Cytolethal Distending Toxin (CDT)-Negative Campylobacter jejuni Strains and Anti-CDT Neutralizing Antibodies Are Induced during Human Infect
http://www.100md.com
感染与免疫杂志 2005年第5期
Veterinary Laboratories Agency (Weybridge), Surrey, United Kingdom
Molecular Microbiology and Genomics Consultants, Zotzenheim, Germany
ABSTRACT
The cytolethal distending toxin (CDT) of Campylobacter jejuni was detectable, using an in vitro assay, in most but not all of 24 strains tested. The reason for the absence of toxin activity in these naturally occurring CDT-negative C. jejuni strains was then investigated at the genetic level. CDT is encoded by three highly conserved genes, cdtA, -B, and -C. In the CDT-negative strains, two types of mutation were identified. The CDT activities of C. jejuni strains possessing both types of mutation were successfully complemented with the functional genes of C. jejuni 11168. The first type of mutation comprised a 667-bp deletion across cdtA and cdtB and considerable degeneration in the remainder of the cdt locus. Using a PCR technique to screen for this deletion, this mutation occurred in fewer than 3% of 147 human, veterinary, and environmental strains tested. The second type of mutation involved at least four nonsynonymous nucleotide changes, but only the replacement of proline with serine at CdtB position 95 was considered important for CDT activity. This was confirmed by site-directed mutagenesis. This type of mutation also occurred in fewer than 3% of strains as determined using a LightCycler biprobe assay. The detection of two CDT-negative clinical isolates raised questions about the role of CDT in some cases of human campylobacteriosis. To determine if anti-CDT antibodies are produced in human infection, a toxin neutralization assay was developed and validated using rabbit antisera. Pooled human sera from infected patients neutralized the toxin, indicating expression and immunogenicity during infection. However, no neutralizing antibodies were detected in colonized chickens despite the expression of CDT in the avian gut as indicated by reverse transcription-PCR.
INTRODUCTION
Campylobacter jejuni and Campylobacter coli are major causes of acute human bacterial enteritis in industrialized countries (35). These Campylobacter species asymptomatically colonize the intestinal tracts of most mammals and birds (24), and one major route of human campylobacteriosis is assumed to be the consumption of contaminated poultry meat products (10). The pathogenic mechanisms by which campylobacters cause diarrhea are as yet unknown, although motility, adhesion, and invasion have been implicated (38). Several toxic activities have been reported, but their roles in disease remain debatable (37).
The best-characterized Campylobacter toxin is the cytolethal distending toxin (CDT). CDT production has been described in several gram-negative bacteria, including Escherichia coli (30, 33), Haemophilus ducreyi (7), Actinobacillus actinomycetemcomitans (21), Shigella dysenteriae (25, 26), and Helicobacter spp. (42). However, not all of these species are implicated in enteric disease. In C. jejuni, CDT causes progressive cellular distension with eventual cell death (16). These morphological changes appear to be a consequence of alterations in the progression of the cell cycle, in particular cell cycle arrest in the G2/M phase (6, 8, 28, 41).
CDT production is dependent on the expression of three tandem genes, cdtA, cdtB, and cdtC (31). The CdtA, CdtB, and CdtC proteins form a tripartite holotoxin complex required for CDT activity (18). Current evidence indicates that cdtB encodes the active/toxic component of the toxin, while cdtA and cdtC are involved with binding to and internalization into the host cell (18, 19).
The role of CDT in human campylobacteriosis is unclear. Although, all C. jejuni strains tested to date appear to possess the cdt genes (11, 12, 31), the levels of toxin activities expressed are strain dependent, with two strains (1.2%) reported to produce no detectable levels of CDT in vitro (2, 12). The explanation for such CDT-negative strains is currently unknown. In this study we have investigated the molecular basis of this using eight C. jejuni CDT-negative strains isolated from human diarrheic stools (n = 2), bacteremia (blood) (n = 2), a sheep (n = 1), a poultry processing plant (n = 1), and a broiler (n = 2). The results indicated that lack of the CDT phenotype was a consequence of either major deletions (51 and 667 bp) in or around cdtB or one or more point mutations within the cdtABC genes. Site-directed mutagenesis and complementation were used to confirm these observations. A PCR assay and a LightCycler BiProbe assay were developed to screen 123 randomly selected veterinary and human Campylobacter isolates for either the deletion or the predominant point mutation. The isolation of CDT-negative strains from cases of human campylobacteriosis raised questions about the role of this potential virulence factor in disease. Therefore, we developed an assay to detect specific anti-CDT neutralizing antibodies in sera from infected individuals. The results of these studies indicated that circulating antibody responses, which neutralize CDT activity, are elicited during human infection but not during chicken colonization with C. jejuni.
MATERIALS AND METHODS
Bacterial strains and growth conditions. A panel of 24 human (n = 10), veterinary (n = 6), and environmental (n = 8) C. jejuni strains were initially tested for CDT activity. In addition, three strains (81116, 11168, and 81-176) previously reported to have CDT activity and a cdtABC mutant of 81-176 (81-176cdt), kindly provided by Carol Pickett (41), were used as controls. The DNAs from an additional 123 strains, isolated from humans, poultry, cattle, and sheep, were tested in the molecular screening assays developed. Chemically competent E. coli TOPO10F' (Invitrogen, Paisley, United Kingdom) and DH5MCR (Invitrogen) strains were used for cloning and site-directed mutagenesis studies, respectively.
C. jejuni strains were grown for 24 h at 42°C under microaerobic conditions (7.5% O2, 7.5% CO2, 85% N2) on either Mueller-Hinton or blood agar plates supplemented with 10% sheep blood, 50 μg/ml Actidione, and selective antibiotics (Oxoid, Basingstoke, United Kingdom) (34). C. jejuni 81-176cdt was grown on medium supplemented with 50 μg/ml of kanamycin. E. coli was grown in Luria-Bertani (LB) medium under atmospheric conditions at 37°C. Where necessary, LB medium was supplemented with 50 μg/ml of ampicillin or 20 μg/ml of chloramphenicol. The C. jejuni and E. coli strains were stored frozen at –80°C in 1% (wt/vol) proteose peptone water containing 10% (vol/vol) glycerol or in LB broth containing 50% (vol/vol) glycerol, respectively.
In vitro HeLa cell CDT assay. The assay used in this study was adapted from previously published CDT assays (14, 31). Cultured HeLa cells (European Collection of Cell Cultures, Porton Down, United Kingdom) were maintained in complete Eagle's minimal essential medium with 10% (vol/vol) fetal bovine serum, 1% (vol/vol) L-glutamine, 1% (vol/vol) nonessential amino acids, and 0.5% (vol/vol) gentamicin at 37°C in 5% CO2. For the assay, HeLa cell concentrations were adjusted to 2 x 104 cells ml–1 and 150 μl of this suspension was added to each well of a flat-bottomed, tissue culture-grade, 96-well plate. The cells were then cultured for 2 to 3 h at 37°C in 5% CO2, prior to the addition of bacterial lysates.
Lysates were prepared from bacteria grown on Mueller-Hinton agar plates for 24 h at 42°C under microaerobic conditions and harvested into complete Eagle's minimal essential medium. The bacterial suspensions were adjusted to an optical density at 550 nm of 1.6 and sonicated on ice (1.5 min at 30 A with 6-s pulses) (Vibra Cell; Sonics and Materials Inc., United States), and cell debris was removed by centrifugation (6,000 x g for 10 min at 4°C). The lysates were sterilized by filtration using 0.22-μm-pore-size filters (Minisart; Sartorius, Germany). Undiluted and 1/10 dilutions of the bacterial lysates were added to HeLa cells and twofold dilutions performed across each plate, which were then incubated for 5 days at 37°C in 5% CO2. The HeLa cell monolayers were then fixed in 10% (vol/vol) formalin, stained with 2.3% (wt/vol) crystal violet, and examined by light microscopy. The toxin titers were expressed as the reciprocal of the highest dilution that caused 50% of cells to become distended and were adjusted by dividing the optical density at 550 nm of bacterial sonic lysates. The toxin titer for each strain was tested blind in at least three independent assays.
CDT antibody neutralization assay. Hyperimmune rabbit antisera were produced against whole organisms of C. jejuni strains 81116 (R12), EF (R43) and C37596 (R42) as previously described (23). Sera from 10 chickens experimentally colonized with C. jejuni 81116 were collected at 9 weeks postchallenge as previously described (3). Sera pooled from six human patients, taken 2 to 4 weeks after campylobacter isolation (pooled positive sera), and from six human donors with no recorded enteric infection (pooled negative sera, were used in the neutralization assay. The collection, storage, and characterization of these sera have been described previously (4, 5).
All sera used were initially treated at 56°C for 2 h to inactivate complement before use in the in vitro HeLa cell CDT assays. Prior to the addition to the HeLa cells, 1/10 dilutions of bacterial lysates were preincubated with R12 (1/50 dilution), R42 (1/30 dilution), R43 (1/30 dilution), pooled chicken sera (1/30 dilution), or pooled human sera (1/30 dilution) at 37°C for 1 h. Treated lysates were subsequently applied to the HeLa cells as described previously.
Statistical analyses. Experimental results from at least 3 independent assays were transformed to log (x + 0.5) for analysis. Mean titers for each treatment were compared to those for controls by two-way analysis of variance, and a P value of <0.05 indicated statistical significance.
Western blotting. Whole-cell protein profiles of C. jejuni 11168 and 81-176 were prepared by separation on 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electrotransferred to nitrocellulose membrane for Western blotting as described previously (3). Membrane strips of each strain were incubated with the human and chicken sera (1/30 dilution) used for the CDT neutralization studies, and bound antibodies were detected using species-specific anti-immunoglobulin antisera conjugated to horseradish peroxidase according to the instructions of the manufacturers (DakoCytomation Ltd., Cambridgeshire, United Kingdom, and Amersham Biosciences UK Ltd., Buckinghamshire, United Kingdom, respectively).
PCR assay for cdt genes. Chromosomal DNA was isolated from strains by the cetyltrimethylammonium bromide-NaCl method (1). Recombinant Taq DNA polymerase (Invitrogen) was used for all PCRs according to the manufacturer's recommendations. Oligonucleotide primers (Sigma Genosys, Haverhill, United Kingdom) used for PCR and sequencing reactions are given in Table 1. PCRs were performed on a PE Applied Biosystems GeneAmp PCR 9700, with an initial denaturation step of 95°C for 5 min, followed by 25 cycles of 95°C for 2 min, 60°C for 2 min, and 72°C for 1.5 min and a final extension step at 72°C for 10 min.
Cloning and DNA sequence analysis of cdt genes. PCRs were repeated using Advantage cDNA polymerase mix (BD Biosciences Clontech, Oxford, United Kingdom), containing 3'-5' proofreading activity, according to the manufacturer's instructions. PCR products were then cloned into pCR2.1-TOPO (TOPO-TA cloning kit; Invitrogen) and transformed into chemically competent E. coli TOPO10F' One Shot cells (Invitrogen) according to the manufacturer's instructions. Plasmid DNA was isolated for analysis using a mini-plasmid preparation kit (Qiagen Ltd., Crawley, United Kingdom) and the DNA inserts confirmed by restriction analysis. Inserts were sequenced using BigDye terminator cycle sequencing chemistry (Applied Biosystems, Warrington, United Kingdom) according to the manufacturer's instructions. Sequenced products were separated on an ABI Prism 377 automated DNA sequencer (Applied Biosystems). DNA sequences were assembled and edited using SeqMan (DNAstar; Lasergene, Madison, United States), and ClustalV alignments to the published C. jejuni 81-176 cdtABC genes (GenBank accession number U51121) were done in MegAlign 5.00 (DNAstar; Lasergene, Madison, United States).
Southern blot analysis. Genomic DNA (10 μg) was digested with the restriction endonuclease HindIII (Promega, Southampton, United Kingdom) overnight at 37°C and used for Southern blot analysis using the CDP-Star chemiluminescent detection reagent (Amersham Biosciences, Bucks, United Kingdom) according to the manufacturer's instructions.
RNA extraction and RT-PCR. Total RNA was isolated from C. jejuni strains and chicken cecal contents by using Tri reagent (Sigma Aldrich, Poole, United Kingdom) according to the manufacturer's instructions. RNA was suspended in an appropriate amount of RNase-free water and RNA concentrations estimated by spectrophotometry. Reverse transcriptase PCR (RT-PCR) was carried out using the high-fidelity ProSTAR HF single-tube RT-PCR system (Stratagene, Texas, United States), according to the manufacturer's instructions. Since the cdt genes of C. jejuni strains are known to be expressed in a single mRNA transcript (14), RT-PCR primers DS15 and DS18 (Table 1) were used to amplify the region between cdtA and cdtB. RT-PCR was performed with an initial incubation step of 42°C for 30 min, during which time cDNA was synthesized from the RNA template. The reverse transcriptase was inactivated and the cDNA denatured by incubation at 95°C for 1 min, followed by an amplification reaction comprising 40 cycles of 95°C for 30 s, 60°C for 30 s, and 68°C for 2 min. RT-PCR products were visualized on 1.2% (wt/vol) agarose gels.
LightCycler BiProbe assay of cdtB polymorphism. For the LightCycler, PCR primers LC-T-F and LC-T-R (Table 1) were designed by alignment of the published sequence of the C. jejuni 11168 cdtB gene with the cdtB gene from C. jejuni strain EF to amplify a 163-bp region encompassing codon 95 of cdtB. The polymorphism was detected by melting curve analysis of the probe LC-T-probe (Table 1) after PCR amplification. The probe was identical to the sequence of C. jejuni EF, containing a serine at codon 95 (TCT). Amplification and probe hybridization were performed in 20-μl reaction mixtures, using the LightCycler DNA Master SYBR Green I kit (Roche Diagnostics Ltd., Lewes, United Kingdom) in a LightCycler instrument (Roche Diagnostics Ltd.). Typically, reaction mixtures comprised 10 to 15 pmol of DNA, 2 pmol of LC-T-F, 5 pmol of LC-T-R, 5 pmol of LC-T-probe, and 3 mM MgCl2 in 1x LightCycler DNA SYBR Green master mix. LightCycler PCR was performed for 50 cycles, with 15 s at 95°C, 5 s at 55°C, and 10 s at 74°C and a transition rate of 20°C/s. Fluorescence was measured at a wavelength of 540 nm (F1 channel) at the end of each amplification step to monitor the accumulation of PCR product. Melting curve analysis was performed immediately after amplification by heating the product to 94°C (20°C/s), cooling to 40°C for 10 s, and then heating to 85°C (0.1°C/s). The final heating step was performed under continuous fluorescence measurement. DNAs from C. jejuni EF and 11168 were included as positive controls, and a no-template control was included in each run.
Complementation studies of CDT-negative strains. Complementation of CDT-negative strains was achieved by the introduction of the Campylobacter shuttle vector pUOA18 (36), expressing a 2.4-kb region from C. jejuni 11168 containing the cdt genes (41). The region, including 206 bp of upstream and 106 bp of downstream sequence, was amplified with ClCdtABC-F and ClCdtABC-R (Table 1) and initially cloned into pCR2.1 TOPO. These primers included BamHI restriction enzyme sites at the 5' ends of both primers. Following BamHI digestion and gel purification, the insert was ligated into dephosphorylated, BamHI-digested pUOA18, to make the construct pCDT, and transformed into DH5MCR (Invitrogen). The construct was isolated from E. coli and electroporated into C. jejuni EF and 99/68, using a method adapted from that of Wassenaar et al. (39). Colonies that appeared on 20 mg/ml of chloramphenicol medium were tested for CDT activity in the in vitro CDT assay.
Site-directed mutagenesis of cdtB codon 95. Pro-95 of CdtB was changed to Ser-95 by using the QuikChange XL site-directed mutagenesis kit (Stratagene), according to the manufacturer's instructions. Mutagenic primers (Table 1) were used to generate pCdtBP95S, and mutants were identified by HaeIII restriction digestion of the cdtABC genes and sequencing of the entire insert.
Chick colonization model. A quantitative chick colonization model was used as previously described (40). Briefly, groups of 10 specific-pathogen-free, 1-day-old chickens (Charles River SPAFAS Inc., Hanover, Germany), housed in isolators, were dosed by oral gavage with C. jejuni 81-176, the 81-176 cdtABC mutant, or C289/6. Doses, administered in 100 μl of 0.1 M phosphate-buffered saline, ranged from 102 to 109 CFU. Doses were prepared by harvesting bacteria, grown overnight on 10% blood agar plates at 42°C, into sterile 0.1 M phosphate-buffered saline. At 5 days postchallenge, colonization levels were determined by plating of dilutions of cecal contents. Colonization levels were determined as CFU per gram of cecal contents for individual birds. All animal experiments were performed in accordance with the local ethics committee and United Kingdom home office license guidelines.
Nucleotide sequence accession numbers. The cdt gene sequences have been submitted to the GenBank database under the following accession numbers: C. jejuni C37596, AY442300; C. jejuni C35926, AY442302; C. jejuni C37533, AY442301; and C. jejuni EF, AY445094.
RESULTS
Identification and characterization of CDT-negative C. jejuni strains. A panel of 24 selected C. jejuni strains were tested for CDT activity. A range of CDT activities, including titers of <50 (n = 13), 50 to 150 (n = 4), 150 to 300 (n = 2), and >300 (n = 1), were observed, but there was no obvious correlation with strain source (data not shown). Of the 24 strains tested, 4 (C37596, C35926, C37533, and EF) had no detectable CDT activity. Two of these CDT-negative strains, C37596 and C35926, were isolated from the blood of independent campylobacteriosis patients (within the United Kingdom in 2000) with underlying medical problems, as discussed below. C. jejuni strain C37533 was isolated from the feces of the same patient as isolate C37596. None of these strains were serotypeable. The remaining CDT-negative strain (EF) was isolated from a chicken carcass in the chilling area of a poultry-processing plant.
The presence of cdtABC genes in all 24 strains was determined by PCR. In 21 of the 24 strains, the expected PCR amplicon size of 2,143 bp, generated with primers CdtF and CdtR, was observed (Fig. 1). However, in strains C37596, C35926, and C37533, which were CDT negative, a smaller product of 1,400 bp was detected. The PCR products of all four CDT-negative strains were cloned and sequenced. ClustalV alignment to the published cdtABC sequence of C. jejuni 81-176 (accession number U51121) revealed a 667-bp deletion between cdtA and cdtB and a separate 51-bp deletion within the cdtB genes of strains C37596, C35926, and C37533, consistent with the reduced PCR amplicon size observed by PCR (Fig. 2). The remainder of the sequences from these three strains had only 51% identity, which was due to additional insertions-deletions and substitutions throughout the remaining cdtABC gene sequence.
We next determined, by Southern blot hybridization, whether the 667-bp deleted region between cdtA and cdtB was the result of recombination with another part of the genome or whether it was completely lost from the genome. Using primers specific for this region (C7-Del), a PCR product was generated from strain 11168 and used to probe HindIII-digested genomic DNAs of strains C37596, C35926, and C37533. The results confirmed the absence of the region between cdtA and cdtB in the genomic DNAs of these strains (data not shown).
To determine the prevalence of the 667-bp deletion in C. jejuni isolates, the DNAs of an additional 123 randomly selected clinical and veterinary strains were screened by PCR with primers Cdt-F and Cdt-R. All but one strain, 99/68, produced a PCR product of 2.14 kb. Strain 99/68, which was isolated from a broiler, produced an amplicon of only 1,400 bp, indicative of a deletion. Cloning and sequencing of this amplicon revealed 96% sequence identity with the cdt genes of C37596, C35926, and C37533, which were previously identified as having the same deletions (data not shown). This result indicates that significant deletions in the cdt genes occur in 2.7% (4 of 147 isolates tested) of C. jejuni strains, but when such deletions in the cdt gene locus are present, they were found between cdtA and cdtB.
Interestingly the sequences of the cdt genes from the CDT-negative strain, EF, identified 18 nucleotide substitutions compared to previously published sequences of C. jejuni cdtABC (strains 11168 and 81-176). Sequence analysis indicated that all three reading frames were open in C. jejuni EF. Of the 18 nucleotide substitutions, only 4 were identified as nonsynonymous: an alanine to a valine at codon 88 in CdtA, a proline to serine and a methionine to threonine at codons 95 and 120, respectively, of CdtB (Fig. 3), and an isoleucine to asparagine at codon 167 of CdtC. In order to determine if the lack of CDT activity in strain EF was due to a loss of gene expression rather than to the substitutions, RT-PCR analysis was undertaken. Total RNA extracted from C. jejuni EF was used as a template, and RT-PCR was carried out, using the DS15 and DS18 primers (Table 1). An mRNA transcript was detected from both C. jejuni EF and 11168 (Fig. 4A), indicating that gene expression occurred and that the lack of toxicity was most likely caused by the production of inactive toxin, probably as a consequence of one or more of the amino acid substitutions.
Alignment of the CdtB amino acid sequences from other CDT-producing bacteria revealed Pro-95 to be a highly conserved residue (Fig. 3), suggesting that this residue is important for the function of the protein. In order to determine the frequency of the mutation at this residue, melting curve analysis using a LightCycler assay with a probe designed to match the EF Ser-95 mutation was developed and performed on the original 24 strains and the additional 123 strains (Fig. 5). The assay identified two strains (99/373 and S58, from a human and a sheep, respectively), in addition to EF, which gave a melting temperature of 60°C, indicating the presence of the Ser-95 mutation. A third strain (99/12), also from a broiler, which had a lower melting temperature of 48°C was identified, suggesting that more than one mutation was present within the region probed. These observations were further confirmed by sequencing of the cdtABC genes. Strains 99/373 and S58 were identical at the amino acid level to C. jejuni EF, except that S58 did not have the substitution at codon 167 in CdtC changing an isoleucine to asparagine. In contrast, the sequence of 99/12 revealed several amino acid substitutions in addition to those found in C. jejuni EF; eight of these occurred in CdtB and six in CdtC. In addition to these substitutions, four nucleotide insertions and one deletion were also identified. Overall the frequency of the Pro-95-Ser mutation, as detected by the LightCycler BiProbe assay, was 2.29% (3 of 147 strains tested). The absence of CDT activity in strains 99/373, S58, and 99/12 was confirmed using the in vitro HeLa cell CDT assay (data not shown). By RT-PCR the cdt genes were transcribed in strains 99/373, S58, and 99/12 (Fig. 4B), providing supporting evidence once again suggesting that these mutations were associated with the lack of CDT activity.
Complementation and site-directed mutagenesis. To verify whether CDT-negative strains could produce active CDT, a complementation vector (pCDT), containing functional cdt genes of C. jejuni 11168, was introduced into C. jejuni EF and 99/68. All transformants tested produced active CDT in the in vitro CDT assays, with levels comparable to those produced by C. jejuni 11168 (Fig. 6).
To ascertain whether the proline residue at CdtB position 95 (CdtB-95) was essential for CDT activity, site-directed mutagenesis was used to mutate proline to serine at this residue in the complementation vector pCDT to generate pCdtBP95S. C. jejuni EF and 99/68 were complemented with pCdtBP95S and transformants tested for CDT activity. The single mutation at CdtB-95 to serine resulted in 98.4% (P = 0.0004) and 97.3% (P < 0.0001) reductions in CDT titer for EF and 99/68, respectively (Fig. 6). These results indicated that the residue at CdtB-95 is critical for the toxicity of CDT in C. jejuni.
Expression and immunogenicity of CDT during colonization. As yet there are no suitable in vivo models of campylobacteriosis (22); however, the 1-day-old chicken is an excellent model of colonization. Challenge of 1-day-old chicks with the CDT-negative mutant of 81-176 gave the same level of colonization, up to 109 CFU per gram of cecal contents, as challenge with the parent strain (data not shown), suggesting that the absence of CDT expression does not affect colonization potential.
Despite this high level of colonization, no clinical signs of disease were discernible in chicks challenged with either the mutant or the wild-type strain. These differences in outcome of colonization in the chicken and human are well recognized but unexplained (22). To determine whether the cdt genes were expressed during chicken colonization, the cecal contents of chicks challenged for up to 10 days with CDT-positive C. jejuni strain C289/6 were investigated by RT-PCR. The results indicated that CDT is expressed during colonization of the avian gut (Fig. 7).
As determining the CDT expression in humans by a similar approach was not feasible, an alternative approach was sought. For other bacteria, the development of host circulating antitoxin antibody responses has provided indicators of both in vivo expression and the virulence potentials of putative toxins (9, 13). Therefore an assay to detect specific anti-CDT antibodies was developed. In this assay, sera from experimentally colonized chickens, hyperimmunized rabbits, and patients with campylobacteriosis were tested for their capacities to neutralize the in vitro CDT activity from three CDT-positive C. jejuni strains, 81116, 81-176, and 11168, which had CDT titers of 23.6 (± 0.8), 16.2 (± 3.7), and 28.6 (± 4.8), respectively. When lysates from these strains were pretreated with rabbit anti-C. jejuni 81116 antisera (R12), the CDT activity was completely or largely neutralized (100%, 100%, and 79%, respectively) (Fig. 8), whereas no neutralization was observed with sera from the preimmunized rabbits, indicating that the CDT expressed during in vitro culture is antigenic in immunized rabbits. To confirm the specificity of this neutralizing activity, rabbit antisera directed against two of the CDT-negative strains, C37596 (R42) and EF (R43), were also tested in the neutralization assay. As expected, R42 did not neutralize the CDT activity of any of the strains tested, which is consistent with the presence of the large deletions and the substantial degeneration of the cdt locus in this strain. In contrast, R43 only partly neutralized the CDT activities of strains 81116, 81-176, and 11168 (71%, 54%, and 46%, respectively) (Fig. 8), probably reflecting changes in the antigenic structure of CDT in strain EF as a consequence of the amino acid sequence variation.
The neutralization assay was further used to determine the presence of anti-CDT neutralizing antibodies in pooled sera from patients convalescing from campylobacteriosis. Previous studies, using enzyme-linked immunosorbent assay (ELISA) and Western blotting (4), demonstrated the presence and specificity of anticampylobacter antibodies in the serum from each individual patient. Complete (100%) neutralization of the CDT activity was observed in all strains (Fig. 8). This level of neutralization was significantly higher (P = 0.001) than the level of neutralization obtained with pooled human sera from blood donors with no history of enteric disease and no demonstrable anticampylobacter antibodies as detected by ELISA and Western blotting (4), indicating that CDT both is expressed during human infection and induces antibody responses.
Finally we investigated the presence of neutralizing anti-CDT antibodies in the pooled sera of chickens experimentally colonized with C. jejuni 81116. The development of humoral responses directed against the antigens of this strain during colonization has been reported previously (3) and observed in two other strains by Western blotting (Fig. 9). Interestingly, in contrast to antibodies from colonized humans and immunized rabbits, these chicken antibodies demonstrated no detectable CDT-neutralizing activity (Fig. 8) against either the homologous strain or the two heterologous strains.
DISCUSSION
The lack of a suitable small animal model of campylobacteriosis continues to be a major limitation to our understanding of the bacterial pathogenic mechanisms involved in campylobacteriosis. Thus, potential C. jejuni virulence factors have been defined largely on the basis of in vitro properties. Although many toxin activities have been observed in C. jejuni strains in vitro, to date only CDT has been sufficiently characterized to suggest that this is a true virulence factor with a role in enteric campylobacteriosis. Our observations confirmed previous reports (12, 31) that the majority of C. jejuni strains express detectable CDT activity. However, clearly the level of CDT activity expressed varied between strains, and moreover, some strains expressed no detectable activity.
Although CDT-negative C. jejuni strains have been described previously (11, 12, 31), the frequency, molecular basis, and potential consequences of this negative phenotype have not been fully investigated. From our studies, two distinct types of naturally occurring mutation have been identified in the cdt locus of CDT-negative strains. The deletion mutation, detected in isolates from two unrelated patients and a broiler, involved a 667-bp deleted region between cdtA and cdtB and a separate 51-bp deletion further downstream in cdtB, as well as several point mutations throughout the remaining cdt genes. Southern blot analysis indicated that these observations were the consequence of a genomic deletion rather than a transition. It therefore seems likely that the major deletion occurred first and was followed over time by the subsequent degeneration of the residual surrounding, now redundant, genetic material.
The recovery of such mutants from patients, albeit immunocompromised, with clinical symptoms calls into question the role of active CDT in the pathogenesis of campylobacteriosis. That CDT is not required for effective colonization of the avian intestine was supported by results from this study using the experimental challenge of 1-day-old chicks with a defined cdt mutant and by previously published work using the mammalian intestine and a cdtB mutant (32). However, without a suitable animal model, the role of CDT in disease is more difficult to determine. Two out of the three clinical infections with CDT-negative strains were in patients with underlying problems, of cirrhosis and/or neutropenia, which could potentially induce an immunocompromised status and thus influence the outcome of that infection. Interestingly, both patients developed campylobacter bacteremia. However, in one of these patients, from whom strain C37596 was isolated, an earlier fecal isolate (C37533) was also recovered during a period of acute enteritis and had the same mutation as the blood isolate. The isolation of CDT-negative fecal strains (C37533 and 99/373) from patients with enteric disease indicates that CDT expression is not essential for the production of enteric disease in humans. Mechanisms other than CDT expression may be involved in the outcome of campylobacteriosis. Similar large deletions in the toxin genes of other bacteria have been described, without obvious consequences for disease outcome. For example, large deletions (1.18 kb and 5.08 kb) have been reported in the vacuolating cytotoxin (vacA) of Helicobacter pylori (15), and yet these isolates were still associated with gastritis and peptic ulceration.
The second type of naturally occurring mutation identified in CDT-negative strains (4 out of 147 strains tested) from various sources (a poultry-processing plant, a broiler, a sheep, and a human) were point mutations resulting in at least three amino acid substitutions. As RT-PCR demonstrated that the cdt genes of these strains were transcribed, we hypothesized that these mutations resulted in the loss of CDT activity. This hypothesis was further supported by complementation and site-directed mutagenesis studies in which the CDT activity was restored by cdt genes from a CDT-positive strain.
Because CDT is a tripartite protein, in which the CdtA and -C subunits are required for the delivery of the CdtB subunit containing the enzymatically active site associated with toxin activity (18), nonsynomonous point mutations in these genes may have different effects on phenotype. The comparisons of the sequences within this group of strains suggest that the minimal set of mutations required to induce naturally occurring CDT negativity were represented in strain EF. In this strain a mutation resulting in the replacement of alanine with valine at codon 88 was detected in CdtA. This change was also conserved in the other two CDT-negative strains within this group. As both amino acids are aliphatic and hydrophobic and have nonpolar side chains, this transition is unlikely to substantially affect the structure or folding of the polypeptide chain. Moreover, alignment of CdtA amino acid sequences from Haemophilus and Actinobacillus species with that from strain EF also showed a valine at this position (data not shown), and as both of these bacterial species produce active CDT, it appears that this polymorphism within the CdtA subunit has little consequence for the activity of this toxin. The mutation at codon 167 of CdtC resulted in the replacement of isoleucine with asparagine and was found in an additional C. jejuni strain, 99/373, a human isolate. This results in a change from a nonpolar hydrophobic residue to a hydrophilic polar residue. Alignment of the CdtC amino acid sequence from Helicobacter hepaticus with that from C. jejuni 11168 shows that the two sequences are identical at this position. This substitution may therefore have a possible role in the lack of CDT activity in C. jejuni EF.
Moreover, there were two potentially important nonsynonomous changes in cdtB in this group of CDT-negative strains. All demonstrated a replacement of methionine by threonine at CdtB-120. Although this could potentially influence the three-dimensional structure of CdtB, alignment of the amino acid sequences from other bacterial species shows considerable variability at this position (Fig. 3), suggesting that this mutation in CdtB would have little, if any, consequence for CDT activity. In contrast, the replacement of proline with serine at CdtB-95 appeared more important. Proline at CdtB-95 appears to be a highly conserved residue (Fig. 3), and its role in CDT activity has now been confirmed using site-directed mutagenesis. Interestingly, CDT is known to have similarities with diphosphodiesterases, the active sites of which have been identified (17, 20). However, neither CdtB-95 nor any other of the mutations identified in C. jejuni EF fell into the conserved or functionally important residues predicted for such enzymes. Although CdtB-95 is clearly essential for CDT activity, whether any of the observed or as-yet-unidentified mutations within the cdt locus could explain the observed variability in levels of CDT activity detectable in vitro is unknown.
For toxins expressed by related organisms, such as VacA of Helicobacter pylori (9, 29), the immunogenicity of the toxin during infection, as determined by Western blotting, is considered evidence of both in vivo expression and importance as a marker of pathogenicity. Unfortunately, Western blotting of C. jejuni cell preparations using anti-CDT antibodies from human campylobacteriosis patient sera proved unsuccessful. Therefore, we developed a CDT neutralization assay to detect serum anti-CDT antibodies. This assay was validated using serum from a rabbit hyperimmunized with the whole cells of CDT-expressing C. jejuni strain 81116. This antiserum completely neutralized the CDT activity of this strain. The specificity of these neutralizing antibodies was then demonstrated using rabbit antiserum directed against strain C37596, which contained large deletions in the cdt locus. This antiserum failed to neutralize the toxin activity of C. jejuni 81116, 11168, and 81-176. In contrast, antiserum directed against strain EF, containing the point mutations, partly neutralized this activity. Comparison of the neutralizing capacities of the anti-C. jejuni 81116 antisera against several strains suggests that the observed variation may be due to a quantitative rather than a qualitative variation in toxin titer (due to coding region differences), which may be under some regulatory control.
This CDT neutralization assay was then used to investigate the presence of neutralizing antibodies in the sera of patients recovering from campylobacteriosis. Because of the limited availability of such human sera and the large volumes of sera required for such studies, a pool of six sera from patients was used and compared with a pool of six sera from donors with no laboratory-confirmed history of enteric disease. Previous studies using ELISA and Western blotting had already demonstrated the presence of specific anticampylobacter antibodies in the sera from the individual patients and the absence of these antibodies in the sera from the individual control donors (4, 5). The results clearly demonstrated the complete neutralization of CDT activity by the pooled sera from the campylobacteriosis patients. In contrast, the level of neutralization by the pooled sera from the control donors was significantly lower (P = 0.001). These results suggest that CDT is expressed by C. jejuni during human enteric infection and is antigenic and that antibodies directed against this antigen can neutralize the toxin activity.
Finally, the neutralization assay was also used to investigate CDT immunogenicity during chicken colonization. This absence of detectable neutralizing antibodies, despite a substantial antibody response in colonized chickens (Fig. 9) (3) and the demonstrable expression of CDT in the avian gut, was surprising but suggests that CDT is not antigenic in chickens. The reason for this is currently under investigation but may reflect host-specific differences in immune responsiveness, particularly as the colonization model uses young chicks. It is possible that nonneutralizing antibodies are induced. However, it is well recognized that colonization in chickens, unlike that in humans, is asymptomatic. Thus, the relationship between the lack of detectable immunogenicity of CDT in the chicken and the absence of a disease outcome of colonization may be a valuable future approach to the investigation of host-specific differences in campylobacter infection.
In summary, CDT remains the only clearly identified toxin in the genome sequence of C. jejuni (27). Nevertheless, the recovery, albeit rarely, of naturally occurring CDT-negative C. jejuni strains raises questions about the role of this toxin in Campylobacter biology. However, infected patients with disease symptoms elicited circulating and neutralizing antibodies directed against this toxin during infection. Moreover, the absence of similar antibodies in colonized, asymptomatic chickens suggests that the role of this toxin in disease requires considerable further investigation.
ACKNOWLEDGMENTS
Many thanks go to Jenny Frost (Health Protection Agency, Colindale), Exeter Public Health Laboratories, and CAMPYNET for providing some of the isolates used in this study.
The work described in this paper was funded by the Department for Environment, Food and Rural Affairs (Defra), Great Britain.
REFERENCES
1. Ausubel, S. F., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl. 1994. Current protocols in molecular biology. John Wiley & Sons, New York, N.Y.
2. Bang, D. D., F. Scheutz, P. Ahrens, K. Pedersen, J. Blom, and M. Madsen. 2001. Prevalence of cytolethal distending toxin (cdt) genes and CDT production in Campylobacter spp. isolated from Danish broilers. J. Med. Microbiol. 50:1087-1094.
3. Cawthraw, S., R. Ayling, P. Nuijten, T. Wassenaar, and D. G. Newell. 1994. Isotype, specificity, and kinetics of systemic and mucosal antibodies to Campylobacter jejuni antigens, including flagellin, during experimental oral infections of chickens. Avian Dis. 38:341-349.
4. Cawthraw, S. A., R. A. Feldman, A. R. Sayers, and D. G. Newell. 2002. Long-term antibody responses following human infection with Campylobacter jejuni. Clin. Exp. Immunol. 130:101-106.
5. Cawthraw, S. A., L. Lind, B. Kaijser, and D. G. Newell. 2000. Antibodies, directed towards Campylobacter jejuni antigens, in sera from poultry abattoir workers. Clin. Exp. Immunol. 122:55-60.
6. Comayras, C., C. Tasca, S. Y. Peres, B. Ducommun, E. Oswald, and J. De Rycke. 1997. Escherichia coli cytolethal distending toxin blocks the HeLa cell cycle at the G2/M transition by preventing Cdc2 protein kinase dephosphorylation and activation. Infect. Immun. 65:5088-5095.
7. Cope, L. D., S. Lumbley, J. L. Latimer, J. Klesney-Tait, M. K. Stevens, L. S. Johnson, M. Purven, R. S. Munson, Jr., T. Lagergard, J. D. Radolf, and E. J. Hansen. 1997. A diffusible cytotoxin of Haemophilus ducreyi. Proc. Natl. Acad. Sci. USA 94:4056-4061.
8. Cortes-Bratti, X., E. Chaves-Olarte, T. Lagergard, and M. Thelestam. 1999. The cytolethal distending toxin from the chancroid bacterium Haemophilus ducreyi induces cell-cycle arrest in the G2 phase. J. Clin. Investig. 103:107-115.
9. Cover, T. L., P. Cao, U. K. Murthy, M. S. Sipple, and M. J. Blaser. 1992. Serum neutralizing antibody response to the vacuolating cytotoxin of Helicobacter pylori. J. Clin. Investig. 90:913-918.
10. Deming, M. S., R. V. Tauxe, P. A. Blake, S. E. Dixon, B. S. Fowler, T. S. Jones, E. A. Lockamy, C. M. Patton, and R. O. Sikes. 1987. Campylobacter enteritis at a university: transmission from eating chicken and from cats. Am. J. Epidemiol. 126:526-534.
11. Eyigor, A., K. A. Dawson, B. E. Langlois, and C. L. Pickett. 1999. Cytolethal distending toxin genes in Campylobacter jejuni and Campylobacter coli isolates: detection and analysis by PCR. J. Clin. Microbiol. 37:1646-1650.
12. Eyigor, A., K. A. Dawson, B. E. Langlois, and C. L. Pickett. 1999. Detection of cytolethal distending toxin activity and cdt genes in Campylobacter spp. isolated from chicken carcasses. Appl. Environ. Microbiol. 65:1501-1505.
13. Gosciniak, G., A. Przondo-Mordarska, B. Iwanczak, and E. Poniewierka. 2000. Neutralization of vacuolating activity of cytotoxin by serum antibodies of Helicobacter pylori infected patients. Acta Microbiol. Pol. 49:113-120.
14. Hickey, T. E., A. L. McVeigh, D. A. Scott, R. E. Michielutti, A. Bixby, S. A. Carroll, A. L. Bourgeois, and P. Guerry. 2000. Campylobacter jejuni cytolethal distending toxin mediates release of interleukin-8 from intestinal epithelial cells. Infect. Immun. 68:6535-6541.
15. Ito, Y., T. Azuma, S. Ito, H. Suto, H. Miyaji, Y. Yamazaki, Y. Kohli, and M. Kuriyama. 1998. Full-length sequence analysis of the vacA gene from cytotoxic and noncytotoxic Helicobacter pylori. J Infect. Dis. 178:1391-1398.
16. Johnson, W. M., and H. Lior. 1988. A new heat-labile cytolethal distending toxin (CLDT) produced by Campylobacter spp. Microb. Pathog. 4:115-126.
17. Lara-Tejero, M., and J. E. Galan. 2000. A bacterial toxin that controls cell cycle progression as a deoxyribonuclease I-like protein. Science 290:354-357.
18. Lara-Tejero, M., and J. E. Galan. 2001. CdtA, CdtB, and CdtC form a tripartite complex that is required for cytolethal distending toxin activity. Infect. Immun. 69:4358-4365.
19. Mao, X., and J. M. DiRienzo. 2002. Functional studies of the recombinant subunits of a cytolethal distending holotoxin. Cell Microbiol. 4:245-255.
20. Matsuo, Y., A. Yamada, K. Tsukamoto, H. Tamura, H. Ikezawa, H. Nakamura, and K. Nishikawa. 1996. A distant evolutionary relationship between bacterial sphingomyelinase and mammalian DNase I. Protein Sci. 5:2459-2467.
21. Mayer, M. P., L. C. Bueno, E. J. Hansen, and J. M. DiRienzo. 1999. Identification of a cytolethal distending toxin gene locus and features of a virulence-associated region in Actinobacillus actinomycetemcomitans. Infect. Immun. 67:1227-1237.
22. Newell, D. G. 2001. Animal models of Campylobacter jejuni colonization and disease and the lessons to be learned from similar Helicobacter pylori models. Symp. Ser. Soc. Appl. Microbiol. 30:57S-67S.
23. Newell, D. G., H. McBride, and A. D. Pearson. 1984. The identification of outer membrane proteins and flagella of Campylobacter jejuni. J. Gen. Microbiol. 130:1201-1208.
24. Newell, D. G., J. E. Shreeve, M. Toszeghy, G. Domingue, S. Bull, T. Humphrey, and G. Mead. 2001. Changes in the carriage of Campylobacter strains by poultry carcasses during processing in abattoirs. Appl. Environ. Microbiol. 67:2636-2640.
25. Okuda, J., M. Fukumoto, Y. Takeda, and M. Nishibuchi. 1997. Examination of diarrheagenicity of cytolethal distending toxin: suckling mouse response to the products of the cdtABC genes of Shigella dysenteriae. Infect. Immun. 65:428-433.
26. Okuda, J., H. Kurazono, and Y. Takeda. 1995. Distribution of the cytolethal distending toxin A gene (cdtA) among species of Shigella and Vibrio, and cloning and sequencing of the cdt gene from Shigella dysenteriae. Microb. Pathog. 18:167-172.
27. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S. Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668.
28. Peres, S. Y., O. Marches, F. Daigle, J. P. Nougayrede, F. Herault, C. Tasca, J. De Rycke, and E. Oswald. 1997. A new cytolethal distending toxin (CDT) from Escherichia coli producing CNF2 blocks HeLa cell division in G2/M phase. Mol. Microbiol. 24:1095-1107.
29. Perez-Perez, G. I., W. R. Brown, T. L. Cover, B. E. Dunn, P. Cao, and M. J. Blaser. 1994. Correlation between serological and mucosal inflammatory responses to Helicobacter pylori. Clin. Diagn. Lab. Immunol. 1:325-329.
30. Pickett, C. L., D. L. Cottle, E. C. Pesci, and G. Bikah. 1994. Cloning, sequencing, and expression of the Escherichia coli cytolethal distending toxin genes. Infect. Immun. 62:1046-1051.
31. Pickett, C. L., E. C. Pesci, D. L. Cottle, G. Russell, A. N. Erdem, and H. Zeytin. 1996. Prevalence of cytolethal distending toxin production in Campylobacter jejuni and relatedness of Campylobacter sp. cdtB gene. Infect. Immun. 64:2070-2078.
32. Purdy, D., C. M. Buswell, A. E. Hodgson, K. McAlpine, I. Henderson, and S. A. Leach. 2000. Characterisation of cytolethal distending toxin (CDT) mutants of Campylobacter jejuni. J. Med. Microbiol. 49:473-479.
33. Scott, D. A., and J. B. Kaper. 1994. Cloning and sequencing of the genes encoding Escherichia coli cytolethal distending toxin. Infect. Immun. 62:244-251.
34. Skirrow, M. B. 1977. Campylobacter enteritis: a "new" disease. Br. Med. J. 2:9-11.
35. Tauxe, R. E. 1992. Epidemiology of Campylobacter jejuni infections in the United States and other industrialized nations, p. 9-19. In I. Nachamkin, M. J. Blaser, and L. S. Tompkins (ed.), Campylobacter jejuni: current status and future trends. American Society for Microbiology, Washington, D.C.
36. Wang, Y., and D. E. Taylor. 1990. Chloramphenicol resistance in Campylobacter coli: nucleotide sequence, expression, and cloning vector construction. Gene 94:23-28.
37. Wassenaar, T. M. 1997. Toxin production by Campylobacter spp. Clin. Microbiol. Rev. 10:466-476.
38. Wassenaar, T. M., and M. J. Blaser. 1999. Pathophysiology of Campylobacter jejuni infections of humans. Microbes Infect. 1:1023-1033.
39. Wassenaar, T. M., B. N. Fry, and B. A. van der Zeijst. 1993. Genetic manipulation of Campylobacter: evaluation of natural transformation and electro-transformation. Gene 132:131-135.
40. Wassenaar, T. M., B. A. van der Zeijst, R. Ayling, and D. G. Newell. 1993. Colonization of chicks by motility mutants of Campylobacter jejuni demonstrates the importance of flagellin A expression. J. Gen. Microbiol. 139:1171-1175.
41. Whitehouse, C. A., P. B. Balbo, E. C. Pesci, D. L. Cottle, P. M. Mirabito, and C. L. Pickett. 1998. Campylobacter jejuni cytolethal distending toxin causes a G2-phase cell cycle block. Infect. Immun. 66:1934-1940.
42. Young, V. B., K. A. Knox, and D. B. Schauer. 2000. Cytolethal distending toxin sequence and activity in the enterohepatic pathogen Helicobacter hepaticus. Infect. Immun. 68:184-191.(Manal AbuOun, Georgina Ma)
Molecular Microbiology and Genomics Consultants, Zotzenheim, Germany
ABSTRACT
The cytolethal distending toxin (CDT) of Campylobacter jejuni was detectable, using an in vitro assay, in most but not all of 24 strains tested. The reason for the absence of toxin activity in these naturally occurring CDT-negative C. jejuni strains was then investigated at the genetic level. CDT is encoded by three highly conserved genes, cdtA, -B, and -C. In the CDT-negative strains, two types of mutation were identified. The CDT activities of C. jejuni strains possessing both types of mutation were successfully complemented with the functional genes of C. jejuni 11168. The first type of mutation comprised a 667-bp deletion across cdtA and cdtB and considerable degeneration in the remainder of the cdt locus. Using a PCR technique to screen for this deletion, this mutation occurred in fewer than 3% of 147 human, veterinary, and environmental strains tested. The second type of mutation involved at least four nonsynonymous nucleotide changes, but only the replacement of proline with serine at CdtB position 95 was considered important for CDT activity. This was confirmed by site-directed mutagenesis. This type of mutation also occurred in fewer than 3% of strains as determined using a LightCycler biprobe assay. The detection of two CDT-negative clinical isolates raised questions about the role of CDT in some cases of human campylobacteriosis. To determine if anti-CDT antibodies are produced in human infection, a toxin neutralization assay was developed and validated using rabbit antisera. Pooled human sera from infected patients neutralized the toxin, indicating expression and immunogenicity during infection. However, no neutralizing antibodies were detected in colonized chickens despite the expression of CDT in the avian gut as indicated by reverse transcription-PCR.
INTRODUCTION
Campylobacter jejuni and Campylobacter coli are major causes of acute human bacterial enteritis in industrialized countries (35). These Campylobacter species asymptomatically colonize the intestinal tracts of most mammals and birds (24), and one major route of human campylobacteriosis is assumed to be the consumption of contaminated poultry meat products (10). The pathogenic mechanisms by which campylobacters cause diarrhea are as yet unknown, although motility, adhesion, and invasion have been implicated (38). Several toxic activities have been reported, but their roles in disease remain debatable (37).
The best-characterized Campylobacter toxin is the cytolethal distending toxin (CDT). CDT production has been described in several gram-negative bacteria, including Escherichia coli (30, 33), Haemophilus ducreyi (7), Actinobacillus actinomycetemcomitans (21), Shigella dysenteriae (25, 26), and Helicobacter spp. (42). However, not all of these species are implicated in enteric disease. In C. jejuni, CDT causes progressive cellular distension with eventual cell death (16). These morphological changes appear to be a consequence of alterations in the progression of the cell cycle, in particular cell cycle arrest in the G2/M phase (6, 8, 28, 41).
CDT production is dependent on the expression of three tandem genes, cdtA, cdtB, and cdtC (31). The CdtA, CdtB, and CdtC proteins form a tripartite holotoxin complex required for CDT activity (18). Current evidence indicates that cdtB encodes the active/toxic component of the toxin, while cdtA and cdtC are involved with binding to and internalization into the host cell (18, 19).
The role of CDT in human campylobacteriosis is unclear. Although, all C. jejuni strains tested to date appear to possess the cdt genes (11, 12, 31), the levels of toxin activities expressed are strain dependent, with two strains (1.2%) reported to produce no detectable levels of CDT in vitro (2, 12). The explanation for such CDT-negative strains is currently unknown. In this study we have investigated the molecular basis of this using eight C. jejuni CDT-negative strains isolated from human diarrheic stools (n = 2), bacteremia (blood) (n = 2), a sheep (n = 1), a poultry processing plant (n = 1), and a broiler (n = 2). The results indicated that lack of the CDT phenotype was a consequence of either major deletions (51 and 667 bp) in or around cdtB or one or more point mutations within the cdtABC genes. Site-directed mutagenesis and complementation were used to confirm these observations. A PCR assay and a LightCycler BiProbe assay were developed to screen 123 randomly selected veterinary and human Campylobacter isolates for either the deletion or the predominant point mutation. The isolation of CDT-negative strains from cases of human campylobacteriosis raised questions about the role of this potential virulence factor in disease. Therefore, we developed an assay to detect specific anti-CDT neutralizing antibodies in sera from infected individuals. The results of these studies indicated that circulating antibody responses, which neutralize CDT activity, are elicited during human infection but not during chicken colonization with C. jejuni.
MATERIALS AND METHODS
Bacterial strains and growth conditions. A panel of 24 human (n = 10), veterinary (n = 6), and environmental (n = 8) C. jejuni strains were initially tested for CDT activity. In addition, three strains (81116, 11168, and 81-176) previously reported to have CDT activity and a cdtABC mutant of 81-176 (81-176cdt), kindly provided by Carol Pickett (41), were used as controls. The DNAs from an additional 123 strains, isolated from humans, poultry, cattle, and sheep, were tested in the molecular screening assays developed. Chemically competent E. coli TOPO10F' (Invitrogen, Paisley, United Kingdom) and DH5MCR (Invitrogen) strains were used for cloning and site-directed mutagenesis studies, respectively.
C. jejuni strains were grown for 24 h at 42°C under microaerobic conditions (7.5% O2, 7.5% CO2, 85% N2) on either Mueller-Hinton or blood agar plates supplemented with 10% sheep blood, 50 μg/ml Actidione, and selective antibiotics (Oxoid, Basingstoke, United Kingdom) (34). C. jejuni 81-176cdt was grown on medium supplemented with 50 μg/ml of kanamycin. E. coli was grown in Luria-Bertani (LB) medium under atmospheric conditions at 37°C. Where necessary, LB medium was supplemented with 50 μg/ml of ampicillin or 20 μg/ml of chloramphenicol. The C. jejuni and E. coli strains were stored frozen at –80°C in 1% (wt/vol) proteose peptone water containing 10% (vol/vol) glycerol or in LB broth containing 50% (vol/vol) glycerol, respectively.
In vitro HeLa cell CDT assay. The assay used in this study was adapted from previously published CDT assays (14, 31). Cultured HeLa cells (European Collection of Cell Cultures, Porton Down, United Kingdom) were maintained in complete Eagle's minimal essential medium with 10% (vol/vol) fetal bovine serum, 1% (vol/vol) L-glutamine, 1% (vol/vol) nonessential amino acids, and 0.5% (vol/vol) gentamicin at 37°C in 5% CO2. For the assay, HeLa cell concentrations were adjusted to 2 x 104 cells ml–1 and 150 μl of this suspension was added to each well of a flat-bottomed, tissue culture-grade, 96-well plate. The cells were then cultured for 2 to 3 h at 37°C in 5% CO2, prior to the addition of bacterial lysates.
Lysates were prepared from bacteria grown on Mueller-Hinton agar plates for 24 h at 42°C under microaerobic conditions and harvested into complete Eagle's minimal essential medium. The bacterial suspensions were adjusted to an optical density at 550 nm of 1.6 and sonicated on ice (1.5 min at 30 A with 6-s pulses) (Vibra Cell; Sonics and Materials Inc., United States), and cell debris was removed by centrifugation (6,000 x g for 10 min at 4°C). The lysates were sterilized by filtration using 0.22-μm-pore-size filters (Minisart; Sartorius, Germany). Undiluted and 1/10 dilutions of the bacterial lysates were added to HeLa cells and twofold dilutions performed across each plate, which were then incubated for 5 days at 37°C in 5% CO2. The HeLa cell monolayers were then fixed in 10% (vol/vol) formalin, stained with 2.3% (wt/vol) crystal violet, and examined by light microscopy. The toxin titers were expressed as the reciprocal of the highest dilution that caused 50% of cells to become distended and were adjusted by dividing the optical density at 550 nm of bacterial sonic lysates. The toxin titer for each strain was tested blind in at least three independent assays.
CDT antibody neutralization assay. Hyperimmune rabbit antisera were produced against whole organisms of C. jejuni strains 81116 (R12), EF (R43) and C37596 (R42) as previously described (23). Sera from 10 chickens experimentally colonized with C. jejuni 81116 were collected at 9 weeks postchallenge as previously described (3). Sera pooled from six human patients, taken 2 to 4 weeks after campylobacter isolation (pooled positive sera), and from six human donors with no recorded enteric infection (pooled negative sera, were used in the neutralization assay. The collection, storage, and characterization of these sera have been described previously (4, 5).
All sera used were initially treated at 56°C for 2 h to inactivate complement before use in the in vitro HeLa cell CDT assays. Prior to the addition to the HeLa cells, 1/10 dilutions of bacterial lysates were preincubated with R12 (1/50 dilution), R42 (1/30 dilution), R43 (1/30 dilution), pooled chicken sera (1/30 dilution), or pooled human sera (1/30 dilution) at 37°C for 1 h. Treated lysates were subsequently applied to the HeLa cells as described previously.
Statistical analyses. Experimental results from at least 3 independent assays were transformed to log (x + 0.5) for analysis. Mean titers for each treatment were compared to those for controls by two-way analysis of variance, and a P value of <0.05 indicated statistical significance.
Western blotting. Whole-cell protein profiles of C. jejuni 11168 and 81-176 were prepared by separation on 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electrotransferred to nitrocellulose membrane for Western blotting as described previously (3). Membrane strips of each strain were incubated with the human and chicken sera (1/30 dilution) used for the CDT neutralization studies, and bound antibodies were detected using species-specific anti-immunoglobulin antisera conjugated to horseradish peroxidase according to the instructions of the manufacturers (DakoCytomation Ltd., Cambridgeshire, United Kingdom, and Amersham Biosciences UK Ltd., Buckinghamshire, United Kingdom, respectively).
PCR assay for cdt genes. Chromosomal DNA was isolated from strains by the cetyltrimethylammonium bromide-NaCl method (1). Recombinant Taq DNA polymerase (Invitrogen) was used for all PCRs according to the manufacturer's recommendations. Oligonucleotide primers (Sigma Genosys, Haverhill, United Kingdom) used for PCR and sequencing reactions are given in Table 1. PCRs were performed on a PE Applied Biosystems GeneAmp PCR 9700, with an initial denaturation step of 95°C for 5 min, followed by 25 cycles of 95°C for 2 min, 60°C for 2 min, and 72°C for 1.5 min and a final extension step at 72°C for 10 min.
Cloning and DNA sequence analysis of cdt genes. PCRs were repeated using Advantage cDNA polymerase mix (BD Biosciences Clontech, Oxford, United Kingdom), containing 3'-5' proofreading activity, according to the manufacturer's instructions. PCR products were then cloned into pCR2.1-TOPO (TOPO-TA cloning kit; Invitrogen) and transformed into chemically competent E. coli TOPO10F' One Shot cells (Invitrogen) according to the manufacturer's instructions. Plasmid DNA was isolated for analysis using a mini-plasmid preparation kit (Qiagen Ltd., Crawley, United Kingdom) and the DNA inserts confirmed by restriction analysis. Inserts were sequenced using BigDye terminator cycle sequencing chemistry (Applied Biosystems, Warrington, United Kingdom) according to the manufacturer's instructions. Sequenced products were separated on an ABI Prism 377 automated DNA sequencer (Applied Biosystems). DNA sequences were assembled and edited using SeqMan (DNAstar; Lasergene, Madison, United States), and ClustalV alignments to the published C. jejuni 81-176 cdtABC genes (GenBank accession number U51121) were done in MegAlign 5.00 (DNAstar; Lasergene, Madison, United States).
Southern blot analysis. Genomic DNA (10 μg) was digested with the restriction endonuclease HindIII (Promega, Southampton, United Kingdom) overnight at 37°C and used for Southern blot analysis using the CDP-Star chemiluminescent detection reagent (Amersham Biosciences, Bucks, United Kingdom) according to the manufacturer's instructions.
RNA extraction and RT-PCR. Total RNA was isolated from C. jejuni strains and chicken cecal contents by using Tri reagent (Sigma Aldrich, Poole, United Kingdom) according to the manufacturer's instructions. RNA was suspended in an appropriate amount of RNase-free water and RNA concentrations estimated by spectrophotometry. Reverse transcriptase PCR (RT-PCR) was carried out using the high-fidelity ProSTAR HF single-tube RT-PCR system (Stratagene, Texas, United States), according to the manufacturer's instructions. Since the cdt genes of C. jejuni strains are known to be expressed in a single mRNA transcript (14), RT-PCR primers DS15 and DS18 (Table 1) were used to amplify the region between cdtA and cdtB. RT-PCR was performed with an initial incubation step of 42°C for 30 min, during which time cDNA was synthesized from the RNA template. The reverse transcriptase was inactivated and the cDNA denatured by incubation at 95°C for 1 min, followed by an amplification reaction comprising 40 cycles of 95°C for 30 s, 60°C for 30 s, and 68°C for 2 min. RT-PCR products were visualized on 1.2% (wt/vol) agarose gels.
LightCycler BiProbe assay of cdtB polymorphism. For the LightCycler, PCR primers LC-T-F and LC-T-R (Table 1) were designed by alignment of the published sequence of the C. jejuni 11168 cdtB gene with the cdtB gene from C. jejuni strain EF to amplify a 163-bp region encompassing codon 95 of cdtB. The polymorphism was detected by melting curve analysis of the probe LC-T-probe (Table 1) after PCR amplification. The probe was identical to the sequence of C. jejuni EF, containing a serine at codon 95 (TCT). Amplification and probe hybridization were performed in 20-μl reaction mixtures, using the LightCycler DNA Master SYBR Green I kit (Roche Diagnostics Ltd., Lewes, United Kingdom) in a LightCycler instrument (Roche Diagnostics Ltd.). Typically, reaction mixtures comprised 10 to 15 pmol of DNA, 2 pmol of LC-T-F, 5 pmol of LC-T-R, 5 pmol of LC-T-probe, and 3 mM MgCl2 in 1x LightCycler DNA SYBR Green master mix. LightCycler PCR was performed for 50 cycles, with 15 s at 95°C, 5 s at 55°C, and 10 s at 74°C and a transition rate of 20°C/s. Fluorescence was measured at a wavelength of 540 nm (F1 channel) at the end of each amplification step to monitor the accumulation of PCR product. Melting curve analysis was performed immediately after amplification by heating the product to 94°C (20°C/s), cooling to 40°C for 10 s, and then heating to 85°C (0.1°C/s). The final heating step was performed under continuous fluorescence measurement. DNAs from C. jejuni EF and 11168 were included as positive controls, and a no-template control was included in each run.
Complementation studies of CDT-negative strains. Complementation of CDT-negative strains was achieved by the introduction of the Campylobacter shuttle vector pUOA18 (36), expressing a 2.4-kb region from C. jejuni 11168 containing the cdt genes (41). The region, including 206 bp of upstream and 106 bp of downstream sequence, was amplified with ClCdtABC-F and ClCdtABC-R (Table 1) and initially cloned into pCR2.1 TOPO. These primers included BamHI restriction enzyme sites at the 5' ends of both primers. Following BamHI digestion and gel purification, the insert was ligated into dephosphorylated, BamHI-digested pUOA18, to make the construct pCDT, and transformed into DH5MCR (Invitrogen). The construct was isolated from E. coli and electroporated into C. jejuni EF and 99/68, using a method adapted from that of Wassenaar et al. (39). Colonies that appeared on 20 mg/ml of chloramphenicol medium were tested for CDT activity in the in vitro CDT assay.
Site-directed mutagenesis of cdtB codon 95. Pro-95 of CdtB was changed to Ser-95 by using the QuikChange XL site-directed mutagenesis kit (Stratagene), according to the manufacturer's instructions. Mutagenic primers (Table 1) were used to generate pCdtBP95S, and mutants were identified by HaeIII restriction digestion of the cdtABC genes and sequencing of the entire insert.
Chick colonization model. A quantitative chick colonization model was used as previously described (40). Briefly, groups of 10 specific-pathogen-free, 1-day-old chickens (Charles River SPAFAS Inc., Hanover, Germany), housed in isolators, were dosed by oral gavage with C. jejuni 81-176, the 81-176 cdtABC mutant, or C289/6. Doses, administered in 100 μl of 0.1 M phosphate-buffered saline, ranged from 102 to 109 CFU. Doses were prepared by harvesting bacteria, grown overnight on 10% blood agar plates at 42°C, into sterile 0.1 M phosphate-buffered saline. At 5 days postchallenge, colonization levels were determined by plating of dilutions of cecal contents. Colonization levels were determined as CFU per gram of cecal contents for individual birds. All animal experiments were performed in accordance with the local ethics committee and United Kingdom home office license guidelines.
Nucleotide sequence accession numbers. The cdt gene sequences have been submitted to the GenBank database under the following accession numbers: C. jejuni C37596, AY442300; C. jejuni C35926, AY442302; C. jejuni C37533, AY442301; and C. jejuni EF, AY445094.
RESULTS
Identification and characterization of CDT-negative C. jejuni strains. A panel of 24 selected C. jejuni strains were tested for CDT activity. A range of CDT activities, including titers of <50 (n = 13), 50 to 150 (n = 4), 150 to 300 (n = 2), and >300 (n = 1), were observed, but there was no obvious correlation with strain source (data not shown). Of the 24 strains tested, 4 (C37596, C35926, C37533, and EF) had no detectable CDT activity. Two of these CDT-negative strains, C37596 and C35926, were isolated from the blood of independent campylobacteriosis patients (within the United Kingdom in 2000) with underlying medical problems, as discussed below. C. jejuni strain C37533 was isolated from the feces of the same patient as isolate C37596. None of these strains were serotypeable. The remaining CDT-negative strain (EF) was isolated from a chicken carcass in the chilling area of a poultry-processing plant.
The presence of cdtABC genes in all 24 strains was determined by PCR. In 21 of the 24 strains, the expected PCR amplicon size of 2,143 bp, generated with primers CdtF and CdtR, was observed (Fig. 1). However, in strains C37596, C35926, and C37533, which were CDT negative, a smaller product of 1,400 bp was detected. The PCR products of all four CDT-negative strains were cloned and sequenced. ClustalV alignment to the published cdtABC sequence of C. jejuni 81-176 (accession number U51121) revealed a 667-bp deletion between cdtA and cdtB and a separate 51-bp deletion within the cdtB genes of strains C37596, C35926, and C37533, consistent with the reduced PCR amplicon size observed by PCR (Fig. 2). The remainder of the sequences from these three strains had only 51% identity, which was due to additional insertions-deletions and substitutions throughout the remaining cdtABC gene sequence.
We next determined, by Southern blot hybridization, whether the 667-bp deleted region between cdtA and cdtB was the result of recombination with another part of the genome or whether it was completely lost from the genome. Using primers specific for this region (C7-Del), a PCR product was generated from strain 11168 and used to probe HindIII-digested genomic DNAs of strains C37596, C35926, and C37533. The results confirmed the absence of the region between cdtA and cdtB in the genomic DNAs of these strains (data not shown).
To determine the prevalence of the 667-bp deletion in C. jejuni isolates, the DNAs of an additional 123 randomly selected clinical and veterinary strains were screened by PCR with primers Cdt-F and Cdt-R. All but one strain, 99/68, produced a PCR product of 2.14 kb. Strain 99/68, which was isolated from a broiler, produced an amplicon of only 1,400 bp, indicative of a deletion. Cloning and sequencing of this amplicon revealed 96% sequence identity with the cdt genes of C37596, C35926, and C37533, which were previously identified as having the same deletions (data not shown). This result indicates that significant deletions in the cdt genes occur in 2.7% (4 of 147 isolates tested) of C. jejuni strains, but when such deletions in the cdt gene locus are present, they were found between cdtA and cdtB.
Interestingly the sequences of the cdt genes from the CDT-negative strain, EF, identified 18 nucleotide substitutions compared to previously published sequences of C. jejuni cdtABC (strains 11168 and 81-176). Sequence analysis indicated that all three reading frames were open in C. jejuni EF. Of the 18 nucleotide substitutions, only 4 were identified as nonsynonymous: an alanine to a valine at codon 88 in CdtA, a proline to serine and a methionine to threonine at codons 95 and 120, respectively, of CdtB (Fig. 3), and an isoleucine to asparagine at codon 167 of CdtC. In order to determine if the lack of CDT activity in strain EF was due to a loss of gene expression rather than to the substitutions, RT-PCR analysis was undertaken. Total RNA extracted from C. jejuni EF was used as a template, and RT-PCR was carried out, using the DS15 and DS18 primers (Table 1). An mRNA transcript was detected from both C. jejuni EF and 11168 (Fig. 4A), indicating that gene expression occurred and that the lack of toxicity was most likely caused by the production of inactive toxin, probably as a consequence of one or more of the amino acid substitutions.
Alignment of the CdtB amino acid sequences from other CDT-producing bacteria revealed Pro-95 to be a highly conserved residue (Fig. 3), suggesting that this residue is important for the function of the protein. In order to determine the frequency of the mutation at this residue, melting curve analysis using a LightCycler assay with a probe designed to match the EF Ser-95 mutation was developed and performed on the original 24 strains and the additional 123 strains (Fig. 5). The assay identified two strains (99/373 and S58, from a human and a sheep, respectively), in addition to EF, which gave a melting temperature of 60°C, indicating the presence of the Ser-95 mutation. A third strain (99/12), also from a broiler, which had a lower melting temperature of 48°C was identified, suggesting that more than one mutation was present within the region probed. These observations were further confirmed by sequencing of the cdtABC genes. Strains 99/373 and S58 were identical at the amino acid level to C. jejuni EF, except that S58 did not have the substitution at codon 167 in CdtC changing an isoleucine to asparagine. In contrast, the sequence of 99/12 revealed several amino acid substitutions in addition to those found in C. jejuni EF; eight of these occurred in CdtB and six in CdtC. In addition to these substitutions, four nucleotide insertions and one deletion were also identified. Overall the frequency of the Pro-95-Ser mutation, as detected by the LightCycler BiProbe assay, was 2.29% (3 of 147 strains tested). The absence of CDT activity in strains 99/373, S58, and 99/12 was confirmed using the in vitro HeLa cell CDT assay (data not shown). By RT-PCR the cdt genes were transcribed in strains 99/373, S58, and 99/12 (Fig. 4B), providing supporting evidence once again suggesting that these mutations were associated with the lack of CDT activity.
Complementation and site-directed mutagenesis. To verify whether CDT-negative strains could produce active CDT, a complementation vector (pCDT), containing functional cdt genes of C. jejuni 11168, was introduced into C. jejuni EF and 99/68. All transformants tested produced active CDT in the in vitro CDT assays, with levels comparable to those produced by C. jejuni 11168 (Fig. 6).
To ascertain whether the proline residue at CdtB position 95 (CdtB-95) was essential for CDT activity, site-directed mutagenesis was used to mutate proline to serine at this residue in the complementation vector pCDT to generate pCdtBP95S. C. jejuni EF and 99/68 were complemented with pCdtBP95S and transformants tested for CDT activity. The single mutation at CdtB-95 to serine resulted in 98.4% (P = 0.0004) and 97.3% (P < 0.0001) reductions in CDT titer for EF and 99/68, respectively (Fig. 6). These results indicated that the residue at CdtB-95 is critical for the toxicity of CDT in C. jejuni.
Expression and immunogenicity of CDT during colonization. As yet there are no suitable in vivo models of campylobacteriosis (22); however, the 1-day-old chicken is an excellent model of colonization. Challenge of 1-day-old chicks with the CDT-negative mutant of 81-176 gave the same level of colonization, up to 109 CFU per gram of cecal contents, as challenge with the parent strain (data not shown), suggesting that the absence of CDT expression does not affect colonization potential.
Despite this high level of colonization, no clinical signs of disease were discernible in chicks challenged with either the mutant or the wild-type strain. These differences in outcome of colonization in the chicken and human are well recognized but unexplained (22). To determine whether the cdt genes were expressed during chicken colonization, the cecal contents of chicks challenged for up to 10 days with CDT-positive C. jejuni strain C289/6 were investigated by RT-PCR. The results indicated that CDT is expressed during colonization of the avian gut (Fig. 7).
As determining the CDT expression in humans by a similar approach was not feasible, an alternative approach was sought. For other bacteria, the development of host circulating antitoxin antibody responses has provided indicators of both in vivo expression and the virulence potentials of putative toxins (9, 13). Therefore an assay to detect specific anti-CDT antibodies was developed. In this assay, sera from experimentally colonized chickens, hyperimmunized rabbits, and patients with campylobacteriosis were tested for their capacities to neutralize the in vitro CDT activity from three CDT-positive C. jejuni strains, 81116, 81-176, and 11168, which had CDT titers of 23.6 (± 0.8), 16.2 (± 3.7), and 28.6 (± 4.8), respectively. When lysates from these strains were pretreated with rabbit anti-C. jejuni 81116 antisera (R12), the CDT activity was completely or largely neutralized (100%, 100%, and 79%, respectively) (Fig. 8), whereas no neutralization was observed with sera from the preimmunized rabbits, indicating that the CDT expressed during in vitro culture is antigenic in immunized rabbits. To confirm the specificity of this neutralizing activity, rabbit antisera directed against two of the CDT-negative strains, C37596 (R42) and EF (R43), were also tested in the neutralization assay. As expected, R42 did not neutralize the CDT activity of any of the strains tested, which is consistent with the presence of the large deletions and the substantial degeneration of the cdt locus in this strain. In contrast, R43 only partly neutralized the CDT activities of strains 81116, 81-176, and 11168 (71%, 54%, and 46%, respectively) (Fig. 8), probably reflecting changes in the antigenic structure of CDT in strain EF as a consequence of the amino acid sequence variation.
The neutralization assay was further used to determine the presence of anti-CDT neutralizing antibodies in pooled sera from patients convalescing from campylobacteriosis. Previous studies, using enzyme-linked immunosorbent assay (ELISA) and Western blotting (4), demonstrated the presence and specificity of anticampylobacter antibodies in the serum from each individual patient. Complete (100%) neutralization of the CDT activity was observed in all strains (Fig. 8). This level of neutralization was significantly higher (P = 0.001) than the level of neutralization obtained with pooled human sera from blood donors with no history of enteric disease and no demonstrable anticampylobacter antibodies as detected by ELISA and Western blotting (4), indicating that CDT both is expressed during human infection and induces antibody responses.
Finally we investigated the presence of neutralizing anti-CDT antibodies in the pooled sera of chickens experimentally colonized with C. jejuni 81116. The development of humoral responses directed against the antigens of this strain during colonization has been reported previously (3) and observed in two other strains by Western blotting (Fig. 9). Interestingly, in contrast to antibodies from colonized humans and immunized rabbits, these chicken antibodies demonstrated no detectable CDT-neutralizing activity (Fig. 8) against either the homologous strain or the two heterologous strains.
DISCUSSION
The lack of a suitable small animal model of campylobacteriosis continues to be a major limitation to our understanding of the bacterial pathogenic mechanisms involved in campylobacteriosis. Thus, potential C. jejuni virulence factors have been defined largely on the basis of in vitro properties. Although many toxin activities have been observed in C. jejuni strains in vitro, to date only CDT has been sufficiently characterized to suggest that this is a true virulence factor with a role in enteric campylobacteriosis. Our observations confirmed previous reports (12, 31) that the majority of C. jejuni strains express detectable CDT activity. However, clearly the level of CDT activity expressed varied between strains, and moreover, some strains expressed no detectable activity.
Although CDT-negative C. jejuni strains have been described previously (11, 12, 31), the frequency, molecular basis, and potential consequences of this negative phenotype have not been fully investigated. From our studies, two distinct types of naturally occurring mutation have been identified in the cdt locus of CDT-negative strains. The deletion mutation, detected in isolates from two unrelated patients and a broiler, involved a 667-bp deleted region between cdtA and cdtB and a separate 51-bp deletion further downstream in cdtB, as well as several point mutations throughout the remaining cdt genes. Southern blot analysis indicated that these observations were the consequence of a genomic deletion rather than a transition. It therefore seems likely that the major deletion occurred first and was followed over time by the subsequent degeneration of the residual surrounding, now redundant, genetic material.
The recovery of such mutants from patients, albeit immunocompromised, with clinical symptoms calls into question the role of active CDT in the pathogenesis of campylobacteriosis. That CDT is not required for effective colonization of the avian intestine was supported by results from this study using the experimental challenge of 1-day-old chicks with a defined cdt mutant and by previously published work using the mammalian intestine and a cdtB mutant (32). However, without a suitable animal model, the role of CDT in disease is more difficult to determine. Two out of the three clinical infections with CDT-negative strains were in patients with underlying problems, of cirrhosis and/or neutropenia, which could potentially induce an immunocompromised status and thus influence the outcome of that infection. Interestingly, both patients developed campylobacter bacteremia. However, in one of these patients, from whom strain C37596 was isolated, an earlier fecal isolate (C37533) was also recovered during a period of acute enteritis and had the same mutation as the blood isolate. The isolation of CDT-negative fecal strains (C37533 and 99/373) from patients with enteric disease indicates that CDT expression is not essential for the production of enteric disease in humans. Mechanisms other than CDT expression may be involved in the outcome of campylobacteriosis. Similar large deletions in the toxin genes of other bacteria have been described, without obvious consequences for disease outcome. For example, large deletions (1.18 kb and 5.08 kb) have been reported in the vacuolating cytotoxin (vacA) of Helicobacter pylori (15), and yet these isolates were still associated with gastritis and peptic ulceration.
The second type of naturally occurring mutation identified in CDT-negative strains (4 out of 147 strains tested) from various sources (a poultry-processing plant, a broiler, a sheep, and a human) were point mutations resulting in at least three amino acid substitutions. As RT-PCR demonstrated that the cdt genes of these strains were transcribed, we hypothesized that these mutations resulted in the loss of CDT activity. This hypothesis was further supported by complementation and site-directed mutagenesis studies in which the CDT activity was restored by cdt genes from a CDT-positive strain.
Because CDT is a tripartite protein, in which the CdtA and -C subunits are required for the delivery of the CdtB subunit containing the enzymatically active site associated with toxin activity (18), nonsynomonous point mutations in these genes may have different effects on phenotype. The comparisons of the sequences within this group of strains suggest that the minimal set of mutations required to induce naturally occurring CDT negativity were represented in strain EF. In this strain a mutation resulting in the replacement of alanine with valine at codon 88 was detected in CdtA. This change was also conserved in the other two CDT-negative strains within this group. As both amino acids are aliphatic and hydrophobic and have nonpolar side chains, this transition is unlikely to substantially affect the structure or folding of the polypeptide chain. Moreover, alignment of CdtA amino acid sequences from Haemophilus and Actinobacillus species with that from strain EF also showed a valine at this position (data not shown), and as both of these bacterial species produce active CDT, it appears that this polymorphism within the CdtA subunit has little consequence for the activity of this toxin. The mutation at codon 167 of CdtC resulted in the replacement of isoleucine with asparagine and was found in an additional C. jejuni strain, 99/373, a human isolate. This results in a change from a nonpolar hydrophobic residue to a hydrophilic polar residue. Alignment of the CdtC amino acid sequence from Helicobacter hepaticus with that from C. jejuni 11168 shows that the two sequences are identical at this position. This substitution may therefore have a possible role in the lack of CDT activity in C. jejuni EF.
Moreover, there were two potentially important nonsynonomous changes in cdtB in this group of CDT-negative strains. All demonstrated a replacement of methionine by threonine at CdtB-120. Although this could potentially influence the three-dimensional structure of CdtB, alignment of the amino acid sequences from other bacterial species shows considerable variability at this position (Fig. 3), suggesting that this mutation in CdtB would have little, if any, consequence for CDT activity. In contrast, the replacement of proline with serine at CdtB-95 appeared more important. Proline at CdtB-95 appears to be a highly conserved residue (Fig. 3), and its role in CDT activity has now been confirmed using site-directed mutagenesis. Interestingly, CDT is known to have similarities with diphosphodiesterases, the active sites of which have been identified (17, 20). However, neither CdtB-95 nor any other of the mutations identified in C. jejuni EF fell into the conserved or functionally important residues predicted for such enzymes. Although CdtB-95 is clearly essential for CDT activity, whether any of the observed or as-yet-unidentified mutations within the cdt locus could explain the observed variability in levels of CDT activity detectable in vitro is unknown.
For toxins expressed by related organisms, such as VacA of Helicobacter pylori (9, 29), the immunogenicity of the toxin during infection, as determined by Western blotting, is considered evidence of both in vivo expression and importance as a marker of pathogenicity. Unfortunately, Western blotting of C. jejuni cell preparations using anti-CDT antibodies from human campylobacteriosis patient sera proved unsuccessful. Therefore, we developed a CDT neutralization assay to detect serum anti-CDT antibodies. This assay was validated using serum from a rabbit hyperimmunized with the whole cells of CDT-expressing C. jejuni strain 81116. This antiserum completely neutralized the CDT activity of this strain. The specificity of these neutralizing antibodies was then demonstrated using rabbit antiserum directed against strain C37596, which contained large deletions in the cdt locus. This antiserum failed to neutralize the toxin activity of C. jejuni 81116, 11168, and 81-176. In contrast, antiserum directed against strain EF, containing the point mutations, partly neutralized this activity. Comparison of the neutralizing capacities of the anti-C. jejuni 81116 antisera against several strains suggests that the observed variation may be due to a quantitative rather than a qualitative variation in toxin titer (due to coding region differences), which may be under some regulatory control.
This CDT neutralization assay was then used to investigate the presence of neutralizing antibodies in the sera of patients recovering from campylobacteriosis. Because of the limited availability of such human sera and the large volumes of sera required for such studies, a pool of six sera from patients was used and compared with a pool of six sera from donors with no laboratory-confirmed history of enteric disease. Previous studies using ELISA and Western blotting had already demonstrated the presence of specific anticampylobacter antibodies in the sera from the individual patients and the absence of these antibodies in the sera from the individual control donors (4, 5). The results clearly demonstrated the complete neutralization of CDT activity by the pooled sera from the campylobacteriosis patients. In contrast, the level of neutralization by the pooled sera from the control donors was significantly lower (P = 0.001). These results suggest that CDT is expressed by C. jejuni during human enteric infection and is antigenic and that antibodies directed against this antigen can neutralize the toxin activity.
Finally, the neutralization assay was also used to investigate CDT immunogenicity during chicken colonization. This absence of detectable neutralizing antibodies, despite a substantial antibody response in colonized chickens (Fig. 9) (3) and the demonstrable expression of CDT in the avian gut, was surprising but suggests that CDT is not antigenic in chickens. The reason for this is currently under investigation but may reflect host-specific differences in immune responsiveness, particularly as the colonization model uses young chicks. It is possible that nonneutralizing antibodies are induced. However, it is well recognized that colonization in chickens, unlike that in humans, is asymptomatic. Thus, the relationship between the lack of detectable immunogenicity of CDT in the chicken and the absence of a disease outcome of colonization may be a valuable future approach to the investigation of host-specific differences in campylobacter infection.
In summary, CDT remains the only clearly identified toxin in the genome sequence of C. jejuni (27). Nevertheless, the recovery, albeit rarely, of naturally occurring CDT-negative C. jejuni strains raises questions about the role of this toxin in Campylobacter biology. However, infected patients with disease symptoms elicited circulating and neutralizing antibodies directed against this toxin during infection. Moreover, the absence of similar antibodies in colonized, asymptomatic chickens suggests that the role of this toxin in disease requires considerable further investigation.
ACKNOWLEDGMENTS
Many thanks go to Jenny Frost (Health Protection Agency, Colindale), Exeter Public Health Laboratories, and CAMPYNET for providing some of the isolates used in this study.
The work described in this paper was funded by the Department for Environment, Food and Rural Affairs (Defra), Great Britain.
REFERENCES
1. Ausubel, S. F., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl. 1994. Current protocols in molecular biology. John Wiley & Sons, New York, N.Y.
2. Bang, D. D., F. Scheutz, P. Ahrens, K. Pedersen, J. Blom, and M. Madsen. 2001. Prevalence of cytolethal distending toxin (cdt) genes and CDT production in Campylobacter spp. isolated from Danish broilers. J. Med. Microbiol. 50:1087-1094.
3. Cawthraw, S., R. Ayling, P. Nuijten, T. Wassenaar, and D. G. Newell. 1994. Isotype, specificity, and kinetics of systemic and mucosal antibodies to Campylobacter jejuni antigens, including flagellin, during experimental oral infections of chickens. Avian Dis. 38:341-349.
4. Cawthraw, S. A., R. A. Feldman, A. R. Sayers, and D. G. Newell. 2002. Long-term antibody responses following human infection with Campylobacter jejuni. Clin. Exp. Immunol. 130:101-106.
5. Cawthraw, S. A., L. Lind, B. Kaijser, and D. G. Newell. 2000. Antibodies, directed towards Campylobacter jejuni antigens, in sera from poultry abattoir workers. Clin. Exp. Immunol. 122:55-60.
6. Comayras, C., C. Tasca, S. Y. Peres, B. Ducommun, E. Oswald, and J. De Rycke. 1997. Escherichia coli cytolethal distending toxin blocks the HeLa cell cycle at the G2/M transition by preventing Cdc2 protein kinase dephosphorylation and activation. Infect. Immun. 65:5088-5095.
7. Cope, L. D., S. Lumbley, J. L. Latimer, J. Klesney-Tait, M. K. Stevens, L. S. Johnson, M. Purven, R. S. Munson, Jr., T. Lagergard, J. D. Radolf, and E. J. Hansen. 1997. A diffusible cytotoxin of Haemophilus ducreyi. Proc. Natl. Acad. Sci. USA 94:4056-4061.
8. Cortes-Bratti, X., E. Chaves-Olarte, T. Lagergard, and M. Thelestam. 1999. The cytolethal distending toxin from the chancroid bacterium Haemophilus ducreyi induces cell-cycle arrest in the G2 phase. J. Clin. Investig. 103:107-115.
9. Cover, T. L., P. Cao, U. K. Murthy, M. S. Sipple, and M. J. Blaser. 1992. Serum neutralizing antibody response to the vacuolating cytotoxin of Helicobacter pylori. J. Clin. Investig. 90:913-918.
10. Deming, M. S., R. V. Tauxe, P. A. Blake, S. E. Dixon, B. S. Fowler, T. S. Jones, E. A. Lockamy, C. M. Patton, and R. O. Sikes. 1987. Campylobacter enteritis at a university: transmission from eating chicken and from cats. Am. J. Epidemiol. 126:526-534.
11. Eyigor, A., K. A. Dawson, B. E. Langlois, and C. L. Pickett. 1999. Cytolethal distending toxin genes in Campylobacter jejuni and Campylobacter coli isolates: detection and analysis by PCR. J. Clin. Microbiol. 37:1646-1650.
12. Eyigor, A., K. A. Dawson, B. E. Langlois, and C. L. Pickett. 1999. Detection of cytolethal distending toxin activity and cdt genes in Campylobacter spp. isolated from chicken carcasses. Appl. Environ. Microbiol. 65:1501-1505.
13. Gosciniak, G., A. Przondo-Mordarska, B. Iwanczak, and E. Poniewierka. 2000. Neutralization of vacuolating activity of cytotoxin by serum antibodies of Helicobacter pylori infected patients. Acta Microbiol. Pol. 49:113-120.
14. Hickey, T. E., A. L. McVeigh, D. A. Scott, R. E. Michielutti, A. Bixby, S. A. Carroll, A. L. Bourgeois, and P. Guerry. 2000. Campylobacter jejuni cytolethal distending toxin mediates release of interleukin-8 from intestinal epithelial cells. Infect. Immun. 68:6535-6541.
15. Ito, Y., T. Azuma, S. Ito, H. Suto, H. Miyaji, Y. Yamazaki, Y. Kohli, and M. Kuriyama. 1998. Full-length sequence analysis of the vacA gene from cytotoxic and noncytotoxic Helicobacter pylori. J Infect. Dis. 178:1391-1398.
16. Johnson, W. M., and H. Lior. 1988. A new heat-labile cytolethal distending toxin (CLDT) produced by Campylobacter spp. Microb. Pathog. 4:115-126.
17. Lara-Tejero, M., and J. E. Galan. 2000. A bacterial toxin that controls cell cycle progression as a deoxyribonuclease I-like protein. Science 290:354-357.
18. Lara-Tejero, M., and J. E. Galan. 2001. CdtA, CdtB, and CdtC form a tripartite complex that is required for cytolethal distending toxin activity. Infect. Immun. 69:4358-4365.
19. Mao, X., and J. M. DiRienzo. 2002. Functional studies of the recombinant subunits of a cytolethal distending holotoxin. Cell Microbiol. 4:245-255.
20. Matsuo, Y., A. Yamada, K. Tsukamoto, H. Tamura, H. Ikezawa, H. Nakamura, and K. Nishikawa. 1996. A distant evolutionary relationship between bacterial sphingomyelinase and mammalian DNase I. Protein Sci. 5:2459-2467.
21. Mayer, M. P., L. C. Bueno, E. J. Hansen, and J. M. DiRienzo. 1999. Identification of a cytolethal distending toxin gene locus and features of a virulence-associated region in Actinobacillus actinomycetemcomitans. Infect. Immun. 67:1227-1237.
22. Newell, D. G. 2001. Animal models of Campylobacter jejuni colonization and disease and the lessons to be learned from similar Helicobacter pylori models. Symp. Ser. Soc. Appl. Microbiol. 30:57S-67S.
23. Newell, D. G., H. McBride, and A. D. Pearson. 1984. The identification of outer membrane proteins and flagella of Campylobacter jejuni. J. Gen. Microbiol. 130:1201-1208.
24. Newell, D. G., J. E. Shreeve, M. Toszeghy, G. Domingue, S. Bull, T. Humphrey, and G. Mead. 2001. Changes in the carriage of Campylobacter strains by poultry carcasses during processing in abattoirs. Appl. Environ. Microbiol. 67:2636-2640.
25. Okuda, J., M. Fukumoto, Y. Takeda, and M. Nishibuchi. 1997. Examination of diarrheagenicity of cytolethal distending toxin: suckling mouse response to the products of the cdtABC genes of Shigella dysenteriae. Infect. Immun. 65:428-433.
26. Okuda, J., H. Kurazono, and Y. Takeda. 1995. Distribution of the cytolethal distending toxin A gene (cdtA) among species of Shigella and Vibrio, and cloning and sequencing of the cdt gene from Shigella dysenteriae. Microb. Pathog. 18:167-172.
27. Parkhill, J., B. W. Wren, K. Mungall, J. M. Ketley, C. Churcher, D. Basham, T. Chillingworth, R. M. Davies, T. Feltwell, S. Holroyd, K. Jagels, A. V. Karlyshev, S. Moule, M. J. Pallen, C. W. Penn, M. A. Quail, M. A. Rajandream, K. M. Rutherford, A. H. van Vliet, S. Whitehead, and B. G. Barrell. 2000. The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403:665-668.
28. Peres, S. Y., O. Marches, F. Daigle, J. P. Nougayrede, F. Herault, C. Tasca, J. De Rycke, and E. Oswald. 1997. A new cytolethal distending toxin (CDT) from Escherichia coli producing CNF2 blocks HeLa cell division in G2/M phase. Mol. Microbiol. 24:1095-1107.
29. Perez-Perez, G. I., W. R. Brown, T. L. Cover, B. E. Dunn, P. Cao, and M. J. Blaser. 1994. Correlation between serological and mucosal inflammatory responses to Helicobacter pylori. Clin. Diagn. Lab. Immunol. 1:325-329.
30. Pickett, C. L., D. L. Cottle, E. C. Pesci, and G. Bikah. 1994. Cloning, sequencing, and expression of the Escherichia coli cytolethal distending toxin genes. Infect. Immun. 62:1046-1051.
31. Pickett, C. L., E. C. Pesci, D. L. Cottle, G. Russell, A. N. Erdem, and H. Zeytin. 1996. Prevalence of cytolethal distending toxin production in Campylobacter jejuni and relatedness of Campylobacter sp. cdtB gene. Infect. Immun. 64:2070-2078.
32. Purdy, D., C. M. Buswell, A. E. Hodgson, K. McAlpine, I. Henderson, and S. A. Leach. 2000. Characterisation of cytolethal distending toxin (CDT) mutants of Campylobacter jejuni. J. Med. Microbiol. 49:473-479.
33. Scott, D. A., and J. B. Kaper. 1994. Cloning and sequencing of the genes encoding Escherichia coli cytolethal distending toxin. Infect. Immun. 62:244-251.
34. Skirrow, M. B. 1977. Campylobacter enteritis: a "new" disease. Br. Med. J. 2:9-11.
35. Tauxe, R. E. 1992. Epidemiology of Campylobacter jejuni infections in the United States and other industrialized nations, p. 9-19. In I. Nachamkin, M. J. Blaser, and L. S. Tompkins (ed.), Campylobacter jejuni: current status and future trends. American Society for Microbiology, Washington, D.C.
36. Wang, Y., and D. E. Taylor. 1990. Chloramphenicol resistance in Campylobacter coli: nucleotide sequence, expression, and cloning vector construction. Gene 94:23-28.
37. Wassenaar, T. M. 1997. Toxin production by Campylobacter spp. Clin. Microbiol. Rev. 10:466-476.
38. Wassenaar, T. M., and M. J. Blaser. 1999. Pathophysiology of Campylobacter jejuni infections of humans. Microbes Infect. 1:1023-1033.
39. Wassenaar, T. M., B. N. Fry, and B. A. van der Zeijst. 1993. Genetic manipulation of Campylobacter: evaluation of natural transformation and electro-transformation. Gene 132:131-135.
40. Wassenaar, T. M., B. A. van der Zeijst, R. Ayling, and D. G. Newell. 1993. Colonization of chicks by motility mutants of Campylobacter jejuni demonstrates the importance of flagellin A expression. J. Gen. Microbiol. 139:1171-1175.
41. Whitehouse, C. A., P. B. Balbo, E. C. Pesci, D. L. Cottle, P. M. Mirabito, and C. L. Pickett. 1998. Campylobacter jejuni cytolethal distending toxin causes a G2-phase cell cycle block. Infect. Immun. 66:1934-1940.
42. Young, V. B., K. A. Knox, and D. B. Schauer. 2000. Cytolethal distending toxin sequence and activity in the enterohepatic pathogen Helicobacter hepaticus. Infect. Immun. 68:184-191.(Manal AbuOun, Georgina Ma)