of all the six strains studied and with the purified flagellin protein (Fig. 6b).



View larger version (37K):
[in this window]
[in a new window]
 
Fig. 6. (a) Purification of C. difficile 79-685 FliC protein. SDS-PAGE showing the protein marker (lane M) as in Fig. 1(a) and FliC eluted from glutathione–Sepharose after digestion of GST–FliC with Prescission protease (lane G). A major band at 39 kDa is observed. (b) Isolated C. difficile surface proteins reacted with a 1:2000 dilution of polyclonal antiserum raised against purified FliC. Proteins were isolated from strains: 79-685 (lane A); 51187 (lane B); W1194 (lane C); B-1 (lane D); 56026 (lane E); Kohn (lane F). Lane G, FliC eluted from glutathione–Sepharose after digestion of GST–FliC with Prescission protease. The arrow indicates the band corresponding to the 39 kDa flagellin.

 
Adherence inhibition assays
Involvement of the flagellin filament protein in adherence of C. difficile to eukaryotic cells was investigated in inhibition assays using antiflagellin antibodies raised against the purified protein. Coincubation of bacteria with antibodies at a dilution of 1:2 demonstrated no inhibition of adherence as compared with control adherence of 100% (incubation with preimmune serum), indicating that the flagellin subunit is not involved in the adherence process of C. difficile (data not shown).


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In this study, we have characterized the C. difficile fliC gene encoding the flagellin filament, which was sequenced from two strains. This gene is one of the few genes that has so far been isolated from this pathogen. The fliC gene seems to be present in a single copy on the chromosome of the strains examined, as in most species studied. The N-terminal region of FliC exhibits an absence of amino acids that are not neutral or hydrophobic, with the exception of one arginine, which is a characteristic of these structural proteins. There is evidence suggesting that amino acids in the N-terminal domain are closely linked to export of these molecules and their subsequent polymerization during biosynthesis of the flagella (Fedorov & Efimov, 1990 ). A number of structural studies have shown that the N- and C-terminal regions of flagellins are well conserved and play important roles in polymerization and polymorphism of bacterial flagellin filaments (Mimori-Kiyosue et al., 1997 ). Unlike many other flagellins, the N-terminal methionine residue of C. difficile flagellin is not removed post-translationally. Flagellin and several other external components of the bacterial flagellum are thought to be exported by a flagellum-specific pathway involving a central channel in the flagellum itself. Thus no signal peptide is present at the N-terminus. Instead, a consensus sequence resembling the export signal of flagellins, LIAN, is evident. The FliC protein contains three imperfect repetitive motifs, a feature often seen in surface-exposed proteins.

Analysis of the promoter structure reveals the presence of a motif resembling the consensus sequence for {sigma}28 regulated promoters. However, the distance between the -10 and -35 motifs is 16 nt instead of the usual 15. {sigma}28, which is involved in transcription of the flagellar and chemotaxis genes, was originally found in B. subtilis (Gilman & Chamberlin, 1983 ). Although transcription of some flagellar genes is initiated by {sigma}54, no consensus motif for this factor is present in the C. difficile fliC promoter.

The flagella of six strains of C. difficile were isolated. The molecular mass of C. difficile flagellin, 39 kDa, is in the middle of the range of other characterized flagellin molecules, which have been reported to have molecular masses ranging from 15 to 62 kDa (Arnold et al., 1998 ; Joys, 1988 ; Sakamoto et al., 1992 ; Wilson & Beveridge, 1993 ). C. difficile flagellins did not display heterogeneity between the different strains studied nor did they have multiple molecular masses. Immunoblotting and immunogold labelling of strains with a polyclonal antiserum raised against purified flagellin demonstrated that all C. difficile strains reacted with the antiserum, in contrast to strains of C. sordellii and B. subtilis. This result suggests that the flagellin of each strain contains cross-reacting epitopes due to the presence of the flagellin monomers. Although flagellins generally are structurally conserved in the N- and C-termini, the internal region is divergent and accounts for serological distinctiveness. Prediction of antigenic determinants in FliC of C. difficile using the Hopp and Woods algorithm (Hopp & Woods, 1981 ) revealed the highest probability for the presence of such motifs in the central, variable region between aa 100 and 130. The central region has been proposed to be the region that is exposed to the outer environment (Sakamoto et al., 1992 ). Like other flagellins, the central portion has a surplus of acidic residues over basic residues and, by implication, a net negative charge. It appears that most bacterial cell surface proteins carry a net negative charge (Wilson & Beveridge, 1993 ). We suspected that the ORF downstream of the fliC gene could encode a glycosyltransferase that could be involved in the in situ glycosylation of flagella. If FliC of C. difficile were glycosylated, this could explain the difference in molecular mass observed on SDS-PAGE (39 kDa) and that estimated from the nucleotide sequence (31 kDa). In fact, using the DIG Glycan Detection kit we demonstrated that C. difficile strain 79-685 flagella are non-glycosylated. The fact that the cloned gene expressed in E. coli produces a flagellin with almost the same molecular mass as the native protein from C. difficile suggests that the protein undergoes post-translational modification other than glycosylation, since E. coli is not able to glycosylate proteins. Flagellar filaments can contain phosphorylated tyrosines or serines, {epsilon}-N-methyl-lysine, or they can be sulfated glycoproteins.

The isolation of the fliC gene and isolation of a monospecific antiserum will allow further investigations as to the role of flagella in the pathogenic process. It has been suggested that the virulence of C. difficile strains is not solely attributable to toxin production; other factors such as presence of flagella could contribute to virulence. The role of flagella in microbial pathogenesis factors such as virulence, adherence, invasiveness or colonization has been demonstrated for numerous bacteria (Grant et al., 1993 ; Scherer et al., 1993 ; Zhang et al., 1993 ; Grossman et al., 1995 ; Pruckler et al., 1995 ; Tamura et al., 1995 ; Milton et al., 1996 ; Mobley, 1996 ; Bosshardt et al., 1997 ; Kennedy et al., 1997 ; Rosalski et al., 1997 ; West et al., 1997 ; Feldman et al., 1998 ). In our experiments in vitro, no inhibition of adherence was shown with antibodies against the recombinant flagellum subunit. However, lack of adherence does not mean there is no role in colonization or virulence. We are planning to investigate further the role of the flagellin cap in adhesion (Arora et al., 1998 ) and that of flagella in colonization and virulence using, for example, animal models.

The gene isolated here could be a potential biomarker to assess intraspecies genetic variation as there appears to be a divergent region in the amino acid sequence of the protein, a feature of surface-located proteins in bacteria (Whittam, 1995 ; Winstanley et al., 1996 ; Winstanley & Morgan, 1997 ). An epidemiological survey using the combined PCR-RFLP method and nucleotide sequencing is in progress on a larger number of C. difficile isolates.


   ACKNOWLEDGEMENTS
 
This work was supported in part by the FAIR Programme of the European Union (Contract CT95-0433), ACC-SV6 programme (Actions Concertées Coordonnées des Sciences du Vivant) of the Ministère de l’Education Nationale, de l’Enseignement Supérieur et de la Recherche of France and a Medical Research Council, Programme Grant (no. G9122850). We thank Professor Delmée (Université Catholique de Louvain, Belgium) for kindly providing the antiflagellin antiserum and C. difficile strains 51187 and 56026 and for serogrouping Kohn and B1 strains. We thank Mrs Nick Powell (Institute of Infection and Immunity, Nottingham, UK) for helping us with electron microscopy.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Arnold, F., Bedouet, L., Batina, G., Robreau, G., Talbot, F., Lecher, L. & Malcoste, R. (1998). Biochemical and immunological analysis of the flagellin of Clostridium tyrobutyricum ATCC 25755.Microbiol Immunol 42, 23-31.[Medline]

Arora, S. K., Ritchings, B. W., Almira, E. C., Lory, S. & Ramphal, R. (1996). Cloning and characterization of Pseudomonas aeruginosa fliF, necessary for flagella assembly and bacterial adherence to mucin.Infect Immun 64, 2130-2136.[Abstract]

Arora, S. K., Ritchings, B. W., Almira, E. C., Lory, S. & Ramphal, R. (1998). The Pseudomonas aeruginosa flagellar cap protein, FliD, is responsible for mucin adhesion.Infect Immun 66, 1000-1007.[Abstract/Free Full Text]

Bartlett, J. G., Chang, T. W., Gurwiyh, M., Gorbach, S. L. & Onderdonk, A. M. (1978). Antibiotic associated pseudomembranous colitis due to toxin producing clostridia. N Engl J Med 298, 531-534.[Abstract]

Borriello, S. P., Davies, H. A. & Barclay, F. E. (1988a). Detection of fimbriae amongst strains of Clostridium difficile.FEMS Microbiol Lett 49, 65-67.

Borriello, S. P., Welch, A. R., Barclay, F. E. & Davies, M. A. (1988b). Mucosal association by Clostridium difficile in the hamster gastrointestinal tract.J Med Microbiol 25, 191-196.[Abstract]

Bosshardt, S. C., Benson, R. F. & Field, B. S. (1997). Flagella are a positive predictor for virulence in Legionella.Microb Pathog 23, 107-112.[Medline]

Brett, P. J., Mah, D. C. & Wood, D. E. (1994). Isolation and characterization of Pseudomonas pseudomallei flagellin proteins.Infect Immun 62, 1914-1918.[Abstract]

Chou, P. Y. & Fasman, G. D. (1978). Prediction of the secondary structure of proteins from their amino acid sequence.Adv Enzymol Relat Areas Mol Biol 47, 45-148.[Medline]

Chung, C. T., Niemala, S. L. & Miller, H. R. (1989). One-step preparation of competent Escherichia coli: transformation and storage of bacterial cells in the same solution.Proc Natl Acad Sci USA 86, 2172-2175.[Abstract]

Davies, H. A. & Borriello, S. P. (1990). Detection of capsule in strains of Clostridium difficile.Microb Pathog 9, 141-146.[Medline]

Delmée, M., Avesani, V., Delferriere, N. & Burtonboy, G. (1990). Characterization of flagella of Clostridium difficile and their role in serogrouping reactions.J Clin Microbiol 28, 2210-2214.[Medline]

Eaton, K. A., Suerbaum, S., Josenhams, C. & Krakowka, S. (1996). Colonization of gnotobiotic piglets by Helicobacter pylori deficient in two flagellin genes.Infect Immun 64, 2445-2448.[Abstract]

Edman, P. (1950). Preparation of phenylhydantoins from some natural amino acids.Acta Chem Scand 4, 277-282.

Eveillard, M., Fourel, V., Barc, M. C., Kerneis, S., Coconnier, M. H., Karjalainen, T., Bourlioux, P. & Servin, A. (1993). Identification and characterization of adhesive factors of Clostridium difficile involved in adhesion to human colonic enterocyte-like Caco-2 and mucus-secreting HT29 cells in culture.Mol Microbiol 7, 371-381.[Medline]

Fedorov, O. V. & Efimov, A. V. (1990). Flagellin as an object for supramolecular engineering.Protein Eng 3, 411-413.[Abstract]

Feldman, M., Bryan, R., Rajan, S., Scheffler, L., Tang, B. S. L. & Prince, A. (1998). Role of flagella in pathogenesis of Pseudomonas aeruginosa pulmonary infection.Infect Immun 66, 43-51.[Abstract/Free Full Text]

Gardel, C. L. & Mekalanos, J. J. (1996). Alterations in Vibrio cholerae motility phenotypes correlate with changes in virulence factor expression.Infect Immun 64, 2246-2255.[Abstract]

Ge, Y., Li, C., Slaughter, C. A. & Charon, N. W. (1988). Structure and expression of the FlaA periplasmic flagellar protein of Borrelia burgdorferi.J Bacteriol 180, 2418-2425.[Abstract/Free Full Text]

George, W. L. (1984). Antimicrobial agent associated colitis and diarrhea: historical background and clinical aspects.Rev Infect Dis 6, 208-213.

Gilman, M. Z. & Chamberlin, M. J. (1983). Development and genetic regulation of the Bacillus subtilis genes transcribed by {sigma}28 RNA polymerase.Cell 35, 285-293.[Medline]

Grant, C. C., Konkel, M. E., Cieplak, W. J. & Tompkins, L. S. (1993). Role of flagella in adherence, internalization and translocation of Campylobacter jejuni in nonpolarized and polarized epithelial cell cultures.Infect Immun 61, 1764-1771.[Abstract]

Grossman, D. A., Witham, N. D., Burr, D., Lesmana, M., Rubin, F. A., Schoolnik, G. K. & Parsonnet, J. (1995). Flagellar serotypes of Salmonella typhi in Indonesia: relationship among motility, invasiveness, and clinical illness.J Infect Dis 171, 212-216.[Medline]

Guerry, P. (1997). Nonlipopolysaccharide surface antigens of Campylobacter species. J Infect Dis 176 (suppl. 2), S122–S124.

Heinzerling, H. F., Olivares, M. & Burne, R. A. (1997). Genetic and transcriptional analysis of flgB flagellar operon constituents in the oral spirochete Treponema denticola and their heterologous expression in enteric bacteria.Infect Immun 65, 2041-2051.[Abstract]

Hopp, T. P. & Woods, K. R. (1981). Prediction of protein antigenic determinants from aminoacid sequences.Proc Natl Acad Sci USA 78, 3824-3828.[Abstract]

Joys, T. M. (1988). The flagellar filament protein.Can J Microbiol 34, 452-458.[Medline]

Karjalainen, T., Barc, M. C., Collignon, A., Trollé, S., Boureau, H., Cotte-Laffite, J. & Bourlioux, P. (1994). Cloning of a genetic determinant from Clostridium difficile involved in adherence to tissue culture cells and mucus.Infect Immun 62, 4347-4355.[Abstract]

Kennedy, M. J., Rosey, E. L. & Yancey, R. J. J. (1997). Characterization of flaA- and flaB- mutants of Serpulina hyodysenteriae: both flagellin subunits, FlaA and FlaB, are necessary for full motility and intestinal colonization.FEMS Microbiol Lett 153, 119-128.[Medline]

Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4.Nature 227, 680-685.[Medline]

Larson, H. E., Honour, P., Price, A. B. & Borriello, S. P. (1978). Clostridium difficile and aetiology of pseudomembranous colitis.Lancet 1, 1063-1066.[Medline]

Liu, S. L., Ezaki, T., Miura, H., Matsui, K. & Yabuuchi, E. (1988). Intact motility as a Salmonella typhi invasion-related factor.Infect Immun 56, 1967-1973.[Medline]

Lyerly, D. M. H., Krivan, H. C. & Wilkins, T. D. (1988). Clostridium difficile: its disease and toxins. Clin Microbiol Rev 1, 1-12.[Medline]

McGee, K., Horstedt, P. & Milton, D. L. (1996). Identification and characterization of additional flagellin genes from Vibrio anguillarum.J Bacteriol 178, 1310-1319.[Abstract]

Milton, D. L., Otoole, R., Horstedt, P. & Wolfwatz, P. (1996). Flagellin A is essential for the virulence of Vibrio anguillarum.J Bacteriol 178, 1310-1319.[Abstract]

Mimori-Kiyosue, Y., Vonderviszt, F. & Namba, K. (1997). Locations of terminal segments of flagellin in the filament structure and their roles in polymerization and polymorphism.J Mol Biol 270, 222-237.[Medline]

Mobley, H. L. (1996). Defining Helicobacter pylori as a pathogen: strain heterogeneity and virulence. Am J Med 100(5A), 2S–9S.

Mobley, H. L. T., Belas, R., Lockatell, V., Chippendale, G., Trifillis, A. L., Johnson, D. E. & Warren, J. W. (1996). Construction of a flagellum-negative mutant of Proteus mirabilis: effect on internalization by human renal epithelial cells and virulence in a mouse model of ascending urinary tract infection.Infect Immun 64, 5332-5340.[Abstract]

Morooka, T., Umeda, A. & Amado, K. (1985). Motility as an intestinal colonization factor for Campylobacter jejuni.J Gen Microbiol 131, 1973-1980.[Medline]

Perelle, S., Gibert, M., Bourlioux, P., Corthier, G. & Popoff, M. (1997). Production of a complete binary toxin (actin-specific ADP-ribosyltransferase) by Clostridium difficile CD196.Infect Immun 65, 1402-1407.[Abstract]

Poilane, I., Karjalainen, T., Barc, M. C., Bourlioux, P. & Collignon, A. (1998). Protease activity of Clostridium difficile strains.Can J Microbiol 44, 157-161.[Medline]

Pruckler, J. M., Benson, R. F., Moyenuddin, M., Martin, W. T. & Fields, B. S. (1995). Association of flagellum expression and intracellular growth of Legionella pneumophila.Infect Immun 63, 4928-4932.[Abstract]

Richardson, K. (1991). Roles of motility and flagellar structure in pathogenicity of Vibrio cholerae: analysis of motility mutants in three animal models.Infect Immun 59, 2727-2736.[Medline]

Ritchings, B. W., Almira, E. C., Lory, S. & Ramphal, R. (1995). Cloning and phenotypic characterization of fleS and fleR, new response regulators of Pseudomonas aeruginosa which regulate motility and adhesion to mucin.Infect Immun 63, 4868-4876.[Abstract]

Rosalski, A., Sidorczyk, Z. & Kotelko, K. (1997). Potential virulence factors of Proteus bacilli.Microbiol Mol Biol Rev 6, 65-89.

Sakamoto, Y., Sutherland, K. J., Tamaoka, J., Kobayashi, T., Kudo, T. & Horikoshi, K. (1992). Analysis of the flagellin (hag) gene of alkalophilic Bacillus sp. C-125.J Gen Microbiol 138, 2159-2166.[Medline]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Scherer, D. C., DeBurron-Connors, C. I. & Minnick, M. F. (1993). Characterization of Bartonella bacilliformis flagella and effect of antiflagellin antibodies on invasion of human erythrocytes.Infect Immun 61, 4962-4971.[Abstract]

Seddon, S. V. & Borriello, S. P. (1992). Proteolytic activity of Clostridium difficile.J Med Microbiol 36, 307-311.[Abstract]

Sneath, P. H. A. (1986). Endospore-forming Gram-positive rods and cocci. In Bergey’s Manual of Systematic Bacteriology, pp. 1104-1207. Edited by P. H. A. Sneath, N. S. Mair, M. E. Sharpe & J. G. Holt. Baltimore: Williams & Wilkins.

Szymanski, C. M., King, M., Haardt, M. & Armstrong, G. T. (1995). Campylobacter jejuni motility and invasion of Caco-2 cells.Infect Immun 63, 4295-4300.[Abstract]

Tamura, Y., Kijima-Tanaka, M., Aoki, A., Ogikubo, Y. & Takahashi, T. (1995). Reversible expression of motility and flagella in Clostridium chauvoei and their relationship to virulence.Microbiology 141, 605-610.[Abstract]

Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F. & Higgins, D. G. (1997). The CLUSTALX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools.Nucleic Acids Res 24, 4876-4882.

West, N. P., Fitter, J. T., Jakubzik, U., Rohde, M., Guzman, C. A. & Walker, M. J. (1997). Non-motile mini-transposon mutants of Bordetella bronchiseptica exhibit altered abilities to invade and survive in eukaryotic cells.FEMS Microbiol Lett 146, 263-269.[Medline]

Whittam, T. S. (1995). Genetic population structure and pathogenicity in enteric bacteria. In Population Genetics of Bacteria (Society for General Microbiology Symposium 52), pp. 217-245. Edited by S. Baumberg, J. P. W. Young, E. M. H. Wellington & J. R. Saunders. Cambridge: Cambridge University Press.

Wilson, D. R. & Beveridge, T. J. (1993). Bacterial flagellar filaments and their component flagellins.Can J Microbiol 39, 451-472.[Medline]

Winstanley, C. & Morgan, J. A. (1997). The bacterial flagellin gene as a biomarker for detection, population genetics and epidemiological analysis.Microbiology 143, 3071-3084.[Free Full Text]

Winstanley, C., Coulson, M., Wepner, B., Morgan, J. A. & Hart, C. (1996). Flagellin gene and protein variation amongst clinical isolates of Pseudomonas aeruginosa.Microbiology 142, 2145-2151.[Abstract]

Zhang, M. Y., Lovgren, A., Low, M. G. & Landen, R. (1993). Characterization of an avirulent pleiotropic mutant of the insect pathogen Bacillus thuringiensis: reduced expression of flagellin and phospholipases.Infect Immun 61, 4947-4954.[Abstract]

Received 22 September 1999; revised 20 December 1999; accepted 7 January 2000.