Key role of poly-gamma-DL-glutamic acid in immune evasion and virulence of Staphylococcus epidermidis

Abstract

Coagulase-negative staphylococci, with the leading species Staphylococcus epidermidis, are the predominant cause of hospital-acquired infections. Treatment is especially difficult owing to biofilm formation and frequent antibiotic resistance. However, virulence mechanisms of these important opportunistic pathogens have remained poorly characterized. Here we demonstrate that S. epidermidis secretes poly-gamma-DL-glutamic acid (PGA) to facilitate growth and survival in the human host. Importantly, PGA efficiently sheltered S. epidermidis from key components of innate host defense, namely antimicrobial peptides and neutrophil phagocytosis, and was indispensable for persistence during device-related infection. Furthermore, PGA protected S. epidermidis from high salt concentration, a key feature of its natural environment, the human skin. Notably, PGA was synthesized by all tested strains of S. epidermidis and a series of closely related coagulase-negative staphylococci, most of which are opportunistic pathogens. Our study presents important novel biological functions for PGA and indicates that PGA represents an excellent target for therapeutic maneuvers aimed at treating disease caused by S. epidermidis and related staphylococci.

Figure 1

Molecular genetic comparison of bacteria with genes encoding a putative PGA synthesis machinery. (A) Phylogenetic trees based on sequence comparisons of the capB (amide ligase), capC (unknown function), and capD (depolymerase) genes. CapA is a putative PGA exporter. We have excluded a comparison of capA genes, because capA homologs were not found in all the organisms and comparison of transporters is normally less indicative of phylogenetic relations. Of the microorganisms shown, production of PGA has been demonstrated previously only in B. anthracis and B. subtilis, and in this study, in S. epidermidis. (B) cap genes and homologs in bacteria for which genetic information is available. The B. anthracis cap gene cluster is located on a plasmid and flanked by IS231 insertion sequences. All other genes are located in the bacterial chromosomes.

Figure 2

PGA production in S. epidermidis. (A) Relative expression of PGA determined by immuno-dot blot analysis. PGA was extracted from bacterial cell surfaces as described in Methods. A calibration curve was obtained by dilution of the most intensive sample obtained from the PGA-overexpressing complemented strain S. epidermidis Δcap (pRBcapBCAD). Results are the mean ± SEM of 4 experiments for the samples and the mean ± SEM of 4 different serial dilutions for the calibration curve. A representative blot is shown at the top. (B) Detection of S. epidermidis PGA with immunoscanning electron microscopy. PGA was detected with anti-PGA antiserum. (C) Analysis of D-glutamic (D-Glu) and L-glutamic (L-Glu) acid in S. epidermidis PGA by stereoselective chromatography and liquid chromatographic–mass spectrometric detection of glutamic acid. To determine D- and L-glutamic acid amounts, the L-glutamic acid background detected in the cap mutant strain (Δcap) was subtracted from PGA expression strains. (A–C) Δcap, isogenic cap deletion strain; capBCAD, complemented strain S. epidermidis Δcap (pRBcapBCAD).

Figure 3

PGA production in S. epidermidis strains. PGA expression in S. epidermidis strains of clinical and commensal origin, and under low and high salt conditions, determined by immuno-dot blot analysis. Cells were grown for 24 hours at 37°C with shaking at 200 rpm. PGA was then purified as described in Methods. Horizontal bars show the group mean. The membrane background was subtracted from each sample. The experiment, including purification and detection, was repeated twice with very similar results.

Figure 4

Role of PGA in osmoprotection and inducibility of cap expression by NaCl. (A) Growth (OD₆₀₀) and (B) viability (CFU) of wild-type and cap mutant strains in Luria-Bertani medium supplemented with 2 M NaCl. Bacteria were inoculated from an overnight preculture (1:1,000) and grown in flasks at 37°C with shaking at 200 rpm. Values are the mean ± SEM of 3 experiments. *P < 0.05; **P < 0.01; ^#P < 0.001 (wild-type versus mutant strain). (C) Quantitative real-time PCR analysis of NaCl inducibility of cap expression. Bacteria were grown as in (A) with the indicated concentrations of NaCl. Cells were harvested after 6 hours of growth, RNA was isolated, and real-time PCR was performed using a capB probe. Values are the mean ± SEM of 3 experiments.

Figure 5

Role of PGA in immune evasion and virulence of S. epidermidis. (A and B) Resistance to cationic antimicrobial peptides. Washed S. epidermidis cells (approximately 10⁵) were incubated with LL-37 (A) or human β-defensin 3 (B) in various concentrations for 2 hours at 37°C. Thereafter, S. epidermidis survivor cells were counted by plating. Results are shown as dose-response curves. The log LD₅₀ values for all strain/peptide combinations are given in the key. Statistical analyses are for each peptide concentration. Values of significance were calculated against the wild-type (for Δcap) and Δcap (for capBCAD) strains. (C) Resistance to neutrophil phagocytosis. Phagocytosis by human neutrophils was determined after 30 minutes of incubation with S. epidermidis at a ratio of 20 bacteria per PMN. (D) Mouse model of subcutaneous device-related infection. Catheter pieces with equal amounts of adhered S. epidermidis cells (2 × 10⁵) were placed under the dorsum of the animals. CFU on implanted devices 1 week after infection were counted. The horizontal bar shows the group mean. *P < 0.05; ***P < 0.01; ***P < 0.001.