The Eukaryotic-Type Serine/Threonine Protein Kinase Stk Is Required for Biofilm Formation and Virulence in Staphylococcus epidermidis

Background Serine/threonine kinases are involved in gene regulation and signal transduction in prokaryotes and eukaryotes. Here, we investigated the role of the serine/threonine kinase Stk in the opportunistic pathogen Staphylococcus epidermidis. Methodology/Principal Findings We constructed an isogenic stk mutant of a biofilm-forming clinical S. epidermidis isolate. Presence of stk was important for biofilm formation in vitro and virulence in a murine subcutaneous foreign body infection model. Furthermore, the stk mutant exhibited phenotypes indicating an impact of stk on metabolic pathways. Using different constructs for the genetic complementation of the stk mutant strain with full-length Stk or specific Stk domains, we determined that the Stk intracellular kinase domain is important for biofilm formation and regulation of purine metabolism. Site-specific inactivation of the Stk kinase domain led to defective biofilm formation, in further support of the notion that the kinase activity of Stk regulates biofilm formation of S. epidermidis. According to immunological detection of the biofilm exopolysaccharide PIA and real-time PCR of the PIA biosynthesis genes, the impact of stk on biofilm formation is mediated, at least in part, by a strong influence on PIA expression. Conclusions Our study identifies Stk as an important regulator of biofilm formation and virulence of S. epidermidis, with additional involvement in purine metabolism and the bacterial stress response.


Introduction
Reversible protein phosphorylation plays a fundamental role in signal transduction pathways in pro-and eukaryotes [1]. Various protein kinases mediate protein phosphorylation, which is generally coupled to dephosphorylation reactions catalyzed by protein phosphatases and enables translation of extracellular signals into cellular responses [2]. Since the first characterization of a eukaryote-like serine/threonine kinase (ESTK) in Myxococcus xanthus, similar ESTKs have been identified in numerous bacteria [3,4,5]. In particular, the Stk serine/threonine kinase has emerged as a critical signaling molecule in prokaryotes. In several bacterial species, Stk and similar proteins have been implicated in various phenotypes, including biofilm formation, cell wall biosynthesis, stress responses, metabolic pathways, autolysis and virulence [6,7]. For example, a PknB-like sensor kinase with similarity to Stk proteins was found essential for growth in Mycobacterium tuberculosis [8] and further Stk-like proteins were reported to be involved in biofilm formation in Streptococcus mutans, Bacillus subtilis and Mycobacterium smegmatis [9,10,11]. Furthermore, Stk was shown to play a role in virulence in streptococci, M. tuberculosis, Yersinia pseudotuberculosis and Staphylococcus aureus [12,13,14,15]. Hence, Stk homologues are widely distributed in bacteria and involved in diverse steps of bacterial pathogenesis.
The opportunistic pathogen Staphylococcus epidermidis, a member of the coagulase-negative staphylococci group, colonizes the skin and mucous membranes of the human body [16]. Unlike S. aureus, which produces a series of aggressive virulence determinants and frequently causes severe acute infections, S. epidermidis mainly causes persistent infections associated with biofilm formation on indwelling medical devices [17]. With the increasingly frequent application of such devices in recent years, S. epidermidis has drawn substantial interest as one of the most common causes of nosocomial infections [18].
Biofilms are multilayered, surface attached agglomerations of microorganisms which have intrinsic resistance to host immune defenses and antibiotic treatment. S. epidermidis biofilm formation occurs in two stages, including initial attachment to a surface and subsequent accumulation of cells, leading to multicellular structures [19]. Several factors have been identified to play a significant role in biofilm formation of S. epidermidis. Determinants that affect primary attachment to abiotic surfaces include the bifunctional adhesin and autolysin AtlE, the biofilm associated protein Bap, the fibrinogen-binding protein Fbe/SdrG and the fibronectin-binding protein Embp [20,21,22,23]. Polysaccharide intercellular adhesin (PIA) is crucial for biofilm formation in vitro and biofilm-associated infections [24]. In particular in PIAnegative S. epidermidis, the accumulation-associated protein (Aap) serves as an intercellular adhesin [25,26]. Other factors, including the ClpP protease, the DNA-binding protein SarZ, the substrate of ClpP, Spx, and extracellular DNA have also been reported to be directly or indirectly impact S. epidermidis biofilm formation [27,28,29,30].
Biofilm formation in S. epidermidis is under the influence of a diverse range of regulatory mechanisms [17]. For example, the quorum-sensing system agr (accessory gene regulator) regulates adhesin factors during the early stage of biofilm formation [31]. The icaADBC locus, encoding the proteins responsible for the synthesis and deacetylation of PIA, is up-regulated by the global regulator sigma factor SigB [32]. The product of the icaR gene, located adjacent to the ica operon, is a negative regulator of the icaADBC operon [33]. However, the signaling networks that control S. epidermidis biofilm formation remain incompletely understood.
In the present study, we identified a putative ESTK proteinencoding gene (designated stk) and a co-transcribed eukaryotic-like serine/threonine phosphatase gene (designated stp) in S. epidermidis. To explore the function of Stk in S. epidermidis, we constructed and characterized an stk mutant strain, which revealed important roles of Stk in biofilm formation, virulence, and metabolism of S. epidermidis.

Results
Genome context of stk and stp in S. epidermidis ESTKs have been identified in a wide range of prokaryotes. We performed a genome search (using BlastP, http://blast.ncbi.nlm. nih.gov/Blast.cgi) of S. epidermidis strain RP62A, which revealed the presence of a gene, SERP0786, whose protein product shows pronounced homology to Stk proteins in other species (36% identity with PrkC of B. subtilis and 67% identity with Stk of S. aureus) (Fig. 1A). We therefore designated the gene stk. The 39 end of stk overlaps with the adjacent gene, which encodes a protein phosphatase (SERP0785, designated stp) (Fig. 1B). Thus, stk and stp are likely co-transcribed [34], which we confirmed by reverse transcription (RT-) PCR analysis using primers 41F and 30R, one of which hybridizes with a sequence within the stk and the other with one within the stp gene. RT-PCR analysis also demonstrated co-transcription of stp and stk with the adjacent up-and downstream genes, encoding a SAM protein and a hypothetical protein, respectively (Fig. 1C). Of note, the 666-amino acid (aa) Stk of S. epidermidis displays conserved motifs that fit the description of a Hank's-type ESTK [35]: it includes an intracellular ATPbinding kinase domain (9-263 aa), a transmembrane spanning domain (347-369 aa) and three extracellular PASTA (penicillin and Ser/Thr kinase associated) repeated domains (376-437 aa, 444-483 aa, 514-572 aa) [36]. Within the N-terminal catalytic kinase domain, the protein harbors an important ATP-binding site (16-39 aa) that is presumably essential for kinase activity ( Fig. 2A).

Inactivation of stk in S. epidermidis results in reduced biofilm formation in vitro
To investigate the role of Stk in S. epidermidis physiology and pathogenesis, we constructed an isogenic deletion mutant of the stk gene in the biofilm-forming clinical isolate S. epidermidis 1457. Then, to determine whether stk influences biofilm formation as a main virulence phenotype in S. epidermidis, we analyzed biofilm formation by a semi-quantitative colorimetric assay. We found that in-vitro biofilm formation of the stk mutant strain was significantly impaired compared to the WT strain (Fig. 2B). The impact of stk on biofilm formation was confirmed by genetic complementation with the entire copy of stk expressed from a plasmid (Dstk (pQG71)) (Fig. 2B). The complemented strains showed much higher biofilm-forming capacity compared to the WT, which is likely mainly due to a plasmid copy effect and the use of a strong promoter (ica) to express the stk gene.
To study which stage of biofilm formation is influenced by stk, we performed a primary attachment assay and found that the stk mutant strain displayed decreased attachment ability to polystyrene compared with the WT strain, while this ability was restored in the complemented mutant strain (Fig. 3A). Furthermore, immunoblot analysis of polysaccharide intercellular adhesin (PIA), a key factor contributing to biofilm accumulation of S. epidermidis [37] demonstrated that stk positively affects PIA production and thus the second stage of biofilm formation (Fig. 3B). These observations indicate that stk is an important regulator of S. epidermidis biofilm formation in both the initial and accumulative stages.

The kinase activity of Stk is crucial for biofilm formation in S. epidermidis
To determine which domains of the Stk protein are important for its influence on biofilm formation, we genetically complemented the stk mutant strain with constructs expressing different Stk domains ( Fig. 2A). Only the strain expressing the kinase domain (Dstk (pQG72)) restored biofilm formation to a level comparable to that of the WT, whereas the strains expressing only the transmembrane segment (Dstk (pQG73)), the PASTA repeated domain (Dstk (pQG74)) or the C-terminal segment (Dstk (pQG75)) did not form a biofilm. Similarly, a strain expressing a truncated kinase domain lacking the conserved ATP-binding site (Dstk (pQG76)) also did not form biofilm (Fig. 2B).
As it has been reported that the conserved aspartic acid residue D 126 is crucial for the catalytic activity of the enzyme [13], we confirmed the role of the kinase activity in biofilm formation in S. epidermidis using a D126E mutant (Dstk (pQG81)). In addition, we mutated the D 133 residue that is in close proximity to D 126 to investigate the specificity of the effect exerted by the D126E exchange. The D126E mutant showed significantly decreased biofilm formation compared to the clone expressing the unaltered kinase domain, while biofilm formation of the D133E mutant (Dstk (pQG80)) was not changed (Fig. 2B). Corresponding results were achieved in the primary attachment and PIA production assays (Fig. 3). These findings indicated that the kinase activity of Stk is crucial for its impact on primary attachment, PIA production, and biofilm formation.
The predicted kinase domain of S. epidermidis Stk has kinase activity To further confirm the involvement of the kinase domain in kinase activity, we produced GST fusion constructs with the entire Co-transcription analysis of the four genes SERP0784 to SERP0787 using reverse transcription (RT-) PCR analysis with genomic DNA (gDNA), RNA or cDNA as templates. Lanes 1, 4, and 7 represent the amplification using primer set 39F and 40R, lanes 2, 5, and 8 represent the amplification using primer set 41F and 30R, lanes 3, 6, and 9 represent the amplification using primer set 35F and 42R. doi:10.1371/journal.pone.0025380.g001 Stk protein, only the kinase domain, or an inactivated kinase domain, in which the ATP-binding cassette was deleted. In kinase assays, only the GST fusion construct with the kinase domain lacking the ATP binding cassette did not exhibit activity, showing that the cassette is crucial for kinase activity (Fig. 4A). Staurosporine, a potent protein kinase inhibitor, inhibited the WT Stk

Effect of the stk mutation on the transcription of biofilmrelated genes
ESTKs commonly serve as global regulators of gene expression [38]. To analyze whether the impact of S. epidermidis Stk on biofilm formation is mediated by altering expression of biofilm-related regulatory or structural genes, we tested the transcriptional levels of several such genes in the stk mutant and wild-type strains by quantitative RT-PCR. We detected a significant positive impact of stk on the expression of the icaB and atlE genes, as expected from our results showing increased PIA production and primary attachment, of which AtlE is a key determinant [21] (Fig. 5A&B). To examine whether stk has an impact on the expression of global regulatory system, we examined transcript levels of sarA and agr, the latter by measuring RNAIII. The transcript levels of agr were up-regulated significantly in the mutant strain compared with WT strain (Fig. 5C). In contrast, stk did not impact expression of sarA (data not shown). Interestingly, we also did not detect an impact on the expression of the icaR regulatory gene (data not shown), indicating that the icaADBC but not the icaR promoter is under the influence of stk-dependent regulation.
The role of stk in the pathogenesis of S. epidermidis biofilm-associated infection To investigate the impact of stk on biofilm-associated infection in vivo, we performed a murine subcutaneous foreign body infection model. Progression of disease was measured by determining bacterial loads on the implanted catheters after sacrifice of the test animals. We detected a significantly lower bacterial load in animals infected with the stk mutant compared to those infected with the WT strain ( Fig. 6A), indicating that stk contributes to pathogen survival during S. epidermidis infection. Furthermore, histopathological analysis demonstrated that inflammatory lesions were stronger in the epithelial tissue around the implanted device of the mice infected with the WT strain compared to those infected with the mutant strain ( Fig. 6B-E). Fibrous structures of epithelial tissues were severely damaged and numerous inflammatory cells were found that had infiltrated into the infected tissues in the WT group. Among those cells, neutrophils were the predominant type by microscopic analysis. In contrast, in the epithelial tissues of mutant-infected mice only mild inflammatory responses (few infiltrated inflammatory cells and slight tissue lesions) were observed. These results indicated that stk plays a significant role in biofilm-associated infection in S. epidermidis.

Phenotypic analysis of an isogenic S. epidermidis stk deletion mutant
We then analyzed whether Stk has an influence on metabolic pathways, as these phenotypes are reportedly under Stk control in other bacteria [7,39,40]. Growth of the isogenic stk mutant in TSB was similar to that of the WT strain (data not shown). However, in defined RPMI 1640 medium growth of the stk mutant strain was slower compared to the WT strain, while supplementing the medium with 200 mM adenine partially reversed the growth defect of the stk mutant strain (Fig. 7A). Interestingly, in addition to fulllength Stk, the Stk kinase domain alone was able to complement the growth defect of the stk mutant, in contrast to the PASTA domains or an inactivated kinase domain (Fig. 7B). To analyze whether the impact of S. epidermidis Stk on purine biosynthesis is mediated by altering expression of purine-related regulatory genes, we tested the transcriptional levels of several such genes in the stk mutant and WT strains by quantitative RT-PCR. We detected a significant positive impact of stk on the expression of the purA gene (Fig. 7C). In contrast, stk did not impact expression of purR (data not shown). These results indicated that Stk in S. epidermidis is involved in purine biosynthesis through the kinase activity, maybe by regulating the expression of purA gene.
Furthermore, the stk mutant was more resistant to Triton X-100 induced lysis than the WT strain, which could be complemented with the PASTA, but not the kinase domain (Fig. 8A). Moreover, culture filtrate of the stk mutant strain exhibited diminished autolytic activity, which also could be complemented with the PASTA but not the kinase domain (Fig. 8B). These results indicated that Stk in S. epidermidis is involved in the bacterial stress response.

Discussion
In this study, we characterized the genetic locus encoding the ESTK-like Stk protein of S. epidermidis and report on the roles of this gene in physiology and virulence. Similar to S. aureus and Streptococcus pyogenes [34,39], the stk gene is adjacent to, and as we show here co-transcribed with, the stp gene encoding the cognate phosphatase. Deletion of stk did not have an effect on the transcription of stp, as shown by quantitative RT-PCR (data not shown). Since biofilm formation is a crucial mechanism by which S. epidermidis escapes human host defenses, we determined whether stk contributes to biofilm formation in S. epidermidis. Our data show that stk is involved in both the initial attachment and subsequent accumulation stages of biofilm formation and this impact is mediated at least in part by altering expression of the global regulator agr, atlE genes and the ica operon, the latter resulting in increased production of the biofilm exopolysaccharide PIA. Likely as a consequence of the impact of stk on these biofilm-and immune evasion-related genes, we also detected a significantly decreased virulence potential of the stk mutant strain in a murine subcutaneous foreign body infection model. This is consistent with reports on the in-vivo role of Stk in other prokaryotes such as S. aureus, S. pneumoniae and M. tuberculosis [12,41,42]. Of note, the effect of stk on S. epidermidis pathogenesis may be largely dependent on the strong regulation of PIA synthesis, since PIA acts both as a matrix and contributes to immune evasion [43,44,45].
Similar to other serine/threonine kinases in Gram-positive bacteria, the Stk protein of S. epidermidis has two domains (four segments), including an amino-terminal kinase domain, a  transmembrane segment, an extracellular PASTA domain (three PASTA repeats) and a carboxyl-terminal part [46]. We showed here that the kinase domain, whose kinase activity we confirmed, is crucial for the impact of Stk on the detected in-vitro biofilm phenotype.
According to our results, S. epidermidis Stk is also involved in the regulation of metabolism. In particular, delayed growth in RPMI medium, which could be overcome by adding adenine and genetic complementation with the kinase domain, indicated an impact of Stk on purine metabolism that is dependent on the Stk kinase domain. Our results are consistent with previous reports showing that a mutant of the purine biosynthesis gene purR of S. epidermidis displayed lower PIA production and was biofilm-negative [47,48] and a study that showed regulation of purine biosynthesis by the ESTK protein PknB of S. aureus, which regulates the adenylosuccinate synthase PurA via reversible phosphorylation [38]. It has been reported that kinase-dependent phosphorylation decreases PurA activity and is responsible for regulating purine biosynthesis in Streptococcus agalactiae [49], which seems to contrast our results.
However, it is conceivable that PurA activity is regulated at the transcriptional and/or translational levels in different organisms. How exactly PurA function is regulated in S. epidermidis remains to be investigated.
Furthermore, we showed that Stk had a significant influence on the bacterial stress response. The influence was mainly dependent on the extracellular PASTA domains, which are highly conserved and comprise three beta sheets and an alpha helix that binds the beta-lactam stem [50]. Stk proteins containing PASTA domains have been shown to play a key role in cell wall biosynthesis and suggested to sense dissociated peptidoglycan subunits [13,36,51]. It is likely that the PASTA repeats of the S. epidermidis Stk protein also recognize unlinked peptidoglycan and binding of those structures activates the Stk cytoplasmic N-terminal kinase domain.
In conclusion, to our knowledge this is the first study to show that Stk influences biofilm formation through its kinase activity and has an impact on virulence in S. epidermidis. Additionally, we showed that stk is involved in purine biosynthesis and stress response in S. epidermidis. While it is likely that Stk exerts its regulatory function directly through phosphorylation of proteins involved in biofilm formation or metabolism, with several such Stk substrates having been identified in other organisms to date [34,52,53], the specific substrates of Stk in S. epidermidis remain to be determined.

DNA manipulation
Genomic DNA of S. epidermidis 1457 was prepared by a standard protocol for gram-positive bacteria [54]. Plasmid DNA from E. coli was extracted using a plasmid purification kit (Axygen). Plasmid DNA from S. aureus and S. epidermidis was extracted using the same kit except that the cells were incubated for at least 30 min at 37uC in solution I (P1, suspension buffer) with lysostaphin (0.8 mg/ml; Sigma) before solution II (P2, dissociation buffer) was added. PrimStar polymerase and restriction enzymes were obtained from New England Biolabs (NEB). S. epidermidis was transformed by electroporation as described previously [55].

Co-transcription assay
At an OD 600 value of 0.5, cells in 1.5-ml TSB cultures were harvested and resuspended in 1.3 ml Trizol (Invitrogen). Cells were disrupted by shaking with a Mini-Beadbeater (Biospec Products) at maximum speed for 30 s. Tubes were then incubated on ice for 5 min. This shaking/cooling cycle was repeated 4 times. Then, the suspension was centrifuged. Total RNA isolation from the supernatant was performed in Trizol (Invitrogen) according to the manufacturer's instructions. After treatment using a TURBO DNA-free TM kit (Ambion), approximately 2 mg total RNA was used to create cDNA using a PrimeScript RT reagent kit (TaKaRa). An identical reaction was performed without reverse transcriptase as a negative control. cDNA with or without reverse transcriptase and genomic DNA (gDNA) were used as templates in PCRs using specific primer sets specific for overlapping (41F/ 30R), and outermost regions of stp and stk (39F/40R and 35F/ 42R), as shown in Fig. 1B. Construction of an isogenic stk deletion mutant and complemented strains To delete the stk gene in S. epidermidis 1457, ,1-kb DNA fragments, upstream and downstream of stk, were amplified by PCR from the chromosomal DNA of S. epidermidis 1457 using the primers listed in Table 2, introducing an overlapping sequence of ,15 bp. The resulting fusion PCR fragment with attB sites at both ends was used for recombination with plasmid pKOR1, yielding plasmid pKOR1stk. It was transferred via electroporation first to S. aureus RN4220 and then to S. epidermidis 1457. Allelic replacement and selection of positive clones were performed as described [56]. The integration site was verified by analytical PCR and sequencing. As the sequence and location of the endogenous promoter that facilitates stk transcription in S. epidermidis are unknown, we used the promoter sequence of the icaADBC operon for the construction of a genetic complementation plasmid. This fragment was amplified from S. epidermidis 1457 genomic DNA by PCR using the primer set Pica1 and Pica2 and cloned into EcoRI and XmaI sites of EcoRI/XmaI-digested pYJ90, yielding pQG70. Plasmid pQG71 (stk full length complementing plasmid) was constructed by cloning the entire coding region of stk (1998 bp) into BamHI/XmaI-digested pQG70, using the primer set stk-F and stk-R. Plasmids pQG72 (stk kinase domain complementing plasmid), pQG73 (stk transmembrane domain complementing plasmid), pQG74 (stk PASTA repeated domain complementing plasmid), pQG75 (stk C-terminal domain complementing plasmid) and pQG76 (stk kinase domain without ATP-binding site complementing plasmid) were constructed by cloning the respective domains into BamHI/XmaI-digested pQG70, using the following primers: 29F and 30R for the kinase domain, 31F and 32R for the transmembrane domain, 33F and 34R for the PASTA domain, 35F and 36R for the C-terminal domain, and 57F and 30R for the kinase domain without ATP-binding segment. All these plasmids (Table 1) were used to transform the stk mutant strain.

Site-directed mutagenesis
Site-directed mutagenesis was performed as described by Ho et al [57]. Briefly, a DNA fragment with the stk gene carrying a distinct single aspartic acid residue codon mutated to glutamic acid at position 133, or 126 was cloned into pQG70, yielding pQG80 and pQG81 respectively. The oligonucleotide primers 43F and 44R were used for mutagenesis at D 133 , while 53F and 54R were used for mutagenesis at D 126 . The presence of the mutation was verified by DNA sequencing. The two plasmids (Table 1) were used to transform the stk mutant strain.

Semi-quantitative biofilm assay
Semi-quantitative biofilm assays were performed as described in our previous work [58]. Briefly, overnight cultures of S. epidermidis strains were diluted 1:100 into fresh TSB. The diluted cultures were pipetted into sterile 96-well flat-bottom tissue culture plates and incubated at 37uC for 24 h. Culture supernatants were gently removed, and wells were washed with phosphate-buffered saline (PBS). The adherent organisms at the bottom of the wells were fixed by Bouin fixative over 1 h. Then the fixative was gently removed, wells were washed with PBS and stained with 0.4% (wt/ vol) crystal violet. Biofilm formation was measured with a MicroELISA autoreader.

Primary attachment assay
The assay was performed as described in our previous work [58]. Briefly, overnight cultures of S. epidermidis strains were diluted in fresh BM to an OD 600 value of 0.02 and grown at 37uC to an OD 600 value of 1.0. The cultures were pipetted into sterile 96-well flat-bottom tissue culture plates and incubated at 37uC for 1 h. The subsequent procedures were the same as those applied for the semi-quantitative biofilm assay.

Immuno-dot blot analysis of PIA production
Immuno-dot blot assays were performed as described in our previous work [59]. PIA samples were isolated from the surface of cells in exponential growth phase (4 h) by boiling with 0.5 M EDTA. PIA production was determined by immuno-dot blot analysis using anti-PIA sera and quantified by photodigital analysis.

Expression and purification of recombinant proteins
Different-length DNA fragments of the 2-kb coding region of stk were produced using genomic DNA of S. epidermidis as a template. The primers are listed in Table 2. The full-length PCR product of stk using primers 17F and 18R was digested with NotI and EcoRI and ligated into pET28a, yielding 6His-tagged protein (pET28a-Stk). As no tagged protein was expressed in pET28a, the stk fragment was acquired from pET28a-Stk plasmid digested with BamHI and XhoI and subcloned into vector pGEX-KG, yielding pQG77 (WT Stk, Stk full length expression plasmid), allowing inframe fusion with glutathione S-transferase (GST) [60]. Plasmids expressing GST fusions were constructed as follows: Stk codons corresponding to amino acid positions 1 to 263 (kinase domain, Stk 1-263 ) were amplified using primers 59F and 60R introducing EcoRI and HindIII sites, and inserted into pGEX-KG (yielding pQG78). Stk codons corresponding to amino acid positions 40-263 (Stk kinase domain without ATP-binding site, Stk 40-263 ) were amplified using primers 58F and 60R introducing EcoRI and HindIII and ligated into pGEX-KG (yielding pQG79). The resulting plasmids (Table 1) were used to chemically transform E. coli BL21 DE3 pLysS competent cells. The recombinant GSTtagged fusion proteins were expressed by the addition of 0.5 mM isopropyl-b-D-thiogalactopyranoside (IPTG) overnight at 16uC . The purification of the GST-tagged proteins were performed using GST-binding resin (Novagen) according to the manufacturer's instruction. Proteins were eluted with 20 mM glutathione prepared in 50 mM Tris-HCl (pH 8.0). Purified proteins were concentrated and dialyzed against PBS, using a 10 K and 30 K molecular weight dialysis cassette (Sangon Co.). Concentrations of proteins were determined using a Bradford Protein Assay kit (Sangon Co.) using bovine serum albumin as standard.

In vitro kinase assay
For the in vitro kinase assay, 2 mg of WT Stk, Stk 1-263 or Stk 40-263 and 1 mg histone 1 (H1) were incubated separately in 50 ml kinase buffer (50 mM Tris-HCl [pH 7.5], 100 mM dithiothreitol [DTT], 1 mM MnCl 2 , 100 mM ATP) for 4 h at 32uC Kinase activity was assayed using a Kinase-Glo TM Luminescent Kinase Assay kit (Promega) and the luminescent signal reflecting the ATP level in the reaction system was measured in a Victor 3 plate reader. Staurosporine (Sangon Co.) was pre-incubated with WT Stk for 30 min at room temperature. The luminescent signal inversely correlates to the amount of kinase activity. All reactions were performed in triplicate.
Quantitative real time polymerase chain reaction (RT-PCR) S. epidermidis strains were grown in BM. Sample collection and RNA extraction were performed as described above. After treatment using a TURBO DNA-free TM kit (Ambion), approximately 2 mg of total RNA were reverse-transcribed with a PrimeScript RT reagent kit (TaKaRa). The cDNA was used as a template for real-time PCR using SYBR-green PCR reagents (TaKaRa). Reactions were performed in a MicroAmp Optical 96-well reaction plate using a 7500 Sequence Detector (Applied Biosystems). Primers used are listed in Table 2. All RT-PCR experiments were performed in triplicate, with gyrase B (gyrB) used as an internal control.

Animal infection model and histopathological analysis
A murine subcutaneous foreign body infection model was performed as described by Kadurugamuwa et al [61]. Briefly, forty male BALB/c mice (Academia Sinica, weight ,20 g) were randomly and evenly divided into two groups and infected with WT or stk mutant bacteria. One-centimeter long silicon catheters (14 gauge) were implanted subcutaneously on each side at the flank of each mouse before injection of 10 7 CFU to the catheter bed. After ten days, the mice were sacrificed. Catheters were aseptically removed and manipulated with the method described previously except with a sonication time of 30 min. Then serial dilutions of the wash fluid were plated on TSB plates and the recovered S. epidermidis colonies were counted. For histological observations, the skin near the catheters were removed and immediately fixed in 2.5% (v/v) glutarildehydepolyoxymethylene solution. These samples were embedded in paraffin and stained with hematoxylin and eosin (HE) and examined by light microscopy.

Autolysis and bacteriolysis assays
Autolysis assays were performed using exponential-phase cells [13]. In brief, cells were cultured in TSB, harvested at an OD 600 of 0.7, washed twice with ice-cold, sterile PBS, resuspended in PBS plus 0.1% Triton X-100, and incubated at 37uC with shaking at 250 rpm, and the optical densities were serially monitored. Bacteriolytic assays were performed as follows. Heatkilled cells were used as substrates and bacterial supernatants as the source of secreted autolysins [41]. Heat-killed cells were resuspended in PBS by heating the bacterial suspension at 65uC for 2 h and normalized to an OD 600 value of 5.0. Secreted autolysins in the culture supernatants were normalized to cells at an OD 600 value of 5, and sterilized by passage through a 0.22-mm filter. Bacteriolytic activity was assessed by mixing 1 ml of heatkilled cells with 5 ml of supernatant (or media as a control), incubation at 37uC with shaking at 250 rpm, and serial monitoring of the optical densities.