Staphylococcus aureus strains that express serotype 5 or serotype 8 capsular polysaccharides differ in virulence

Abstract

Most isolates of Staphylococcus aureus produce a serotype 5 (CP5) or 8 (CP8) capsular polysaccharide. To investigate whether CP5 and CP8 differ in their biological properties, we created isogenic mutants of S. aureus Reynolds that expressed CP5, CP8, or no capsule. Biochemical analyses of CP5 and CP8 purified from the isogenic S. aureus strains were consistent with published structures. The degree of O acetylation of each polysaccharide was similar, but CP5 showed a greater degree of N acetylation. Mice challenged with the CP5(+) strain showed a significantly higher bacteremia level than mice challenged with the CP8(+) strain. Similarly, the CP5(+) strain survived preferentially in the bloodstream and kidneys of infected mice challenged with a mixed inoculum containing both strains. The enhanced virulence of the CP5(+) strain in vivo correlated with its greater resistance to in vitro killing in whole mouse blood. Likewise, in vitro opsonophagocytic killing assays with human neutrophils and sera revealed greater survival of the Reynolds (CP5) strain, even though the kinetics of opsonization by C3b and iC3b was similar for both the CP5(+) and CP8(+) strains. Electron micrographs demonstrated C3 molecules on the cell wall beneath the capsule layer for both serotype 5 and 8 strains. Purified CP5 and CP8 stimulated a modest oxidative burst in human neutrophils but failed to activate the alternative complement pathway. These results indicate that CP5 and CP8 differ in a number of biological properties, and these differences likely contribute to the relative virulence of serotype 5 and 8 S. aureus in vivo.

FIG. 1.

Transmission electron micrographs of S. aureus Reynolds (CP5) A), Reynolds (CP8) B), and Reynolds (CP⁻) C). Prior to dehydration and embedding, the bacteria were incubated with capsular antibodies to stabilize and visualize the capsule. Samples incubated with heterologous antiserum appeared acapsular (data not shown).

FIG. 2.

Percentage of Reynolds (CP5) and Reynolds (CP8) in the blood of mice challenged i.p. with a mixed inoculum containing 5 × 10⁷ CFU of each strain. Separate groups of eight mice were bled at each time point. The time zero point indicates the percentage of each strain in the inoculum. The mean (± SEM) log CFU per ml blood is shown beneath the figure for mice bled at 20, 120, or 200 min, respectively.

FIG. 3.

Opsonophagocytic killing of Reynolds (CP⁻) (grey bars), Reynolds (CP5) (open bars), and Reynolds (CP8) (black bars) by human PMNs after 2 h in the presence of absorbed NHS. Reynolds (CP⁻) was not tested with AltaStaph antibodies. The data represent the means ± SEM of 3 to 10 independent experiments.

FIG. 4.

S. aureus strains were opsonized with 10% absorbed NHS and incubated with human PMNs for 20 or 60 min. Following treatment with lysostaphin to kill extracellular staphylococci, the numbers of viable intracellular S. aureus cells were determined by plate counts. The data represent the means of three or four separate determinations.

FIG. 5.

Respiratory burst of 10⁶ human PMNs in response to isogenic S. aureus strains (n = 5) (A) or purified polysaccharides (n = 2 or 3) (B). Significantly more PMNs underwent an oxidative burst in response to Reynolds (CP⁻) compared with the two encapsulated strains (P < 0.001 by ANOVA). The relative PMN activation in response to 1 μM fMLP was 161. PGA, polygalacturonic acid; TA, teichoic acid.

FIG. 6.

Kinetics of C3b (A) and iC3b (B) deposition on Reynolds (CP5) (grey shaded) and Reynolds (CP8) (black line) in 10% absorbed NHS. The inset represents unlabeled S. aureus.

FIG. 7.

Immunogold labeling of C3 deposition on S. aureus incubated in 40% NHS. The capsule was visualized with rabbit serum containing capsule-specific antibodies. (A) Reynolds (CP5), (B) Reynolds (CP8). Magnification, ×40,000. Bar, 0.2 μm.