Enterococcal leucine-rich repeat-containing protein involved in virulence and host inflammatory response

Abstract

Enterococcus faecalis is an important nosocomial pathogen associated with high morbidity and mortality for patients who are immunocompromised or who have severe underlying diseases. The E. faecalis genome encodes numerous surface-exposed proteins that may be involved in virulence. This work describes the characterization of the first internalin-like protein in E. faecalis, ElrA, belonging to the recently identified WxL family of surface proteins. ElrA contains an N-terminal signal peptide for export, a leucine-rich repeat domain that may interact with host cells, and a C-terminal WxL domain that interacts with the peptidoglycan. Disruption of the elrA gene significantly attenuates bacterial virulence in a mouse peritonitis model. The elrA deletion mutant also displays a defect in infection of host macrophages and a decreased interleukin-6 response in vivo. Finally, elrA expression is induced in vivo. Altogether, these results demonstrate a role for ElrA in the E. faecalis infectious process in vivo and suggest that this surface protein may contribute to E. faecalis virulence by stimulating the host inflammatory response.

FIG. 1.

Principal characteristics of the amino acid sequence of ElrA. The predicted signal peptide is underlined. N-terminal cap and immunoglobulin (Ig)-like domains are indicated on a gray background. Conserved residues of the WxL C-terminal domain are in bold. LRR residues corresponding to LRR consensus sequences are highlighted in gray. Amino acid positions of residues are indicated on the left.

FIG. 2.

Expression analysis of the putative elrA operon. (A) Genetic organization of the putative elrA operon. Each ORF is named according to its TIGR gene identification. Solid black spacers indicate predicted intergenic regions. Putative terminators are indicated as lollipops. (B) RT-PCR of junctions in the putative operon comprising ORFs elrA and EF_2682. Each sample was run in three reactions with three different templates: RNA (−), genomic DNA (+), and cDNA (C). (C) Schematic representation of elrA promoter region organization. Arrows indicate the start of EF_2687 and elrA coding sequences. The ribosome binding site is in italics. Transcriptional initiation nucleotides (T1 and T2) are indicated as +1 in bold. The −35 and −10 boxes are underlined in bold. The 55-bp inverted repeat is underlined.

FIG. 3.

Effect of elrA inactivation on virulence. (A) Kaplan-Meier survival analysis in a mouse peritonitis model with OG1RF (squares), isogenic ΔelrA mutant (circles), and the complemented mutant strain (inverted triangles). Mice were infected intraperitoneally with 8.5 × 10⁸, 9.0 × 10⁸, and 9.3 × 10⁸ CFU of OG1RF, the ΔelrA strain, and the complemented mutant strain, respectively. (B) Dissemination levels of OG1RF (white bars), the ΔelrA strain (gray bars), and the complemented mutant (black bars) were compared in mouse spleen and liver. Mice were infected with ∼10⁸ CFU of each strain. The results represent the means ± standard deviations of the number of bacteria able to colonize the spleen and liver at 20 h postinfection.

FIG. 4.

Time course of intracellular survival of OG1RF and ΔelrA strains within murine peritoneal macrophages. Results correspond to the means ± standard variations of viable intracellular bacteria per 10⁵ macrophages of four independent experiments. White bars, OG1RF; gray bars, ΔelrA strain.

FIG. 5.

Cytokine expression profile of peritoneal fluid of mice infected with OG1RF and ΔelrA strains. Peritoneal fluid was recovered from mice infected intraperitoneally with ∼10⁸ CFU of the OG1RF (gray bars) or ΔelrA (black bars) strain. Results show the average cytokine concentration ± standard deviation in pg/ml for three mice. IFN g, gamma interferon; GMCSF, granulocyte-macrophage colony-stimulating factor.