Genomic analysis reveals the molecular basis for capsule loss in the group B Streptococcus population

Abstract

The human and bovine bacterial pathogen Streptococcus agalactiae (Group B Streptococcus, GBS) expresses a thick polysaccharide capsule that constitutes a major virulence factor and vaccine target. GBS can be classified into ten distinct serotypes differing in the chemical composition of their capsular polysaccharide. However, non-typeable strains that do not react with anti-capsular sera are frequently isolated from colonized and infected humans and cattle. To gain a comprehensive insight into the molecular basis for the loss of capsule expression in GBS, a collection of well-characterized non-typeable strains was investigated by genome sequencing. Genome based phylogenetic analysis extended to a wide population of sequenced strains confirmed the recently observed high clonality among GBS lineages mainly containing human strains, and revealed a much higher degree of diversity in the bovine population. Remarkably, non-typeable strains were equally distributed in all lineages. A number of distinct mutations in the cps operon were identified that were apparently responsible for inactivation of capsule synthesis. The most frequent genetic alterations were point mutations leading to stop codons in the cps genes, and the main target was found to be cpsE encoding the portal glycosyl transferase of capsule biosynthesis. Complementation of strains carrying missense mutations in cpsE with a wild-type gene restored capsule expression allowing the identification of amino acid residues essential for enzyme activity.

Fig 1. GBS SNP-based and MST-based phylogenetic trees.

(A) SNP-based Neighbor-joining phylogenetic tree. The tree was generated using 14,092 polymorphic sites extracted from the alignment of 0.42 Mbp non-duplicated core regions shared by all 373 strains aligned to the reference strain 2603 V/R. CCs assigned to the 12 major clusters are indicated in colored ribbons. Dots represent single strains and are colored according to their capsular genotype. Asterisks indicate strains where the CC assigned by MLST (CC-1 or CC-6-8-10) differs from that assigned to strains belonging to the same SNP clade. (B) Minimum Multilocus Sequence Typing spanning tree of GBS strains. Each node represents one ST and STs differing by only one allele are connected by a line. Node dimensions refer to the relative number of strains belonging to each ST. Colored dots represent the assigned 17 CCs after refinement based on the SNP analysis. CCs included in the dotted line circle are linked by transition STs.

Fig 2. GBS SNP-based Neighbor-joining phylogenetic trees highlighting GBS hosts and NT strains.

(A) Phylogenetic tree where non-typeable or capsulated phenotypes are indicated by colored dots. (B) Phylogenetic tree where strain host origin is indicated by colored dots.

Fig 3. Relative frequency of the different types of mutations possibly responsible for the lack of capsule expression in GBS.

(A) Distribution of 126 genetic alteration events detected in 89 isolates. (B) Number of strains bearing each of the cps mutation types among to the five selected capsular genotypes.

Fig 4. Different kinds of genetic alterations detected in the GBS cps operon.

(A) Distribution of 37 transposable elements (blue), 24 insertion or deletion (indels) targeting a single gene (light green) and 54 point mutations (dark green) scattered across the cps operon. Genes that are missing in one or more capsular types are shown with red font color in the operon diagram. (B) Deletions comprising more than a single gene are indicated with dotted lines; empty spaces represent the absence of cps genes in a specific capsular genotype. (C) Gene targets of capsule inactivation mutations that appeared as single events (blue arrows); the figures below the target genes indicate the number of strains presenting the individual mutation.

Fig 5. Complementation of cpsE missense mutations in selected NT GBS strains.

Flow cytometry analysis of NT GBS strains carrying missense mutations in cpsE and of their counterpart after transformation with pAM-cpsE. Bacteria were incubated with mouse monoclonal antibodies specific for the five capsular polysaccharides, and then treated with labeled secondary antibodies. Fluorescence after incubation with the secondary antibody alone is indicated by empty histograms, while staining after treatment with type the specific antibodies is shown by the colored histograms.

Fig 6. Analysis of the influence of genetic alterations on the transcription of the cps operon in NT isolates.

(A) Mutations in the cps promoter or cpsA-D detected in 18 NT strains. Their positions are indicated as +/- numbers in relation to the first base pair of the cpsA coding sequence. ISs are represented by colored triangles; point mutations in the -10 sequence are marked in bold; deleted regions are represented by dotted lines and stop codons by crosses. Numbers within parentheses identify the strains reported in the x-axis of panel B. (B) Transcription of the cps operon in the 18 NT strains measured with primers cpsAup-F/R (filled bars) and to cpsE-F/R (hatched bars). The relative fold expression for each strain was estimated in comparison to the expression of cpsA in strain 515. For strains 25, 26, 52 and 53 the transcript in cpsE gene was not determined.