Population Structure and Antimicrobial Resistance Profiles of Streptococcus suis Serotype 2 Sequence Type 25 Strains

Abstract

Strains of serotype 2 Streptococcus suis are responsible for swine and human infections. Different serotype 2 genetic backgrounds have been defined using multilocus sequence typing (MLST). However, little is known about the genetic diversity within each MLST sequence type (ST). Here, we used whole-genome sequencing to test the hypothesis that S. suis serotype 2 strains of the ST25 lineage are genetically heterogeneous. We evaluated 51 serotype 2 ST25 S. suis strains isolated from diseased pigs and humans in Canada, the United States of America, and Thailand. Whole-genome sequencing revealed numerous large-scale rearrangements in the ST25 genome, compared to the genomes of ST1 and ST28 S. suis strains, which result, among other changes, in disruption of a pilus island locus. We report that recombination and lateral gene transfer contribute to ST25 genetic diversity. Phylogenetic analysis identified two main and distinct Thai and North American clades grouping most strains investigated. These clades also possessed distinct patterns of antimicrobial resistance genes, which correlated with acquisition of different integrative and conjugative elements (ICEs). Some of these ICEs were found to be integrated at a recombination hot spot, previously identified as the site of integration of the 89K pathogenicity island in serotype 2 ST7 S. suis strains. Our results highlight the limitations of MLST for phylogenetic analysis of S. suis, and the importance of lateral gene transfer and recombination as drivers of diversity in this swine pathogen and zoonotic agent.

Fig 1

A) Genome atlas of Canadian S. suis ST25 strain NSUI060. Depicted data from innermost to outer most circles represent genome size in Mbp (circle 1); percent G+C content (circle 2); GC skew (G—C)/(G + C) averaged over a moving window of 10,000 bp, with excess G and excess C shown in green and purple, respectively (circle 3); annotated coding sequences (CDSs) encoded on the forward/direct (circle 4, red), and reverse/complementary (circle 5, blue) chromosomal strands; TBLASTN comparisons of the CDSs predicted in ST25 strain NSUI060 and ST1 strain P1/7 (circle 6; percent identity defined in the body of the figure); TBLASTN comparisons of the CDSs predicted in ST25 strains NSUI060 and 89/1591 (circle 7; percent identity defined in the body of the figure); TBLASTN comparisons of the CDSs predicted in ST25 strain NSUI060 and ST28 strain NSUI002 (circle 8; percent identity defined in the body of the figure); reference genome landmarks (circle 9): ribosomal RNAs, capsule and pilus related genes, and mobile genetic elements are labelled in black, genes used in the S. suis MLST scheme are labelled in orange; genes encoding resistance to antimicrobial agents are labelled in red. MGE: Mobile genetic element. B) Venn diagram depicting unique and shared gene clusters in S. suis strains as identified by ortholog analysis. The ST28 strain NSUI002 is represented by the blue circle, the ST25 strain NSUI060 is represented by the red circle, and the ST1 strain P1/7 is represented by the green circle. Numbers in the intersectional regions indicate gene clusters shared by strains. Colinearity of the genomes of S. suis ST25 strain NSUI060 and C) ST28 strain NSUI002 or D) ST1 strain P1/7. The genomes of the strains were aligned using progressiveMauve. Sequence alignments that are free of rearrangements are shown as colored local collinear blocks (LCBs). Sequence inversions are denoted by differential positioning of LCBs relative to a reference axis. Pilus related genes, rRNAs, genes used in MLST typing are labelled and other landmarks are indicated.

Fig 2. Sliding window analysis of single-nucleotide polymorphism (SNP) distribution, recombination analysis and phylogenetic relationships among S. suis ST25 strains.

A) In a group of 23 North American strains (defined as North American variant 1, NA V1 Strains) SNPs were distributed across the genome of the reference strain. Sliding windows were created using R-software and custom scripts and a window of 5 kbp. B) In a group of 19 North American strains (NA V2 Strains), SNPS distribution was non-random, and an overabundance of polymorphisms was identified between positions 1 to 1.035 Mbp of the NSUI060 genome. C) In all 8 Thai strains, SNP distribution was non-random, with two distinct peaks (positions ~740 kbp to ~760 kbp and ~790 kbp to ~860 kbp of the reference genome) showing overabundance of SNPs. D) SNP distribution in a single Canadian strain (NSUI069) also identified overabundance of SNPs in discrete areas of the reference genome. E) The left panel depicts results of Bayesian analysis of recombination for 51 ST25 S. suis strains. The colored bars denote recombination events identified for each strain relative to the core ST25 genome. The coloring of the bars at a specific genomic location reflects the clustering of the recombination events into groups, and is unrelated to other bars at distant genomic locations. The right panel shows a Neighbor-joining phylogenetic tree constructed using 3,023 non redundant SNP loci identified among all ST25 S. suis strains relative to the core ST25 genome devoid of areas having undergone recombination. The different clades are indicated by the different colors used to label strain names, circles represent strains that were isolated from a swine host, and triangles represent strains that were isolated from a human host. Variants defined by sliding window analysis are also indicated.

Fig 3. Presence of ICEs and antimicrobial resistance encoding genes in the 51 ST25 S. suis genomes.

The left panel shows inferred genetic relationships between strains, determined as described in Fig 2. The right panel depicts host, country of origin, presence or absence of ICEs and/or resistance genes in the genomes of the strains under investigation. Strains isolated from pigs are depicted by a circle, while strains isolated from humans are depicted by a triangle. Strains isolated in Canada are shown as a white circle or triangle, strains isolated in the United States of America are shown in grey, and strains isolated in Thailand are shown in black. ICEs are depicted in light blue, tetracycline resistance genes are depicted in yellow, erythromycin resistance genes are depicted in light green, all other antimicrobial resistance genes are depicted in pink.

Fig 4. Resistance against tetracycline and macrolides is encoded by genes carried by integrative and conjugative elements (ICEs) in most ST25 S. suis strains.

Genomic comparison identified at least five ICEs carrying genes encoding resistance among the 51 ST25 strains under investigation. The majority of strains, including the reference strain NSUI060 possessed two ICEs, one (ICENsui60t) carrying gene tetO-1 (top, left), and a second (ICENsui60e) carrying gene ermB-20 (top, right). In most other strains a single ICE carried genetic determinants for both resistances, encoded by different tetO and ermB alleles. Homology between ICE genes is depicted by the different shades of grey.