The SrtA Sortase of Streptococcus agalactiae is required for cell wall anchoring of proteins containing the LPXTG motif, for adhesion to epithelial cells, and for colonization of the mouse intestine

Abstract

Streptococcus agalactiae (group B streptococcus [GBS]) is the leading cause of neonatal pneumonia, sepsis, and meningitis. An in silico genome analysis indicated that GBS strain NEM316 encodes 35 proteins containing an LPXTG motif which are thought to be covalently linked to the peptidoglycan by an enzyme called sortase. The role of these cell wall-anchored proteins in GBS pathogenesis was evaluated on a global level by inactivating the srtA gene. This gene encodes the major sortase SrtA that anchors most of the LPXTG-containing proteins. We chose the C5a peptidase (ScpB) and Alp2, an abundant immunogenic protein, as prototypical LPXTG-containing proteins. As expected, the SrtA knockout mutant was unable to anchor the C5a peptidase (ScpB) and Alp2 to the cell wall. Complementation with plasmid-borne srtA inserted into the chromosome restored the correct surface localization of both ScpB and Alp2. Interestingly, the SrtA mutant was impaired for binding to the major extracellular matrix components fibronectin and fibrinogen and displayed a significant reduction in adherence to human (A549, HeLa, and Caco-2) and murine (L2) epithelial cells compared to the parental wild-type strain. Surprisingly, the inactivation of srtA had no effect on the virulence of the type III strain of GBS in a neonatal rat model (measured by the 50% lethal dose and lung colonization) but strongly impaired the capacity of the strain to colonize the intestines of gnotobiotic mice in a competition assay. These results demonstrate that LPXTG-containing proteins are involved in cell adhesion and GBS persistence in vivo.

FIG. 1.

Amino acid sequence comparisons of SrtA proteins from various gram-positive bacteria. Identical residues which are present in the seven sequences are marked by white letters on a black background. Conserved amino acid residues which are present in five of the seven sequences are written in black on a gray background. The critical cysteyl residue within the catalytic TLXTC motif is indicated with a black arrow. The single letter code is that recommended by IUPAC/IUB. Dots represent gaps introduced into the sequences to ensure optimal homology.

FIG. 2.

Western blot analysis of GBS proteins with anti-R28/Alp2 and anti-ScpB sera. (A) Proteins attached to the cell wall (CW) and culture supernatant proteins (CS) were purified from the wild-type strain NEM316 (lanes 1), the SrtA⁻ mutant NEM2135 (lanes 2), and the SrtA⁻/SrtA⁺ complemented strain NEM2136 (lanes 3). (B) Proteins were extracted by a hot SDS treatment from NEM316 (lanes 1), NEM2135 (lanes 2), and the complemented strain NEM2136 (lanes 3). Note that Alp2 and ScpB were massively released by SDS treatment of the SrtA⁻ mutant. In contrast, only a fraction of Alp2 was solubilized in the wild-type and complemented strains, and only a faint release of ScpB was observed for the complemented strain.

FIG. 3.

Display of the LPXTG-containing protein Alp2 on the cell surfaces of S. agalactiae NEM316 (wild-type strain), NEM2135 (SrtA⁻ mutant), and NEM2136 (SrtA⁻/SrtA⁺ complemented strain) cells. Bacteria were grown in TH broth at 37°C to an optical density at 600 nm of 0.5, washed twice with PBS, and incubated for 20 min at room temperature in PBS containing 2% SDS. The cells were analyzed by immunofluorescence with affinity-purified polyclonal anti-R28/Alp2 antibodies and revealed with anti-IgG coupled to Alexa 488 (A), and the same samples were observed by phase-contrast microscopy (B).

FIG. 4.

Adherence of S. agalactiae strains to immobilized fibronectin. Microtiter wells were coated with various concentrations of fibronectin, and 10⁷ CFU of the wild-type strain NEM316 (○), the SrtA⁻ mutant NEM2135 (♦), or the complemented strain NEM2136 (•) was added. The wells were washed and bound bacteria were assayed as described in Materials and Methods. The results are presented as mean values (± standard deviations [SD]) of one experiment performed in triplicate. The curves are representative of three independent experiments.

FIG. 5.

Mortality curves for rat pups infected with the wild-type strain NEM316 (○) and the SrtA⁻ mutant NEM2135 (♦). Groups of 20 neonatal Sprague-Dawley rat pups (48 h old) were inoculated i.p. with 5 × 10⁶ bacteria, and mortality was observed over a 7-day period. The difference in virulence of the two strains was considered not quite significant, with a two-tailed P value of 0.0849 by the Mann-Whitney test, as calculated with GraphPad Instat (version 3.0).

FIG. 6.

Competitive index analysis of mixed cultures and colonizations with the SrtA⁺ strain NEM2093 (Sp^r SrtA⁺) and the SrtA⁻ mutant NEM2135 (Km^r SrtA⁻). For in vitro experiments (•), 20 ml of TH broth was inoculated with 10⁶ CFU each of NEM2093 and NEM2135, and the bacterial mixture was subcultivated over a 23-day period. The two bacterial populations were enumerated daily on agar plates containing spectinomycin (NEM2093) or kanamycin (NEM2135). For in vivo experiments (○), germfree mice were inoculated on day zero with 10⁸ CFU each of the SrtA⁺ strain NEM2093 and the SrtA⁻ mutant NEM2135. The bacterial numbers in homogenates were determined at various intervals by plating on TH agar plates containing the appropriate antibiotic. The competitive indices were calculated by dividing the CFU of NEM2093 by the CFU of NEM2135. The in vitro competitive index was the average of values from three independent experiments, whereas the in vivo competitive index was the average of values from five individual mouse experiments. The vertical bars represent one SD.