Genomic Signatures of Campylobacter Adaptation and Source Attribution Studies

Campylobacter jejuni and Campylobacter coli are the biggest causes of bacterial gastroenteritis in the developed world, with human infections typically arising from zoonotic transmission associated with infected meat. Because Campylobacter is not thought to survive well outside the gut, host-associated populations are genetically isolated to varying degrees. Therefore, the likely origin of most strains can be determined by host-associated variation in the genome. This is instructive for characterizing the source of human infection. However, some common strains, notably isolates belonging to the ST-21, ST-45, and ST-828 clonal complexes, appear to have broad host ranges, hindering source attribution. Whole-genome sequencing has the potential to reveal fine-scale genetic structure associated with host specificity, but this is dependent upon the extent to which lineages are true generalists. Switching hosts too regularly may prevent the development of host-specific genomic signatures in allopatry. Promising approaches, pioneered in human genetics, include genome-wide association studies where DNA sequence variation across the genome is related to particular phenotypes. This has been challenging to apply to bacteria because of their strong population structure resulting from clonal reproduction. Recent work, including Sheppard et al. (PNAS, 2013), presents new methods that identify homologous and non-homologous sequence variation in Campylobacter. This revealed some of the genetic changes associated with adaptations in birds and mammals that could be used as markers of host association. However, we have found that rates of zoonotic transmission among animal host species in generalist clonal complexes are so high that the signal of host association is all but obliterated; estimating one zoonotic transmission event every 1.6, 1.8 and 12 years in the ST-21, ST-45 and ST828 complexes, respectively. Therefore, the weak signal of host association within these complexes presents a challenge for pinpointing the source of clinical infections and underlines the view that whole-genome sequencing, powerful though it is, cannot substitute for intensive sampling of suspected transmission reservoirs.