The Vi capsular polysaccharide prevents complement receptor 3-mediated clearance of Salmonella enterica serotype Typhi

Abstract

Capsular polysaccharides are important virulence factors of invasive bacterial pathogens. Here we studied the role of the virulence (Vi) capsular polysaccharide of Salmonella enterica serotype Typhi (S. Typhi) in preventing innate immune recognition by complement. Comparison of capsulated S. Typhi with a noncapsulated mutant (ΔtviBCDE vexABCDE mutant) revealed that the Vi capsule interfered with complement component 3 (C3) deposition. Decreased complement fixation resulted in reduced bacterial binding to complement receptor 3 (CR3) on the surface of murine macrophages in vitro and decreased CR3-dependent clearance of Vi capsulated S. Typhi from the livers and spleens of mice. Opsonization of bacteria with immune serum prior to intraperitoneal infection increased clearance of capsulated S. Typhi from the liver. Our data suggest that the Vi capsule prevents CR3-dependent clearance, which can be overcome in part by a specific antibody response.

FIG. 1.

Vi capsule expression detected by flow cytometry. Cells of the S. Typhi wild type (A), an S. Typhi ΔtviB-vexE mutant (B), E. coli strain W3110 (C), E. coli strain W3110 carrying the cloned viaB locus (pDC5) (D), an S. Typhi vexE mutant (E), or an S. Typhi vexE mutant complemented with the cloned vexE gene (pAS1) (F) were labeled with rabbit anti-Vi serum/goat-anti rabbit FITC conjugate (Vi expression, y axis), and fluorescence intensities were determined for 10,000 particles. Each experiment was repeated three times independently with similar outcomes, and a representative example is shown.

FIG. 2.

The Vi capsule reduces C3 fixation and increases complement resistance. (A and B) Fixation of C3 after incubation of capsulated (wild-type) and noncapsulated (ΔtviB-vexE) S. Typhi strains in C5-depleted human serum (A) or in C3-depleted human serum (B) detected by flow cytometry using an anti-human C3 FITC conjugate. (C and D) Survival of capsulated and noncapsulated S. Typhi strains (C) or capsulated (viaB on plasmid pDC5) and noncapsulated (wild-type [W3110]) E. coli strains (D) in normal human serum. The experiment was repeated three times independently, and data points represent averages ± standard deviations. (E) Fixation of C3 after incubation of capsulated and noncapsulated S. Typhi strains in murine serum detected by flow cytometry using an anti-murine C3 FITC conjugate. (F) Fixation of C3 after incubation of capsulated (vexE mutant complemented with pAS1) and noncapsulated (vexE mutant) S. Typhi strains in murine serum detected by flow cytometry using an anti-murine C3 FITC conjugate. The experiments whose results are presented in panels A, B, E, and F were repeated three times independently with similar outcomes each time, and a representative examples are shown.

FIG. 3.

The Vi capsule reduces CR3-mediated interaction with macrophages. (A and B) Binding to (after incubation at 4°C) (A) and phagocytosis of (gentamicin protection assay) (B) capsulated bacteria (the S. Typhi wild type or the S. Typhi vexE mutant complemented with pAS1) and noncapsulated bacteria (an S. Typhi ΔtviB-vexE mutant or an S. Typhi vexE mutant) to human macrophage-like THP-1 cells in the absence (black bars) or presence (open bars) of blocking anti-human CD11b antibodies. (C and D) Binding to (after incubation at 4°C) (C) and phagocytosis of (gentamicin protection assay) (D) the indicated bacterial strains by murine BMDMs from wild-type mice (C57BL/6, black bars) or CR3-deficient mice (open bars). Each experiment was repeated at least three times independently, and data are shown as averages ± standard errors. The statistical significance of the differences is indicated above. NS, not significant.

FIG. 4.

The Vi capsule reduces CR3-dependent clearance of S. Typhi from organs of mice. Recovery of capsulated (wild-type) and noncapsulated (ΔtviB-vexE) S. Typhi strains from the livers (A), spleens (B), or blood (C) of wild-type mice (C57BL/6, black bars) or CR3-deficient mice (open bars) 4 h after intraperitoneal infection. Bars represent averages ± standard deviations from four animals. The statistical significance of the differences is indicated above each graph. NS, not significant.

FIG. 5.

Opsonization with immune serum increases clearance of capsulated S. Typhi from the liver. (A to C) Deposition of IgG (A and C) or C3 (B) on the surface of the capsulated S. Typhi wild type (Ty2) or a noncapsulated S. Typhi strain (ΔtviB-vexE mutant) after incubation in naïve mouse serum (C) or immune mouse serum (A and B) detected by flow cytometry using an anti-mouse C3 FITC conjugate (B) or anti-mouse IgG FITC conjugate (A and C). (D and E) The indicated bacterial strains were opsonized in naïve mouse serum (black bars) or immune mouse serum (open bars) and then injected intraperitoneally into mice (C57BL/6). Bars represent average numbers of bacteria recovered from the spleens (D) or the livers (E) of four animals collected 4 h after infection ± standard deviations. The statistical significance of the differences is indicated above each graph. NS, not significant.

FIG. 6.

Repeat unit structure of the Vi capsule and the S. Typhi O antigen. Structures of the O-antigen repeat unit in S. Typhi LPS (A) (18) and the repeat unit of the Vi capsular polysaccharide (B) (13) according to previous reports. Note the presence in the O antigen and the absence in the Vi capsule of free hydroxyl groups available for covalent binding of C3b.