Glycosylation of Staphylococcus aureus cell wall teichoic acid is influenced by environmental conditions

Abstract

Wall teichoic acid (WTA) are major constituents of Staphylococcus aureus (S. aureus) cell envelopes with important roles in the bacteria's physiology, resistance to antimicrobial molecules, host interaction, virulence and biofilm formation. They consist of ribitol phosphate repeat units in which the ribitol residue is substituted with D-alanine (D-Ala) and N-acetyl-D-glucosamine (GlcNAc). The complete S. aureus WTA biosynthesis pathways was recently revealed with the identification of the two glycosyltransferases, TarM and TarS, respectively responsible for the α- and β-GlcNAc anomeric substitutions. We performed structural analyses to characterize WTAs from a panel of 24 S. aureus strains responsible for invasive infections. A majority of the S. aureus strains produced the β-GlcNAc WTA form in accordance with the presence of the tarS gene in all strains assessed. The β-GlcNAc anomer was preferentially expressed at the expense of the α-GlcNAc anomer when grown on stress-inducing culture medium containing high NaCl concentration. Furthermore, WTA glycosylation of the prototype S. aureus Newman strain was characterized in vivo in two different animal models, namely peritonitis and deep wound infection. While the inoculum used to infect animals produced almost exclusively α-GlcNAc WTA, a complete switch to β-glycosylation was observed in infected kidneys, livers and muscles. Overall, our data demonstrate that S. aureus WTA glycosylation is strongly influenced by environmental conditions and suggest that β-GlcNAc WTA may bring competitive advantage in vivo.

Figure 1

Glycosylation pattern of the Newman, Wood 46 and ATCC55804 WTAs. (A) 500 MHz ¹H NMR spectra of purified WTAs in D₂O at 293 K; (B) HPAEC-PAD chromatograms of GlcNAc-ribitol disaccharides obtained from S. aureus WTA HF-hydrolysis of the Wood 46 strain (β-D-GlcNAc-(1→4)-ribitol), ATCC55804 strain (β-D-GlcNAc-(1→3)-ribitol) and Newman (α-D-GlcNAc-(1→4)-ribitol) strains. GlcNAc: N-acetyl-glucosamine, Ribol: Ribitol, Tris: Tris buffer in samples.

Figure 2

Distribution of WTA structures determined by HPAEC-PAD carbotyping of a panel of 24 S. aureus strains grown in TSB. The structures were determined directly from cell growth. The proportion of WTA structures in each strain, calculated as percentage, was determined from a single or two independent experiments for 18 and 6 strains, respectively, as indicated by an asterisk. For the 6 strains, the average percentage values are represented. The percentage was calculated from peak areas using the formulas described in the Supplementary Methods.

Figure 3

Comparison of WTA structure distribution of 8 S. aureus strains grown in (A) TSB and (B) SATA media. The structures were determined by HPAEC-PAD carbotyping directly from cell growth in the same HPAEC-PAD analysis. The proportion of WTA structures in each strain (calculated as percentage) was determined from single experiments and calculated from peak areas using the formulas described in the Supplementary Methods.

Figure 4

HPAEC-PAD analyses of WTA from infected mouse organs: kidneys (A,D), livers (B,E), and muscles (C). Chromatograms of GlcNAc-ribitol disaccharides obtained from mouse infected (solid line) with the S. aureus Newman (A–C) and HT2005 742 (D,E) strains compared to chromatograms of disaccharides obtained from the strains grown in TSB and used for mouse infection (dashed line). Samples from kidney of naïve mice were analyzed in parallel as negative reference (dotted line).