Comparative genomics of Campylobacter concisus isolates reveals genetic diversity and provides insights into disease association

Abstract

Background: In spite of its association with gastroenteritis and inflammatory bowel diseases, the isolation of Campylobacter concisus from both diseased and healthy individuals has led to controversy regarding its role as an intestinal pathogen. One proposed reason for this is the presence of high genetic diversity among the genomes of C. concisus strains.

Results: In this study the genomes of six C. concisus strains were sequenced, assembled and annotated including two strains isolated from Crohn's disease patients (UNSW2 and UNSW3), three from gastroenteritis patients (UNSW1, UNSWCS and ATCC 51562) and one from a healthy individual (ATCC 51561). The genomes of C. concisus BAA-1457 and UNSWCD, available from NCBI, were included in subsequent comparative genomic analyses. The Pan and Core genomes for the sequenced C. concisus strains consisted of 3254 and 1556 protein coding genes, respectively.

Conclusion: Genes were identified with specific conservation in C. concisus strains grouped by phenotypes such as invasiveness, adherence, motility and diseased states. Phylogenetic trees based on ribosomal RNA sequences and concatenated host-related pathways for the eight C. concisus strains were generated using the neighbor-joining method, of which the 16S rRNA gene and peptidoglycan biosynthesis grouped the C. concisus strains according to their pathogenic phenotypes. Furthermore, 25 non-synonymous amino acid changes with 14 affecting functional domains, were identified within proteins of conserved host-related pathways, which had possible associations with the pathogenic potential of C. concisus strains. Finally, the genomes of the eight C. concisus strains were compared to the nine available genomes of the well-established pathogen Campylobacter jejuni, which identified several important differences in the respiration pathways of these two species. Our findings indicate that C. concisus strains are genetically diverse, and suggest the genomes of this bacterium contain respiration pathways and modifications in the peptidoglycan layer that may play an important role in its virulence.

Figure 1

Synteny based similarities and differences across Campylobacter concisus strains. A: Pairwise comparisons between the reference BAA-1457 genome from NCBI and the C. concisus strains sequenced by our group revealed similarities and differences in syntenical placement of genomic blocks. The Mauve alignment tool was used to show paired synteny alignments, which indicate some genome shuffling of C. concisus strains. Inversions are indicated by syntenic blocks placed bellow the main axis. B: Circular layout of seven assembled genomes against the BAA-1457 reference. The CGView tool used with BAA-1457 as the reference (outer 2 rings) highlights that the reference strain contains regions with high diversity (marked by prominent gaps) when compared to the other seven C. concisus strains.

Figure 2

Enrichment of Gene Ontologies for (A) CORE and (B) PAN genes across Campylobacter concisus genomes.

Figure 3

Phylogenetic analysis of (A) 16S and (B) 23S ribosomal RNA genes for the eight Campylobacter concisus strains.

Figure 4

Syntenic placement of specific conserved plasmid genes in the genomes of invasive strains. All genes were placed on a single scaffold in each species.

Figure 5

Analyses of the peptidoglycan biosynthesis pathway within the eight C. concisus strains. A: Phylogenetic analysis of the concatenated sequences of the peptidoglycan biosynthesis pathway within the eight C. concisus strains. B: Genes in the peptidoglycan biosynthesis pathway with SNPs are highlighted in red.

Figure 6

Comparative genomic analyses of Campylobacter concisus and Campylobacter jejuni. A: The core genomes of C. concisus and C. jejuni were identified. B: Enrichment of genes specific to C. concisus core genome. C: Enrichment of genes specific to C. jejuni core genome. Enrichment tests of all Gene Ontology terms were analysed using Fisher’s Exact Test with Multiple Testing Correction of FDR (Benjamini and Hochberg).