Structural, Genetic, and Serological Elucidation of Streptococcus pneumoniae Serogroup 24 Serotypes: Discovery of a New Serotype, 24C, with a Variable Capsule Structure

Abstract

Pneumococcal capsules are important in pneumococcal pathogenesis and vaccine development. Although conjugate vaccines have brought about a significant reduction in invasive pneumococcal disease (IPD) caused by vaccine serotypes, the relative serotype prevalence has shifted with the dramatic emergence of serotype 24F in some countries. Here, we describe 14 isolates (13 IPD and 1 non-IPD) expressing a new capsule type, 24C, which resembles 24F but has a novel serological profile. We also describe the antigenic, biochemical, and genetic basis of 24F and 24C and the related serotypes 24A and 24B. Structural studies show that 24B, 24C, and 24F have identical polysaccharide backbones [β-Ribf-(1→4)-α-Rhap-(1→3)-β-GlcpNAc-(1→4)-β-Rhap-(1→4)-β-Glcp] but with different side chains, as follows: 24F has arabinitol-phosphate and 24B has ribitol-phosphate. 24C has a mixture of 24F and 24B repeating units, with the ratio of ribitol to arabinitol being strain dependent. In contrast, the 24A capsule has a backbone without β-Ribf but with arabinitol-phosphate and phosphocholine side chains. These structures indicate that factor-sera 24d and 24e recognize arabinitol and ribitol, respectively, which explains the serology of serogroup 24, including those of 24C. The structures can be genetically described by the bispecificity of wcxG, which is capable of transferring arabinitol or ribitol when arabinitol is limiting. Arabinitol is likely not produced in 24B but is produced in reduced amounts in 24C due to various mutations in abpA or abpB genes. Our findings demonstrate how pneumococci modulate their capsule structure and immunologic properties with small genetic changes, thereby evading host immune responses. Our findings also suggest a potential for new capsule types within serogroup 24.

Keywords: Streptococcus pneumoniae; capsule polysaccharide; genetic heterogeneity; serogroup; vaccine.

FIG 1

(A) Diagrammatic representation of the capsular biosynthetic (cps) loci of serotypes 24X (MNY586), 24F (CR931688 and MNY576), 24B (CR931687 and MNY583), and 24A (CR931686). The direction of the arrows indicates the transcriptional orientation. The symbol “Ø” designates a pseudogene, and tnp denotes transposase. The five-digit numbers at the 3′ end indicate the size (in bp) of the cps loci. CR931686, CR931687, CR931688 represent the GenBank accession no. of published cps sequences of serotypes 24A, 24B, and 24F, respectively. The cps loci from representative German strains (MNY586, MNY576, and MNY583) are shown for comparison. Since the cps locus of the German 24A strains is similar to the published sequence, only the cps locus of CR931686 is shown. The previously known abp1 and abp2 genes were renamed as abpA and abpB, respectively. (B) Multiple amino acid alignment of the abpA encoded proteins from multiple strains representing serotypes 24X, 24F, 24B, and 24A. CR931686, CR931687, and CR931688 are the GenBank accession no. of the previously published sequences. MNY-labeled sequences are from German strains. The serotype of the strain/sequence is provided in parenthesis next to the strain name/GenBank accession no. In the multiple alignment, the consensus amino acids are represented by dots. Amino acid differences are shown by one-letter symbols. The complete consensus amino acid sequence is shown at the top of the multiple alignment. The numbers on top of the consensus amino acid sequence indicate the position of an amino acid. An asterisk represents a stop codon. GGxxG and DxD conserved motifs are boxed. Multiple amino acid alignment was performed by Multiple Alignment using Fast Fourier Transform (MAFFT) with scoring matrix 200 PAM/K of 2 and a gap open penalty of 1.5.

FIG 2

(A) Selected region (3.60 to 4.00 ppm) of the ¹H NMR spectra shows the difference in the structure of 24F, 24B, and 24C capsule polysaccharides. STREP24F represents serotype 24F and is from Nahm laboratory bacterial strain collection. SSISP24B represents serotype 24B and is from Statens Serum Institut (SSI). MNY585, MNY586, and MNY589 represent serotype 24C, and the strains originated from Germany. Similarly, PHE194136 represents serotype 24C, and the strain originated from the United Kingdom. The selected region shows the mixture of side chain peaks corresponding to ribitol and arabinitol in serotype 24C. The peak at 3.735 ppm corresponds to Tris. The spectra were obtained by the Carr-Purcell-Meiboom-Gill sequence (CPMG) method. (B) ¹H NMR spectra (chemical shifts 1.0 to 5.5 ppm). The spectra show that serotype 24A has a different PS structure from serotypes 24F, 24B, and 24C. The peak at 5.3 ppm, which corresponds to Ribf, is absent in SSISP24A (serotype 24A) and is shown in a black dotted box. SSISP24A represents serotype 24A and was obtained from Statens Serum Institut (SSI). A peak at 3.3 ppm corresponds to phosphocholine (P-Cho), and a peak at 2.1 ppm corresponds to dimethyl sulfoxide (DMSO). Asterisks denote impurities from the cell wall polysaccharide.

FIG 3

(A) Selected region (3.6 to 4.2 ppm) of HSQC spectra of 24C (black) and 24F (red) capsule polysaccharides showing superimposed (red and black) arabinitol (Ara-ol) and unique (black only) ribitol (Rib-ol) peaks. (B) Selected region (3.6 to 4.2 ppm) of HSQC spectra of 24C (black) and 24B (pink) capsule polysaccharides showing superimposed (pink and black) ribitol (Rib-ol) and unique (black only) arabinitol (Ara-ol) peaks. (C) 1D ³¹P spectra of the ribitol and arabinitol phosphate regions in 24C capsule polysaccharide. The spectra show two phosphate (PO₄) groups in the 24C polysaccharide, one each linked to arabinitol and ribitol. Asterisks denote impurities from cell wall polysaccharide. In the peak labels, the letter indicates the sugar residue and the number denotes the carbon position (Table 3). Rib-ol is labeled F1 and Ara-ol is labeled F2 in Table 3. The 24C, 24F, and 24B capsule polysaccharides shown in the figure were purified from PHE194136, 24F/2, and 24B/2 strains, respectively.

FIG 4

Structures of the serogroup 24 (serotypes 24A, 24F, 24B, and 24C) capsule polysaccharides. The structure of the repeating unit of the capsule polysaccharide from each serotype is shown along with the cps genes (shown in red) responsible for the biosynthesis of the corresponding linkages. The structure of the 24C capsule polysaccharide shows the mixture of 24F and 24B repeat units; however, the ratio (i.e., ribitol to arabinitol) is strain dependent and may vary. The structural similarity between serotype 24C and serotypes 24F, 24B, and 24A explains the serological cross-reactivity with FS24d (recognizes arabinitol in serotypes 24F and 24A) and FS24e (recognizes ribitol in serotype 24B).

FIG 5

(A) Genetic manipulation diagram of 24C cps loci to study the function of the wcxG, wcxK_s, and wzy_s genes. The FG204 strain was created by transferring the complete cps locus from MNY586 (serotype 24C) to an unencapsulated TIGR4 genetic background (TIGR4-JS). Mutant strains (FG205, FG206, and FG207) were derived from FG204 by deleting the wcxG, wcxK_s, and wzy_s genes, respectively. The Sweet Janus cassette composed of sacB^s, Kan^r, and rpsL^s was used for gene deletion. (B) Biochemical characterization of the effect of the wcxG gene in serotype 24C. The superimposed HSQC spectra (chemical shifts, 3.6 to 4.2 ppm) of FG205 capsule polysaccharide (blue peaks) and MNY586 capsule polysaccharide (red peaks) show arabinitol (Ara-ol)- and ribitol (Rib-ol)-specific peaks in MNY586 but not in FG205 (red peaks without accompanying blue peaks). In the peak labels, the letter indicates the sugar residue, and the number denotes the carbon position (Table 3). Rib-ol is labeled F1 and Ara-ol is labeled F2 in the chemical shift table (Table 3). Asterisks denote impurities from cell wall polysaccharide. The # symbol denotes newly generated unknown peaks in FG205, which possibly are cell wall contaminants.