jejuni and C. coli isolates was 23.9% (188/176 samples) while the prevalence of erythromycin (currently recommended for the treatment of laboratory-confirmed

campylobacteriosis) resistance in C. coli was 13.3% (28/210 samples). These levels of resistance are likely to represent an unacceptable frequency of therapeutic failure of the drugs indicated for the treatment of human campylobacteriosis. The high levels of antimicrobial resistance cannot be accounted for exclusively by high numbers of a particular group of resistant genotypes. Rather, there is evidence for widespread acquisition of resistance among LY294002 cost relatively distantly related lineages from retail poultry. This is consistent with a small-scale study of C. jejuni isolated from chicken meat in Senegal where quinolone resistant phenotypes were present in three out of four tested lineages, and also dispersed throughout singleton STs [24]. It is possible that mutations that confer antimicrobial resistance have occurred in multiple lineages. However, bacteria can acquire genetic material, including antimicrobial resistance genes, from relatively distantly related lineages through horizontal gene Daporinad mouse transfer [25, 26]. Horizontal Gene Transfer (HGT) can involve recombination

between lineages, or acquisition of plasmids, which has been demonstrated to be the main mechanism of tetracycline resistance in Campylobacter[27]. There is also evidence that plasmid acquisition may mediate resistance to chloramphenicol and aminoglycosides [28, 29]. Resistance to macrolides is conferred by a 2 bp change in the putative erythromycin binding site. Resistance to fluoroquinolones is most usually the result of a single mutation in the gyrA region [30]. The widespread antimicrobial resistance in the Campylobacter populations, is likely to be the result of horizontal gene transfer as well as multiple independent mutation events. When conditions are such that antimicrobial resistance confers a strong selective advantage,

lineages that trace ancestry to resistant isolates will increase as a proportion of the population [31]. Under these circumstances a phylogenetic tree will show clusters of resistant lineages that have expanded clonally. Consistent with this, statistical analyses of the ClonalFrame tree Ketotifen of retail poultry isolates indicated that resistant phenotypes were not randomly distributed but showed some clustering within lineages. At the highest level there was a species-specific association with erythromycin resistance correlated with C. coli, as in previous studies of Campylobacter in pigs, turkeys and chickens [32–35]. Resistance to tetracycline, quinolones and chloramphenicol showed no association with either Campylobacter species, but all were non-randomly distributed among C. jejuni lineages. This could indicate that antimicrobial resistance, having arisen in an ancestral lineage, is propagated clonally by vertical transmission.