Identification of a novel Staphylococcus aureus two-component leukotoxin using cell surface proteomics

Abstract

Staphylococcus aureus is a prominent human pathogen and leading cause of bacterial infection in hospitals and the community. Community-associated methicillin-resistant S. aureus (CA-MRSA) strains such as USA300 are highly virulent and, unlike hospital strains, often cause disease in otherwise healthy individuals. The enhanced virulence of CA-MRSA is based in part on increased ability to produce high levels of secreted molecules that facilitate evasion of the innate immune response. Although progress has been made, the factors that contribute to CA-MRSA virulence are incompletely defined. We analyzed the cell surface proteome (surfome) of USA300 strain LAC to better understand extracellular factors that contribute to the enhanced virulence phenotype. A total of 113 identified proteins were associated with the surface of USA300 during the late-exponential phase of growth in vitro. Protein A was the most abundant surface molecule of USA300, as indicated by combined Mascot score following analysis of peptides by tandem mass spectrometry. Unexpectedly, we identified a previously uncharacterized two-component leukotoxin-herein named LukS-H and LukF-G (LukGH)-as two of the most abundant surface-associated proteins of USA300. Rabbit antibody specific for LukG indicated it was also freely secreted by USA300 into culture media. We used wild-type and isogenic lukGH deletion strains of USA300 in combination with human PMN pore formation and lysis assays to identify this molecule as a leukotoxin. Moreover, LukGH synergized with PVL to enhance lysis of human PMNs in vitro, and contributed to lysis of PMNs after phagocytosis. We conclude LukGH is a novel two-component leukotoxin with cytolytic activity toward neutrophils, and thus potentially contributes to S. aureus virulence.

Figure 1. Workflow diagram for surface proteomics.

Figure 2. Relative abundance of proteins identified on the surface of USA300.

PSORT v2.0 was used to assign identified proteins to subcellular compartments. (A) Total number of peptides assigned to proteins in each cellular compartment. A total of 2559 peptides were matched to proteins in the RSUE database. (B) Each data point in the scatter plot represents one protein. The line for each cellular location represents the median Mascot score for that compartment. Mascot scores were evaluated by a Kruskal-Wallis test with Dunn's post-test to compare the relative confidence of protein identifications between subcellular compartments.

Figure 3. Membrane association of LukG and predicted topology for putative membrane proteins.

(A) Immunoblot analysis of LukG secreted into the culture media (extracellular) or associated with subcellular fractions of USA300. Subcellular fractions were prepared as described in Materials and Methods. (B) Predicted membrane proteins identified by surface proteomics. Blue regions indicate areas of peptide coverage.

Figure 4. Construction and characterization of isogenic USA300 lukGH deletion (ΔlukGH) strains.

(A) Schematic of deletion strategy for generation of ΔlukGH and (B) confirmation of isogenic ΔlukGH and ΔlukGH/Δpvl strains by PCR as described in Methods. (C) Growth of LAC wild-type, ΔlukGH and ΔlukGH/Δpvl strains. (D) Exoprotein profile of LAC culture supernatants from early-stationary phase of growth (6 h). Proteins were resolved by SDS-PAGE and stained with Sypro Ruby. Black arrows correspond to LukG and LukH and yellow arrows denote LukF-PV and LukS-PV. (E) Immunoblot analysis of LukG and LukF-PV from culture supernatants of LAC wild-type-, isogenic gene deletion-, and complemented gene deletion strains as indicated.

Figure 5. LukGH contributes to PMN plasma membrane pore formation and lysis.

(A) Permeability of human PMNs exposed to culture supernatants from LAC wild-type-, isogenic gene deletion-, or complemented gene deletion strains. (B) Capacity of LAC wild-type and mutant CCY culture supernatants to cause PMN lysis. Dilution of culture supernatants was 1∶250. (C) PMNs were cultured with LAC strains for 6 h and PMN lysis was determined by release of LDH. Statistical analyses for panels (A, B, and C) were performed using repeated-measures ANOVA or one-way ANOVA and Tukey's posttest. *P<0.05 versus LAC or †P<0.05 as indicated. Results are the mean ± standard deviation of 4 (panel A), 3 (panel B) or 4–10 (panel C) PMN donors.