Sortase-catalysed anchoring of surface proteins to the cell wall of Staphylococcus aureus

Authors


Abstract

Many surface proteins of Gram-positive bacteria are anchored to the cell wall envelope by a transpeptidation mechanism, requiring a C-terminal sorting signal with a conserved LPXTG motif. Sortase, a membrane protein of Staphylococcus aureus, cleaves polypeptides between the threonine and the glycine of the LPXTG motif and catalyses the formation of an amide bond between the carboxyl-group of threonine and the amino-group of peptidoglycan cross-bridges. S. aureus mutants lacking the srtA gene fail to anchor and display some surface proteins and are impaired in the ability to cause animal infections. Sortase acts on surface proteins that are initiated into the secretion (Sec) pathway and have their signal peptide removed by signal peptidase. The S. aureus genome encodes two sets of sortase and secretion genes. It is conceivable that S. aureus has evolved more than one pathway for the transport of 20 surface proteins to the cell wall envelope.

Introduction

Staphylococcus aureus is an important human pathogen that causes abscesses in many organ tissues, septicaemia and several other diseases (Lowy, 1998). Molecular biologists have examined staphylococci for many years in an effort to explain the extraordinary pathogenic potential of this microorganism (Novick, 1991). S. aureus is capable of binding several human serum factors (Foster and Höök, 1998). Multiple pathogenic strategies have been suspected, and staphylococcal resistance to phagocytic killing seems to be the underlying goal in binding serum proteins. Staphylococci also adhere to several different human tissues, properties that may explain why these microbes can cause abscesses at many anatomical sites (Flock, 1999). In all cases examined, the adhesive and serum-binding properties of staphylococci have been traced back to the expression of surface proteins (Foster and Höök, 1998). Many of these surface proteins are anchored to the cell wall envelope by a mechanism that requires a C-terminal cell wall sorting signal (Cws) (Navarre and Schneewind, 1999).

Protein A – a cell wall-anchored surface protein

Protein A, the first identified surface protein of S. aureus, binds to the Fc portion of antibodies and causes precipitation of immunoglobulin (Ig) (Jensen, 1958; Sjödahl, 1977). The N-terminal region of protein A, including five Ig-binding domains, is displayed on the bacterial surface, whereas the C-terminal region X is buried in the cell wall envelope (Sjöquist et al., 1972a; Guss et al., 1984). Sjöquist and colleagues showed that cleavage of the staphylococcal peptidoglycan with lysostaphin, a glycyl-glycine endopeptidase that cuts cell wall cross-bridges, solubilizes protein A as a species with uniform mass (Sjöquist et al., 1972b). Digestion of the glycan strands with muramidase releases protein A as a spectrum of polypeptide fragments with linked cell wall (Sjöquist et al., 1972a, b). Molecular cloning and DNA sequencing of the spa gene revealed that protein A is synthesized as a precursor, containing an N-terminal signal peptide that directs the polypeptide into the secretory (Sec) pathway (Löfdahl et al., 1983; Uhlén et al., 1984). The predicted C-terminus of protein A harbours the 35-residue Cws that is conserved in surface proteins of Gram-positive bacteria (Fischetti et al., 1990; Schneewind et al., 1992). An LPXTG motif is followed by a hydrophobic domain and a tail of mostly positively charged residues (Fischetti et al., 1990). Although Cws resemble known membrane anchor peptides, the hydrophobic domain of protein A and other Cws cannot insert fused reporter proteins into the lipid bilayer (Schneewind et al., 1992). Truncation of the Cws charged tail causes staphylococci to secrete mutant protein A into the extracellular medium (Schneewind et al., 1992). In contrast, deletion of the LPXTG motif prevents cleavage and cell wall anchoring of protein A; mutant polypeptides fractionate with the cell wall and cytoplasmic membrane compartments (Schneewind et al., 1992). When tethered to the C-terminus of polypeptides bearing N-terminal signal peptides, the Cws promotes anchoring of hybrid polypeptides to the cell wall envelope (Pozzi et al., 1992; Schneewind et al., 1993; Medaglini et al., 1995; Navarre and Schneewind, 1996; Strauss and Götz, 1996; Piard et al., 1997).

Sorting signals and surface proteins

Ten Cws of surface proteins from different Gram-positive bacteria have been fused to the C-terminus of either truncated protein A lacking its own Cws or staphylococcal enterotoxin B (Seb) (Schneewind et al., 1993). Five Cws were functional in staphylococci and caused cell wall anchoring of hybrid proteins. Some of the non-functional Cws were altered by mutation and examined again for anchoring. In all cases examined, the mutant Cws could gain function, which required an alteration in the residue spacing between the LPXTG motif and the positively charged tail (Schneewind et al., 1993). Mutational analysis of the protein A Cws revealed that two arginines (R), positioned 31 and 32 residues downstream of the LPXTG motif, function as a signal to retain the polypeptide within the secretory pathway. It appears that positive charge is a signal for retention, as lysine, but not histidine, can substitute for arginines (Schneewind et al., 1993).

The staphylococcal databases were searched for surface protein genes using the protein A Cws as a blast search query followed by other searches with subsequently identified Cws. This procedure identified 19 surface protein genes (Table 1). Nine genes encode the well-characterized surface proteins Spa, FnbA, FnbB (fibrinonectin-binding proteins), ClfA, ClfB (fibrinogen-binding clumping factors), SdrC, SdrD, SdrE and Pls [containing serine (S) aspartate (D) repeat regions upstream of the Cws] (Uhlén et al., 1984; Flock et al., 1987; Jönsson et al., 1991; McDevitt et al., 1994; Josefsson et al., 1998a; Ní Eidhin et al., 1998). The gene encoding S. aureus collagen adhesin (Cna), a Cws-bearing adhesin for bone tissue (Patti et al., 1992), is not present in the four S. aureus genome sequences. Cna is found in staphylococcal isolates from bone and connective tissue infections (Patti et al., 1994). Ten genes encode unknown surface proteins, herein referred to as sas (S.aureussurface protein). All surface proteins appear to be exported by an N-terminal signal peptide. The Cws is invariably located at the C-terminal end and contains LPXTG sequences with variable residues at the X and T positions. One Cws has a replacement of alanine (A) for threonine (T). The X position can be occupied by acidic [glutamate (E), aspartate (D)], uncharged [alanine (A), glutamine (Q), asparagine (N)] and basic residues [lysine (K)].

Table 1. S. aureus surface protein genes (sas).
sas CodonsLigandPredicted sorting signalReference
  • a . A variant of SdrE binds bone sialoprotein ( Tung et al., 2000).

  • b

    . GenBank accession number,

spa 508IgG/vWBFLPETGEENPFIGTTVFGGLSLALGAALLAGRRREL Uhlén et al. (1984)
fnbA 1018Fibronectin/
fibrinogen
LPETGGEESTNKGMLFGGLFSILGLALLRRNKKNHKA Signäs et al. (1989)
fnbB 914Fibronectin/
fibrinogen
LPETGGEESTNNGMLFGGLFSILGLALLRRNKKNHKA Jönsson et al. (1991)
clfA 933FibrinogenLPDTGSEDEANTSLIWGLLASIGSLLLFRRKKENKDKK McDevitt et al. (1994)
clfB 913FibrinogenLPETGDKSENTNATLFGAMMALLGSLLLFRKRKQDHKEKA Ní Eidhin et al. (1998)
cna 1183CollagenLPKTGMKIITSWITWVLGLYLILRKRFNS Patti et al. (1992)
sdrC 947UnknownLPETGSENNNSNNGTLFGGLFAALGSLLSFGRRKKQNK Josefsson et al. (1998a)
sdrD 1315CalciumLPETGNENSGSNNATLFGGLFAALGSLLLFGRRKKQNK Josefsson et al. (1998a, b)
sdrE a 1166UnknownLPETGSENNGSNNATLFGGLFAALGSLLLFGRRKKQNK Josefsson et al. (1998a)
pls 1637UnknownLPDTGNDAQNNGTLFGSLFAALGGLFLVGRRRKNKNNEEKAF115379b
sasA 2261UnknownLPDTGDSIKQNGLLGGVMTLLVGLGLMKRKKKKDENDQDDSQA 
sasB 937UnknownLPDTGMSHNDDLPYAELALGAGMAFLIRRFTKKDQQTEE de Wu and Lencastre (1999);
 Komatsuzawa et al. (2000)
sasC 2186UnknownLPNTGSEGMDLPLKEFALITGAALLARRRTKN 
sasD 241UnknownLPAAGESMTSSILTASIAALLLVSGLFLAFRRRSTNK 
sasE 354UnknownLPKTGLTSVDNFISTVAFATLALLGSLSLLLFKRKESKAB042826b
sasF 637UnknownLPKAGETIKEHWLPISVIVGAMGVLMIWLSRRNKLKNKA 
sasG 1117UnknownLPKTGLESTQKGLIFSSIIGIAGLMLLARRRKN 
sasH 308UnknownLPKTGTNQSSSPEAMFVLLAGIGLIATVRRRKAS 
sasI 895UnknownLPKTGETTSSQSWWGLYALLGMLALFIPKFRKESK 
sasJ 645UnknownLPQTGEESNKDMTLPLMALLALSSIVAFVLPRKRKN 

Cell wall anchor structure

During cell wall anchoring, the Cws of protein A is cleaved between the threonine and the glycine of the LPXTG motif (Navarre and Schneewind, 1994). The cell wall anchor structure of surface proteins released from peptidoglycan via N-acetylmuramidase, N-acetylglucosaminidase, N-acetylmuramyl-l-Ala amidase, d-Ala-Gly endopeptidase (Φ11 enzyme) and lysostaphin showed that the surface proteins of S. aureus are linked via an amide bond between the carboxyl-group of threonine and the amino-group of the pentaglycine cell wall cross-bridge (Schneewind et al., 1995; Ton-That et al., 1997; Navarre et al., 1998; 1999). Anchoring occurs mostly to cross-linked tetrapeptide murein subunits [MurNac-(l-Ala-d-iGln-l-Lys(surface protein-Gly5)-d-Ala)-GlcNac]n but also to free unsubstituted pentapeptide subunits [MurNac-(l-Ala-d-iGln-l-Lys(surface protein-Gly5)-d-Ala-d-Ala)-GlcNac] (Ton-That et al., 1997; Navarre et al., 1998). The cross-linking of cell wall anchor peptides is extensive: as many as 11 cell wall peptide subunits and five glycan (disaccharide) subunits were found linked to surface proteins (Navarre et al., 1998; 1999). Using a similar strategy, the cell wall anchor structure of surface proteins in Listeria monocytogenes was determined: the C-terminal threonine is amide linked to the side-chain amino-group of m-diaminopimelic acid within cell wall peptides (Dhar et al., 2000). Although several aspects of peptidoglycan structure in Gram-positive bacteria are variable (Schleifer and Kandler, 1972), the principles of surface protein anchoring appear to be conserved (Navarre and Schneewind, 1999).

Sortase (SrtA)

A screen for S. aureus mutants that fail to cleave the protein A Cws identified srtA, for surface protein sorting A (Mazmanian et al., 1999). srtA encodes sortase, an enzyme of 206 amino acids with a putative N-terminal membrane-spanning domain and a C-terminal catalytic domain that is presumably translocated across the cytoplasmic membrane (S. K. Mazmanian and O. Schneewind, unpublished). Knock-out mutations of the srtA gene do not seem to affect growth of staphylococci on laboratory media (Mazmanian et al., 2000). Sortase mutants cannot cleave protein A at the LPXTG motif, and the precursor accumulates in the cytoplasm, membrane and cell wall fractions (missorted phenotype) (Mazmanian et al., 2000). Cws of fibronectin-binding proteins (FnbA and FnbB) and clumping factor (ClfA) also remain uncleaved, suggesting that the srtA mutation causes a defect in the anchoring of at least some of the other surface proteins (Mazmanian et al., 2000). Consistent with a more general defect in cell wall sorting is the finding that srtA mutant strains cannot display Ig, fibronectin and fibrinogen adhesins on the staphylococcal surface (Mazmanian et al., 2000). Genes that are homologous to S. aureus srtA are found in all Gram-positive bacterial genomes sequenced to date, including Actinomyces naeslundii, Bacillus anthracis, Bacillus subtilis, Clostridium acetabutylicum, Corynebacterium diphtheriae, Enterococcus faecalis, Listeria monocytogenes, Streptococcus mutans, Streptococcus pneumoniae and Streptococcus pyogenes (Mazmanian et al., 1999). The genomes of Gram-positive organisms encode at least two, sometimes even more than two, sortase genes (Pallen et al., 2001). The identification of two sortase genes in Methanobacterium thermoautotrophicum represents a curiosity, as this organism synthesizes pseudopeptidoglycan, i.e. N-acetylglucosamine-(β1–3)-N-acetyltalosaminurate polymer, instead of peptidoglycan (Kandler and König, 1998; Pallen et al., 2001). It is conceivable that sortase may link proteins to the amino-group of lysine, as the side-chain of this residue is engaged in cross-linking with γ-glutamyl of neighbouring peptide subunits (Kandler and König, 1998). However, no genes encoding a surface protein with a C-terminal Cws have been identified thus far in Methanobacterium (Pallen et al., 2001).

The ability of srtA mutant staphylococci to cause an infection was assessed using the murine renal abscess model. When compared with the human clinical isolate S. aureus Newman, isogenic srtA knock-out mutants displayed a 2- to 3-log reduction in the number of viable staphylococci recovered from renal abscesses (Mazmanian et al., 2000). As a model for acute infection, the number of sortase mutant staphylococci required to produce a lethal infection after intraperitoneal injection of mice was measured. Compared with the wild-type human isolate Newman, the LD50 of srtA mutant staphylococci is reduced by 1.5 logs (Mazmanian et al., 2000).

The role of srtA in anchoring surface proteins to the cell wall envelope appears to be conserved in Gram-positive pathogens. Knock-out mutations of the srtA gene in Streptococcus gordonii also interfere with the cell wall anchoring and display of surface proteins, which is accompanied by a loss of bacterial adhesive properties (Bolken et al., 2001). Actinomyces species elaborate fimbriae that are composed of Cws-bearing subunit proteins (Yeung and Cisar, 1990). Knock-out mutations of the Actinomyces srtA homologue abolish C-terminal processing of fimbrial proteins at the LPXTG motif (Yeung et al., 1998). It is not yet clear whether the fimbrial proteins are linked to the cell wall peptidoglycan and/or polymerized to form a proteinaceous multisubunit filament (Yeung et al., 1998; Pallen et al., 2001).

Sortase-catalysed cleavage at the LPXTG motif

Treatment of staphylococci with the strong nucleophile hydroxylamine releases surface protein with C-terminal threonine hydroxamate into the extracellular medium (Ton-That et al., 1999). This observation suggests that sortase captures cleaved surface protein as acyl-enzyme intermediates. A model was proposed in which cysteine 184 (C184) of sortase performs a nucleophilic attack on the carbonyl carbon at the scissile peptide bond between threonine and glycine, followed by the formation of a thioester with the carboxyl-group of threonine (Ton-That et al., 1999). During this reaction, the amino-group of the cleaved LPXTG is released, and the Cws is subsequently degraded. This model is supported by several observations. (i) Methyl-methanethiosulphonates, reagents that form disulphide with sulphydryl, act as inhibitors of sortase (Ton-That and Schneewind, 1999). The inhibited enzyme is regenerated by treatment with the reducing reagent dithiothreitol (DTT) (Ton-That et al., 1999). (ii) Mutational replacement of C184 with alanine abolishes the enzymatic activity of sortase (Ton-That et al., 1999). (iii) C184 is conserved within sortase genes (Mazmanian et al., 1999). (iv) Thioester-linked acyl-enzyme intermediates are sensitive to hydroxylaminolysis (Ton-That et al., 2000). (v) In the three-dimensional structure of sortase, C184 and several other conserved residues cluster into a surface-exposed pocket that seems to represent the enzyme active site (Ilangovan et al., 2001).

To obtain a soluble sortase and examine its properties, the N-terminal membrane anchor segment of SrtA (residues 2–25) was replaced with a six-histidine tag (SrtAΔN) (Ton-That et al., 1999). In the absence of the peptidoglycan substrate, SrtAΔN catalyses peptide bond hydrolysis and cleaves LPETG between the threonine and the glycine (Ton-That et al., 2000). Glycine, NH2-Gly2 and NH2-Gly3 can be used as substitutes for peptidoglycan substrate and increase the rate of sortase-mediated cleavage at LPETG peptides (Ton-That et al., 2000). In the presence of NH2-Gly3, sortase catalyses only the transpeptidation reaction of cell wall sorting, forming LPET-Gly3 and NH2-G, but there is no hydrolysis at LPETG peptides (Ton-That et al., 2000).

Nuclear magnetic resonance (NMR) spectroscopy revealed the three-dimensional structure of the catalytically active SrtAΔ2−59 as a β-barrel fold with eight β-strands aligned in a unique antiparallel and parallel manner (Ilangovan et al., 2001). The surface of the barrel is decorated with two short helices and two large loops, which define the enzyme active site. The active-site pocket is open at two sides, forming an elongated tunnel with a hydrophobic surface that can accommodate an unfolded polypeptide chain (Ilangovan et al. 2001). This is in agreement with the presumed scanning function of sortase, examining newly translocated polypeptides for LPXTG motif sequences. Once the C-terminal Cws arrests polypeptide translocation, sortase cuts at the LPXTG motif and transfers surface protein to the cell wall.

Peptidoglycan substrate of the sorting reaction

During bacterial cell wall synthesis, a soluble cytoplasmic peptidoglycan precursor (UDP-MurNac-l-Ala-d-iGln-l-Lys-d-Ala-d-Ala; Park's nucleotide) is linked to undecaprenolpyrophosphate, generating lipid I (Matsuhashi, 1994). The membrane-bound intermediate is modified further by the addition of GlcNac and pentaglycine [undecaprenolpyrophosphate-MurNac(-l-Ala-d-iGln-(NH2-Gly5)l-Lys-d-Ala-d-Ala)-(β1–4)-GlcNac; lipid II] and translocated across the cytoplasmic membrane (Matsuhashi et al., 1965). Lipid II serves as a substrate for the transglycosylation reaction, polymerizing the glycan strands of the bacterial cell wall to yield the repeating disaccharide (MurNac-GlcNac)n (Nakagawa et al., 1984). Cell wall pentapeptides (l-Ala-d-iGln-(NH2-Gly5)l-Lys-d-Ala-d-Ala) of nascent peptidoglycan strands are cross-linked via the transpeptidation reaction, thereby generating a three-dimensional cell wall network (Tipper and Strominger, 1965). Penicillin is an inhibitor of the transpeptidation reaction, but does not interfere with the transglycosylation reaction (Strominger et al., 1967). Vancomycin binds to the d-Ala-d-Ala moiety of lipid II (Bugg et al., 1991) and prevents both transglycosylase and transpeptidase reactions (Tipper and Strominger, 1965), whereas moenomycin is an inhibitor of transglycosylation alone (van Heijenoort et al., 1987). The addition of vancomycin leads to a steady accumulation of surface protein precursor with uncleaved Cws, indicating that this compound causes a reduction in the sorting reaction (Ton-That and Schneewind, 1999). A similar, albeit weaker, effect is observed when moenomycin is added to staphylococcal cultures (Ton-That and Schneewind, 1999). In contrast, penicillin G does not alter the rate of cell wall sorting. These results suggests that sortase may use lipid II as a substrate for surface protein anchoring. This is corroborated by the finding that staphylococcal protoplasts, in which the mature assembled cell wall has been removed by digestion with muralytic enzyme, catalyse cleavage of surface proteins at the LPXTG motif similar to bacteria with an intact cell wall (Ton-That and Schneewind, 1999).

Does staphylococcal sortase recognize mutationally altered cell wall cross-bridges? The FemA, FemB (factor essential for methicillin resistance) and FmhB proteins (Fem homologue) are thought to catalyse the addition of Gly to the ε-amino of l-Lys within lipid I (Berger-Bächi, 1994; Rohrer et al., 1999). Mutant staphylococci lacking the FemB protein synthesize NH2-Gly3 but no NH2-Gly5 cross-bridges, whereas femA and femX (fmbH) mutant strains synthesize NH2-Gly1 or NH2-Gly1/NH2l-Lys cross-bridges respectively (Kopp et al., 1996). The femA, femB, femAB and femAX (femA, fmbH) mutant staphylococci display a reduction in the rate of surface protein anchoring, consistent with the notion that NH2-Gly5 is the preferred substrate of the sorting reaction (Ton-That et al., 1998). Nonetheless, surface protein was found linked to NH2-Gly3, NH2-Gly1 and NH2-Gly4Ser1 cross-bridges, but not to the ε-amino group of l-Lys (Ton-That et al., 1998). It appears that sortase recognizes some, but not all, cross-bridges as substrates for the sorting reaction. This notion is also corroborated by the observation that S. carnosus cells, engineered to synthesize Gly3/Ser2 cross-bridges, are capable of anchoring surface protein to the cell wall peptidoglycan (Strauss et al., 1998).

A second sortase gene (srtB)

A second sortase gene, srtB, was identified by blast searches using the srtA gene as a query (Pallen et al., 2001). All S. aureus strains examined thus far harbour both srtA and srtB genes. What can be the role of two sortase enzymes in staphylococci? Replacement of the srtB gene of S. aureus Newman with the ermC marker does not cause a defect in the cell wall anchoring of protein A, FnbA, FnbB or ClfA (S. K. Mazmanian and O. Schneewind, unpublished). Nevertheless, it is possible that srtB mutant staphylococci display a sorting defect for other surface proteins. Clearly, SrtA and SrtB cannot fulfil redundant functions, but it is conceivable that the two sortases catalyse similar reactions using different surface protein substrates.

Which genes act on the sorting pathway?

The staphylococcal cell wall is composed of peptidoglycan, a covalently linked macromolecule spanning 40 nm in diameter (Giesbrecht et al., 1998). This envelope structure is impermeable for large and small (≥ 10 kDa) polypeptides (Schneewind et al., 1992). We presume that the secretion of proteins into the extracellular medium requires genes whose sole function may be to promote polypeptide translocation across the cell wall. Such cell wall translocation would obviously have to be co-ordinated with the translocation of proteins across the plasma membrane. Protein secretion has not been studied in S. aureus or other Gram-positive pathogens; however, these organisms harbour several genes implicated in protein secretion in Escherichia coli.

The protein translocation pore of E. coli is composed of three membrane proteins, SecYEG, which can accept substrates in two ways (Duong et al., 1997; Pohlschröder et al., 1997). Signal peptide-bearing precursor proteins are maintained in a secretion-competent state by binding to chaperones, for example SecB (Randall, 1992). B. subtilis makes do without SecB, and other chaperones presumably function as substitutes (Kunst et al., 1997). Translocation substrate is transferred to SecA, an ATPase that undergoes conformational rearrangements upon interacting with the secretion machinery, a process that is thought to push precursor proteins through the translocation pore (Economou and Wickner, 1994). Once polypeptides have been translocated, signal peptidase (lepB– leader peptidase in E. coli) removes the signal peptide from the precursor, and mature protein is released into the periplasmic space (Dalbey and Wickner, 1985). A second mechanism of protein secretion involves the co-translational translocation of membrane proteins (Ulbrandt et al., 1997). The signal recognition particle (SRP), Ffh and 4.5S RNA (ffs) in E. coli, binds to signal peptide bearing nascent polypeptides, an interaction that is thought to stall ribosomal translation (Poritz et al., 1990). Once the SRP complex is bound to its receptor (FtsY in E. coli) and the ribosome has docked on the translocation pore, translation resumes and presumably provides the force to translocate polypeptides across the membrane (Miller et al., 1994). Three membrane proteins, SecD, SecF and YajC, associate with the SecYEG pore and appear to regulate SecA-dependent translocation activity (Duong and Wickner, 1997).

We have searched staphylococcal genome databases for the presence of secretion genes (secABDEFGY, ffh, ftsY and lepB) (Table 2). The S. aureus genome encodes secAYEG and yajC similar to E. coli and B. subtilis (Blattner et al., 1997; Kunst et al., 1997). A secB gene could not be found; however, a second set of secretion genes, secA-2 and secY-2, was identified. As observed for B. subtilis, the S. aureus genome encodes for a secDF fusion gene but not for single secD and secF genes (Bolhuis et al., 1998). Two signal peptidase genes (spsA and spsB) are present in the S. aureus chromosome (Cregg et al., 1996), whereas ffh and ftsY are present in single copy only. The duplicate secretion genes (secA2 and secY2) are located in a single large operon inserted into the genome between sasA and sasB. What can be the function of additional secretion genes in the staphylococcal genome? One possibilty is the assembly of two secretion pathways that transport polypeptides to distinct locations, perhaps membrane and extracellular medium? Many other models can of course be proposed. If one assumes that two secretion pathways do indeed exist in S. aureus, additional questions come to mind. Do surface proteins travel only one or both secretion pathways? Can any one of the sec genes be removed without a loss of cell wall sorting? The list of ‘Gedanken Experimente’ could go on, but one can say only this with certainty – very little is known about protein secretion and cell wall sorting pathways in S. aureus.

Table 2. Secretion genes in S. aureus, B. subtilis and E. coli.
Gene S. aureus B. subtilis E. coli
secA secA-1, secA-2 secA secA
secB secB
secD secDF secDF secD
secE secE secE secE
secF secF
secG secG secG secG
secY secY-1, secY-2 secY secY
yajC yajC yajC yajC
ffh ffh ffh ffh
ftsY ftsY ftsY ftsY
lepB spsA, spsB sipSTUVW lepB

Conclusions and prospects

Based on the evidence described above, cell wall sorting of surface proteins can be divided into five steps (Fig. 1). (i) Export. Precursor proteins with an N-terminal signal peptide are initiated into the secretory (Sec) pathway, and the signal peptide is removed. (ii) Retention. The C-terminal Cws retains polypeptides within the secretory pathway. (iii) Cleavage. Sortase cleaves between the threonine and the glycine of the LPXTG motif, resulting in the formation of a thioester enzyme intermediate. (iv) Linkage. Nucleophilic attack of the free amino-group of lipid II at the thioester bond resolves the acyl-enzyme intermediate, generating the amide bond between surface protein and the pentaglycine cross-bridge. (v) Cell wall incorporation. Lipid-linked surface protein is incorporated into the cell wall via the transglycosylation and transpeptidation reactions of bacterial cell wall synthesis. The sorting pathway appears to be composed of Sec machinery, signal peptidase and sortase, but will probably require several additional genes that promote surface protein retention and translocation across the cell wall envelope. The presence of multiple sets of secretion, signal peptidase and sortase genes in the chromosome of S. aureus suggests that this microorganism may have evolved more than one pathway for surface protein transport. Some 20 different polypeptides seem to be anchored by the sortase-catalysed pathway and play important roles during the pathogenesis of S. aureus infections.

Figure 1.

Surface protein anchoring in Staphylococcus aureus. (i) Export. Precursor proteins with an N-terminal signal peptide are initiated into the secretory (Sec) pathway, and the signal peptide is removed. (ii) Retention. The C-terminal sorting signal retains polypeptides within the secretory pathway. (iii) Cleavage. Sortase cleaves between the threonine and the glycine of the LPXTG motif, resulting in the formation of a thioester enzyme intermediate. (iv) Linkage. Nucleophilic attack of the free amino group of lipid II at the thioester bond resolves the acyl-enzyme intermediate, synthesizing the amide bond between surface proteins and the pentaglycine cross-bridge and regenerating the active-site sulphydryl. (v) Cell wall incorporation. Lipid-linked surface protein is first incorporated into the cell wall via the transglycosylation reaction. The murein pentapeptide subunit with attached surface protein is then cross-linked to other cell wall peptides, generating the mature murein tetrapeptide.

Acknowledgements

S.K.M. is supported by the Predoctoral Training Program in Genetics (GM07104), and H.T.-T. by the Microbial Pathogenesis Training Grant (AI07323). Work in O.S. laboratory is supported by a grant from the NIH-NIAID, Infectious Disease Branch AI33987. We apologize to the many authors whose work could not be adequately described because of space constraints.

Ancillary