Introduction Campylobacter concisus is an oral commensal, which has historically been linked to gingivitis and periodontitis but more recently has been linked with gastrointestinal diseases including inflammatory bowel disease and gastroenteritis. Functional and genomic data for C. concisus is limited with little evidence to define potential differences, in terms of pathogenicity, between oral and intestinal isolates. The aim of the current study was to generate robust genome sequence data for a number of clinical C. concisus strains and compare pathogenic potential between oral, faecal and mucosal biopsy derived isolates.

Method Genomes from 53 C . concisus strains were sequenced, assembled and annotated including thirty-four strains from gastroenteritis patients, fourteen strains from Crohn’s disease patients, 4 strains from colitis patients (3 collagenous colitis and 1 lymphocytic colitis), and one from a colonic adenoma patient. Six patients had both oral and faecal isolates available. Library construction was performed using the Illumina Nextera XT DNA library prep kit and sequenced using Illumina MiSeq. Selected strains were also subjected to long-read sequencing (PacBio). Genome assemblies were annotated with Prokka and quality assessment tools including QUAST, BUSCO and REAPR were used.

Results Genome assemblies compared very favourably to existing publicly available C. concisus genome assemblies (10 available from previous studies). Isolates clustered into two main groups/genomospecies (GS) based on comparison of the 16S rRNA and 23S rRNA gene sequences. Clinical characterisation or location of the strains (oral vs intestinal) did not stratify according to GS typing. Of the 6 sets of paired faecal-oral strains, derived from Crohn’s patients most paired samples were separated based on GS typing. Interestingly the oral strains did not form a GS and did not strongly cluster together relatively to the faecal and biopsy strains in the study. PCR investigation for the virulence factor exotoxin 9 indicated that positive strains were more likely to belong to genomospecies II. Comparison of exotoxin 9 positivity between patient paired samples indicated that oral strains were PCR positive whilst faecal strains were predominantly negative for the virulence factor exotoxin 9.

Conclusion Our findings indicate that C. concisus strains are phenotypically and genetically diverse, and suggest that there may well be differences in virulence potential depending on where strains are located.

Disclosure of Interest None Declared