Identification of genetic differences between two Campylobacter jejuni strains with different colonization potentials

If H. Ahmed1, Georgina Manning1, Trudy M. Wassenaar1,2, Shaun Cawthraw1 and Diane G. Newell1

Veterinary Laboratories Agency (Weybridge), New Haw, Addlestone, Surrey KT15 3NB, UK1
Molecular Microbiology and Genomics Consultants, Zotzenheim, Germany2

Author for correspondence: Diane G. Newell. Tel: +44 1932 357547. Fax: +44 1932 357595. e-mail: dnewell.cvl.wood{at}gtnet.gov.uk


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The consumption of poultry meat contaminated with Campylobacter jejuni is considered to be a risk factor for human campylobacteriosis. The development of targeted strategies to control campylobacters in broilers would benefit from knowledge of those bacterial factors important in colonization of the avian gut. During preliminary studies it was noted that C. jejuni NCTC 11168 was a poorer colonizer of chickens than strain 81116. This poor colonization could not be fully restored by in vivo passage, suggesting that it was a genetically endowed property of strain 11168. As the genome sequence is available for this strain, the technique of subtractive hybridization was used to identify gene fragments of strain 81116 not present in strain 11168. After two screening cycles, 24 out of 42 clones were identified as having DNA inserts specific for strain 81116. Six of these 24 clones contained gene fragment inserts with similarities to restriction–modification enzymes found in other bacteria. Two inserts had similarity to arsenic-resistance genes, whereas four others had similarities to cytochrome c oxidase III, dTDP-glucose 4,6-dehydratase, {gamma}-glutamyl transpeptidase and an abortive phage-resistance protein. At least some of these genes may be involved with colonization. A further six inserts had weak similarities to hypothetical proteins or to proteins with assigned functions from strain 11168. The remaining six clones had gene-fragment inserts with no database matches. Southern-blot analysis confirmed that strain-dependent variation existed for each of these DNA inserts. These results indicate that subtractive hybridization can successfully identify genes that are absent from the only C. jejuni strain for which the genome sequence is currently available.

Keywords: subtractive hybridization

Abbreviations: AP, alkaline phosphatase; R–M, restriction–modification

The GenBank accession numbers for the sequences reported in this paper can be found in Table 2.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Campylobacter jejuni is a major cause of human acute bacterial enteritis worldwide (Tauxe, 1992 ). The sources and routes of infection are debatable, but the consumption of contaminated poultry meat is considered to be a significant risk factor. Surveys have shown that broilers frequently carry large numbers of these organisms in their intestinal contents (Newell & Wagenaar, 2000 ). This carriage is both chronic and asymptomatic. The spillage of gut contents during processing can contaminate the retail poultry product and the abattoir environment (Newell et al., 2001 ). Currently, the reduction or elimination of campylobacters in the food chain is largely focussed at the poultry farm level. Unfortunately, enhanced biosecurity has largely failed to prevent campylobacter entering the broiler house from the environment. In part, this failure is a reflection of the low infectious dose, rapid bacterial growth within the avian gut and subsequent rapid transmission to adjacent animals. All of the evidence suggests that C. jejuni has evolved to efficiently colonize the avian gut as a commensal. Therefore, a better understanding of the colonization determinants involved in this process may ultimately lead to more targeted and effective control strategies, such as probiotic administration or vaccination.

To date, only a few potential colonization factors, specific for the avian host, have been identified using defined mutants in experimental oral models of chick colonization. A role for the flagellin genes in chicken colonization has been demonstrated (Wassenaar et al., 1993 ; Nachamkin et al., 1993 ); however, non-motile and aflagellate bacteria are still able to colonize (Wassenaar et al., 1993 ). Other genes apparently important in avian-gut colonization identified so far include sod (Purdy et al., 1999 ), racR (Bras et al., 1999 ), pglH (A. Karlyshev, unpublished data) and cadF (Ziprin et al., 1999 ).

Epidemiological studies on large numbers of C. jejuni isolates have shown a high degree of phenotypic and genotypic diversity (Owen et al., 1997 ; Lindstedt et al., 2000 ; Dingle et al., 2001 ; Wassenaar & Blaser, 1999 ). It seems likely that such diversity is reflected in phenotypic variation in the colonization and disease-causing potentials of strains. Certainly strain variation in virulence properties has been observed in both in vitro (Gorelov et al., 1990 ) and in vivo models, though such models may poorly reflect the spectrum of disease seen in humans (Pang et al., 1987 ; Black et al., 1988 ; Wassenaar & Blaser, 1999 ). Understanding the molecular basis and biological consequences of genetic diversity among strains is important for disease control. To this end the recent genome sequencing of C. jejuni NCTC 11168 (Parkhill et al., 2000 ) now enables, by DNA array technology (Lan & Reeves, 2000 ), the absence of genes in other C. jejuni strains to be detected. However, the identification of the presence of additional genes in other strains is not feasible using this approach.

C. jejuni NCTC 11168 was derived from a case of human campylobacteriosis in 1977 (M.B. Skirrow, personal communication). Preliminary studies indicated that this strain displayed a relatively poor colonization potential in chickens compared to that previously reported for the well-characterized strain 81116 (Cawthraw et al., 1996 ). The inability of chicken passage of strain 11168 to fully restore its colonization potential indicated that this phenotype had a genetic basis. Therefore, the method of subtractive hybridization was used to identify genetic material, potentially associated with colonization, present in C. jejuni strain 81116 but not present in the genome sequence of strain 11168.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains.
C. jejuni NCTC 11168 and strains 81116 and 81-176, were all derived from human cases of campylobacteriosis (Parkhill et al., 2000 ; Palmer et al., 1983 ; Korlath et al., 1985 , respectively) and are all routinely used in laboratory investigations. Other isolates included in this investigation were randomly selected from the collection of C. jejuni strains held at the Veterinary Laboratories Agency: strains 99/189, 99/322 and 99/388 were from the faeces of patients with diarrhoea; strains 94/194, 99/346 and 99/385 were from the caeca of broilers in unrelated poultry flocks; strain EX114 was an environmental strain from a poultry farm; and strain 99/312 was from the faeces of a cow.

Media and growth conditions.
C. jejuni strains were routinely grown for 24 h at 42 °C under microaerobic conditions (85% N2, 7·5% CO2, 7·5% O2) on blood agar plates supplemented with 10% sheep blood, 50 µg Actidione ml-1 and selective antibiotics (Oxoid) (Skirrow, 1977 ). Escherichia coli was maintained on Luria–Bertani (LB) agar or in Luria–Bertani broth (Difco) at 37 °C in atmospheric conditions. When required the media were supplemented with 100 µg ampicillin ml-1 or 50 µg kanamycin ml-1. C. jejuni and E. coli were stored as frozen cultures at -80 °C in 1% (w/v) proteose peptone water (Difco) containing 10% (v/v) glycerol or in Luria broth (LB) containing 50% glycerol (Ausubel et al., 1994 ), respectively.

Chicken-colonization model.
A quantitative oral chick-colonization assay was performed as previously described (Wassenaar et al., 1993 ). Briefly, groups of 10 specific pathogen-free chickens (SPAFAS, Charles River), housed in isolators, were dosed at day 1 of age by oral gavage. Doses ranged from 102 to 109 c.f.u. (100 µl 0·1 M PBS, pH 7·2)-1 of C. jejuni NCTC 11168, administered in 100 µl 0·1 M PBS (0·0234 M disodium hydrogen phosphate; 0·0066 M potassium dihydrogen orthophosphate; 0·07 M NaCl; pH 7·2). Doses were prepared by harvesting bacteria, grown overnight on blood agar plates at 42 °C, into sterile PBS. At 5 days post-challenge, the levels of colonization were determined by plating out dilutions of caecal contents. Chick-colonization levels were given as c.f.u. (g caecal contents)-1 for individual birds. The limit of detection of colonization was 100 c.f.u. (g caecal contents)-1. The dose response for the colonization potential of C. jejuni 81116 has been previously described (Wassenaar et al., 1993 ) and has been consistent in over 20 experiments.

Subtractive hybridization.
A schematic diagram of the subtractive-hybridization technique is shown in Fig. 1. The DNA sequences of adapters and primers used in this study are shown in Table 1. Initially, genomic DNA preparations from C. jejuni 81116 (tester) and strain 11168 (driver) were isolated using the CTAB method for DNA isolation (Ausubel et al., 1994 ). Each preparation was digested with AluI and DraI, according to the methods described in the PCR-Select Bacterial Subtraction kit (Clonetech), to give a range of blunt-ended fragments of between 0·1 and 2·0 kbp, as determined by gel electrophoresis. The blunt-ended DNA fragments from C. jejuni 81116 were subdivided into two aliquots and each aliquot was ligated to a different adapter (1 or 2R). This resulted in double-stranded DNA fragments with an extended, single-stranded adapter sequence on the 5' end of each complementary strand. Both aliquots were denatured and combined with an excess of denatured driver DNA and the mixture was incubated at 63 °C for 1·5 h (Fig. 1, hybridization one). In the second hybridization (Fig. 1, hybridization two) the two aliquots were combined and incubated at 63 °C overnight with more fresh denatured driver DNA. During this stage tester-specific DNA (ligated to adapter 1) annealed to tester-specific DNA (ligated to adapter 2R). The recessed ends of the adapters were filled and the primary PCR was carried out to exponentially amplify tester-specific DNA (type f molecules). In the secondary PCR, nested primers to adapters 1 and 2R were used to further enrich the tester-specific DNA (type f molecules). The first genomic subtraction was carried out with a hybridization temperature of 63 °C, as recommended by the manufacturer. A second round of subtractive hybridization was performed with 10 µl of the secondary PCR products, but this time a hybridization temperature of 61 °C was used.



View larger version (26K):
[in this window]
[in a new window]
 
Fig. 1. The strategy adopted for subtractive hybridization. Digested tester DNA (81116) is subdivided into two portions and the 5' ends are ligated to an adapter. Each aliquot is denatured and hybridized to excess denatured driver DNA (11168). In the second hybridization, the aliquots are combined and excess denatured driver is again added. Amplification with PCR primer 1 exponentially concentrates the type f molecules, which represent DNA present in the tester but which is absent from the driver. A second PCR with nested primers 1 and 2R further enriches tester-specific DNA.

 

View this table:
[in this window]
[in a new window]
 
Table 1. Summary of the bacterial strains, plasmids and primers used in subtractive hybridization, cloning and sequencing

 
Preparation of the subtractive-hybridization library.
The secondary PCR products were ligated into pCR2.1-TOPO vector using the TOPO TA Cloning kit (Invitrogen). The ligated products were transformed into TOP10F' competent E. coli cells, as recommended, and transformants were selected for kanamycin and ampicillin resistance. PCR amplification was performed directly on the transformants using 50 pmol µl-1 of M13 forward and reverse primers together with HotStarTaq master mix (Qiagen). PCR parameters were as follows: 95 °C for 15 min followed by 25 cycles of 94 °C for 45 s, 60 °C for 45 s and 72 °C for 1 min 45 s, with a final incubation at 72 °C for 10 min. Agarose-gel electrophoresis confirmed the presence of DNA inserts (data not shown).

Dot-blot and Southern-blot hybridizations.
For the dot blots, 50 ng of each PCR product was denatured by incubating in a boiling water bath for 5 min and spotted onto a Hybond-N+ membrane (Amersham). DNA was cross-linked to the membrane by UV light. For probe labelling, 200 ng of genomic DNA digested with AluI and DraI was labelled using alkaline phosphatase (AP) according to AlkPhosDirect labelling kit guidelines (Amersham). The blots were prehybridized in AlkPhos Direct Hybe buffer for 1 h at 60 °C and then hybridized overnight at 60 °C with 200 ng of the AP probe. For Southern-blot analysis, 10 µg of each C. jejuni genomic DNA was digested with either HindIII or HinfI, electrophoresed through a 1% (w/v) agarose gel and transferred onto a Hybond-N+ membrane. Each DNA insert (50 ng) of interest was labelled with AP; hybridizations and stringency washes were carried out at 60 °C as before.

Sequencing and analysis.
Strain-81116-specific clones were cultured in 1 ml LB broth with kanamycin selection, purified using a QIAprep spin miniprep kit (Qiagen) and sequenced by using the CEQ DTCS sequencing system (with 3·2 pmol M13 primer µl-1) on a CEQ 2000 DNA sequencer (Beckman Coulter). The DNA sequences were edited using the EDITSEQ program (DNAstar, Lasergene) and homology searches were performed using the BLAST suite of programs (Altschul et al., 1997 ).


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
C. jejuni 81116 colonizes the chick model in a dose-dependent manner, with a dose of 105 c.f.u. maximally colonizing 100% of the birds (Wassenaar et al., 1993 ). This differs from other freshly isolated strains tested, which display all-or-nothing-type colonization patterns and maximally colonize all the birds with doses of 102–103 c.f.u. (Cawthraw et al., 1996 ). However, this apparent lower colonization potential of strain 81116 can be overcome by a single passage through chicks, so that the passaged strain (81116P) maximally colonizes birds with a dose as low as 35 c.f.u. (Cawthraw et al., 1996 ). In comparison with strain 81116, strain 11168 was a poor colonizer (Fig. 2). A dose of 108 c.f.u. gave a detectable colonization level in only two out of 10 birds. A dose of 109 c.f.u. was required to colonize 100% of birds challenged, and even then colonization was at a mean level below the maximum level seen with freshly isolated campylobacter strains. Chick-passaged strain 11168 (11168P) did show some increased ability to colonize, but this was still lower than that of the unpassaged strain 81116 (Fig. 2).



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 2. Caecal colonization levels of chicks given a range of doses of C. jejuni strains 81116 ({square}), 11168 ({triangledown}) and chick-passaged 11168 (11168P) ({diamondsuit}). The broken line represents the detection limit of the experiment.

 
Genomic subtractive hybridization was applied to identify DNA sequences present in strain 81116 and absent in strain 11168 (Fig. 1). The initial round of subtractive hybridization at 63 °C indicated that subtraction had occurred at only a low efficiency, given that relatively few transformants were recovered and that many of the clones contained DNA inserts with sequences present in 11168. Consequently, the second round of subtraction was performed at a lower hybridization temperature (61 °C), which greatly enhanced the annealing of homologous driver and tester DNA, and the subsequent amplification of tester-specific DNA. Cloning of these PCR products into pCR2.1-TOPO resulted in 255 colonies, of which 167 contained inserts. Dot-blot hybridization was performed to test the specificity of the DNA fragments using the genomic DNA (labelled with AP) of strains 81116 and 11168 as probes (Fig. 3). Forty-two of the 167 inserts were found to strongly hybridize to strain 81116 and either weakly or not at all to strain 11168. DNA sequencing revealed that 24 of these 42 clones contained inserts that were absent in strain 11168 (Table 2), of which one DNA sequence was represented twice (inserts 126 and 136, only the latter has been included in the remaining analysis). The remaining 18 DNA inserts (43%) were false positives, i.e. analysis using the BLAST suite of programs revealed high sequence similarities in strain 11168 and also predicted similarities of 75–100% at the amino acid level when each DNA sequence was translated in all six reading frames. Most of these false-positive inserts were present in the cloning plasmid as chimeras (13 inserts), in which two different inserts had ligated into one vector joined by adapter sequences. Five inserts (46, 141, 149, 209 and 212) had weaker, but significant, similarities (40–61% identity at the amino acid level) with genes present in strain 11168. Twelve sequences had no similarity to any gene in the strain 11168 genome sequence, but they did have significant similarity to genes of other organisms. The remaining six sequences (inserts 30, 50, 65, 117, 121 and 243) had no significant similarity with any sequence present in the NCBI databases searched. The 23 strain-81116-unique DNA fragments (mean size of 272 bp; range 124–644 bp) were further analysed for the presence of potential surface structures and conserved motifs in protein families by PREDICT PROTEIN (http://dodo.cpmc.columbia.edu/predictprotein/submit_def.html). The prevalence of these unique sequences, among a diverse range of other C. jejuni strains, was determined by Southern-blot analysis (Table 3). The chromosomal DNA of each strain, digested with either HindIII or HinfI, was probed with each of the 23 strain-81116-specific DNA inserts. The results confirmed the presence of all 23 sequences in strain 81116 and their absence from strain 11168 DNA (Table 3). Only one of the 23 inserts (insert 121) was unique to strain 81116 (Fig. 4a). In contrast, only insert 50 hybridized to all other strains tested, with the exception of strain 11168 (Fig. 4b). The remaining 21 sequences gave variable distributions and banding patterns, of which two examples are shown (Fig. 4c, d, insert 246 and 135, respectively). The results are summarized in Table 3. Insert 46 is an example of a chimeric insert, in which two DNA fragments (excinuclease ABC subunit A and putative oxidoreductase) were ligated together (Fig. 5a). Removal of the excinuclease sequence and generation of a new AP-labelled probe, containing only the putative oxidoreductase, removed those bands that were common to all the strains (~2 kbp band for HindIII profile, Fig. 5b), leaving a 1·1 kbp band in strains 99/385, 81116, 99/388 and 99/346, and a 2·4 kbp band for strain EX114 (data not shown). For the other chimeric inserts, the individual strain-81116-specific sequences were also amplified by PCR from strain 81116 genomic DNA and then used to probe the Southern blots.



View larger version (95K):
[in this window]
[in a new window]
 
Fig. 3. Representative dot-blot hybridization of products from the subtractive hybridization. The subtractive-hybridization products were cloned and PCR, using the M13 forward and reverse primers, was used to amplify the inserts. Each PCR product was spotted onto a Hybond-N+ membrane and probed with (a) AP-labelled strain 81116 genomic DNA fragments and (b) AP-labelled strain 11168 genomic DNA fragments. PCR products that hybridized strongly to strain 81116 DNA and weakly or not at all with strain 11168 DNA were chosen for sequencing. An example of a strain-81116-specific insert has been circled.

 

View this table:
[in this window]
[in a new window]
 
Table 2. Summary of the sequence analysis of clone inserts specific for C. jejuni 81116 and absent from C. jejuniNCTC 11168

 

View this table:
[in this window]
[in a new window]
 
Table 3. Southern-blot analysis of each insert against randomly chosen C. jejuni strains

 


View larger version (57K):
[in this window]
[in a new window]
 
Fig. 4. The presence of DNA, from selected inserts, among a range of C. jejuni strains, as demonstrated by Southern-blot analysis. Two random isolates of strain 11168 were tested. Southern hybridization of HindIII-digested genomic DNA from each strain was carried out with AP-labelled subtractive-hybridization-DNA inserts as probes. (a) Insert 121 hybridized to strain 81116 only. (b) Insert 50 hybridized to all the strains tested except strain 11168. (c) Insert 246 hybridized to strains 99/189, 99/385, 81116, 99/388 and 99/346. (d) Insert 135 hybridized to strains 99/189, 99/322, 99/312, 99/385, 81116, 99/388 and 99/346.

 


View larger version (37K):
[in this window]
[in a new window]
 
Fig. 5. An example of a chimeric clone, insert 46. (a) The DNA sequence between 426 and 16 has an amino acid identity of 61% to a putative oxidoreductase in C. jejuni 11168; the sequence between 431 and 595 has 98% amino acid identity to excinuclease ABC subunit A from C. jejuni 11168. The amino acid consensus is shown in between the amino acid sequence for insert 46 and the putative oxidoreductase or excinuclease ABC subunit A; conservative substitutions are indicated by +. (b) Southern-blot hybridization using insert 46 as a probe. The oxidoreductase part of chimeric insert 46 was shown to hybridize to strains 99/385, 81116, 99/388 and 99/346 (1·1 kbp band) and to EX114 (2·4 kbp band) in a separate experiment (data not shown). The excinuclease part of insert 46 was shown to hybridize to all strains, including strain 11168, as shown by the band in all lanes at 2 kbp.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
All freshly isolated C. jejuni strains, i.e. those with minimal laboratory passage, tested to date maximally colonize [108–1010 c.f.u. (g caecal contents)-1] one-day-old chicks within 5 days with oral doses of 102–104 c.f.u. (Cawthraw et al., 1996 ). Although the laboratory strain 81116 has a comparatively low colonization potential, this can be upregulated by in vivo passage (81116P) to a level similar to that of freshly isolated strains. In contrast, strain 11168 appears to be a very poor colonizer, and even in vivo passage will not upregulate its colonization potential to that seen in freshly isolated strains. It therefore seems likely that the observed lower colonization potential of C. jejuni 11168, compared with strain 81116, is a reflection of genetic difference(s). In this study subtractive hybridization has been successfully applied to identify genetic material present in C. jejuni 81116 but absent in strain 11168. This approach was rendered easier because strain 11168 has recently had its genome sequenced.

The low percentage of genuine strain-81116-specific sequences, obtained from the initial subtractive-hybridization experiment, may have been due to the inadequate annealing of homologous DNA sequences from strain 11168 to 81116. Decreasing the temperature from 63 °C (optimal for bacteria with a GC ratio of 40–51 mol%) to 61 °C improved this efficiency by allowing the preferential amplification of 81116-specific sequences. Improved optimization might be achieved by decreasing the hybridization temperature further. The presence of a number of chimeric clones also confounded the overall efficiency, as these were detected as positive clones. Nonetheless, a total of 23 DNA sequences present in strain 81116 and absent from strain 11168 have been identified, confirming that intra-species variation exists between strains of C. jejuni (Dorrell et al., 2001 ). Of the 23 strain-81116-specific DNA inserts identified six sequences had no significant similarities to any entry in the databases searched. In several cases, two or more of the strain-81116-specific inserts appeared to have been derived from the same genomic location, as indicated by Southern blotting, suggesting that such strain-specific genes may be clustered; for example, the HinfI profiles of inserts 60, 243 and 246 were identical.

The appearance of identical banding profiles on Southern blots for the two C. jejuni strains 99/312 and 99/322, when probed with inserts 50, 68, 135 and 202, suggests that the genetic arrangement of these strains is similar. Indeed, PFGE profiles of these two strains are also identical, indicating that they are clonally related (A.M. Ridley, personal communication).

Although subtractive hybridization was developed to screen for the presence or absence of genes, it can also be used to detect polymorphisms between two strains. This was apparent for inserts 205 and 141. These inserts have similarities to dTDP-glucose 4,6-dehydratase (rmlB) from Legionella pneumophila, an enzyme involved in the glycosylation of the cell wall of pathogenic bacteria (Allard et al., 2000 ), and a putative glycosyltransferase from C. jejuni 11168, respectively. Both inserts had identical HinfI Southern-blot profiles and both were either present or absent in the C. jejuni strains tested. This suggests that the functions of these two genes, involved in polysaccharide biosynthesis, are coordinated. That inserts were identified that do not hybridize to strain 11168, despite significant amino acid similarity as in the case of insert 141, suggests that (at least some) polymorphic genes can be detected by subtractive hybridization. These results confirm DNA microarray observations indicating that strain 81116 differs from strain 11168 within this genetic region (Dorrell et al., 2001 ). Thus, subtractive hybridization not only complements DNA microarray technology but provides the additional benefit of detecting novel sequences in test strains.

It has been reported that sequences within the flagellar modification region and the capsule locus vary between strains. In particular, it has been reported that a region between Cj1318 and Cj1336 of the flagella modification locus has been lost from strain 81116 (Dorrell et al., 2001 ). However, subtractive hybridization identified insert 209 to have similarity to Cj1333. One explanation of this anomaly would be a genetic rearrangement in strain 81116, so that Cj1333 was no longer located where expected relative to strain 11168. However, the identification of a second insert (149) with similarity to Cj1337, and the identical Southern-blot profiles using inserts 149 and 209 as probes suggests that these two DNA fragments are in close proximity on the 81116 genome.

A number of DNA inserts were found to have similarities to restriction–modification (R–M) genes. Similarities to both type I and II R–M systems were identified in inserts 60, 68, 135, 136 (all type I from Haemophilus influenzae Rd, Salmonella enterica, S. enterica and Streptococcus thermophilus, respectively) and 202 (type II from Helicobacter pylori J99) (Table 2). Southern-blot profiles of inserts 68 and 135 were identical with both HindIII and HinfI, suggesting that they are closely located in all strains tested. Analysis of these two inserts revealed a substantial amount of overlapping DNA sequence, suggesting that they might be part of the same R–M gene. There is increasing evidence for the existence of multiple R–M systems in C. jejuni (Miller, 2001 ). The functions of those R–M genes found in strain 81116 are currently under investigation. Interestingly, the chromosomal DNA of this strain does not cut with SmaI (G. Manning, personal communication), even though SmaI sites have been shown to be present (I.H. Ahmed, unpublished data). It is therefore interesting to speculate that the R–M genes detected have a role in this enzyme restriction.

Two more inserts (31 and 186), which were absent from the strain 11168 genome, have similarities with genes involved in arsenic resistance in several bacteria. Insert 31 has 64% amino acid identity to ArsB from Sinorhizobium sp. As4, an integral membrane protein needed for the extrusion of arsenite. Insert 186 has 35% amino acid identity to ArsC from Aquifex aeolicus. This gene encodes arsenate reductase, which reduces intracellular arsenate to arsenite. The Southern-blot profiles of inserts 31 and 186 were different (data not shown), suggesting that these two genes do not form part of an operon. However, insert 186 had identical HindIII and HinfI profiles (data not shown) to those of insert 212, which has similarity to Cj1560 ( a putative membrane protein). This suggests that ArsB and Cj1560 are co-located in the genome of all strains tested. Generally, in Gram-negative bacteria the arsenic-efflux pump comprises ArsA, the ATPase subunit, ArsB, which anchors the former to the membrane, and ArsC in the cytosol (Cervantes et al., 1994 ). The observation of arsenite-metabolizing genes in strain 81116 and not in strain 11168 is interesting especially since phenylarsonic compounds have been used, in the past, in poultry feed for controlling caecal coccidiosis and to promote growth (Pergantis et al., 1997 ). The use of such compounds may have led to the contamination of some agricultural lands on which poultry litter has been used as manure (Gupta & Charles, 1999 ). The ars operon is widespread among the Enterobacteriaceae and arsenic has also been shown to induce the synthesis of heat-shock proteins, indicative of the stress response triggered due to exposure to this toxic chemical (Chang et al., 1989 ). Thus, the biological role of these genes in the colonization of chickens by campylobacters requires further investigation.

Insert 236 has 72% amino acid identity to the H. pylori {gamma}-glutamyl transpeptidase (ggt) gene. Southern-blot analysis of insert 236 revealed an identical HinfI, but a different HindIII pattern, to that of insert 65, for which no significant protein similarities could be identified. This suggests that these inserts are located in close proximity in the genome of strain 81116 and that they may be linked. As well as functions in the antioxidant glutathione pathway (Carmel-Harel & Storz, 2000 ), GGT appears to have a role in the colonization of the gastric mucosa by H. pylori (Chevalier et al., 1999 ; McGovern et al., 2001 ). Whether this gene product has a similar role in colonization of the chicken intestinal tract by C. jejuni will require further investigation.

The successful application of subtractive hybridization, to genetically compare two strains of C. jejuni with differing colonization potentials, has revealed a number of strain-specific differences, some of which may be potential colonization factors. It should, however, be noted that the sequences identified in this study are only fragments of much larger genes and that expression of the corresponding genes has not yet been proven. Moreover, future work involving the mutagenesis of the 81116-specific sequences in tandem with further chick-colonization experiments will be necessary to determine the importance of such genes in colonization. Insert 50 is particularly interesting since it was universally present in the other C. jejuni strains tested (apart from strain 11168) and for one strain (94/194) was the only insert hybridizing. As this strain colonizes the chick model to the same extent as freshly isolated strains (K.J. Clow, personal communication), insert 50 may be especially important in chicken colonization.

In conclusion, the 23 novel DNA segments identified in this study represent a selection of genes present in strain 81116 and absent in strain 11168, and they indicate extensive genetic variation even among a small number of C. jejuni strains. These results imply that we have, to date, found only a portion of the total gene pool existing within the C. jejuni species. While C. jejuni genome sequence data are limited techniques like subtractive hybridization can be used to identify novel genes, which can then be used to construct more comprehensive DNA arrays to accurately determine genomic diversity.


   ACKNOWLEDGEMENTS
 
This work was funded by the Department for Environment, Food and Rural Affairs (DEFRA, formerly MAFF), UK.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Allard, S. T., Giraud, M. F., Whitfield, C., Messner, P. & Naismith, J. H. (2000). The purification, crystallization and structural elucidation of dTDP-D-glucose 4,6-dehydratase (RmlB), the second enzyme of the dTDP-L-rhamnose synthesis pathway from Salmonella enterica serovar typhimurium. Acta Crystallogr D Biol Crystallogr 56, 222-225.[Medline]

Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D. J. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25, 3389-3402.[Abstract/Free Full Text]

Ausubel, S. F., Brent, R., Kingston. R. E. and 7 other authors (1994). Current Protocols in Molecular Biology. New York, NY: Wiley.

Black, R. E., Levine, M. M., Clements, M. L., Hughes, T. P. & Blaser, M. J. (1988). Experimental Campylobacter jejuni infection in humans. J Infect Dis 157, 472-479.[Medline]

Bras, A. M., Chatterjee, S., Wren, B. W., Newell, D. G. & Ketley, J. M. (1999). A novel Campylobacter jejuni two-component regulatory system important for temperature-dependent growth and colonization. J Bacteriol 181, 3298-3302.[Abstract/Free Full Text]

Carmel-Harel, O. & Storz, G. (2000). Roles of the glutathione- and thioredoxin-dependent reduction systems in the Escherichia coli and Saccharomyces cerevisiae responses to oxidative stress. Annu Rev Microbiol 54, 439-461.[Medline]

Cawthraw, S. A., Wassenaar, T. M., Ayling, R. & Newell, D. G. (1996). Increased colonization potential of Campylobacter jejuni strain 81116 after passage through chickens and its implication on the rate of transmission within flocks. Epidemiol Infect 117, 213-215.[Medline]

Cervantes, C., Ji, G., Ramirez, J. L. & Silver, S. (1994). Resistance to arsenic compounds in microorganisms. FEMS Microbiol Rev 15, 355-367.[Medline]

Chang, E. C., Kosman, D. J. & Willsky, G. R. (1989). Arsenic oxide-induced thermotolerance in Saccharomyces cerevisiae. J Bacteriol 171, 6349-6352.[Medline]

Chevalier, C., Thiberge, J. M., Ferrero, R. L. & Labigne, A. (1999). Essential role of Helicobacter pylori gamma-glutamyltranspeptidase for the colonisation of the gastric mucosa of mice. Mol Microbiol 31, 1359-1372.[Medline]

Dingle, K. E., Colles, F. M., Wareing, D. R. A. and 7 other authors (2001). A multi locus sequence typing scheme for Campylobacter jejuni. J Clin Microbiol 39, 14–23.[Abstract/Free Full Text]

Dorrell, N., Mangan, J. A., Laing, K. G. and 9 other authors (2001). Whole genome comparison of C. jejuni human isolates using a low-cost microarray reveals extensive genetic diversity. Genome Res 11, 1706–1715.[Abstract/Free Full Text]

Gorelov, A. V., Gorskaia, E. M., Zhukhovitskii, V. G. & Bondarenko, V. M. (1990). The adhesion of clinical isolates of Campylobacter jejuni to intestinal epithelial cells in vitro. Zh Mikrobiol Epidemiol Immunobiol (in Russian) pt 7, 3–6.

Gupta, G. & Charles, S. (1999). Trace elements in soils fertilized with poultry litter. Poult Sci 78, 1695-1698.[Medline]

Korlath, J. A., Osterholm, M. T., Judy, L. A., Forfang, J. C. & Robinson, R. A. (1985). A point-source outbreak of campylobacteriosis associated with consumption of raw milk. J Infect Dis 152, 592-596.[Medline]

Lan, R. & Reeves, P. R. (2000). Intraspecies variation in bacterial genomes: the need for a species genome concept. Trends Microbiol 8, 396-401.[Medline]

Lindstedt, B. A., Heir, E., Vardund, T., Melby, K. K. & Kapperud, G. (2000). Comparative fingerprinting analysis of Campylobacter jejuni subsp. jejuni strains by amplified-fragment length polymorphism genotyping. J Clin Microbiol 38, 3379-3387.[Abstract/Free Full Text]

McGovern, K. J., Blanchard, T. G., Gutierrez, J. A., Czinn, S. J., Krakowka, S. & Youngman, P. (2001). Gamma-glutamyltransferase is a Helicobacter pylori virulence factor but is not essential for colonization. Infect Immun 69, 4168-4173.[Abstract/Free Full Text]

Miller, W. G. (2001). Characterization of multiple Campylobacter jejuni type I restriction–modification loci. In 11th International Workshop on Campylobacter, Helicobacter and Related Organisms, pp. 79. Edited by J. Hacker and others. Freiburg: Urban & Fischer.

Nachamkin, I., Yang, X. H. & Stern, N. J. (1993). Role of Campylobacter jejuni flagella as colonization factors for three-day-old chicks: analysis with flagellar mutants. Appl Environ Microbiol 59, 1269-1273.[Abstract]

Newell, D. G. & Wagenaar, J. A. (2000). Poultry infections and their control at the farm level. In Campylobacter , pp. 497-509. Edited by I. Nachamkin & M. J. Blaser. Washington, DC:American Society for Microbiology.

Newell, D. G., Shreeve, J. E., Toszeghy, M., Domingue, G., Bull, S., Humphrey, T. & Mead, G. (2001). Changes in the carriage of Campylobacter strains by poultry carcasses during processing in abattoirs. Appl Environ Microbiol 67, 2636-2640.[Abstract/Free Full Text]

Owen, R. J., Slater, E., Telford, D., Donovan, T. & Barnham, M. (1997). Subtypes of Campylobacter jejuni from sporadic cases of diarrhoeal disease at different locations in England are highly diverse. Eur J Epidemiol 13, 837-840.[Medline]

Palmer, S. R., Gully, P. R., White, J. M., Pearson, A. D., Suckling, W. G., Jones, D. M., Rawes, J. C. & Penner, J. L. (1983). Water-borne outbreak of Campylobacter gastroenteritis. Lancet 1, 287-290.[Medline]

Pang, T., Wong, P. Y., Puthucheary, S. D., Sihotang, K. & Chang, W. K. (1987). In vitro and in vivo studies of a cytotoxin from Campylobacter jejuni. J Med Microbiol 23, 193-198.[Abstract]

Parkhill, J., Wren, B. W., Mungall, K. and 18 other authors (2000). The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403, 665–668.[Medline]

Pergantis, S. A., Heithmar, E. M. & Hinners, T. A. (1997). Speciation of arsenic animal feed additives by microbore high-performance liquid chromatography with inductively coupled plasma mass spectrometry. Analyst 122, 1063-1068.[Medline]

Purdy, D., Cawthraw, S., Dickinson, J. H., Newell, D. G. & Park, S. F. (1999). Generation of a superoxide dismutase (SOD)-deficient mutant of Campylobacter coli: evidence for the significance of SOD in Campylobacter survival and colonization. Appl Environ Microbiol 65, 2540-2546.[Abstract/Free Full Text]

Skirrow, M. B. (1977). Campylobacter enteritis: a new disease. Br Med J 2, 9-11.[Medline]

Tauxe, R. E. (1992). Epidemiology of Campylobacter jejuni infections in the United States and other industrialized nations. In Campylobacter jejuni: Current Status and Future Trends , pp. 9-19. Edited by I. Nachamkin, M. J. Blaser & L. S. Tompkins. Washington, DC:American Society for Microbiology.

Wassenaar, T. M. & Blaser, M. J. (1999). Pathophysiology of Campylobacter jejuni infections of humans. Microbes Infect 1, 1023-1033.[Medline]

Wassenaar, T. M., van der Zeijst, B. A., Ayling, R. & Newell, D. G. (1993). Colonization of chicks by motility mutants of Campylobacter jejuni demonstrates the importance of flagellin A expression. J Gen Microbiol 139, 1171-1175.[Medline]

Ziprin, R. L., Young, C. R., Stanker, L. H., Hume, M. E. & Konkel, M. E. (1999). The absence of cecal colonization of chicks by a mutant of Campylobacter jejuni not expressing bacterial fibronectin-binding protein. Avian Dis 43, 586-589.[Medline]

Received 31 August 2001; revised 5 November 2001; accepted 22 November 2001.