AEM
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wittwer, M.
Right arrow Articles by Bissig-Choisat, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wittwer, M.
Right arrow Articles by Bissig-Choisat, B.
Agricola
Right arrow Articles by Wittwer, M.
Right arrow Articles by Bissig-Choisat, B.

 Previous Article  |  Next Article 

Applied and Environmental Microbiology, June 2005, p. 2840-2847, Vol. 71, No. 6
0099-2240/05/$08.00+0     doi:10.1128/AEM.71.6.2840-2847.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.

Genetic Diversity and Antibiotic Resistance Patterns in a Campylobacter Population Isolated from Poultry Farms in Switzerland

M. Wittwer,1,{dagger} J. Keller,1 T. M. Wassenaar,3* R. Stephan,4 D. Howald,1 G. Regula,1 and B. Bissig-Choisat1

Swiss Federal Veterinary Office, Bern, Switzerland,1 Molecular Microbiology and Genomics Consultants, Zotzenheim, Germany,3 Institute for Food Safety and Hygiene, University of Zurich, Zurich, Switzerland4

Received 1 June 2004/ Accepted 16 December 2004


   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
The diversity and genetic interrelation of Campylobacter jejuni and C. coli isolated from Swiss poultry were assessed by three independent typing methods. Samples were derived prior to slaughter from 100 randomly selected flocks (five birds per flock) raised on three different farm types. The observed flock prevalence was 54% in total, with 50% for conventional and 69% for free-range farms. Birds held on farms with a confined roaming area had the lowest prevalence of 37%. Campylobacter isolates were characterized by amplified fragment length polymorphism (AFLP), restriction fragment length polymorphism of flaA PCR fragments (flaA-RFLP), and disk diffusion testing for eight antimicrobial agents that are commonly used in veterinary or human medicine in Switzerland. Analysis of the genotypic results indicates that the Campylobacter population in Swiss poultry is genetically highly diverse. Nevertheless, occasionally, isolates with identical or nearly identical characteristics were isolated from different farms or farm types in different locations. Genetic typing by AFLP and flaA-RFLP was found to be complementary. The majority of isolates (67%) were susceptible to all tested antibiotics; however, single, double, and triple resistances were observed in 7%, 23%, and 2% of the strains, respectively. There was no correlation between genotype and antibiotic resistance. Surprisingly, sulfonamide resistance was frequently found together with streptomycin resistance. Our findings illustrate the results of common genetic exchange in the studied bacterial population.


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
The importance of Campylobacter jejuni as a human pathogen is generally accepted. This gram-negative, obligate microaerophilic, thermophilic bacterial species is ubiquitous and possesses the ability to colonize the intestinal mucosa of most warm-blooded animals. Birds are the preferred host for C. jejuni, where colonization occurs asymptomatically. Human campylobacteriosis is the most common food-borne bacterial disease in many developed and developing countries (7), and the consumption or handling of poultry meat is a major risk factor contributing to approximately half of the infections (12). The reduction of the prevalence of Campylobacter in chicken products is, therefore, an effective way to reduce the risk of infection. To achieve this goal, the factors influencing Campylobacter prevalence at different points during its propagation from "stable to table" need to be clearly identified and their relevance needs to be evaluated using risk assessment models (13, 25).

The various DNA-based typing methods which have been developed for Campylobacter (32) are effective tools for identifying contamination sources and describing the population dynamics of any given microorganism. None of these methods is superior to describe the "true" evolutionary and epidemiological interrelations of a bacterial population. Studies addressing the pros and cons of bacterial typing schemes suggest that a combination of at least two methods is desirable in order to obtain significant discrimination and reliable strain identification (4, 32). This recommendation applies specifically to any bacterial species with a high diversity and a known weakly clonal population structure, such as C. jejuni.

The aim of our study was to describe the genetic interrelations of the Campylobacter population isolated from Swiss broilers. Taking the weak clonality of C. jejuni into account, we used three typing methods: amplified fragment length polymorphism (AFLP), based on the whole genome of a given organism; restriction fragment length polymorphism of flaA PCR fragments (flaA-RFLP), assessing polymorphisms in a single gene of C. jejuni/C. coli; and the phenotypic determination of antibiotic (AB) resistance. The latter phenotype can result from point mutations or the presence of (plasmids harboring) resistance genes. AFLP is considered to give relatively stable genotypes in C. jejuni, as it is insensitive to genome rearrangement and point mutation (5). RFLP analysis of flaA is valuable in short-term epidemiology of Campylobacter (26). Antibiotic resistance is of increasing concern and may be rapidly shared between populations with different genotypes. As C. jejuni/C. coli is naturally capable of taking up DNA, this applies to both plasmid-borne and chromosomally encoded resistance. The grouping based on cluster analysis carried out for each typing scheme was related to parameters identified as being relevant for risk assessment at the farm level. We focused on three factors: the housing system for broiler production, broiler flock size, and geographic location of the farm. The strains under investigation were derived from a study to evaluate the antimicrobial resistance situation of Campylobacter (as a model organism) in Switzerland, in which a representative sample group was collected from a population covering 90% of Swiss broiler production.


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Sampling.
In the year 2002, 100 chicken flocks were randomly selected from companies which together cover 90% of commercial Swiss poultry production. Three housing systems were represented in our samples. Of the 100 farms tested, 20 had a conventional indoor broiler production system (animals held at ground level), 35 had an indoor production unit with a roofed outdoor roaming area with bedded concrete floors (hereafter called confined roaming area type), and 45 produced their broilers exclusively in a free-range outdoor system (free-range type). Birds in the free-range-type farms could come in contact with wild animals, as the outdoor facilities (grass soil) are fenced with open-wire fences. The size of the flocks tested ranged from 1,000 to 16,000 animals: 36 farms were in the range of 1,000 to 4,000 birds, 57 farms were in the range of 4,000 to 8,000 birds, and 7 farms reared more than 8,000 birds per flock. Flock size was independent of farm type.

Since Campylobacter prevalence is seasonally dependent, samples were collected between March and June and between October and November, when a medium prevalence is expected. Five cloacal swabs were taken from live birds from each flock on delivery to the abattoir.

Once a sample proved positive, one single colony of Campylobacter spp. was taken for further characterization. The presence of multiple strains per bird was not assessed.

Laboratory testing.
Cloacal swabs (BIOSWAB Biolife) were transported in Cary Blair agar prior to bacteriological testing. The swabs were inoculated into 10 ml of selective enrichment broth (Brucella bouillon [product no. 0495-17-3; Difco]) from Becton Dickinson, with Campylobacter growth supplement (product no. SR84; Oxoid) and Skirrow Campylobacter-selective supplement (product no. SR69; Oxoid), and incubated at 42°C for 24 h under microaerobic conditions (5% O2, 10% CO2, and 85% N2) provided by commercial gas packs (BBL product no. 271045; Becton Dickinson). The enrichment samples were then streaked onto selective agar medium (Brucella agar [product no. 0964-17-5; Difco]) with 6% horse blood (product no. SR48; Oxoid) and Butzler Campylobacter-selective supplement (product no. SR85; Oxoid) and incubated at 42°C for 36 h under microaerobic conditions. Suspect colonies were identified as Campylobacter spp. by typical morphology, gram-negative stain, catalase and oxidase positivity, characteristic motion, and intrinsic resistance to cephalotin.

Twofold identification of C. coli and C. jejuni was carried out using hippurate hydrolysis and PCR (22). Using PCR, C. jejuni was identified by the detection of the gene coding for the species-specific membrane protein MapA, according to a previously published protocol (28). A PCR specific for C. coli was based on random primers (27) which had been shown to be specific for the species in a comparative study (22). Hippurate hydrolysis was used on multiple Campylobacter colonies, transferred to tubes containing hippurate medium, and incubated at 37°C for 2 h. Following incubation, 0.2 ml of ninhydrin (30 g/liter) was added to the tubes. Samples showing a deep blue coloring after 5 to 10 min were considered to be C. jejuni. It was crucial to evaluate the coloring within 25 min of adding ninhydrin in order to prevent false-positive results. Hippurate-negative, nalidixic acid-sensitive, and indoxylacetate hydrolysis-positive colonies were considered to be C. coli.

DNA extraction.
Campylobacter isolates were incubated on nonselective 5% (vol/vol) sheep blood agar for 48 h at 37°C under microaerobic conditions. Cells were harvested and resuspended in 1 ml distilled water. Extraction of total DNA was performed with the Puregene DNA purification kit for genomic DNA according to the manufacturer's protocol for gram-negative bacteria. Visual inspection and quantification of DNA were performed using a 1.5% agarose gel.

Genotyping.
The sampled strains were characterized by two different genotyping approaches. AFLP was performed according to a method described previously by Duim et al. (5) in which bands are generated from multiple locations on the genome. For flaA-RFLP (following the CAMPYNET protocol described previously [10]), we used primers described previously (32) in order to determine polymorphisms in a highly variable, single gene.

Antibiotic susceptibility typing.
At the time of this study, there was no internationally standardized procedure available for susceptibility testing of Campylobacter. A good correlation has been reported between disk diffusion and agar dilution methods for testing the susceptibility of Campylobacter spp. to erythromycin, ciprofloxacin, and tetracycline (15), and comparison of two agar dilution methods and three agar diffusion methods showed that reliable results can be achieved with disk diffusion (8). For practical reasons, we chose the disk diffusion method. The following antibiotic-impregnated disks (bioMérieux SA, France) were used: erythromycin (15 µg), ciprofloxacin (5 µg), and tetracycline (30 µg) (considered relevant in human medicine) as well as gentamicin (10 µg), streptomycin (10 µg), ampicillin (10 µg), amoxicillin (25 µg), and sulfonamide (20 µg) (all registered for use in veterinary medicine and permitted in commercial poultry production). Three to five isolated colonies of the same morphological type were selected from each agar plate culture and transferred into trypticase soy broth (catalog no. 211768; Becton Dickinson). After incubation at 42°C for 24 h under microaerobic conditions, a sterile cotton swab was dipped into the suspension and streaked onto the entire surface of a Mueller-Hinton agar plate (catalog no. CM 337; Oxoid) with 5% sheep blood. Four antibiotic disks were placed onto each plate, and the diameter of the inhibition zone was measured with callipers after 48 h of microaerobic incubation at 42°C. Escherichia coli ATCC 25922 and Staphylococcus aureus ATCC 25923 were used as reference strains. Zones of growth inhibition were evaluated in accordance with standards published by the National Committee for Clinical Laboratory Standards which were available at the time of analysis (17), although we were aware of the limitations (9). Standards specified for C. jejuni subsequently became available (18). For binary analysis with 0 as sensitive and 1 as resistant, the following cutoff values were used: erythromycin, ≤13 mm; ciprofloxacin, ≤15 mm; tetracycline, ≤14 mm; streptomycin, ≤11 mm; ampicillin, ≤13 mm, gentamicin, ≤12 mm; amoxicillin, ≤13 mm; and sulfonamide, ≤12 mm. The reproducibility of the disk diffusion method was assessed by testing five Campylobacter isolates from each sample. Zone diameters for all antibiotics tested were found to vary ±2 mm or less.

Data analysis.
The raw AFLP profiles collected with Genescan (PE Applied Biosystem) were imported into BioNumerics, version 3.0 (Applied Maths, Kortrijk, Belgium), and subsequently transformed into banding patterns. flaA-RFLP gels were photographed with a Gel Doc 2000 system (Bio-Rad), and images in TIF format were imported into BioNumerics, version 3.0. After pattern normalization, the similarity matrix was calculated with the Pearson product-moment correlation algorithm. The unweighted-pair group with mathematical average method was used for dendrogram construction.


   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Of the 500 fecal swabs collected from birds from 100 farms, 139 (29% of samples) originating from 54 farms tested positive for Campylobacter spp. Of these, 24 (17%) were identified as C. coli. The prevalence of infection at the flock level was 54%. Fifty percent of conventional housing farms, 37% of confined roaming-area-type farms, and 69% of free-range-type farms tested positive for Campylobacter. Single-colony isolated strains were subjected to AFLP genotyping. The calculated similarity values obtained from all 139 samples covered a range from 18% to 98.98%, whereas at the 90% similarity level considered epidemiologically significant for microorganisms (20), 48 different AFLP patterns could be identified. In spite of the heterogeneity of the banding patterns, three major clusters could be identified at a within-cluster similarity level of >65% for cluster 1, 58% for the major subset of cluster 2, and 42% for cluster 3 (Fig. 1). The similarity values ranged from 38% between clusters 2 and 3 to as low as 14% between cluster 1 and the other two clusters. Based on hippurate hydrolase activity and species-specific PCR, the strains grouped in cluster 1 were identified as C. coli. At a similarity level of >97%, isolates were generally grouped according to the farm they were derived from. A farm designated farm 31' is highlighted in Fig. 1 as an example. Nevertheless, in some cases (8 of 24 clusters), isolates from two independent farms were joined together within a group with a >97% similarity level, as indicated with asterisks in Fig. 1.



View larger version (16K):
[in this window]
[in a new window]
  		FIG.1. Dendrogram of AFLP banding patterns obtained with all 139 Campylobacter sp. isolates obtained in this study. For a description of clusters 1, 2, and 3, see the text. The dotted vertical line indicates the 90% similarity threshold. Hashed horizontal lines separate the isolates from farm 31. Isolates indicated with an asterisk display >97% similarity but were isolated from different farms.

 
Parallel to the AFLP analysis, the 139 Campylobacter isolates were genotyped by flaA-RFLP. One of the 139 strains was not typeable with this method since no DNA was amplified. The remaining 138 showed the expected fragment size (1,719 bp) after amplification with the flaA primers, and these yielded 29 different restriction patterns comprising 2 to 7 distinct bands. The number of isolates sharing an identical pattern type varied from 1 to 36. In comparison to the 48 banding patterns obtained with AFLP, flaA-RFLP has a lower discrimination capacity. Isolates with AFLP banding patterns with >97% similarity usually (in 22 of 24 cases) shared identical flaA-RFLP banding patterns.

Antibiotic susceptibility testing was carried out for ampicillin, amoxicillin, ciprofloxacin, erythromycin, tetracycline, gentamicin, streptomycin, and sulfonamide. All isolates were susceptible to gentamicin and erythromycin (Fig. 2). Ninety-three isolates were susceptible to all antibiotics tested. Of the remaining 46 isolates, 10 were resistant to one antibiotic, 33 were resistant to two antibiotics, and 3 were resistant to three antibiotics.



View larger version (32K):
[in this window]
[in a new window]
  		FIG. 2. Numbers of strains representing each antibiotic resistance type. Shaded boxes below the columns indicate resistance.

 
To avoid a bias on data analysis due to multiple sampling, each isolate originating from fecal swabs of animals of the same flock was assessed for identity based on its AFLP, flaA-RFLP, and antibiotic resistance profile. After excluding redundant samples with identical characteristics, 78 of 138 isolates were used for further analysis. Although this set of isolates comprises unique combinations of the three characteristics tested, shared genotypes or AB phenotypes exist within this sample set.

Analysis of 78 strains with unique characteristics.
Based on cluster analysis algorithms, each of the three typing methods used provides a dendrogram which describes the genetic interrelation of the tested strains. The correlation of the similarity matrixes calculated for each experiment is a measure of the congruence of the obtained dendrograms. Correlation coefficients calculated for the similarity matrices of each pair of experiments were as follows: R2 = 0.21 for AFLP versus flaA-RFLP, R2 = 0.16 for AFLP versus AB typing, and R2 = 0.04 for flaA-RFLP versus AB typing, suggesting that there is a better correlation between the two genotyping methods AFLP and flaA-RFLP than between either the genotyping or AB typing method.

At a cutoff of 90%, clustering of the 78-strain subset yielded, as above, 48 different AFLP types and 29 different fla types. Twenty-nine of the 48 AFLP types were represented by a single isolate. The 29 different flaA-RFLP types comprise 12 types with a single isolate, 9 types with 2 isolates, 3 types with 3 isolates, 2 types with 4 isolates, 2 types with 7 isolates, and 1 type with 17 isolates.

Figure 3 shows the relationship between AFLP and flaA-RFLP genotypes. No statements concerning the coherence of the typing methods can be made for the 29 AFLP types and the 12 flaA-RFLP types which are represented by one single isolate. However, 10 of the remaining 19 AFLP genotypes encompassing 23 (47%) samples showed a 100% concordance between the AFLP and flaA-RFLP genotypes (these isolates have different AB susceptibilities and are thus included in the sample set). Some AFLP types are more stable with regard to their flaA-RFLP type than others. All members of AFLP type 18 and 23 were associated with flaA-RFLP type 4. A similar but weaker correlation was seen for AFLP types 14 (four out of five strains have flaA-RFLP type 4) and 24 (three out of four have flaA-RFLP type 17). In contrast to a shared AFLP and flaA genotype, these strains had variable AB profiles and did not share isolation characteristics such as geographic origin, housing, retailer, or flock size. This may indicate that these isolates represent stable clones within the diverse C. jejuni population. Other AFLP types were more variable in their flaA-RFLP type. For example, the three strains grouped in AFLP type 26 and the four strains grouped in AFLP type 6 were further subdivided into three different flaA-RFLP types each.



View larger version (51K):
[in this window]
[in a new window]
  		FIG. 3. AFLP/flaA-RFLP relationship. Two dendrograms with collapsed branches (horizontal axis, flaA-RFLP type; vertical axis, AFLP type) show the interrelation of the strains based on each typing method. The shades of gray indicate the frequency of a given AFLP/flaA-RFLP type.

 
In AB typing, two strains were assigned to identical groups when there was a 100% match in their binary (0 = sensitive; 1 = resistant; for cutoff values, see Materials and Methods) antimicrobial susceptibility pattern. Under this presupposition, the 78 isolates were separated into 13 antibiotic resistance types. Twenty-seven (35%) out of 78 isolates showed resistance to at least one of the six substances, and 20 (26%) of the samples showed concomitant resistance to two or three of the compounds tested. In 8 (30%) out of the 27 resistant isolates, we found a combined resistance to streptomycin and sulfonamide (Fig. 2). Six of these eight strains (75%) were identified as C. coli.

With respect to the AFLP/flaA-RFLP types, which were represented more than once in the subset, there was no correlation between genotype and AB resistance patterns. For instance, of three strains sharing an AFLP/flaA-RFLP genotype, one was sensitive to all of the tested compounds, one was resistant to ciprofloxacin, and one showed a concomitant resistance to ciprofloxacin and sulfonamide. Similar observations were made for isolates sharing other combined genotypes. This suggests, as expected, that the AB phenotype is more variable within a bacterial population than the combined AFLP/flaA-RFLP genotype.

Strain diversity and prevalence at the farm level.
The combined application of three typing methods allowed us (using the complete data set of 139 isolates) to investigate strain diversity at the farm level, although it should be noted that only five birds per flock and one colony per bird were analyzed. Nevertheless, in 19 (35%) of the 54 Campylobacter-positive farms, we identified two or three different AFLP/flaA-RFLP/AB types. The setup of this study did not allow us to determine whether multiple strains could be present in individual birds. Analyzing these 19 farms according to the housing system resulted in 4 (20%) of the conventional farms, 5 (14%) of the confined roaming-area-type farm systems, and 10 (22%) of the free-range producers showing multiple infection. All of the four farms harboring three different AFLP/flaA-RFLP/AB types had a free-range broiler production system.

In 8 (42%) of 19 farms showing multiple infection, none of the AFLP/flaA-RFLP/AB types matched. The degree of accordance for the three variables within farms was determined for the remaining 11 farms. When AB-sensitive isolates (found in six farms) were ignored completely, strains were identical in their resistance patterns but not identical in AFLP genotype (farms F24 and F87 in Table 1) at only two farms. Various combinations of accordance between the parameters were observed. In one farm (F97), three isolates formed pairs sharing two out of three (AFLP, flaA-RFLP, and AB type) identical characteristics, but none of the isolates was identical. In one case (farm F24), the AFLP and AB types matched, but in other farms (F18 and F97), identical AB types were observed with different AFLP types. Not one example was found of complete identity of all parameters tested from isolates derived from one farm. This indicates a high level of diversity between isolates present per farm. At one farm (F24), C. coli was exclusively detected. The confined roaming-area-type farms were overrepresented in the set of farms with multiple isolates, but not enough samples were available to test the significance of this observation.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Strain diversity at the farm level and accordance between AFLP, flaA-PCR, and AB types

 
Assessment of determining factors for cluster formation.
For every cloacal sample, the parameters flock size, housing system, and geographic location of the farm were collected, and the typing results were broken down for these parameters. Contingency tables summarizing the frequency of the parameters of all members of a given cluster were formed based on the dendrogram calculated for the similarity of AFLP patterns. The contingency tables were calculated for the within-cluster similarity values of 65%, 80%, and 90%. At a similarity value above 90%, the only parameter correlating with the AFLP-based grouping of the samples was geographic location of the farm. Lower similarity values did not result in correlation of parameters.

The same analysis was performed based on the dendrogram calculated for flaA-RFLP restriction patterns and AB typing. Using this genotypic characteristic, none of the parameters correlated with the flaA-RFLP/AB-derived groups.


   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
We found a 54% flock prevalence for Campylobacter in Switzerland, which is comparable to that found in France (24), Germany (1), and Denmark (33). Our findings indicate no significant difference between production systems when conventional housing, available confined outdoor roaming areas, and fenced-only free-range farms were compared. This finding contrasts with data obtained in Denmark, where a flock prevalence of 100% was observed in organic systems compared to 36.7% in conventional housing (14). As reviewed previously (19), the comparison of prevalence is influenced by differences in sampling and isolation methods. A possible explanation for our finding of relatively high prevalence in conventional housing may be that conventional production is the most established housing type in Switzerland, with facilities and infrastructure of the farms probably older and therefore more prone to hygiene security loopholes (6). This may counteract the beneficial influence that a confined environment has on Campylobacter prevalence.

The diversity of the C. jejuni and C. coli populations found in Swiss broilers was investigated by means of AFLP and flaA-RFLP. Our findings are comparable with those of similar studies carried out in other countries; for instance, the reported within-cluster similarity levels of 37% for C. jejuni and 70% for C. coli strains are in agreement with work reported previously by others (11, 21). The ability of AFLP to cluster clonal samples at the farm level, together with the capability to separate C. jejuni from C. coli strains, emphasizes that AFLP is a suitable method to reflect the taxonomic and, at least at farm level, geographic interrelations in a given population. However, in spite of the apparently higher discriminatory capacity of AFLP compared to that of flaA-RFLP, 8 (40%) of the 20 AFLP types that included more than one isolate could be further subdivided by flaA-RFLP typing, resulting in 59 different combined AFLP/flaA-RFLP types. Analogous to flaA-RFLP, AB typing refines the picture of strain diversity by further separating the 59 AFLP/flaA-RFLP types into 65 AFLP/flaA-RFLP/AB subtypes.

The high number of AFLP and, to a lesser extent, flaA-RFLP types together with the observed variety in combinations of both typing schemes are illustrative of the high diversity seen within the C. jejuni population. This high diversity is possibly a consequence of its genomic instability. Since the variability in the C. jejuni genome is not limited to genes involved in the adaptation to environmental conditions and other selective pressures (for which we take flaA-RFLP and AB typing to be illustrative), the selective mechanisms driving the genomic rearrangement are still unknown. Lack of DNA repair and SOS response genes have been mentioned as the cellular mechanism resulting in variation (23). Some genes or loci are probably more prone to instability than others. The flagellin gene locus may be highly sensitive to recombination and rearrangement due to the repetitive nature of the flagellin genes and due to the exposure of the Campylobacter flagella to the intestinal environment of the host organism, which may result in selection of antigenic variants. Both DNA recombination between the tandem flaA and flaB genes within the same genome and (partial) flaAB gene exchange between different strains (or even species) by means of horizontal gene transfer have been described previously (11, 30). Although it is known that at least a part of the C. jejuni population is naturally capable of transformation in vitro (29), the impact of horizontal gene transfer on the genetic heterogeneity under naturally occurring conditions is still being debated (3, 16). Our study delivers observations that could point to horizontal gene transfer events. We found 3 out of 19 farms that were infected with more than one AFLP/flaA-RFLP/AB type, where individual isolates shared particular genotypic characteristics (either AFLP banding pattern or flaA genotype) or phenotypic AB resistance patterns. In farm B45, for example, AFLP type 42 was represented in two isolates in combination with either flaA-RFLP type 13 or 16. Another strain from this farm also had flaA-RFLP type 13 but in combination with AFLP type 38. Furthermore, we observed a correlation only between AFLP-based grouping of samples and the parameter geographic location of the farm. That AFLP patterns do not correlate well with flaA genotype or AB resistance patterns in the studied population suggests that typing based on flaA or AB sensitivity determines factors which may be lost more quickly with increasing time-distance than the factors determining AFLP type.

These observations point out two important characteristics of Campylobacter colonization in broilers. On the one hand, mixed infection at the flock level is probably not uncommon (note that our study was not set up to specifically detect this or to observe mixed infections in single birds). On the other hand, genetic exchange may be the underlying mechanism for particular genetic or phenotypic characteristics being shared between isolates occurring in mixed infections.

The interpretation of AB resistance data against the highly variable genetic background of C. jejuni is troublesome. We are aware of the limitations of our study, where the phenotypic sensitivity to antibiotics is addressed without knowledge of the possible molecular mechanisms involved to achieve resistance to a given antimicrobial substance. Work is ongoing to determine the genetic basis leading to a given resistance for each strain. Furthermore, data on antibiotic use during fattening or in previous flocks on a particular farm were not available. In spite of these limitations, our AB resistance data outline the high adaptive potential of C. jejuni/C. coli against antimicrobial pressure.

The high incidence of strains doubly resistant to sulfonamide and streptomycin raises the question whether these resistances are genetically linked. We were unable to detect any of the genes known (2, 34) to underlie the resistance to each of the two compounds using PCR analysis (data not shown). Further experiments are required to identify the mechanisms causing the double resistance to streptomycin and sulfonamide. These will reveal a possible linkage of the resistances, for instance, by colocalization of yet-unknown resistance genes on a mobile DNA element or involvement of an efflux pump or porin that affects the intracellular concentration of both compounds.

A low correlation of the dendrograms calculated for the three typing methods was observed, as is typical for a weakly clonal population. Each typing approach provides a different picture of the genetic interrelations within the investigated Campylobacter population. Available evidence suggests (5, 16) that AFLP patterns can be relatively stable over a prolonged time span. Supporting this observation, we found that the AFLP pattern of some C. jejuni strains isolated from human diarrhea patients in 1993 matched >95% of strains from 2002 presented in this study (unpublished data). In contrast, flaA-RFLP typing and AB typing rely solely on genes which play a crucial role in the adaptation of the bacteria to its (hostile) environment. Therefore, these typing results reflect rather recent events in the evolution of the population that result from host adaptation or antimicrobial treatment schemes. With this in mind, the question of which typing scheme reflects the "true" evolutionary pathway of an organism cannot be answered, since each method highlights a different aspect of the development of the weakly clonal population.

In conclusion, our observations on genetic diversity of campylobacters isolated from Swiss broilers illustrate the complexity of this population structure. The findings point out the limitations of typing to interpret offspring lineages from randomly selected populations, even within a narrow time span, when dealing with a weakly clonal species.


   ACKNOWLEDGMENTS
 
This work was funded by the Swiss Veterinary Office and by National Research Project 49, entitled "Antibiotic Resistance," of the Swiss National Foundation.


   FOOTNOTES
 
* Corresponding author. Mailing address: Molecular Microbiology and Genomics Consultants, Tannenstrasse 7, D-55576 Zotzenheim, Germany. Phone: 49 6701 8531. Fax: 49 6701 901803. E-mail: mmgc.de{at}t-online.de. Back

{dagger} Present address: ZLB Behring AG, Bern, Switzerland. Back


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Atanassova, V., and C. Ring. 1999. Prevalence of Campylobacter spp. in poultry and poultry meat in Germany. fsInt. J. Food Microbiol. 51:187-190.
  2. Baumgartner, A., J. Nicolet, and R. Braun. 1983. Plasmid fingerprints of E. coli 0149:91:K 88 ac pathogenic to swine: characterization of a plasmid for streptomycin and sulfonamide resistance. Zentralbl. Veterinarmed. B 30:288-296. (In German.)[Medline]
  3. de Boer, P., J. A. Wagenaar, R. P. Achterberg, J. P. Putten, L. M. Schouls, and B. Duim. 2002. Generation of Campylobacter jejuni genetic diversity in vivo. Mol. Microbiol. 44:351-359.[CrossRef][Medline]
  4. de Boer, P., B. Duim, A. Rigter, J. van Der Plas, W. F. Jacobs-Reitsma, and J. A. Wagenaar. 2000. Computer-assisted analysis and epidemiological value of genotyping methods for Campylobacter jejuni and Campylobacter coli. J. Clin. Microbiol. 38:1940-1946.[Abstract/Free Full Text]
  5. Duim, B., T. M. Wassenaar, A. Rigter, and J. A. Wagenaar. 1999. High-resolution genotyping of Campylobacter strains isolated from poultry and humans with AFLP fingerprinting. Appl. Environ. Microbiol. 65:2369-2375.[Abstract/Free Full Text]
  6. Evans, S. J., and A. R. Sayers. 2000. A longitudinal study of campylobacter infection of broiler flocks in Great Britain. Prev. Vet. Med. 46:209-223.[CrossRef][Medline]
  7. Friedman, C. R., J. Neimann, H. C. Wegener, and R. V. Tauxe. 2000. Epidemiology of Campylobacter jejuni infections in the United States and other industrialized nations. p. 121-138. In I. Nachamkin and M. J. Blaser (eds.), Campylobacter, 2nd ed. American Society for Microbiology, Washington, D.C.
  8. Gaudreau, C., and H. Gilbert. 1997. Comparison of disc diffusion and agar dilution methods for antibiotic susceptibility testing of Campylobacter jejuni subsp. jejuni and Campylobacter coli. J. Antimicrob. Chemother. 39:707-712.[Abstract/Free Full Text]
  9. Hakanen, A., P. Huovinen, P. Kotilainen, A. Siitonen, and H. Jousimies-Somer. 2002. Quality control strains used in susceptibility testing of Campylobacter spp. J. Clin. Microbiol. 40:2705-2706.[Free Full Text]
  10. Harrington, C. S., L. Moran, A. M. Ridley, D. G. Newell, and R. H. Madden. 2003. Inter-laboratory evaluation of three flagellin PCR/RFLP methods for typing Campylobacter jejuni and C. coli: the CAMPYNET experience. J. Appl. Microbiol. 95:1321-1333.[CrossRef][Medline]
  11. Harrington, C. S., F. M. Thomson-Carter, and P. E. Carter. 1997. Evidence for recombination in the flagellin locus of Campylobacter jejuni: implications for the flagellin gene typing scheme. J. Clin. Microbiol. 35:2386-2392.[Abstract]
  12. Harris, N. V., N. S. Weiss, and C. M. Nolan. 1986. The role of poultry and meats in the etiology of Campylobacter jejuni/coli enteritis. Am. J. Public Health 76:407-411.[Abstract/Free Full Text]
  13. Hartnett, E., L. Kelly, D. Newell, M. Wooldridge, and G. Gettinby. 2001. A quantitative risk assessment for the occurrence of Campylobacter in chickens at the point of slaughter. Epidemiol. Infect. 127:195-206.[Medline]
  14. Heuer, O. E., K. Pedersen, J. S. Andersen, and M. Madsen. 2001. Prevalence and antimicrobial susceptibility of thermophilic Campylobacter in organic and conventional broiler flocks. Lett. Appl. Microbiol. 33:269-274.[CrossRef][Medline]
  15. Huang, M. B., C. N. Baker, S. Banerjee, and F. C. Tenover. 1992. Accuracy of the E test for determining antimicrobial susceptibilities of staphylococci, enterococci, Campylobacter jejuni, and gram-negative bacteria resistant to antimicrobial agents. J. Clin. Microbiol. 30:3243-3248.[Abstract/Free Full Text]
  16. Manning, G., B. Duim, T. Wassenaar, J. A. Wagenaar, A. Ridley, and D. G. Newell. 2001. Evidence for a genetically stable strain of Campylobacter jejuni. Appl. Environ. Microbiol. 67:1185-1189.[Abstract/Free Full Text]
  17. National Committee for Clinical Laboratory Standards. Performance standards for antimicrobial disc susceptibility tests. Approved standards, 8th edition. NCCLS, Wayne, Pa.
  18. National Committee for Clinical Laboratory Standards. 2004. Performance standards for antimicrobial disk and dilution susceptibility tests for bacteria isolated from animals. Informational supplement M31-S1. NCCLS, Wayne, Pa.
  19. Newell, D. G., and C. Fearnley. 2003. Sources of Campylobacter colonization in broiler chickens. Appl. Environ. Microbiol. 69:4343-4351.[Free Full Text]
  20. Olive, D. M., and P. Bean. 1999. Principles and applications of methods for DNA-based typing of microbial organisms. J. Clin. Microbiol. 37:1661-1669.[Free Full Text]
  21. On, S. L., and C. S. Harrington. 2000. Identification of taxonomic and epidemiological relationships among Campylobacter species by numerical analysis of AFLP profiles. FEMS Microbiol. Lett. 193:161-169.[CrossRef][Medline]
  22. On, S. L., and P. J. Jordan. 2003. Evaluation of 11 PCR assays for species-level identification of Campylobacter jejuni and Campylobacter coli. J. Clin. Microbiol. 41:330-336.[Abstract/Free Full Text]
  23. 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.[CrossRef][Medline]
  24. Refregier-Petton, J., N. Rose, M. Denis, and G. Salvat. 2001. Risk factors for Campylobacter spp. contamination in French broiler-chicken flocks at the end of the rearing period. Prev. Vet. Med. 50:89-100.[CrossRef][Medline]
  25. Rosenquist, H., N. L. Nielsen, H. M. Sommer, B. Norrung, and B. B. Christensen. 2003. Quantitative risk assessment of human campylobacteriosis associated with thermophilic Campylobacter species in chickens. Int. J. Food Microbiol. 83:87-103.[CrossRef][Medline]
  26. Shreeve, J. E., M. Toszeghy, A. Ridley, and D. G. Newell. 2002. The carry-over of Campylobacter isolates between sequential poultry flocks. Avian Dis. 46:378-385.[CrossRef][Medline]
  27. Stonnet, V., L. Sicinschi, F. Mégraud, and J.-L. Guesdon. 1995. Rapid detection of Campylobacter jejuni and Campylobacter coli isolated from clinical specimens using the polymerase chain reaction. Eur. J. Clin. Microbiol. Infect. Dis. 14:355-359.[CrossRef][Medline]
  28. Stucki, U., J. Frey, J. Nicolet, and A. P. Burnens. 1995. Identification of Campylobacter jejuni on the basis of a species-specific gene that encodes a membrane protein. J. Clin. Microbiol. 33:855-859.[Abstract]
  29. Wang, Y., and D. E. Taylor. 1990. Natural transformation in Campylobacter species. J. Bacteriol. 172:949-955.[Abstract/Free Full Text]
  30. Wassenaar, T. M., B. N. Fry, and B. A. van der Zeijst. 1995. Variation of the flagellin gene locus of Campylobacter jejuni by recombination and horizontal gene transfer. Microbiology 141:95-101.[Abstract]
  31. Wassenaar, T. M., B. Geilhausen, and D. G. Newell. 1998. Evidence of genomic instability in Campylobacter jejuni isolated from poultry. Appl. Environ. Microbiol. 64:1816-1821.[Abstract/Free Full Text]
  32. Wassenaar, T. M., and D. G. Newell. 2000. Genotyping of Campylobacter spp. Appl. Environ. Microbiol. 66:1-9.[Free Full Text]
  33. Wedderkopp, A., K. O. Gradel, J. C. Jorgensen, and M. Madsen. 2001. Pre-harvest surveillance of Campylobacter and Salmonella in Danish broiler flocks: a 2-year study. Int. J. Food Microbiol. 68:53-59.[CrossRef][Medline]
  34. Willson, P. J., H. G. Deneer, A. Potter, and W. Albritton. 1989. Characterization of a streptomycin-sulfonamide resistance plasmid from Actinobacillus pleuropneumoniae. Antimicrob. Agents Chemother. 33:235-238.[Abstract/Free Full Text]

Applied and Environmental Microbiology, June 2005, p. 2840-2847, Vol. 71, No. 6
0099-2240/05/$08.00+0     doi:10.1128/AEM.71.6.2840-2847.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:


This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Wittwer, M.
Right arrow Articles by Bissig-Choisat, B.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wittwer, M.
Right arrow Articles by Bissig-Choisat, B.
Agricola
Right arrow Articles by Wittwer, M.
Right arrow Articles by Bissig-Choisat, B.

Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
J. Bacteriol. Microbiol. Mol. Biol. Rev. Eukaryot. Cell All ASM Journals