The enzyme activity of sortase A is regulated by phosphorylation in Staphylococcus aureus

ABSTRACT In many Gram-positive bacteria, the transpeptidase enzyme sortase A (SrtA) anchors surface proteins to cell wall and plays a critical role in the bacterial pathogenesis. Here, we show that in Staphylococcus aureus, an important human pathogen, the SrtA is phosphorylated by serine/threonine protein kinase Stk1. S. aureus SrtA can also be phosphorylated by small-molecule phosphodonor acetyl phosphate (AcP) in vitro. We determined that various amino acid residues of S. aureus SrtA are subject to phosphorylation, primarily on its catalytic site residue cysteine-184 in the context of a bacterial cell lysate. Both Stk1 and AcP-mediated phosphorylation inhibited the enzyme activity of SrtA in vitro. Consequently, deletion of gene (i.e. stp1) encoding serine/threonine phosphatase Stp1, the corresponding phosphatase of Stk1, caused an increase in the phosphorylation level of SrtA. The stp1 deletion mutant mimicked the phenotypic traits of srtA deletion mutant (i.e. attenuated growth where either haemoglobin or haem as a sole iron source and reduced liver infections in a mouse model of systemic infection). Importantly, the phenotypic defects of the stp1 deletion mutant can be alleviated by overexpressing srtA. Taken together, our finding suggests that phosphorylation plays an important role in modulating the activity of SrtA in S. aureus.

Sortases are transpeptidases found in many Grampositive bacteria [14,19,20]. Among them, S. aureus sortase A (SrtA) is the archetypal sortase that cleaves secreted protein substrates between the threonine and glycine residues of the conserved C-terminal LPXTG sorting motif and links the newly released threonine residue to the pentaglycine cross-bridge of S. aureus peptidoglycan, thereby effectively immobilizing substrates into the cell wall [14]. In S. aureus, 20 proteins with LPXTG sorting signals fulfill diverse functions such as iron acquisition from the host, bacterial adhesion, and immune evasion during the infectious process [14]. SrtA plays a critical role in the pathogenesis of Gram-positive bacterial infections and a S. aureus srtA knockout mutant is unable to cause bacteraemia or sepsis in mice [11,14,21].
SrtA is not required for bacterial viability in vitro and belongs to an exclusively bacterial family of enzymes, suggesting that specific SrtA inhibitors may have a limited selective pressure to promote the rise of resistance traits and are unlikely to exhibit particularly deleterious side effects [16]. SrtA is thus considered an ideal target for anti-virulence drug development [14][15][16][17][18][19][20][21][22][23][24]. The transpeptidase activity of SrtA has also been exploited in a range of industrial applications, including protein conjugation and targeted labelling of cell surface macromolecules [25][26][27][28][29][30][31][32]. To date, S. aureus SrtA is the most understood of all the sortases and its molecular features (i.e. structures, functions, and mechanisms) have been extensively studied [14,20] subsequent to its discovery in 1999 [11].
SrtA contain three conserved residues within its active site: His120, Cys184, and Arg197 [33]. It is clear that Cys184 forms a covalent linkage to the sorting signal while the functions of His120 and Arg197 are controversial [34]. Currently, the specificity and kinetic/catalytic mechanism of sortases remains an active area of research and a reverse protonation chemical mechanism of S. aureus SrtA has been proposed recently [34]. In this mechanism, C184 thiolate is the active site nucleophile, H120 imidazolium is a general acid, and R197 is a transition state stabilizer [34].
Although insight into the biological function, structure, specificity and mechanism of S. aureus SrtA is significant at present [14,20,34], the mechanisms that may enable S. aureus to regulate the activity of SrtA have not been described. Here, we show that S. aureus SrtA is subject to post-translational regulation. We found that phosphorylation markedly decreased the enzyme activity of S. aureus SrtA. We also showed that SrtA is a key factor for the Stk1/Stp1 signalling pathway in S. aureus.

SrtA is a substrate for Stk1 in vitro
Eukaryote-type Ser/Thr protein kinases (STKs) and phosphatases (STPs) have emerged as critical signalling molecules in prokaryotes [35,36]. In S. aureus, Stk1 and Stp1 provide an additional level of regulation for a variety of biological functions, including cell wall biosynthesis, drug resistance and virulence regulation [36][37][38][39]. To examine whether SrtA is a substrate for Stk1, the catalytic domain of S. aureus Stk1, termed Stk1 KD , was cloned, expressed, and purified from E. coli. Using SDS-PAGE and autoradiography, we found that SrtA ΔN24 , a truncated tagged form of S. aureus SrtA [40], was phosphorylated by Stk1 KD in vitro ( Figure 1A). Similar results were obtained using the Pro-Q Diamond and Coomassie blue stains with purified SrtA ΔN59 ( Figure 1B), another truncated S. aureus SrtA lacking the N-terminal 59 residues [40].
In the presence of Stk1 KD , the SrtA ΔN59 band stained intensely with Pro-Q Diamond but not with Coomassie blue ( Figure 1B). These results clearly suggest that SrtA is a substrate of Stk1 in vitro. Either phosphorylated SrtA ΔN24 ( Figure 1A) or phosphorylated SrtA ΔN59 ( Figure 1B) was dephosphorylated by Stp1, indicating that phosphorylated SrtA is a substrate of Stp1.
To further examine the roles of Stk1 and Stp1 in the phosphorylation of SrtA, we performed phosphorylation assays by incubating SrtA ΔN24 proteins respectively with cell lysates of a wild-type (WT) Newman strain, a Δstp1 stain, and a Stk1overproducing strain (stk1 +++ , the WT Newman strain carrying pYJ335:stk1). When cell lysates of either Δstp1 or stk1 +++ strain was used, a higher phosphorylation level of SrtA ΔN24 was observed compared to that obtained using the cell lysates of WT Newman strain ( Figure 1C). Thus, both stp1 deletion and over-expression of stk1 facilitate the phosphorylation of SrtA in the context of a cell lysate. These results are also consistent with in vitro reversible phosphorylation of SrtA mediated by purified recombinant Stk1 and Stp1 ( Figure 1A and 1B).

Phosphorylation of recombinant SrtA expressed in S. aureus
To probe the existence of phosphorylation of SrtA in vivo, we purified SrtA ΔN59 from Δstp1 and WT S. aureus Newman strain respectively, and dephosphorylation assay was performed by incubating the SrtA ΔN59 protein with Stp1. As shown in ( Figure 1D), Stp1-treatment decreased the intensity of SrtA ΔN59 band stained with Pro-Q Diamond but not with Coomassie blue, indicating that SrtA is phosphorylated in vivo. SrtA ΔN59 proteins purified from the Δstp1 strain displayed higher intensity of Pro-Q Diamond stained band than those purified from the WT Newman strain (Coomassie staining as loading control) ( Figure 1E). We also noted that Stp1treatment decreases the phosphorylation level of SrtA ΔN59 purified from the WT Newman strain ( Figure 1E), but to a much lesser extent than it did for the SrtA ΔN59 that had been purified from the Δstp1 strain ( Figure 1D). Based on these data, we conclude that SrtA is very likely to be subject to reversible phosphorylation in vivo, and this effect appears to be mediated by at least Stk1 and Stp1. Supporting this view, Co-Immunoprecipitation (Co-IP) assays demonstrated the interactions between WT, cell lysates from wild-type Newman strain (harbouring pYJ335); Δstp1, cell lysates from a Δstp1 strain (harbouring pYJ335); stk1 +++ , cell lysates from S. auerus strain overexpressing the stk1 (Newman/pYJ335:stk1). (D and E) Dephosphorylation of recombinant SrtA ΔN59 expressed and purified from either Δstp1 or WT Newman strain, band intensities from Pro-Q Diamond chemiluminescence were normalized against the corresponding coomassie brilliant blue stained band, which were set to 1.0 for purified recombinant SrtA ΔN59 from the Δstp1 strain. In A-E, Reactions were stopped by addition of SDS -PAGE loading buffer (without DTT), samples were separated by SDS-PAGE, visualized by either autoradiography ( 32 P) or pro-Q diamond stained gel (Pro-Q), and stained with coomassie blue. (F) Co-IP between Stk1 and SrtA. IP was performed with anti-Stk1 antibody followed by immunoblotting with anti-SrtA antibody (SrtA) and anti-Stk1 antibody (Stk1). The beads control was obtained in the absence of Stk1 antibody. (G) Co-IP between Stp1 and SrtA. IP was performed with anti-Stp1 antibody followed by immunoblotting with anti-SrtA antibody (SrtA) and anti-Stp1 antibody (Stp1). The beads control was obtained in the absence of Stp1 antibody. (H) Co-IP between SrtA and Stk1 or Stp1. IP was performed with anti-SrtA antibody followed by immunoblotting with anti-Stk1 antibody (Stk1), anti-Stp1 antibody (Stp1), and anti-SrtA antibody (SrtA). The beads control was obtained in the absence of SrtA antibody. In

SrtA contains multiple phosphorylation sites
To further verify the phosphorylation of SrtA and to identify the phosphorylation sites, we performed mass spectrometric characterization on the SrtA protein phosphorylated by the cell lysates of stk1 +++ strain. Searching tandem mass spectrometry (MS/MS) spectra for neutral loss peaks of 80 daltons revealed the presence of at least 28 phosphorylation sites and that the SrtA protein can be phosphorylated simultaneously at multiple residues (Figure 2A, Fig. S1 A to E, and Table  S3). These phosphorylation sites are mainly located in and around strands ß1, ß3 and ß4, strand ß6 and several amino acid residues close to the end of it, and strand ß7 of the crystal structure of SrtA ( Figure 2A) [41]. The catalytic residues of S. aureus SrtA, Cys184, and His120 [14], were subject to phosphorylation ( Figure 2B, Fig. S1). Asp112, which is important for calcium binding [42] and SrtA activity [43], was also a potential phosphorylation site ( Figure 2A, Table S3). These observations indicate that SrtA contains multiple phosphorylation sites.

Cysteine phosphorylation of SrtA
As aforementioned, Cys184 of S. aureus SrtA was subject to phosphorylation ( Figure 2B). Because Cys184 is essential for the catalytic activity of SrtA and this residue is absolutely conserved and essential for sortase activity from Gram-positive bacteria [14], we next sought to determine the role of Cys184 in the phosphorylation of SrtA. To this end, we analysed the phosphorylation levels of SrtA ΔN24 and its mutant protein SrtA ΔN24 C184A (in which Cys184 was replaced by an alanine residue) in the cell lysates of stk1 +++ strain in the presence of [ 32 P] ATP using autoradiography. Our results showed that phosphorylation occurred exclusively to the monomeric reduced forms of SrtA ΔN24 , but not to its dimeric oxidized forms and its mutant forms (SrtA ΔN24 C184A ) ( Figure 2C). These observations suggest that Cys184 might be a key site responsible for the phosphorylation of SrtA in the context of a bacterial cell lysate.
Additionally, we observed that, in the presence of 1 mM dithiothreitol (DTT), phosphorylation of SrtA ΔN24 by the cell lysates was almost abolished ( Figure 2C). This result support the notion that Cys184 was very likely the major site of the observed phosphorylation ( Figure 2C, lane 2) since cysteine phosphorylation is DTT-reversible [36]. Using Pro-Q Diamond and Coomassie blue stains, we found that the phosphorylation of SrtA ΔN59 that had been purified from Δstp1 stain displayed liability towards DTT ( Figure 2D). Taken together, these biochemical data, along with the LC-MS/MS data ( Figure 2B), suggest that Cys184 is a primary site for the phosphorylation of SrtA in the context of a bacterial cell lysate and cellular.

Different phosphorylation events on SrtA
Since several aspartic acid residues of S. aureus SrtA were subject to phosphorylation (Figure 2A, Fig. S1, and Table  S3), we next sought to analyse the effect of acetyl phosphate (AcP) on the phosphorylation of SrtA, given that AcP can be used as a phosphodonor for the acceptor aspartic acid residue of some bacterial proteins such as response regulators [44][45][46][47][48][49]. Using Pro-Q Diamond and Coomassie blue stains, we found that SrtA ΔN24 , but not Caseinolytic protease proteolytic subunit (ClpP) that serves as a control, can be efficiently phosphorylated by AcP ( Figure 2E). We also observed that sample heating, which promotes extensive hydrolysis of phospho-Asp residues [50], caused essentially loss of phosphorylation of SrtA ΔN24 that has been treated by AcP ( Figure 2F). These results indicate that certain aspartic acid residue of S. aureus SrtA may serve as acceptor for the phosphodonor AcP. Additionally, Stp1-treatment also decreased the phosphorylation level of SrtA ΔN24 mediated by AcP ( Figure 2F), indicating that Stp1 may possess aspartic acid dephosphorylation activity.
AcP was able to efficiently phosphorylate not only the SrtA ΔN24 ( Figure 2E andF) but also the SrtA ΔN59 C184A ( Figure 2G). Additionally, the cysteine residue does not play an important role in the in vitro phosphorylation of SrtA by purified Stk1 KD ( Figure 2H). Thus, the Cys184 is unlikely a primary phosphorylation site of SrtA under these experimental conditions, although it did in the context of a bacterial cell lysate ( Figure 2C) and cellular ( Figure 2D). Moreover, we observed that phosphorylated SrtA ΔN24 that had been phosphorylated by purified Stk1 KD was not susceptible to either DTT treatment ( Figure 3A) or sample heating ( Figure 3B). In all, these data suggest that SrtA can contain different types of phosphorylation with distinct chemical properties including DTT-labile ( Figure 2C and 2D), heat-labile ( Figure 2F), DTT-stable ( Figure 3A), and heat-stable ( Figure 3B). These distinct chemical properties are probably due to phosphorylation occurring on different forms of amino acid residues, which is in line with the observation that various forms of amino acid  Table S3).

Phosphorylation inhibits the enzymatic activity of SrtA
Cys184 is essential for the functions of SrtA [14], we thus reasoned that its phosphorylation may abolish the enzyme activity of SrtA. We performed a fluorescence resonance energy transfer (FRET) assay with phosphorylated SrtA ΔN24 (P~SrtA ΔN24 ) that has been phosphorylated by the cell lysates of the stk1 +++ strain, where cysteine residue is the primary phosphorylation site ( Figure 2C and 2D). In this assay, Stp1 and DTT were respectively added to a similar reaction in order to dephosphorylate the P~SrtA ΔN24 . Our results showed that the treatment of purified P~SrtA ΔN24 with Stp1 obviously increased its ability to cleave the abz-LPETG-dnp substrate ( Figure 4A). DTT-treatment also gave rise to an obvious increase of SrtA enzyme activity ( Figure 4A). Thus, cysteine phosphorylation appears to inhibit the enzyme activity of SrtA. However, it was a little unexpected that the purified P~SrtA ΔN24 still exhibited enzyme activity ( Figure 4A) and the reasons for this are currently unknown. It may be due to the chemical liability of the S-linked phosphorylation of SrtA and/or the fact that the purified protein sample is a mixture of phosphorylated SrtA ΔN24 with different phosphorylation events, except for the cysteine phosphorylation.
AcP -mediated phosphorylation also inhibited the enzyme activity of SrtA. As shown, AcP-treated SrtA ΔN24 displayed almost no enzyme activity in our FRET assay while retreatment of the AcP-treated SrtA ΔN24 with Stp1 obviously increased its ability to cleave the abz-LPETGdnp substrate ( Figure 4B). AcP might have a role in the regulation of SrtA-mediated anchoring of surface proteins in S. aureus. However, this hypothesis awaits further investigations. Nonetheless, these data clearly suggest that the inhibitory effect of AcP on the in vitro enzyme activity of SrtA is mediated at least in part by protein phosphorylation.
Additionally, we observed that the purified E. coliexpressed S. aureus SrtA ΔN24 can cleave the abz-LPETG -dnp substrate efficiently ( Figure 4B andC) and as expected [12] while it failed to do this in the presence of Stk1 KD ( Figure 4C). When Stp1 was subsequently added to the phosphorylation reaction mixture, the enzyme activity of SrtA ΔN24 was restored ( Figure 4C), indicating that Stk1-mediated phosphorylation inhibits the enzyme activity of SrtA. Based on these data, we conclude that the enzyme activity of SrtA is regulated by phosphorylation, at least in vitro.

Deletion of stp1 in S. aureus Newman produces srtA-deficient phenotypes
Deletion of stp1 caused an increase in the phosphorylation level of SrtA in S. aureus ( Figure 1E), we thus reasoned that stp1 deletion may cause an inhibition of the function of SrtA because phosphorylation inhibited the in vitro enzymatic activity of SrtA ( Figure 4). Supporting this notion, we found that, like the srtA deletion mutant [11,14], Δstp1 strain exhibited severe defects in bacterial growth in CRPMI (an ironrestricted medium) supplemented with either haemoglobin ( Figure 5A) or haem ( Figure 5B) as the sole iron source. Δstp1 strain displayed slight growth defect only when FeSO 4 was supplemented as the sole iron source, and as expected, all the testing S. aureus strains fail to grow in the iron-restricted CRPMI medium ( Figure 5C). The introduction of plasmid p-srtA can alleviate the growth defects of Δstp1 strain in the CRPMI medium supplemented with either haemoglobin or haem the sole iron source ( Figure 5A and 5B).
To examine the role of SrtA in Stp1-mediated signalling pathway during infections, we subjected the WT Newman, Δstp1, and Δstp1/p-srtA to a murine model of abscess formation and measured the bacterial survival in livers. As shown in (Figure 5D), the Δstp1 strain exhibited a significantly decreased colonization of liver compared to that of mice infected with the wildtype Newman (~4.5 log decrease in the CFU), whereas the introduction of plasmid p-srtA significantly increased the bacterial survival (~2.1 log increase in the CFU) ( Figure 5D), indicating that ectopic expression of SrtA can suppress the virulence defect of Δstp1. Taken together, these results suggest that SrtA appears to be involved in the Stp1/Stk1 signalling pathway of S. aureus ( Figure 5E).

Discussion
Here, we provide the first evidence that SrtA, an attractive anti-infective target for Gram-positive infections, is subject to phosphorylation that can inhibit its function ( Figure 5E). Notably, the catalytic site residue of S. aureus SrtA [14], Cys184, is a primary phosphorylation site in a cellular context (Figure 2 B to 2D). Because Cys184 is essential for the enzyme activity and the function of SrtA [14], it is thus difficult or impossible to interpret experiments involving mutations at this site unequivocally. However, consistent with previous reports that covalently modify this cysteine thiol abolished SrtA activity [14], the enzymatic activity of P~SrtA that has been primarily phosphorylated at Cys184 is severely reduced when compared to that of Stp1-treated P~SrtA ( Figure 4A). Moreover, the enzyme activity of P~SrtA was obviously increased in the presence of DTT ( Figure 4A), which is in line with the observation that Cys-phosphorylation is DTT-labile ( Figure 2C) [36]. S. aureus SrtA is highly resistant to oxidative inhibition and maintaining a high reduction potential of the Cys184 [51]. It has also been reported that DTT fails to increase the enzymatic activity of SrtA even at a concentration of up to 10 mM [52]. Given these, it is very likely that phosphorylation at Cys184 will introduce additional charge and conformational change, resulting in reduced enzyme activity of SrtA. We should, however, keep in mind that phosphorylation at other residue(s) may also contribute to the inhibition of the enzyme activity of SrtA ( Figure 4B andC).
S. aureus SrtA protein contains multiple phosphorylatable amino acids (Figure 2A). Hence, in S. aureus the properties of SrtA (i.e. activity and specificity) may be shaped by different phosphorylation events, given that multisite phosphorylation is an important mechanism for fine-tuned regulation of protein function [53,54]. Meanwhile, phosphorylation occurring at various amino acid residues implies that SrtA may act as a mediator in a variety of signalling pathways in S. aureus through integrating diverse phosphorylation events. Indeed, we found that, at least in vitro, environmental conditions determine the phosphorylation events of SrtA (Figure 2,3), and that phosphorylation inhibited its enzyme activity (Figure 4). Unraveling these aspects of signal transduction mechanisms will open up the new horizon towards understanding of the regulation of surface display of staphylococcal proteins.
In this study, we used an established FRET-based assay to monitor the enzyme activity of SrtA in vitro [52,55,56]. Although FRET has been extensively applied for probing the enzyme activity of SrtA, it is associated with several limitations [17]. For example, as the assay depends on the measurements of fluorescence, the quencher product would probably interfere with the results due to the inner filter effect [17]. Additionally, the presence of the cleaved product NH 2 -G-Dnp would decrease the fluorescence of Abz-LPET-OH when its concentration above 20 µM [52]. Besides, the FRETbased assay can only provide information on the initial cleavage of LPXTG-containing peptides, but not the kinetic parameters associated with the transpeptidation step, since the fluorophore Abz group has already been departed from the Dnp quencher group after the completion of the first step [56,57].
In sum, to our best knowledge, this is the first report of the phosphorylation of a sortase enzyme in bacteria. Further studies on the detailed mechanism by which the phosphorylation events affect the enzyme activity of SrtA will provide new insights into how this protein is regulated in vivo and contribute to the development of potent SrtA inhibitors which aim to combat Gram-positive infections. Moreover, in this study, we determined that Stk1 and Stp1 play a role in the phosphorylation/de-phosphorylation of SrtA in vitro (Figure 1 A to C) and these events are expected to happen in vivo (Figure 1 E to H). Since phosphorylation events inhibit SrtA enzyme activity (Figure 4), the inhibition of the Stp1 phosphatase might be a future pharmaceutical tool to reduce the activity of SrtA in S. aureus.

Ethics statement
The animal study protocols were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of the Shanghai Public Health Clinical Center (Permit Number: 2013P201) and were performed in strict accordance with the Regulations for the Administration of Affairs Concerning Experimental Animals approved by the State Council of People's Republic of China [11-14-]1988). The laboratory animal usage licence number is SYXK-HU -2010-0098, certificated by Shanghai Committee of Science and Technology. The animals were randomized but the investigators were not blinded to the experimental conditions.

Construction of plasmids
To construct the plasmid for the constitutive expression of srtA, a ca. 700 bp DNA covering 39 bp of srtA upstream region, the srtA gene, and a 40 bp downstream of srtA was amplified from S. aureus Newman genomic DNA with primers srtA-F and srtA-R (Table  S2), and then cloned into pYJ335. The construct with srtA in the same orientation as the tetracyclineinducible xyl/tetO promoter was confirmed by DNA sequencing, yielding plasmid pYJ335:srtA.
To construct pYJ335:6his-srtA ΔN59 , a DNA fragment was amplified from pET28a:6his-srtA ΔN59 with primers T7 promoter and T7 terminator (Table S2), and cloned into pYJ335. Primer pairs of pYJ335-F (Table S2) and T7 terminator were used to select the plasmid clones in which the genes are located downstream of the tetracycline-inducible xyl/tetO promoter.
To construct pET28a:srtA ΔN59 , primer srtA ΔN59 -F and srtA ΔN59 -R (Table S2) were used. To introduce a single amino acid substitution (Cys to Ala substitution at C184) to srtA gene, primer pair srtA C184A -F/srtA C184A -R and plasmid DNA of either pET28a:srtA ΔN24 or pET28a:srtA ΔN59 were used as template. All the selected plasmid clones were sequenced to confirm that no additional mutations were introduced by PCR reactions.

Construction of S. aureus ΔsrtA mutant
The gene replacement vector pKOR1 was used to construct an in-frame, unmarked srtA deletion mutant (ΔsrtA). PCRs were performed to amplify flanking sequences of the intended deletion. The upstream fragment was amplified from S. aureus strain Newman genomic DNA using primers srtA-delete-up-F and srtA-delete-up-R (with EcoRI site), and the downstream fragment was amplified with primers srtA-delete-down -F (with EcoRI site) and srtA-delete-down-R. PCR products were digested with EcoRI, mixed together and ligated by T4 DNA ligase (New England Biolabs). The ligation product was amplified with primers srtA-delete -up-F and srtA-delete-down-R. Next, the PCR product was used for recombination with pKOR1, and the product was introduced to E. coli DH5α. The construction was sequenced to ensure that no additional mutations resulted. The resulting plasmid, pKOR1:ΔsrtA, was electroporated to S. aureus RN4220, and subsequently into S. aureus Newman. The allelic replacement was performed as described previously [36] and the deletion of srtA was further confirmed by PCR and DNA sequencing.

Recombinant protein purification from E. coli
Transformed bacteria were sub-cultured into 1000 ml LB medium supplemented with 50 μg/ml kanamycin at 37°C to an optical density of 0.5 at 600 nm. After lowering the temperature to 16°C (for Stp1 and SrtA expression) or 20°C (for Stk1 KD expression), 1 mM isopropyl-thiogalactopyranoside (IPTG) was added in order to induce the expression of protein and the bacteria were harvested by centrifugation after an overnight of shaking (250 rpm of aeration).

Protein phosphorylation by purified 6his-Stk1 KD
For the detection of phosphoproteins with autoradiography, in vitro kinase assay were performed by incubating 0.5 μg 6His-Stk1 KD without or with 4 μg of the desired proteins (6His-SrtA ΔN24 or 6His-SrtA ΔN24 C184A , as indicated) in phosphorylation reaction buffer [50 mM Tris-HCl, pH 7.5; 0.1 mM EDTA, 1 mM MnCl 2 , 5 mM MgCl 2 ] in the presence of 10 μCi [γ-32 P]ATP at room temperature for 30 min. Likewise, for dephosphorylation assay, reaction was incubated with 0.5 μg Stp1 for an additional 30 min at room temperature. Reactions were performed in volumes of 20 μl, and an equal volume of protein storage buffer was added as needed in order to maintain a constant reaction condition. Reactions were stopped by the addition of 5 μl 5 × SDS loading buffer (without DTT), and the samples were resolved on a 12% SDS-PAGE, visualized by autoradiography.
Similarly, for the detection of phosphoproteins by Pro-Q Diamond stain, the in vitro kinase assay was performed by incubating 2 μg 6His-Stk1 KD without or with 4 μg of the desired proteins in phosphorylation reaction buffer supplemented with 1 mM ATP at room temperature for 30 min. Likewise, the dephosphorylation assay was performed by incubating the reactions with 1 μg Stp1 for an additional 30 min at room temperature. Reaction was stopped by the addition of 5 μl 5 × SDS loading buffer (with or without DTT, as indicated). Samples were resolved on a 12% SDS-PAGE. Following electrophoresis, proteins were immersed 30 min in fixing solution [50% methanol, 10% acetic acid]. The gel was subsequently washed in distilled water and treated for 90 min with Pro-Q Diamond Phosphoprotein Gel Stain. After destaining with destain buffer [20% acetonitrile, 50 mM sodium acetate, pH 4.0], phosphorylated species were visualized by using Tanon-5200 multi.
All the gels were subsequently stained with Coomassie blue to ensure the quantity of the loaded proteins.

Protein phosphorylation by bacterial cell lysates
S. aureus overnight cultures were diluted 100-fold in fresh TSB medium, and grown in a 250-ml flask with a flask volume to medium volume ratio of 5:1, shaking with 250 rpm at 37°C for 6 h. 10 ml cultures were harvested by centrifugation, washed once and suspended in 1 ml of phosphorylation buffer [50 mM Tris-HCl, pH 7.5; 0.1 mM EDTA, 1 mM MnCl 2 , 5 mM MgCl 2 ] supplied with 1 μl protease inhibitor cocktail (Thermo) and 1 μl phosphatase inhibitor cocktail 2 (Sigma). The mixture was homogenized by mechanical disruption (Fast Prep FP2400 instrument; Qbiogene) and then removed debris by centrifuging and filtrating through 0.45 μm filter.
For phosphorylation assay, 30 μg cell lysates were incubated with or without the desired 6His-proteins (~2 μg) in a phosphorylation buffer in the presence of 10 μCi of [γ-32 P] ATP. Reactions were performed in volumes of 20 μl at room temperature for 30 min, and stopped by adding 5 μl 5 × SDS loading buffer (with or without DTT, as indicated). The phosphorylation signal was visualized by autoradiography and the gel was subsequently stained with Coomassie blue to ensure the quantity of the loaded proteins.

Protein phosphorylation by acetyl phosphate
Recombinant SrtA proteins (~4 µg) purified from E. coli were incubated with or without lithium potassium acetyl phosphate (AcP) (a final concentration of 25 mM) in a phosphorylation reaction buffer. Reactions were performed in volumes of 20 μl at 37°C for 1 h. 6his-tagged Pseudomonas aeruginosa ClpP protein (6His-ClpP) [58], which contains 13 aspartic acid residues, was included as a recombinant protein controls, where appropriate. Reactions were stopped by adding 5 μl 5 × SDS loading buffer (with 100 mM DTT), and the samples were resolved on SDS-PAGE, and visualized by Pro-Q Diamond stain, subsequently stained with Coomassie blue to ensure the quantity of the loaded proteins. The signal intensity was determined by densitometric analysis using ImageQuant software.
For the dephosphorylation assay, recombinant proteins that have treated by AcP were loaded to a desalting spin column (BIORAD Bio-Spin® 6 column) for buffer exchange with 10 mM Tris buffer (10 mM Tris-HCl, pH 7.4) containing 0.02% sodium azide. After that, AcP-treated proteins (~4 µg) were incubated with or without Stp1 (~1 µg) at 37°C for 30 min, or subject to heating (95°C for 10 min). An equal volume of protein storage buffer was added as needed in order to maintain a constant reaction condition and reactions were performed in a phosphorylation reaction buffer in volumes of 20 μl.
For protein dephosphorylation, 6His-SrtA ΔN59 proteins (~4 μg) were incubated with Stp1 (~1 μg) or DTT in a phosphorylation buffer for final volume of 20 μl. In the control reaction (6His-SrtA ΔN59 alone), an equal volume of protein storage buffer for Stp1 was added. The reaction mixtures were maintained at room temperature for 30 min and then were stopped by adding 5 μl 5 × SDS loading buffer (without DTT). The samples were resolved on SDS-PAGE, and visualized by Pro-Q Diamond stain, subsequently stained with Coomassie blue to ensure the quantity of the loaded proteins. The signal intensity was determined by densitometric analysis using ImageQuant software.

LC-MS/MS identification of phosphorylation sites on SrtA
6His-SrtA ΔN59 (~60 μg) purified from E. coli was added into cell free extracts (~300 µg) of stk1 +++ (Newman/ pYJ335:stk1) and incubated with 1 mM ATP for 30 min at room temperature. The trypsin digested sample was analysed using an LTQ-Orbitrap Elite or Q exactive mass spectrometer using a nanospray source. After phosphopeptide enrichment by Fe-NTA spin column, 50 µg sample was loaded onto the MudPIT column. Full MS spectra (from m/z 300-1800) were acquired by the precursor ion scan using the Orbitrap analyser with resolution R = 60, 000 at m/z 400, followed by 20 MS/MS events. The twenty most intense ions were sequentially isolated and subjected to CID or HCD for fragmentation.
Protein identification analysis was done with Integrated Proteomics Pipeline (IP2, Integrated Proteomics Applications, Inc. San Diego, CA. http:// www.integratedproteomics.com) using ProLuCID/ Sequest, DTASelect2. Tandem mass spectra were extracted into ms2 files from raw files using RawExtract 1.9.9 (http://fields.scripps.edu/downloads. php) and were searched against Uniprot Staphylococcus aureus Newman local database. Database search criteria are as follows: mass tolerances for precursor ions were set at 12 ppm and for MS/MS were set at 600 ppm. Cysteine, serine, threonine, tyrosine, histidine, arginine, lysine and aspartic phosphorylation were considered as variable modifications and we require one peptide per protein and at least one tryptic terminus for each peptide identification.

Co-immunoprecipitation
Overnight cultures of S. aureus were diluted 1:100 and grown in TSB in a 250-ml flask with a tube volume-to-medium volume ratio of 5:1, shaking with 250 rpm at 37°C for 6 h. 10 ml cultures were harvested by centrifugation, washed once and suspended in 1 ml of lysis buffer provided in the CO-IP kit (Pierce) supplied with 1 μl protease inhibitor cocktail (Thermo). The mixture was homogenized by mechanical disruption (Fast Prep FP2400 instrument; Qbiogene) and then the debris was removed by centrifuging and filtrating through 0.45 μm filters. 500 µl clarified lysates were incubated with the resins which immobilized with specific antibodies or not for 12 h at 4°C. The bound materials were washed and eluted following the manufacturer's recommendation. Total samples (clarified lysates) and the immunocomplexes eluted from the resins were analysed by western blot with specific antibodies.

Western blot analysis
15 μl of 5 × SDS loading buffer [250 mM Tris-HCl, pH 6.8; 10% SDS; 0.5% bromophenol blue; 50% glycerol; 100 mM DTT] was added into 60 μl eluted immunocomplexes and then heated at 100°C for 10 min. SDS-PAGE was carried out subsequently. Thereafter, proteins were transferred to a PVDF (Bio-Rad) membrane at 350 mA for 60 min. The membrane was incubated with the primary antibody in 10 ml of 5% (wt/vol) skim milk at 4°C overnight following the blocking step [10 ml of 5% (wt/vol) skim milk] at room temperature for 2 h, and then washed three times at room temperature for 30 min in TBST buffer [10 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 0.04% Tween 20]. Then, membranes were incubated with the secondary antibody for 2 h at room temperature and washed three times for 30 min in TBST. The chemiluminescent detection reaction was performed and detected by Tanon-5200 multi. The following primary antibodies were used: anti-Sortase A antibody (Abcam, #ab13959) with a 1:3000 dilution, anti-Stk1 polyclonal antibody (prepared by immunizing rabbits with the purified 6His-Stk1 KD , Shanghai Immune Biotech CO., Ltd) with a 1:1000 dilution, anti-Stp1 polyclonal antibody (prepared by immunizing rabbits with the purified Stp1 protein, Shanghai Immune Biotech CO., Ltd) with a 1:500 dilution. For detection of the primary antibodies, anti-rabbit IgG (GE Healthcare, #NA934) with a 1:5000 dilution was used.

Enrichment of phosphorylated SrtA
6His-SrtA ΔN24 proteins (~15 mg) purified from E. coli was incubated with the cell lysates (~75 mg) of S. aureus Δstp1 strain in the presence of 1 mM ATP in phosphorylation buffer. Reactions were performed in volumes of 40 ml at room temperature for 30 min. 6His-SrtA ΔN24 was purified with HisTrap HP 5 ml column and the phosphoprotein purification kit (Qiagen) was used to further enrichment of the phosphorylated forms of 6His-SrtA ΔN24 (P ~ 6His-SrtA ΔN24 ) following the manufacturer's instruction. The enriched protein was concentrated to 1 mg/ml in phosphoprotein elution buffer [2.92 g NaCl, 7.42 g K 2 HPO 4 , 1.15 g KH 2 PO 4 , pH 7.5, ddH 2 O to 1,000 ml].

Enzymatic activity of SrtA
A FRET assay was used to monitor the in vitro activity of SrtA as described previously using LPETG fluorescent peptides (abz-LPETG-dnp) as a substrate [52]. In a typical FRET assay, a 100 μl reaction mixture contained 3 μM LPETG fluorescent peptides (abz-LPETGdnp), 4 μg 6His-SrtA ΔN24 in assay buffer [50 mM Tris-HCl, pH 7.5; 150 mM NaCl, 5 mM CaCl 2 ]. An equal volume of protein storage buffer (and/or phosphorylation buffer) was added as needed in order to maintain constant reaction conditions.
Briefly, for the dephosphorylation of 6His-SrtA ΔN24 that has been phosphorylated by the cell lysates of S. aureus Δstp1, 12 µg phosphorylated proteins were incubated with either Stp1 (~3 µg) or DTT (final concentration of 1 mM) at 37°C for 30 min in phosphorylation buffer in a total volume of 60 µl and after that, 20 μl sample mixture is used for subsequent FRET assay. For the dephosphorylation of 6His-SrtA ΔN24 that has been phosphorylated by AcP, AcP-terated 6His-SrtA ΔN24 (12 µg) was incubated with 3 µg Stp1 in phosphorylation buffer in a total volume of 60 µl and incubated at 37°C for 30 min, and 20 μl sample mixture is then used for subsequent FRET assay.
For the phosphorylation of 6His-SrtA ΔN24 by Stk1, in vitro kinase assay was performed in 60 μl of phosphorylation buffer containing 6His-SrtA ΔN24 (~12 μg) and 6His-Stk1 KD (~6 μg) in the presence of 1 mM ATP at 37°C for 30 min. Likewise, for the dephosphorylation assay, reaction was incubated with 3 μg Stp1 for an additional 30 min at room temperature. 20 μl sample mixture is then used for subsequent FRET assay, respectively.
Reaction sample was distributed in 96-well black plates (PerkinElmer), and the increase in fluorescence intensity was monitored every 1 min for 40 min using Synergy 2 (Biotek) plate reader at an excitation wavelength of 360 nm and emission wavelength of 460 nm. Triplicate measurements were taken for each sample, and the data are reported as mean.

Mouse infection models
S. aureus overnight cultures were diluted 1:100 and grown in fresh TSB in a 250-ml flask with a tube volume-to-medium volume ratio of 5:1, shaking with 250 rpm at 37°C for 3 h. Bacteria were harvested and washed twice with ice-cold PBS. The CFU per millilitre was determined before mice were inoculated. Mouse infection models were carried out as described previously [7,59]. 6 to 8-week-old female BALB/c mice were obtained from SIPPR-BK Lab Animal ltd (http:// www.slarc.org.cn) and housed under specified pathogen-free conditions.
Mice were anaesthetized with pentobarbital sodium (intra-peritoneal injection, 80 mg/kg) and were challenged with either 4 × 10 6 CFU of S. aureus Newman and its variants. Animals were euthanized 5 days post infection. Livers were aseptically removed and homogenized in PBS in order to obtain single-cell suspensions, and serial dilutions of each organ were plated on TSA plates for the enumeration of CFU. In this experiment, mice were also received intra-peritoneal injections every 24 h with a total dose of 40 mg/kg erythromycin hydrochloride (in order to maintain the plasmid) and 10 μg/kg anhydrotetracycline (in order to induce srtA expression) dissolved in sterile doubledistilled water (ddH 2 O) for 120 h (5 d). The statistical significance was determined by the Mann-Whitney Test (two-tailed).