Anchoring of Surface Proteins to the Cell Wall of Staphylococcus aureus

Surface proteins of Staphylococcus aureus are anchored to the cell wall envelope by a mechanism requiring a C-terminal sorting signal with an LPXTG motif. Sortase A cleaves surface proteins between the threonine (T) and the glycine (G) residues of the LPXTG motif and catalyzes the formation of an amide bond between the carboxyl group of threonine at the C-terminal end of polypeptides and the amino group of pentaglycine cross-bridges of cell wall peptidoglycan. Previous work showed that Cys184 and His120 of sortase A are absolutely essential for catalysis; however an active site thiolateimidazolium ion pair may not be formed. The three-dimensional crystal structure of sortase A revealed that Arg197 is located in close proximity to both the active site Cys184 and the scissile peptide bond between threonine and glycine. We show here that substitution of Arg197 with alanine, lysine, or histidine severely reduced sortase A function both in vivo and in vitro, whereas Asn98, which had earlier been implicated in hydrogen bonding to His120, was found to be dispensable for catalysis. As the structural proximity of Arg197 and Cys184 is conserved in sortase enzymes and as ionization of the Cys184 sulfhydryl group seems required for sortase activity, we propose that Arg197 may function as a base, facilitating thiolate formation during sortase-mediated cleavage and transpeptidation reactions.

The cell wall envelope of Gram-positive microbes functions as a surface organelle that allows bacterial pathogens to interact with their environment (1,2). Gram-positive bacteria use cell wall-anchored surface proteins to invade, colonize, and replicate within their hosts (3)(4)(5), as polypeptides can promote bacterial attachment to specific host molecules or prevent escape from host immune responses (6,7). Sortase A (SrtA) 1 provides one mechanism for the attachment of proteins to the cell wall envelope of Gram-positive bacteria (8 -11). The enzyme recognizes and processes cell wall sorting signals of surface proteins (12,13), C-terminal peptides with an LPXTG motif, a hydrophobic domain, and a positively charged tail (14,15). Sortase A cleaves surface protein substrates between the threonine (T) and glycine (G) residues of the LPXTG motif (16). The C-terminal end of cleaved surface proteins is then tethered to the cell wall envelope, i.e. the pentaglycine cell wall crossbridges of staphylococci. Lipid II, GlcNAc-(␤1-4)-MurNAc-[L-Ala-D-iGln-L-Lys-(NH 2 -Gly 5 )-D-Ala-D-Ala]-P-P-undecaprenol, is the peptidoglycan substrate of Staphylococcus aureus sortase A (17). The product of the sorting reaction, surface protein linked to lipid II, is incorporated into the cell wall envelope via the transglycosylation and transpeptidation reactions of cell wall biosynthesis (18 -20).
The side chain thiol of Cys 184 is presumed to attack the carbonyl group of the scissile peptide bond, thereby generating a thioacyl intermediate (12). Supporting this model are three different observations. First, sortase A can be inhibited both in vivo and in vitro with thiolate reagents such as methylmethane thiosulfonate (e.g. MTSET) or para-hydroxymercuribenzoic acid, but not with a sulfhydryl reagent such as iodoacetate or iodoacetamide (12,18). Second, surface protein can be released from staphylococci with C-terminal threonine hydroxamate after treatment of bacteria with hydroxylamine (12), a strong nucleophile that is known to attack thioacyl intermediates (27). Third, sortase A-mediated hydroxylaminolysis or cell wall anchoring of surface proteins in vivo requires Cys 184 as well as His 120 , another charged residue that is conserved in all sortases (26). The sortase A thioacyl intermediate is thought to be resolved by the nucleophilic attack of the amine of the pentaglycine moiety within staphylococcal lipid II (17,19).
Cys 184 and His 120 are absolutely required for sortase A-mediated LPXTG peptide cleavage or transpeptidation in vitro (13,26). Analysis of the three-dimensional structure of sortase A with or without bound substrate revealed that His 120 is positioned at a distance (7 Å) from Cys 184 (24). This measurement is not consistent with the formation of a thiolate-imidazolium ion bond as is known to occur in papain and papain-like cysteine proteases (28 -30). Experimental determination of the pK a for Cys 184 thiol (9.4) and for the His 120 imidazolium (7.0) further corroborates the notion that sortase A catalysis may occur by a mechanism that does not involve a thiolate-imidazolium ion bond (31). The three-dimensional structure of sor-tase A bound to the LPETG substrate revealed the presence of Arg 197 on the ␤8-strand, in close proximity to Cys 184 and to the scissile peptide bond (Fig. 1B). We show here that substitution of Arg 197 with alanine, lysine or histidine severely reduced sortase A function, whereas Asn 98 was found to be dispensable for catalysis. As the structural proximity of Arg 197 and Cys 184 is conserved in sortase enzymes (32,33) and as ionization of the Cys 184 sulfhydryl group appears to be required for sortase activity (12,26), we propose that Arg 197 may act as a base, facilitating thiolate formation during sortase A cleavage and transpeptidation reactions.

EXPERIMENTAL PROCEDURES
Bacterial Strains and Plasmids-Plasmid pHTT27, with an in-frame insertion of SrtA⌬ N59 coding sequence into the six histidyl tag expression vector pQE30 (Qiagen), has been previously described (22,26). pHTT27 and its mutant derivatives were transformed into E. coli XL-1 Blue and selected on Luria agar containing 100 g/ml ampicillin. The sortase A mutant S. aureus strains SKM1 (pGL4), SKM1 (pGL4, pSrtA), and SKM1 (pGL4, pOS1) have also been previously described (9). Mutant derivatives of the pSrtA plasmid were transformed into SKM1 (pGL4) and selected on tryptic soy agar supplemented with 10 g/ml chloramphenicol and 2.5 g/ml tetracycline. pGL4, a pT181-type shuttle vector, provides for the expression of Seb-Spa 490 -524 , a model surface protein substrate for analysis of the S. aureus cell wall sorting pathway (8).
Pulse-chase Experiments-Staphylococcal cultures were grown overnight in tryptic soy broth media supplemented with 10 g/ml chloramphenicol and 2.5 g/ml tetracycline at 37°C. Overnight cultures were diluted into fresh media and grown until OD 600 reached 0.5. Cells were harvested by centrifugation, washed, and suspended in minimal medium lacking methionine and cysteine (14). Cells were labeled with 100 Ci of [ 35 S]Promix (Amersham Biosciences) for 2 min. After labeling, 50 l of chase solution were added (100 mg/ml casamino acids, 10 mg/ml methionine) and at timed intervals (0, 2, and 10 min after the chase), 250 l of cells were removed and transferred into an Eppendorf tube containing an equal volume of 15% trichloroacetic acid to quench all further processing of surface proteins. After 30 min of incubation on ice, cells and precipitated proteins were collected by centrifugation at 14,000 ϫ g for 10 min, washed in ice-cold acetone, sedimented again by centrifugation at 14,000 ϫ g for 10 min, and then dried. Samples were suspended in 1 ml of 0.5 M Tris-HCl, pH 6.8 and the peptidoglycan digested by adding either 150 g mutanolysin or 100 g of lysostaphin with incubation for 4 hours at 37°C and intermittent mixing of samples. Digests were precipitated by the addition of 7.5% trichloroacetic acid and incubation on ice for 30 min. The precipitate was collected by FIG. 1. The three-dimensional structure and active site of sortase. A, ribbon drawing of the structure of SrtA ⌬N59 bound to its peptide substrate (LPETG, shown in orange). ␤-Strands and ␣-helices are colored blue and pink, respectively. Carbon, sulfur, and nitrogen atoms are shown in gray, yellow, and blue, respectively. The active site Cys 184 is positioned at the end of ␤7, whereas Arg 197 is located on the ␤8-strand and His 120 at the end of the ␤6-strand. B, expanded view of the active site of sortase with bound peptide substrate. Cys 184 and Arg 197 confront the scissile peptide bond from opposing sides, while His 120 is positioned at a 7-Å distance from Cys 184 . The three-dimensional views of the sortase A enzyme with its substrate were generated from the x-ray crystallography data provided by Zong et al. (24). centrifugation at 14,000 ϫ g for 10 min, washed in ice-cold acetone, and then sedimented by centrifugation at 14,000 ϫ g for 10 min and dried. Samples were solubilized by boiling in 50 l of 0.5 M Tris-HCl, 4% SDS, pH 8.0. 40 l aliquots were transferred to 1 ml of RIPA buffer containing 1 l of rabbit ␣-Seb antibody. Antigen-antibody complexes were captured on 100 l of preswollen protein A CL 4B-Sepharose, washed five times with RIPA buffer, and solubilized by boiling in sample buffer. Immunoprecipitates were separated on 12% SDS-PAGE, dried, and analyzed by phosphorimager.
Measuring Sortase Expression in Vivo-A 1-ml aliquot of staphylococci was grown in tryptic soy broth until an OD 600 of 0.5. Bacteria and soluble proteins were sedimented by the addition of trichloroacetic acid to 7.5% and incubated 30 min on ice. The precipitate was collected by centrifugation, washed with ice-cold acetone, and again centrifuged. The sediment was suspended in sample buffer, boiled, and subjected to SDS-PAGE followed by immunoblotting with a rabbit ␣-sortase antibody.
Hydroxylaminolysis of Surface Proteins in Vivo-Cells were grown and pulse-labeled for 1 min and chased for 5 min as described above, however this reaction was performed either in the presence or absence of 0.1 M NH 2 OH. A 0.5-ml aliquot was centrifuged at 15,000 ϫ g for 5 min, and the supernatant was precipitated with 0.5 ml of 15% trichloroacetic acid. The remaining 0.5-ml culture aliquot was also precipitated but then suspended in 1 ml of 0.5 M Tris-HCl (pH 7.5) containing 100 g of lysostaphin to digest the peptidoglycan for 1 h at 37°C. All four samples were treated with 7.5% trichloroacetic acid to precipitate soluble proteins. The sediment was washed in acetone, dried, and then boiled in 50 l of 0.5 M Tris-HCl, 4% SDS, pH 8.0. 40-l aliquots were immunoprecipitated with ␣-Seb as described above, subjected to SDS-PAGE, and then analyzed by phosphorimager.
Purification and Characterization of Recombinant Sortases-E. coli XL-1 blue harboring plasmids that provide for the expression of SrtA ⌬N59 and wild-type or mutant sortases, were grown at 37°C in Luria broth supplemented with 100 g/ml ampicillin until an OD 600 of 0.7. Gene expression was the induced by the addition of 1 mM isopropyl-1-thio-␤-D-galactopyranoside, and cultures were grown for an additional 2 hours. Cells were collected by centrifugation at 6000 ϫ g, suspended in 30 ml of C buffer (50 mM Tris-HCl, 150 mM NaCl, 10% glycerol, pH 7.5), and lysed in a French pressure cell at 14,000 psi. The extract was centrifuged at 29,000 ϫ g for 30 min, and the supernatant applied to a 1.5-ml Ni-NTA resin pre-equilibrated with C buffer. The column was washed with 40 ml of C buffer, and bound protein was eluted in 4 ml of C buffer containing 0.5 M imidazole. To characterize purified enzymes, the average mass of 1 g of wild-type or mutant sortase was determined in an ion trap mass spectrometry instrument (Agilent 1100 LC/MSD XCT). The average observed mass of all enzymes was within the error rate (0.01%) of the calculated average masses. To exclude the possibility that the introduced mutations produced structurally unstable enzymes, mutant as well as wild-type sortases were subjected to circular dichroism. Buffer was changed to 10 mM phosphate, pH 7.5, and CD spectra from 260 to 190 nm were recorded with an AVIV 202 Circular Dichroism Spectrometer.
In Vitro Analysis of Sortase Enzymes-Sortase activity was assayed in buffer R (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 5 mM CaCl 2 ) at 37°C for 30 min. The Abz-LPETG-Dnp peptide was dissolved in dimethyl sulfoxide and added to the reaction at a final concentration of 5 M. Purified SrtA⌬ N59 or its mutant variants were added to the reaction at a final concentration of 10 M. Reactions were quenched by boiling samples for 5 min and peptide cleavage was monitored by fluorescence at 420 nm after excitation at 320 nm. The mean and standard deviation of three independent measurements are reported. Initial velocity assays were carried out under the same conditions and reaction rates were determined by monitoring fluorescence over 100 s or 2-h time periods. The average of three independent experiments is reported.
Kinetic parameters of wild-type and mutant enzymes were measured in the conditions above detailed, varying the concentration of Abz-LPETG-Dpn substrate in a range from 5 to 160 M and at a constant concentration sortase of 5 M. The fluorescence increment detected was used to calculate the initial velocity as previously reported (13). Average and error of three independent measures were used to obtain V max and K m values from Lineweaver Burk plots (34).
HPLC Purification of Cleaved Products-A reaction containing 10 M Abz-LPETG-Dnp and 15 M recombinant enzyme in a final volume of 1 ml of buffer R (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 5 mM CaCl 2 ) was incubated either in the presence or absence of 5 mM of triglycine at 37°C for 16 h. The reaction was stopped by centrifugation in a Centricon-10 unit (Millipore) at 7,500 ϫ g to separate enzyme from substrate and products. The filtrate was subjected to RP-HPLC separation on C-18 column (2 ϫ 250-mm, C18 Hypersil, Keystone Scientific). Elution of cleaved products was monitored at 215 nm, and 1-min fractions were collected. Vacuum dried fractions were stored at 4°C for MALDI-TOF mass spectrometry and tandem mass spectrometry analysis (ABI 4700, operated in reflection mode) to verify the structure of transpeptidation products.

RESULTS
Arg 197 Substitutions in Sortase A Interfere with S. aureus Surface Protein Anchoring in Vivo-Plasmids encoding either wild-type srtA or srtA variants with nucleotide substitutions in codon 197 were generated with site-directed mutagenesis. Four plasmids were transformed into S. aureus strain SKM1 [⌬(srtA)]: pSrtA, encoding the wild-type srtA gene; pLM-R197A, encoding an R197A substitution; pLM-R197K; encoding an R197K substitution; and pLM-R197H, encoding an R197H substitution. As a control for adequate expression of sortase genes, cell extracts from equal numbers of staphylococci were analyzed for sortase expression by immunoblotting with ␣-SrtA (Fig. 2). To measure surface protein anchoring, staphylococci were pulse-labeled with [ 35 S]methionine/cysteine for 2 min, followed by the addition of an excess of non-radioactive methionine and cysteine to quench the incorporation of all radio-labeled amino acids into polypeptides. During the pulse (0 min) or 2 and 10 min after the addition of the chase, all surface protein processing was quenched by the addition of 7.5% ice-cold trichloroacetic acid. The cell wall envelope of staphylococci was digested with lysostaphin, a glycyl-glycine endopeptidase that cuts the anchor structure of surface proteins at the pentaglycine cell wall cross-bridge (35), and Seb-Spa 490 -524 surface protein was immunoprecipitated using polyclonal ␣-Seb and protein A-Sepharose, followed by SDS-PAGE and phosphorimager analysis (15) (Fig. 2).
When analyzed for S. aureus strain SKM1 harboring pSrtA (9), Seb-Spa 490 -524 was synthesized as a P1 precursor carrying both an N-terminal signal peptide and a C-terminal sorting signal. P1 precrusor was rapidly cleaved to generate first the P2 precursor, lacking the signal peptide but still bearing the sorting signal, and then the mature cell wall anchored product without signal peptide and sorting signal ( Fig. 2A). Ten minutes after the addition of the chase, essentially all Seb-Spa 490 -524 was processed to the mature anchored surface protein species. Treatment of the envelope of S. aureus SKM1 (pSrtA) with muramidase, an enzyme that cuts MurNAc-(␤1-4)-GlcNAc glycosidic bonds of peptidoglycan (36), released Seb-Spa 490 -524 as a spectrum of polypeptides with linked cell wall fragments of variable mass ( Fig. 2A). Transformation of S. aureus strain SKM1 [⌬(srtA)] with empty vector pOS1 did not restore the sorting defect of the parent strain, as pulse-labeled staphylococci accumulated P2 precursor. Furthermore, lysostaphin and muramidase treatments released surface protein P2 precursor, indicating that this species had not been linked to the cell wall envelope (Fig. 2A).
Transformation of SKM1 with pLM-R197A did not restore Seb-Spa 490 -524 surface protein anchoring to wild-type levels and led to the accumulation of P2 precursor ( Fig. 2A). However, in contrast to the srtA mutant strain, expression of sortase with the R197A substitution allowed about 50% P2 precursor processing to generate mature, anchored species. This surface protein species is indeed covalently linked to the cell wall envelope, as the mobility of cell wall-anchored Seb-Spa 490 -524 released by lysostaphin treatment differed from that of the muramidase-solubilized counterpart ( Fig. 2A). We wondered whether the nature of the amino acid side chain at position 197 impacts on surface protein anchoring and tested both lysine and histidine residues at this position. We reasoned that substitutions of arginine with either lysine or histidine represent conservative replacements with two other basic residues. Furthermore, the histidine replacement interrogates whether the active site cysteine of sortase can be impacted by a residue that can engage in thiolate-imidazolium ionization. Sortase with R197K catalyzed very little if any in vivo cell wall anchoring of surface proteins, indicating that the conservative replacement between charged residues caused a major reduction in activity. In contrast, the isogenic variant with an R197H substitution anchored all pulse-labeled Seb-Spa 490 -524 as evidenced by the complete processing to mature, anchored surface protein species and by the mobility differences on SDS-PAGE after lysostaphin or muramidase treatment of the cell wall ( Fig. 2A).
Cys 184 and Arg 197 are positioned within the ␤7and ␤8strands of sortase A. To test whether reciprocal substitution of both residues, C184R and R197C, generated a functional enzyme, surface protein processing was measured in vivo. Switching the position of Cys 184 and Arg 197 within the active site abolished all sortase processing in vivo, indicating that the overall architecture of the active site as well as other residues surrounding the active site of sortase A may be essential for enzyme activity. The NMR structure of sortase A suggested that Asn 98 may be positioned in sufficient proximity to His 120 to allow for the formation of a hydrogen bond between the two residues (21). To determine whether Asn 98 plays a role in catalysis, we analyzed sortase variants with amino acid substitutions after transforming the plasmids pLM-N98A and pLM-98Q into S. aureus strain SKM1. Substitution of Asn 98 with alanine or glutamine did not affect surface protein anchoring, indicating that Asn 98 of sortase is dispensable for in vivo catalysis of sortase A ( Fig. 2A).
Arg 197 Substitutions Interfere with in Vivo Thioacyl Formation of Sortase A-During the first step of surface protein anchoring, Cys 184 of the active site of sortase cleaves its substrate via nucleophilic attack at the scissile peptide bond, resulting in the formation of a thioacyl intermediate. To determine whether Arg 197 is required for the first step of sortase catalysis, the ability of sortase to cleave substrate and form thioacyl intermediates in vivo was measured (Fig. 3). Treatment of staphylococci with hydroxylamine (NH 2 OH), a strong nucleophile that attacks thioacyl intermediates (27), releases surface proteins carrying a C-terminal threonine hydroxamate into the culture medium (12). S. aureus strain SKM1 was subjected to pulse-labeling studies in the presence or absence of NH 2 OH. By choosing a short period of hydroxylaminolysis (5 min hydroxylaminolysis compared with 10 min during the pulse-chase analysis) we hoped to reveal differences in thioacyl formation. Surface protein processing was quenched and proteins precipitated with ice-cold trichloroacetic acid. Samples were then either digested with lysostaphin or left untreated and then analyzed by immunoprecipitation, SDS-PAGE, and phosphorimager analysis.
As expected, treatment of S. aureus SKM1 (pSrtA) with hydroxylamine caused the release of surface protein hydroxamate into the medium as evidence by the immunoprecipitation of Seb-Spa 490 -524 from pulse-labeled cultures even without lysostaphin cleavage of bacterial peptidoglycan (Fig. 3). As a control, transformation of SKM1 with empty vector plasmid (pOS1) did not restore the ability of staphylococci to catalyze surface protein hydroxylaminolysis (Fig. 3). Substitution of Arg 197 of sortase with alanine or lysine also failed to restore hydroxylaminolysis, suggesting that the sorting defect of both variant enzymes may be due to a reduction in thioacyl formation. However, substitution of Arg 197 with histidine at least in part restored the ability of staphylococci to catalyze hydroxylaminolysis, while reciprocal substitutions of C184R and R197C failed to produce active enzyme. In contrast to Arg 197 , amino acid substitutions at Asn 98 did not affect the ability of sortase A to catalyze hydroxylaminolysis of surface proteins (Fig. 3).
Amino Acid Substitutions of Arg 197 Inhibit Sortase Activity in Vitro-Purified sortase A catalyzes cleavage of LPXTG polypeptide substrates in vitro (12). To assess the catalytic properties of sortase A enzymes carrying amino acid substitutions at Arg 197 , both wild-type and mutant enzymes were purified from the cytoplasm of E. coli by affinity chromatography on Ni-NTA Sepharose. The purity of enzyme preparations was assessed on Coomassie Brilliant Blue-stained SDS-PAGE and ion trap mass spectrometry (Fig. 4A). To determine if the introduced amino acid substitutions affected the secondary structure folding of the mutants, each polypeptide was subjected to circular dichroism spectroscopy, which revealed near identical spectra for wild-type and mutant enzymes (data not shown). Thus, amino acid substitutions at Arg 197 or Asn 98 do not seem to affect the abundance, purification or the secondary structure folding of sortase enzymes.
The fluorescence of Abz-LPETG-Dnp is quenched due to the close proximity of the Abz fluorophore and the Dnp quencher. However, sortase-mediated cleavage of the FRET substrate separates the two functional groups and leads to an increase in fluorescence (26). Previous work showed that in the absence of peptidoglycan substrate, purified SrtA ⌬N59 cleaves Abz-LPETG-Dnp between the threonine and the glycine residues (13). As expected, incubation of SrtA ⌬N59 with Abz-LPETG-Dnp substrate for a 30-min time interval led to a large increase in fluorescence that did not occur when the substrate was incubated in the absence of sortase A (Fig. 4B). All enzymes carrying amino acid substitutions at Arg 197 , i.e. R197A, R197K, and R197H, displayed only small amounts of sortase A activity in cleaving Abz-LPETG-Dnp when measured over a 30-min interval. In contrast, sortase enzymes with substitutions at Asn 98 (N98A or N98Q) showed little or no reduction in the ability to cleave the polypeptide substrate in vitro (Fig. 4B).
To obtain a better resolution of sortase A catalysis for the Arg 197 mutants, the catalytic parameters of each enzyme were determined. While the wild-type enzyme cleaved most of the substrate within 2 min, sortase A variants with substitutions at Arg 197 required continuous monitoring for 2 h to detect enzyme activity (Fig. 5, A and B). The slope of the increase in fluorescence was used to calculate the initial velocity values for different concentrations of substrate as described before (13), and the kinetic parameters of each enzyme were obtained from double reciprocal plots (Table I) tion of Arg 197 with histidine displayed a 300-fold decrease in k cat, whereas substitutions with either alanine or lysine caused a 1,500-fold reduction in k cat values. Consistent with previous results, the Cys 184 Arg/Arg 197 Cys sortase variant displayed little or no activity. Thus, the in vitro cleavage of Abz-LPETG-Dnp substrate is severely reduced by the R197H, R197A or R197K substitutions, whereas amino acid substitutions at Cys 184 essentially abolish all enzymatic activity. Interestingly, the Arg 197 substitutions displayed little change in the K m values as compared with the k cat measurements. On balance, the Arg 197 mutants showed an ϳ2-fold increase in this parameter, with the exception of the double mutant, with K m values similar to those observed for wild-type enzyme. Formation of the sortase:Abz-LPETG-Dnp complex is presumed to be the ratelimiting step of the hydrolysis reaction, while the subsequent formation and hydrolysis of the acyl intermediate proceeds quickly (22). The K m values can therefore be considered an approximation of the sortase:Abz-LPETG-Dnp dissociation constant. The kinetic parameters obtained here indicate, then, that while the binding affinities were somewhat perturbed, the major impact of Arg 197 substitutions occurs on the catalysis of the mutant enzymes. The kinetic parameters of N98A and N98Q sortase mutants were also determined ( Fig. 5C and Table I). Consistent with previous results, substitutions of Asn 98 caused only small, insignificant changes in enzyme activity.
We wondered whether sortase mutants with Arg 197 substitutions cleaved polypeptide substrates between the threonine and the glycine residues of Abz-LPETG-Dnp. To achieve sufficient amounts of cleavage product, all reactions were allowed to proceed for 16 h. Sortase enzymes were separated from substrate and product compounds by filtration. The filtrate was then subjected to RP-HPLC on C18 column and eluted peaks were analyzed by mass spectrometry to verify the molecular identity of compounds. Substrate cleavage by wild-type sortase generated two product peaks that eluted at 42 min and 47.2 min (83.2% acetonitrile), as mass spectrometry experiments revealed the products G-Dnp and Abz-LPET, respectively ( Fig.  6A and Table II). Substrate cleavage by the three sortase variants R197A, R197H, and R197K produced only small amounts of product peaks (Fig. 6A). Nevertheless, MALDI-TOF mass spectrometry analysis of the RP-HPLC product peaks revealed the same m/z signals as observed after cleavage with the wildtype enzyme (Table II). Thus, it appears that substitution of Arg 197 with any one of three other amino acids does not interfere with the substrate cleavage specificity of sortase A. The double mutant enzyme C184R/R197C did not generate significant product peaks, indicating that the mutant enzyme is essentially inactive in vitro (data not shown). The observed small amount of fluorescence signal increase after incubation of C184R/R197C sortase A with Abz-LPETG-Dnp is attributed to the binding of polypeptide substrate to the mutant enzyme ( Fig. 4). Analysis of the hydrolysis products of the Asn 98 mutants showed similar abundance and compound structures as compared with the products generated by wild-type sortase ( Fig. 6 and Table II).
Amino Acid Substitutions of Arg 197 Inhibit in Vitro Sortase Transpeptidation Activity-To test whether amino acid substitutions at Arg 197 affected the ability of sortase to catalyze a complete transpeptidation reaction, purified enzymes were incubated in the presence of polypeptide (Abz-LPETG-Dnp) and peptidoglycan substrates (NH 2 -Gly 3 ). All reactions were quenched by filtration, thereby separating the small molecular weight substrate and product compounds from the high molecular weight sortase enzymes. The filtrate was subjected to RP-HPLC on C18 column and the elution of peptides was monitored as an increase in absorbance at 215 nm (Fig. 6, C  and D). Abz-LPETG-Dnp substrate eluted at 59 min, whereas NH 2 -Gly 3 eluted shortly after the flow through eluate (data not shown). As expected, incubation of Abz-LPETG-Dnp and NH 2 -Gly 3 with wild-type SrtA ⌬N59 generated the transpeptidation product Abz-LPET-Gly 3 , which eluted after 45.8 min chromatography (74.8% acetonitrile) from C18 column (Fig. 6C). Substitution of Asn 98 with either alanine or glutamine did not affect the ability of mutant sortase enzymes to cleave Abz-LPETG-Dnp between the threonine and glycine residues (Fig.  6B) or to catalyze the transpeptidation reaction ( Fig. 6D and Table II). As observed above for the cleavage of polypeptide substrates, sortase A enzymes with amino acid substitutions at Arg 197 , R197A, R197H, and R197K, produced only small amounts of transpeptidation product. Nonetheless, the transpeptidation product Abz-LPET-Gly 3 eluted at the same time as that generated by the wild-type enzyme and generated the same m/z signal during mass spectrometry experiments (Table II). Together these data suggest that amino acid substitutions at Arg 197 of sortase A affect neither the specificity of the reaction nor the binding of the substrate, but severely interfere with the nucleophilic attack at the scissile peptide bond and subsequent thioacyl intermediate formation. In contrast, Asn 98 is dispensable for sortase transpeptidase activity ( Fig. 6 and Table II). DISCUSSION This work attempted to characterize the role of Arg 197 of S. aureus sortase A during catalysis. The crystal structure of the sortase A enzyme with its polypeptide substrate revealed that the side chain of Arg 197 is positioned in close proximity and on a neighboring ␤-strand to the active site Cys 184 , with the scissile bond located in between the two residues (24) (Fig.  1B). It was shown here that Arg 197 greatly contributes to the efficiency of sortase A-mediated substrate cleavage. Arg 197 may universally play an important role in catalysis, as analysis of the crystal structure of sortase B of S. aureus also revealed a Cys-Arg dyad in the active site (25) and as Arg 197 is absolutely  Fig. 5. The slopes of the fluorescence increase, calculated from averaged, triplicate data sets for each experiment, were used to determine the initial velocity at different Abz-LPETG-Dpn concentrations for each mutant enzyme as previously described (13). The values obtained were plotted in a double-reciprocal manner to calculate K m and V max . In this report, both in vivo ( Fig. 2A) and in vitro (Fig. 4B) assays for sortase activity showed that substitutions of Arg 197 severely diminished, but not completely eliminated, sortase activity. Analysis of the reaction products indicated that this residual activity was identical to wild-type sortase A (Fig. 6). Hydroxylaminolysis experiments interrogated the formation of a thioacyl intermediate during the sortase A-mediated transpeptidation reaction. The data presented in Fig. 3 suggest that this step of sortase catalysis was greatly impaired when Arg 197 was replaced with either alanine or lysine. If the third hypothesis (see above) were correct, one would assume that an unstable thioacyl intermediate would be released more readily by water molecules, causing hydrolysis and the release of surface proteins into the extracellular medium. This phenomenon was not observed, however, suggesting that Arg 197 may not be solely responsible for stabilizing the thioacyl intermediate. Analysis of the kinetic parameters obtained for the Arg 197 mutants (Table I) argues against the second hypothesis (see above). If substrate binding were indeed the rate-limiting step of the sortase catalysis (22), the observed K m values could be approximated as the substrate dissociation constants of the mutant enzymes. While the K m values suggested a modest decrease in substrate binding affinity for the Arg 197 mutants, the rate of catalysis, as judged by calculation of k cat , was dramatically decreased. Thus, it appears that the major function of Arg 197 may be the ionization of Cys 184 .
Replacement of Arg 197 with histidine produced the most active enzyme from a set of amino acid substitutions both in vivo ( Figs. 2A and 3) and in vitro (Fig. 5B and Table I). However, these experiments showed also that the R197H variant  a The substrate peptide Abz-LPETG-Dnp was incubated with wildtype or mutant SrtA⌬ N59 sortase enzymes, and reaction products were separated by RP-HPLC. Eluted fractions after 47.2 min chromatography (83.2% acetonitrile) were analyzed by MALDI-TOF. The calculated m/z of the Abz-LPET product is 578.62.
b The substrate peptide Abz-LPETG-Dnp and the peptidoglycan substrate NH 2 -Gly 3 were incubated with wild-type or mutant SrtA ⌬N59 sortase enzymes, and reaction products were separated by RP-HPLC. Eluted fractions after 45.8 min chromatography (74.8% acetonitrile) analyzed by MALDI-TOF. The calculated m/z of the Abz-LPET-Gly 3 transpeptidation product is 749.78. was much less efficient than the wild-type enzyme. Although we do not yet know whether the Cys 184 -His 197 variant encompasses an engineered thiolate-imidazolium ion bond, it is conceivable that sporadic ionization and/or orientation of the histidine side chain can produce a favorable arrangement in the active site for thiol ionization and catalysis. In contrast, substitution of Arg 197 with alanine and lysine rendered sortase much less active.
Members of the papain/cathepsin family of proteases also employ an active site cysteine residue to cleave substrate. Papain serves as the paradigm for cysteine proteases (28). A catalytic triad composed of Cys 25 -His 159 -Asn 175 is responsible for activating the active site thiol by a mechanism that is thought to involve formation of a thiolate-imidazolium ion bond as well as the formation of a hydrogen bond between the imidazolium ring of histidine and the side chain oxygen of asparagine (38,39). Amino acid substitutions in either Cys 25 or His 159 completely abrogated all papain activity, however amino acid substitutions at Asn 175 caused only a minor effect (40). Using the NMR structure of sortase A as a template, significant similarities between sortase A and papain were detected (21). Two conserved residues of sortase A, Cys 184 and His 120 , were located in close proximity and amino acid substitutions at these residues indeed abrogated all sortase A activity (26). Further, the NMR structure of sortase A suggested a hydrogen bond between the N⑀2 proton of the His 120 imidazolium ring and the O␦1 of the Asn 98 amide (21). It is shown here that Asn 98 is not involved in catalysis as amino acid substitutions at this residue do not impair sortase A-mediated surface protein anchoring. This view is consistent with the recent high resolution structure provided by x-ray crystallography, in which Asn 98 is positioned at a distance to His 120 and cannot engage in the above mentioned hydrogen bond formation, as well as with the observation that Asn 98 is not universally conserved among sortase enzymes (24).
Any argument on the mechanism of sortase A catalysis must take into account the fact that Cys 184 and His 120 are the only two residues that are absolutely essential for catalysis (26). Notwithstanding these observations, recent reports enlisted several arguments whether the absolutely conserved His 120 residue plays a role in catalysis (24,31). His 120 resides in the beginning of the loop that connects the ␤6 and ␤7 sheets. If His 120 were directly involved in catalysis, it would seem highly unlikely that the residue can accomplish such task unless sortase A underwent major structural rearrangements to position the residue closer to the active site. Even though the x-ray crystallographic structure of sortase A and sortase B revealed the close proximity of the conserved residues Cys 184 and Arg 197 and despite the fact that Arg 197 makes several important contributions to catalysis, one must not overlook the current enigma presented by the uncertain role of His 120 in catalysis. It seems plausible that, while Arg 197 forms a thiolate-guanidinium dyad with Cys 184 , His 120 may act as a general proton donor/acceptor for the acylation and de-acylation steps of the sortase reaction.
By anchoring surface proteins to the peptidoglycan moiety of bacterial cell wall, sortase enzymes assume a central role in implementing interactions between bacterial pathogens and their environment (11,41). Because a great many of these interactions are absolutely essential for the pathogenesis of bacterial infections, a clear view of the mechanisms of sortase action would greatly aid in the design of sortase inhibitors that may also function as therapeutics for human infections (2). Much progress has been made in defining the substrate properties, the active site residues, and the three-dimensional structure of sortases (13, 21, 24 -26, 31, 37); however, the mechanistic details of sortase catalysis have yet to be unveiled. Therefore, new and innovative approaches will need to be implemented in order to obtain clear pictures and ultimately detailed views of a sequence of reactions that allow understanding of how this fascinating enzyme tethers proteins to the cell wall envelope.