Genetic characterisation of Campylobacter concisus: Strategies for improved genomospecies discrimination

Although at least two genetically distinct groups, or genomospecies, have been well documented for Campylobacter concisus, no phenotype has yet been identified for their differentiation and thus formal description as separate species. C. concisus has been isolated from a variety of sites in the human body, including saliva and stool samples from both healthy and diarrhoeic individuals. We evaluated the ability of a range of whole genome-based tools to distinguish between the two C. concisus genomospecies (GS) using a collection of 190 C. concisus genomes. Nine genomes from related Campylobacter species were included in some analyses to provide context. Analyses incorporating sequence analysis of multiple ribosomal genes generated similar levels of C. concisus GS discrimination as genome-wide comparisons. The C. concisus genomes formed two groups; GS1 represented by ATCC 33237T and GS2 by CCUG 19995. The two C. concisus GS were separated from the nine genomes of related species. GS1 and GS2 also differed in G+C content with medians of 37.56% and 39.51%, respectively. The groups are consistent with previously established GS and are supported by DNA reassociation results. Average Nucleotide Identity using MUMmer (ANIm) and Genome BLAST Distance Phylogeny generated in silico DNA-DNA hybridisation (isDDH) (against ATCC 33237T and CCUG 19995), plus G+C content provides cluster-independent GS discrimination suitable for routine use. Pan-genomic analysis identified genes specific to GS1 and GS2. WGS data and genomic species identification methods support the existence of two GS within C. concisus. These data provide genome-level metrics for strain identification to genomospecies level.