Characterization of Camplyobacter jejuni and C. coli from Swedish patients and chickens - antibiotic resistance, genomic diversity and detection

Abstract: Campylobacter jejuni is a Gram-negative, non-spore forming bacterium, which is currently the most common pathogen causing reported cases of food poisoning in Sweden, as well as in most other industrialized countries. Eating or drinking something containing these bacteria is often determined to be the source of infection and as few as 500 bacteria can be enough to cause illness. The cases are often sporadic and hard to trace. Chicken is a well-documented source for Campylobacter infections in humans, and the birds do not get sick by colonization of these bacteria. The focus of the thesis was on the characterization of C. jejuni and C. coli from Swedish patients and chickens and to evaluate antibiotic resistance, genomic diversity, and to facilitate detection in food. The frequency of antibiotic resistance in C. jejuni/C. coli, in patients with campylobacteriosis, from 1996-2005, to the most commonly used antibiotics for treatment of Campylobacter infection in Sweden was studied. The antibiotic resistance among isolates from Swedish chickens at slaughter was also evaluated. Furthermore, the relatedness between strains originating from chickens and humans in Sweden was determined, and a comparison was made with strains originating from other parts of the world. Finally, a method was developed to find C. jejuni on chicken carcasses faster and more specifically. Resistance to norfloxacin increased (7 to 30 %), among the strains tested from patients acquiring the infection in Sweden. Among domestic isolates, the resistance to tetracycline varied between 4 % and 13 % during the decade analysed. The antibiotic resistance to erythromycin was consistently low among the domestic isolates tested (up to 4 %). It was concluded that antibiotic resistance among C. jejuni/coli strains is relatively uncommon in humans infected in Sweden (numbers in % earlier mentioned), as well as in Swedish chicken (only 3 %) for the most commonly used antibiotics for human treatment. Erythromycin was concluded to remain as the current drug of choice for the treatment of most patients infected abroad with C. jejuni/coli, since resistance to this drug was uncommon among all tested isolates. It was further concluded that a large increase in antibiotic resistance has developed for the quinolones (norfloxacin and ciprofloxacin) among domestic human C. jejuni/coli strains. A map of the antibiotic resistance over the time period of this study showed an ongoing emergence of resistance worldwide to norfloxacin for C. jejuni/coli. When PCR/REA was used, 100 % of the human Campylobacter strains and 98 % of the chicken isolates were identified as C. jejuni. Five percent of the human domestic isolates and 10 % of the chicken isolates were falsely negative by hippurate, using the PCR/REA assay as a reference. C. jejuni was observed to be a diverse taxon, based on PFGE profiling and cluster analysis. Over 50 % of the clones from Swedish patients and healthy Swedish chickens were not similar with respect to PFGE groups, which suggests that human infections can occur from sources other than chickens. Only 33 % of domestic isolates from human infection and Swedish chicken strains belonged to the same PFGE groups, indicating that other sources than chicken can be of importance for human campylobacteriosis. More isolates are needed to prove this indication. Similarities were observed between Swedish and Thai strains, suggesting global contamination pathways. In the quantification study, the number of bacteria on contaminated chicken meat was concluded to vary and that easily performed quantification methods, real-time quantitative PCR (q-PCR) and plate counting on selective media, with a highly standardized procedure, has the potential to sort out highly-contaminated meat (for direct heat treatment) from less- or non-contaminated meat. The results from the q-PCR was somewhat more stable than results from plate counting, indicating that small differences in the shape of the bacteria can influence the results when working with living bacteria grown on plates.