Cross-linked peptidoglycan mediates lysostaphin binding to the cell wall envelope of Staphylococcus aureus

Abstract

Staphylococcus simulans bv. staphylolyticus secretes lysostaphin, a bacteriocin that cleaves pentaglycine cross bridges in the cell wall of Staphylococcus aureus. The C-terminal cell wall-targeting domain (CWT) of lysostaphin is required for selective binding of this bacteriocin to S. aureus cells; however, the molecular target for this was unknown. We used purified green fluorescent protein fused to CWT (GFP-CWT) to reveal species-specific association of the reporter with staphylococci. GFP-CWT bound S. aureus cells as well as purified peptidoglycan sacculi. The addition of cross-linked murein, disaccharides linked to interconnected wall peptides, blocked GFP-CWT binding to staphylococci, whereas murein monomers or lysostaphin-solubilized cell wall fragments did not. S. aureus strain Newman variants lacking the capacity for synthesizing polysaccharide capsule (capFO), poly-N-acetylglucosamine (icaAC), lipoprotein (lgt), cell wall-anchored proteins (srtA), or the glycolipid anchor of lipoteichoic acid (ypfP) bound GFP-CWT similar to wild-type staphylococci. A tagO mutant strain, defective in the synthesis of polyribitol wall teichoic acid attached to the cell wall envelope, displayed increased GFP-CWT binding. In contrast, a femAB mutation, reducing both the amount and the length of peptidoglycan cross-linking (monoglycine cross bridges), showed a dramatic reduction in GFP-CWT binding. Thus, the CWT domain of lysostaphin directs the bacteriocin to cross-linked peptidoglycan, which also serves as the substrate for its glycyl-glycine endopeptidase domain.

FIG. 1.

GFP-CWT binds to the surface of S. aureus cells. Bacteria were incubated with purified 50 nM GFP or GFP-CWT, harboring a C-terminal fusion to the lysostaphin CWT. Bacteria with or without bound protein were sedimented by centrifugation and GFP or GFP-CWT binding was visualized by fluorescence microscopy. The top panel shows images captured via charge-coupled-device camera by phase-contrast microscopy; the bottom panel displays images captured via fluorescence microscopy.

FIG. 2.

GFP-CWT binding to receptors on the surface of S. aureus. A fluorescence plate assay was developed to quantify GFP or GFP-CWT binding to bacterial surfaces. (A) S. aureus Newman cells were washed and adjusted to an OD₆₀₀ of 10, and twofold dilutions of staphylococci were mixed with GFP or GFP-CWT. Bacteria were sedimented by centrifugation and fluorescence of supernatant measured. (B) GFP-CWT binding to S. aureus Newman or Enterococcus faecalis strain FA2-2. (C) GFP-CWT binding to S. aureus Newman without (mock) or with the addition of a 2.5-fold, 5-fold, or 10-fold excess of purified GST-CWT protein or a 10-fold excess of GST.

FIG. 3.

GFP-CWT binding to purified peptidoglycan. (A) Highly purified peptidoglycan (PG) sacculi (carbohydrate, lipid, protein, and teichoic acid removed) from S. aureus Newman were adjusted to an OD₆₀₀ of 10, and twofold dilutions of sacculi were mixed with GFP or GFP-CWT. Peptidoglycan sacculi were sedimented by centrifugation and supernatants analyzed for fluorescence. (B) Purified peptidoglycan was cleaved with mutanolysin and concentration of soluble amino sugars (NAG) determined with the Morgan-Elson reaction. S. aureus cells were mixed with solubilized peptidoglycan (or twofold serial dilutions) and GFP-CWT. Inhibition of GFP-CWT binding with solubilized peptidoglycan (PG) and mock inhibition, as well as maximal inhibition (GFP-CWT incubated with solubilized peptidoglycan in the absence of sedimentable bacteria), were analyzed.

FIG. 4.

Peptidoglycan fragments inhibit GFP-CWT binding to S. aureus. (A) HPLC chromatogram of mutanolysin-digested S. aureus peptidoglycan. AU, absorbance units. (B) HPLC chromatogram of mutanolysin/lysostaphin-digested S. aureus peptidoglycan. AU, absorbance units. (C) S. aureus cells were mixed with soluble, reduced peptidoglycan fragments (mutanolysin digested) at an A₂₀₆ of 3 (or twofold serial dilutions thereof) and GFP-CWT. Following sedimentation of bacteria, GFP-CWT fluorescence was measured in supernatants. Low-molecular-weight peptidoglycan fragments (murein monomers and dimers in fractions 56 to 60; gray diamonds) did not inhibit binding of GFP-CWT to S. aureus. High-molecular-weight peptidoglycan (cross-linked fragments in fractions 91 through 95; black triangles) displayed strong inhibitory activity. As controls, no-inhibition-fluorescence values (filled circles) were determined by incubating bacteria and GFP-CWT in the absence of solubilized peptidoglycan fragments. Maximal-inhibition values (black squares) were determined by incubating GFP-CWT in the absence of bacteria and HPLC fractions. (D) S. aureus cells were mixed with soluble, reduced peptidoglycan fragments (mutanolysin/lysostaphin digested) at an A₂₀₆ of 3 (or twofold serial dilutions thereof) and GFP-CWT. No inhibitory activity was found as shown for fractions 29 and 30 (gray diamonds) or 51 through 60 (black triangles). Filled circles and filled squares represent no-inhibition and maximal-inhibition curves, respectively.

FIG. 5.

GFP-CWT binding to S. aureus mutants with altered cell wall envelope properties. Wild-type (wt) S. aureus Newman and Phoenix library transposon mutants with insertions in defined genes were subjected to GFP-CWT binding assays. (A) Capsular polysaccharide mutants (Newman wt, ΦΗΕ5448 [locus tag SAV0154, capF], and ΦΗΕ5283 [locus tag SAV0163, capO]). (B) Poly-N-acetylglucosamine mutants (Newman wt, ΦΗΕ4722 [locus tag SAV2666, icaA], and ΦΗΕ14557 [locus tag SAV2669, icaC]). (C) Peptidoglycan mutants (Newman wt, ΦΗΕ8500 [locus tag SAV2567, oatA], and ΦΗΕ11552 [locus tag SAV0642, pbp4]). (D) Sortase mutants (Newman wt and ΦΗΕ3486 [locus tag SAV2528, srtA]). (E) Lipoprotein and d-alanine modification of secondary wall polymer mutants (Newman wt, ΦΗΕ106 [locus tag SAV0761, lgt], and ΦΗΕ12076 [locus tag SAV0933, dltB]).

FIG. 6.

GFP-CWT binding to S. aureus mutants with altered LTA properties. (A and C) Detection of LTA by immunoblotting. S. aureus cells were disintegrated with glass beads and bacterial lysates boiled in SDS buffer to solubilize LTA. Samples were separated by SDS-PAGE and LTA detected by immunoblotting with a polyglycerol phosphate-specific LTA antibody. Bars and numbers at the right of the panel indicate the positions and sizes of protein standards in kDa. wt, wild type. (B and D) GFP-CWT binding assays and measurement of fluorescence in supernatants. (B) GFP-CWT binding to isogenic wild-type (wt) and ypfP mutant S. aureus strains. (D) GFP-CWT binding to S. aureus Newman, SA113, and RN4220.

FIG. 7.

GFP-CWT binding to S. aureus mutants with altered WTA properties. (A) GFP-CWT binding assays to the surface of S. aureus Newman wild-type (wt) and tagO mutant strains was measured. (B) GFP-CWT binding to the surface of S. aureus Newman wt and tagO mutant strains was viewed by fluorescence microscopy. Increased GFP-CWT binding to the surface of an S. aureus strains lacking WTA is reflected in a steeper binding curve (A) as well as brighter fluorescence of images captured with identical exposure times (B).

FIG. 8.

GFP-CWT binding to a S. aureus mutant with altered peptidoglycan cross bridge structure. GFP-CWT binding to the surface of S. aureus BB270 wild type (wt) and isogenic femAB mutant strain AS145 without or with complementing (compl.) plasmid pBBB64 was measured.