Biochemical, genetic, and serological characterization of two capsule subtypes among Streptococcus pneumoniae Serotype 20 strains: discovery of a new pneumococcal serotype

Abstract

The bacterial pathogen Streptococcus pneumoniae expresses one of over 90 structurally distinct polysaccharide (PS) capsule serotypes. Prior PS structural analyses of the vaccine-associated serotype 20 do not agree with reports describing the genes that mediate capsule synthesis. Furthermore, using immunized human sera-based assays, serological differences were recently noted among strains typed as serotype 20. We examined the capsule structures of two serologically dissimilar serotype 20 strains, 20α and 20β, by extensive biochemical analysis. 20α PS was composed of the previously described serotype 20 hexasaccharide repeat unit, whereas the 20β PS was composed of a novel heptasaccharide repeat unit containing an extra branching α-glucose residue. Genetic analysis of the subtypes revealed that 20α may have arisen from a 20β progenitor following loss of function mutation to the glycosyltransferase gene whaF. Conventional serotyping methods using rabbit polyclonal or mouse monoclonal antibodies were unable to distinguish the subtypes. However, genetic analysis of multiple "serotype 20" clinical isolates revealed that all strains contain the 20β genotype. We propose naming bacteria that express the previously described 20α capsule structure 20A and bacteria that express the novel 20β capsule structure 20B, a new pneumococcal serotype.

FIGURE 1.

Published serotype 20 PS structure and cps locus. A, schematic and written representation of the de-O-acetylated PS structure of serotype 20 according to Ref. . The circled R′ denotes the intersubunit glycosidic linkage. Although the serotype 20 PS structure has been previously drawn with the phosphate on the reducing end of the repeat unit, the reducing residue of the repeat unit according to current models of the capsule biosynthetic process (reviewed in Ref. 28) is βGlc, as shown in this present representation. B, the serotype 20 cps locus according to Ref. (GenBank^TM accession number CR931679), with genes encoding putative glycosyl transferases denoted with gray arrows. Vertical lines indicate locations of the polymorphisms identified in this study compared with the published sequence. Amino acid mutations resulting from point mutations are noted in parentheses. The asterisk denotes a mutation found only in strain 5931-06 (GenBank^TM accession number JQ653093). Pound signs denote mutations found only in strain 6320 (GenBank^TM accession number JQ653094). All other polymorphisms are shared by both 5931-06 and 6320. C, alignment of the last 35 amino acids encoded by whaF in strain 5931-06 and the respective pseudogene in strain 6320. The gray box denotes encoded amino acid differences resulting from a slip strand mutation in whaF of 6320.

FIGURE 2.

Differences in glycosidic content between 20α and 20β. The mass spectrometery spectra (A) and GC chromatogram (B) of PS purified from strains 6320 (top panels) and 5931-06 (bottom panels) are shown. The mass/charge values of the three most intense signals are listed in each spectrum in A. In B, signals corresponding to Gal, Glc, and GlcNAc are labeled in the top chromatogram. Dashed lines are drawn at the highest intensity values of Gal in each sample, for comparison.

FIGURE 3.

Anomeric region of ¹H NMR spectrum of native 20α, de-O-acetylated 20α, native 20β, and de-O-acetylated 20β PS. The arrow identifies the signal attributed to an additional hexose found in 20β polysaccharide. Anomeric protons of each sugar residue of de-O-acetylated 20β PS are labeled A–G. For the identity of each residue, refer to Table 1.

FIGURE 4.

¹H-¹³C HSQC NMR spectrum of de-O-acetylated 20β PS. Individual cross-peaks are labeled following the annotation of sugar residue and carbon numbers (e.g., A1 denotes carbon 1 of residue A). The cross-peak in the oval is due to residual TrisCl. Boxed signals are from unidentified impurities.

FIGURE 5.

Inter-residue connectivity of 20β PS according to HMBC spectra. A, ¹H-¹³C HMBC partial spectrum of 20β PS showing the interglycosidic correlations (circled) that were used to establish the sequence of connectivities between the sugar residues. The arrow denotes the novel connectivity of αGlc H-1 (DH1) to αGlcNAc C-6 (AC6) absent in 20α PS. B, the 20β PS ¹H-³¹P HMBC spectrum consistent with correlations between phosphate and αGlcNAc H-1 (A1), βGlc H-6 (E6), and long range correlation to αGlcNAc H-2 (A2).

FIGURE 6.

¹H-¹³C HMBC (A) and HSQC (B) partial spectra of native 20β polysaccharide. A shows long range correlation from ¹H attached to carbons bearing OAc to the carbonyl carbon of each OAc group. Peaks corresponding to carbons containing OAc substitutions are labeled in B.

FIGURE 7.

The schematic and written representation of the 20β (20B) PS repeat unit. A black arrow points to the additional αGlc residue contained in this repeat unit compared with 20α (20A). White arrows point to OAc substitutions. The circled R′ denotes the intersubunit glycosidic linkage. Galf, galactofuranose. Glycosidic residues are labeled with bold italic capital letters (A–G) according to their assignments in Table 1.