Genomics and Proteomics Provide New Insight into the Commensal

For a more comprehensive list of citations to this article, users are encouraged to ... Emily Amor Stander , Bronwyn Kirby-McCollough , Roland Jourdai...
0 downloads 0 Views 5MB Size
Article pubs.acs.org/jpr

Genomics and Proteomics Provide New Insight into the Commensal and Pathogenic Lifestyles of Bovine- and Human-Associated Staphylococcus epidermidis Strains Kirsi Savijoki,†,‡ Antti Iivanainen,*,§,○ Pia Siljamak̈ i,†,‡,○ Pia K. Laine,∥ Lars Paulin,∥ Taru Karonen,§ Satu Pyöral̈ a,̈ ⊥ Matti Kankainen,# Tuula A. Nyman,‡ Tiina Salomak̈ i,§ Patrik Koskinen,∇ Liisa Holm,∇ Heli Simojoki,⊥ Suvi Taponen,⊥ Antti Sukura,§ Nisse Kalkkinen,‡ Petri Auvinen,*,∥ and Pekka Varmanen*,† †

Department of Food and Environmental Sciences, ‡Institute of Biotechnology, Proteomics Unit, §Department of Veterinary Biosciences, ∥Institute of Biotechnology, DNA Sequencing and Genomics Laboratory, ⊥Department of Production Animal Medicine, and ∇Institute of Biotechnology, Bioinformatics Group, University of Helsinki, FI-00014 Helsinki, Finland # CSC − IT Center for Science Ltd., FI-02101 Espoo, Finland S Supporting Information *

ABSTRACT: The present study reports comparative genomics and proteomics of Staphylococcus epidermidis (SE) strains isolated from bovine intramammary infection (PM221) and human hosts (ATCC12228 and RP62A). Genome-level profiling and protein expression analyses revealed that the bovine strain and the mildly infectious ATCC12228 strain are highly similar. Their genomes share high sequence identity and synteny, and both were predicted to encode the commensal-associated fdr marker gene. In contrast, PM221 was judged to differ from the sepsis-associated virulent human RP62A strain on the basis of distinct protein expression patterns and overall lack of genome synteny. The 2D DIGE and phenotypic analyses suggest that PM221 and ATCC12228 coordinate the TCA cycle activity and the formation of small colony variants in a way that could result in increased viability. Pilot experimental infection studies indicated that although ATCC12228 was able to infect a bovine host, the PM221 strain caused more severe clinical signs. Further investigation revealed strain- and condition-specific differences among surface bound proteins with likely roles in adhesion, biofilm formation, and immunomodulatory functions. Thus, our findings revealed a close link between the bovine and commensal-type human strains and suggest that humans could act as a reservoir of bovine mastitis-causing SE strains. KEYWORDS: Genomics, 2D DIGE, surfacome shaving, Staphylococcus epidermidis, intramammary infection, adaptation, virulence



INTRODUCTION Coagulase-negative Staphylococcus (CoNS) species including Staphylococcus epidermidis (SE) are important causative species of intramammary infection (IMI) in dairy cattle and are a known cause of persistent IMIs.1−4 SE is rarely found in the normal microbiota of the bovine skin or on mucous membranes,5,6 whereas it is common in the barn environment.7 SE strains isolated from bovine IMI share the same genotype with those isolated from the milker’s skin, which has led to the hypothesis that SE isolated from the bovine host may originate from humans.5,6 Multigenome screening analyses of commensal and clinical SE human strains suggest that the ability of this species to generate subpopulation variants may be an additional factor affecting commensal and pathogenic characteristics of this species.8 The mechanisms modulating this interplay are not known. In humans, SE is part of the normal and balanced skin microbiota, which may block the colonization of potentially harmful microbes such as Staphylococcus aureus (SA).9 Although SE is generally considered to be an indolent and benign © 2014 American Chemical Society

inhabitant of the healthy host, in immune-compromised people, neonates, and patients with internal prosthetic devices, it can act as an opportunistic pathogen10 and cause chronic infections, indicating the capacity of this species to evade host defenses. Biofilm formation, involving the synthesis of protective matrix polymers or other surface and secreted components, is one of the important factors contributing to immune evasion.11 In SE, biofilms can be generated by various mechanisms, but the polysaccharide intercellular adhesin/poly-N-acetylglucosamine (PIA/PNAG)-mediated biofilm formation is most common and seems to be present in many disease-associated human strains.12 In addition to PIA/PNAG, SE can also produce an exopolymer, poly-γ-glutamic acid (PGA), which is an essential compound of biofilm matrix and functions as a key player in promoting immune evasion during commensal and infectious states of the bacterium.12 Instead of producing destructive toxins, SE is suggested to exploit passive strategies involving, Received: March 29, 2014 Published: July 11, 2014 3748

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762

Journal of Proteome Research

Article

were mapped to the genome sequence of the commensal strain ATCC12228.16 The majority of the gaps in the sequence were closed by PCR; linker PCR with specific primers was used to map the remaining contigs and to join them, if possible. Low molecular weight plasmid DNA was purified separately and used for identifying a single plasmid (ca. 4.3 kb). In addition, the assembly phases revealed the presence of three additional plasmids. Protein coding sequences were predicted using Glimmer, version 3.02,20 as well the RAST (rapid annotation using subsystem technology) server.21 The automated annotation results made by RAST were manually adjusted based on information from different databases including, KEGG, TIGR, Prosite and InterPro, and COG (clusters of orthologous groups).22−24 Prophages were predicted with Prophinder.25 Whole-genome nucleotide alignments were generated with BLAST and visualized with the Artemis Comparison Tool (ACT)26 and DNAplotter.27 IS elements were predicted by blasting the IS database (https://www-is. biotoul.fr).28 Twenty eight RNA genes were predicted using RAST pipeline.21 The genome and plasmids sequences of PM221 were deposited in NCBI GenBank under following accession numbers: chromosome, HG813242; plasmids, HG813243, HG813244, HG813245, and HG813246. Genome and Gene Comparisons. Orthologues and inparalogues among the genes of PM221, ATCC12228, and RP62A were identified using two complementary tools and on the basis of protein sequences from 36 genomes of interest. Genomes included in the analysis are listed in Supporting Information Table 1 and represent nine different species: SE, SA, Staphylococcus capitis, Staphylococcus carnosus, Staphylococcus hemolyticus, Staphylococcus hominis, Staphylococcus lugdunensis, Staphylococcus saprophyticus, and Staphylococcus warneri. First, protein sequences extracted from annotated genome sequence files were compared against each other with BLASTP,29 and orthologue groups were predicted with OrthoMCL.30 OrthoMCL and BLASTP were run with default settings, except for a percent match threshold of 35 and BLASTP set to print up to 10 000 alignments. Second, reciprocal best BLAST hits (RBBH) were extracted, RBBHs were merged with sequences that are reciprocally more similar to each other than to any sequence from the other strain, and fully interconnected RBBH groups were identified using Cliquer.31 In both cases, results were modified to include only proteins from PM221, ATCC12228, and RP62A. Additional protein sequence searches were performed against the entire nonredundant protein sequence database using BLAST with default parameters.29 For constructing the phylogenetic tree, orthologous proteins in the PM221, ATCC12228, RP62A, and S. carnosus TM300 strains were detected using OrthoMCL,30 and orthologue groups were populated by known r-proteins32 (ribosomal proteins) selected. S. carnosus TM300, representing the same genus but a different species, was in the analysis used as an outgroup species. It was noted that an unambiguous orthologue group was built for each of the 54 r-protein classes having exactly one sequence per organism. Protein sequences within these orthologue groups were aligned with MUSCLE,33 and poorly aligned positions were eliminated using GBLOCKS.34 Multiple alignments were concatenated together, and a phylogenetic tree was reconstructed using PhyML35 with parameters appropriate for protein sequences. Proteome Characterization. Theoretical molecular weights (MW) and isoelectric points (pI) for predicted

e.g., defense against antimicrobial peptides to escape the host immune response.13,14 The goal of the present study is to explore the mechanisms by which SE can infect the bovine host and persist in the mammary gland. For this purpose, we determined the whole genome and four plasmid sequences of the SE strain PM221. PM221 originates from spontaneous persistent bovine IMI15 and has previously been tested in an experimental mastitis model.3 The genomic data of PM221 was next compared with those of ATCC12228 and RP62A to reveal conserved genes and genes that were unique to each strain. In addition, 2D DIGE, cell-surface shaving proteomics, and experimental infections in a bovine model were applied to complement the genome-level findings. To the best of our knowledge, the present study is the first to report the genome sequence of a bovine mastitis-causing SE strain, providing new insight into the pathways that contribute to adaptation to and the pathogenic state in the bovine host.



EXPERIMENTAL PROCEDURES

Bacterial Strains and Culture Conditions

SE strains selected for the study included the bovine IMIcausing PM221 strain,4,6 ATCC12228, representing a mildly infectious and nonbiofilm-forming16 strain, and the virulent ATCC35984 (hereafter RP62A)17 strain. These SE strains were routinely cultivated on tryptic soy (TS) agar (TSA) or TS broth (TSB) at 37 °C with shaking at 220 rpm. For 2D DIGE experiments, the strains were grown in 100 mL of TSB as four biological replicates, and 2 mL samples were taken from the midlogarithmic (OD600 ∼ 1.0 ± 0.1) phase and at the onset of the stationary (OD600 ∼ 4.0 ± 0.1) phase (Supporting Information Figure 1). Cells were harvested by centrifugation at 4 °C and washed once with ice-cold 100 mM Tris-HCl, pH 8. For cell-surface shaving proteomics, two biological replicate cultures of PM221 and ATCC12228 were grown under aerobic (220 rpm) and microaerophilic (incubated without agitation) conditions at 37 °C. Bacterial cell samples (3.5 mL) for trypsin treatment were harvested at the transition between the midexponential and late-exponential phases of growth (aerobic cultures, OD600 ∼ 0.9−1.0; microaerophilic cultures, OD600 ∼ 0.7−0.8) by centrifugation (3000g, 10 min, 4 °C) and used immediately for downstream analysis. For whole-genome sequencing analysis, cells were allowed to grow overnight. Cell pellets were stored at −80 °C until required. Extraction of Genomic and Plasmid DNA

Genomic and plasmid DNAs were extracted from planktonic SE cultures using the Illustra Bacteria genomicPrep Mini Spin Kit (GE Healthcare) in combination with lysostaphin (1 μg/ mL) (Sigma) to ensure complete cell lysis and to increase genomic DNA yield. The concentration and purity of the recovered DNA was examined by the NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies Inc., Wilmington, DE, USA) and by agarose (0.8%) gel electrophoresis. Bioinformatics

Genome Sequencing, Assembly, and Annotation. The genome of PM221 was sequenced using the method based on 454 pyrosequencing.19 90.5 Mb of raw sequence data from 486 000 reads was generated on the Roche 454 GS-FLX instrument corresponding to 36× genome coverage. The reads were assembled into 287 contigs longer than 1000 bp using GS De Novo Assembler Version 2.0.00 (Roche). Most of these contigs 3749

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762

Journal of Proteome Research

Article

PM221, ATCC12228, and RP62A proteins were acquired with ProMoST (protein modification screening tool).36 The GRAVY values (grand average hydropathy) were calculated by the Protein GRAVY calculation tool.37 The cellular location for each protein was predicted using the PSORTb.38 The SignalP 4.0 server was used to detect potential signal sequences in each protein sequence, whereas proteins exploiting nonclassical secretion mechanisms were predicted using the SecretomeP 2.0 server.39,40 The presence of potential transmembrane domains was identified using the TMHMM Server v. 2.0.41 Lipoproteins were identified by scanning for a lipobox with the PRED-LIPO prediction program,42 and a lipoprotein consensus sequence LASAGC43 was identified with the ScanProsite tool.44 The presence of the following were obtained using the ScanProsite tool: cell wall motifs and domains; leusine-rich repeat region (LRR) involved in protein−protein and protein−ligand interactions; LysM, a Cterminal lysine that mediates noncovalent attachment to peptidoglycan; motifs for sortase-mediated protein anchoring (LPxTG, NPQTN and EVPTG) (C-terminal anchoring signal motifs); and TLxTC, a sortase signature motif with a catalytic cysteine residue.45−50

significantly changed to compare the differential protein expression pattern between different strains and conditions. Two-way ANOVA was applied to indicate statistically significant strain-specific and/or growth-specific changes in spot volume values. The data were filtered using the average spot volume ratios ≥1.3 (one-way ANOVA, p < 0.05) to assign significant changes in least one of the strains and/or growth conditions tested, and those appearing in at least four of the 12 analyzed gels (i.e., 12 of the 36 gel images) were picked for mass spectrometric (MS) identification. LC−MS/MS Identification. The fluorescent 2DE gels were subjected to MS-compatible silver staining to visualize the protein spots. Selected spots were cut out for in-gel digestion with trypsin followed by LC−MS/MS identification using an Ultimate 3000 nano-LC (Dionex) and QSTAR Elite hybrid quadrupole TOF mass spectrometer (Applied Biosystems/ MDS Sciex) with nano-ESI ionization.52−54 The samples were first concentrated and desalted on a C18 trap column (10 mm × 150 μm, 3 μm, 120 Å; PROTECOL, SGE Analytical Science, Griesheim, Germany) followed by peptide separation on a PepMap100 C18 analytical column (15 cm × 75 μm, 5 μm, 100 Å; LC Packings, Sunnyvale, CA) at 200 nL/min. The separation gradient consisted of 0−50% B in 20 min, 50% B for 3 min, 50−100% B in 2 min, and 100% B for 3 min (buffer A, 0.1% formic acid; buffer B, 0.08% formic acid in 80% acetonitrile). MS data were acquired automatically with Analyst QS 2.0 Software (Applied Biosystems, USA) using SMART IDA. The information-dependent acquisition method consisted of a 0.5 s TOF-MS survey scan of m/z 400−1400. From every survey scan, the two most abundant ions with charge states 2+ to 4+ were selected for product ion scans. Once an ion was selected for MS/MS fragmentation, it was put on an exclusion list for 60 s. The LC−MS/MS data were searched using the Mascot search algorithm (Matrix Science, version 2.2.03) against the inhouse database composed of the unpublished ORF set of PM221 (2509 protein entries) and the published ORF sets of ATCC12228 (2485 protein entries)16 and RP62A (2526 entries)17 through the ProteinPilot3.0 interface (version 2.0.1, Applied Biosystems/MDS SCIEX). The search criteria for Mascot searches were as follows: trypsin digestion with one missed cleavage allowed, carbamidomethyl modification of cysteine as a fixed modification, and oxidation of methionine as variable modification. For the LC−MS/MS spectra, both the maximum precursor ion mass tolerance and MS/MS fragment ion mass tolerance were 0.2 Da, and peptide charge states of +1, +2, and +3 were used. A successful identification was reported when a significant match (p < 0.05) with more than two peptide hits (one peptide with ion score > 45) was obtained. Trypsin-Shaving LC−MS/MS Analysis. Cell-surface bound proteins from PM221 and ATCC12228 were released by trypsin-mediated shaving of intact living cells using a method optimized for SA55 with minor modifications. The pelleted bacterial cells were washed once with 50 mM Tris-HCl containing 17% sucrose (pH 7.0), centrifuged at 3000g for 5 min (4 °C), and suspended gently in 100 μL of TEAB (triethylammonium bicarbonate) containing 17% sucrose (pH 8.5). Enzymatic digestion in TEAB was conducted by adding trypsin to a final concentration of 50 ng/μL (sequencing grade modified porcine trypsin, Promega) (0.047 μg/μL) at 37 °C for 15 min. Released peptides and trypsin was recovered by filtration through a 0.2 μm pore size acetate membrane by

Proteomic Analyses

2D-DIGE Analyses. Four replicate biological samples of 2.0 mL each from the independent cultures of PM221, ATCC12228, and RP62A were withdrawn at two time points of growth (Supporting Information Figure 1). The cells were broken with glass beads, and the extracted and purified proteins (50 μg) were differentially labeled with Cy2, Cy3, or Cy5 dye (CyDye DIGE Fluor minimal dyes; GE Healthcare) using the experimental design outlined in Supporting Information Table 2. The labeled proteins (150 μg in total) were separated using isoelectric focusing (IEF) with rehydrated IPG strips (24 cm, pH 3−10 nonlinear; Bio-Rad) in a Protean IEF Cell (Bio-Rad) using the linear ramping mode, as previously described.51 After IEF and equilibration in buffers containing 50 mM Tris-HCl at pH 6.8, 6 M urea, 2% SDS, 20% glycerol, and alternatively either 2% DTT or 2.5% iodoacetamide (25 min each), the strips were loaded on 12% acrylamide gels. The second dimension separation of proteins was conducted using the Ettan DALTsix Electrophoresis Unit (GE Healthcare), and the fluorescent proteomes after electrophoresis were detected using the FLA-5100 laser scanner (Fujifilm) at wavelengths of 473 nm (for Cy2), 532 nm (Cy3), and 635 nm (Cy5) using voltages of 420, 410, and 400 V, respectively, with 100 μm resolution. The gel images were cropped to an identical size with the aid of the ImageQuant TL 7.0 software (GE Healthcare). A total of 36 gel images were subjected to data analysis using the DeCyder 2D 7.0 software (GE Healthcare). The gels were analyzed using a differential in-gel analysis (DIA) module, which normalized the Cy2, Cy3, and Cy5 images from each gel. The DIA data sets were submitted to biological variation analysis (BVA), which allowed intergel matching and calculation of average spot abundances for each protein spot across the 12 gels. To test for significant differences in the expression of proteins between the experimental groups, oneway analysis of variance (ANOVA) was performed at a significance level p < 0.05. Unsupervised principal component analysis (PCA) and K-means clustering using average linkage were performed using the DeCyder Extended Data Analysis (EDA) module on the group of spots identified as being 3750

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762

Journal of Proteome Research

Article

Figure 1. Schematic representation of the circular chromosome of PM221 (A). The tracks correspond to (from outside in) map location in base pairs (black), protein coding genes on the plus strand (black), and protein coding genes on the minus strand (black), protein coding genes for which at least one paralogue was found in PM221 (red), rRNA genes, tRNA genes, prophage-like regions, potential IS (insertion sequence) elements, GC plot (green/purple), and GC skew (green/purple). (B) Comparison of the whole genomes of ATCC12228, PM221, and RP62A. The corresponding alignments were identified by pairwise bl2seq comparison with a cutoff at a bit score of 3700. Red matches are in the same orientation; blue matches are in the opposite orientation. (C) PHYML analysis of one-to-one protein orthologues from PM221, ATCC12228, and RP62A, with branch lengths representing evolutionary times. Distance scale is displayed above the plot.

centrifugation (16 000g, 2 min, 4 °C), and the recovered trypsin digestions were incubated for 16 h at 37 °C. Digestions were blocked by the addition of trifluoracetic acid to a final concentration of 0.6% (Applied Biosystems). The protein concentration in the supernatant was measured with a NanoDrop (ND 1000, Fisher Scientific) at 280 nm. Viability of the trypsin-digested cells was determined by plating serial dilutions (10−12, 10−14, and 10−16) of cells in 1× phosphate-buffered

saline (PBS) onto TSA, and CFUs (colony forming units/mL) were determined after incubation of 24 h at 37 °C under aerobic conditions. Two biological replicates were performed for each shaving experiment. Trypsin-digested peptides were purified using ZipTips (μC18) (Millipore). Peptides were identified using LC−MS/ MS as described above with the following change: the tryptic peptides were eluted in a linear gradient of acetonitrile from 0 3751

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762

Journal of Proteome Research

Article

to 40% in 120 min (in 0.1% formic acid) at a flow rate of 200 nL/min. For maximizing the reliability and the number of protein identifications, the MS/MS data were searched against the PM221 and ATCC12228 databases described above using the Mascot (version 2.4.0) and Paragon search engines (version 4.0.0.0) through ProteinPilot (version 4.0). The parameters for Paragon searches included the rapid search mode, carbamidomethyl modification of cysteine as a fixed modification, and oxidation of methionine as a variable modification. The Compid tool56 was used to parse significant hits from both data searches. Proteins with Mascot [ms] scores ≥ 50 and p < 0.05 and/or Paragon Unused ProtScores [pg] ≥ 1.3 and p < 0.05 were considered to be reliable protein identifications. The mass spectrometry proteomics data were deposited to the ProteomeXchange Consortium (http://proteomecentral. proteomexchange.org) via the PRIDE partner repository57 with the data set identifier PXD000683. The MS/MS spectra, including fragment ion assignments, for the proteins with single peptide matches are shown in Supporting Information Figure 2. To estimate the false discovery rates (FDRs), all Mascot and Paragon searches were repeated using identical search parameters and validation criteria against PM221 and ATCC12228 decoy databases containing all protein sequences in both the forward and reverse orientations. Sequences were reversed using the Compid tool, and the FDR percentages were calculated using the formula 2nreverse/(nreverse + nforward) given by Elias and co-workers.58 The calculated FDRs were 0% in each of the three strains.

systemic clinical signs were recorded. Infection was defined as persistent if the challenge strain was detected in the milk sample at the final sampling time 2 weeks postchallenge. The Ethics Committee of the University of Helsinki approved the study protocol.



RESULTS AND DISCUSSION

Comparative Genomics Reveals Close Relationship with the Bovine Mastitis and Commensal-Type SE Strains

Epidemiological studies suggest that humans can act as reservoirs of bovine mastitis-causing SE strains.5,6 Yet, the major question of how human skin SE isolates could turn into an important bovine mammary gland-infecting species has remained unclear. To identify the mechanisms underlying adaptation and pathogenicity of SE-mediated bovine IMI, we sequenced the whole genome and plasmid sequences of a bovine mastitis-causing SE strain (hereafter PM221) isolated from a persistent IMI,15 and we compared the defined sequence with the published genome sequences determined for the human SE strains, RP62A and ATCC12228. RP62A represents a biofilm-positive strain that is also multidrug resistant and capable of causing chronic human skin infection,17 and ATCC12228 is a nonbiofilm-forming and less invasive strain with low infectious potential.16 The sequence analyses revealed that the PM221 genome comprises a single 2 490 012 bp long circular chromosome with an average GC content of 32% that encodes 2393 proteins, altogether covering ca. 83% of the double-stranded chromosome (Figure 1A). The average protein size encoded in the genome is 282 amino acids, which corresponds to 847 bp. In addition, this strain carries four plasmids with sizes ranging from 4439 to 58 811 bp and encoding for a total of 116 proteins (Table 1). The overall gene

Phenotypic Characterization

Assessing Formation of Small Colony Variants (SCVs). The SE strains were cultured in TSB at 37 °C with shaking (220 rpm) for 24, 36, and 48 h. At the indicated time points, cell samples were serially diluted (10−8 to 10−15) in 1× PBS and plated onto TSA. Plates were incubated at 37 °C under aerobic conditions for 1 day to obtain colonies. Slow-growing subpopulations of bacteria that appear as significantly smaller colonies than the parent strain18 on the plate were considered SCVs. The experiment was repeated three times. Catalase Activity Measurement. Logarithmic-phase cells from each three strains were tested for catalase activity essentially as described previously.59 Briefly, the strains were grown in TSB until OD600 reached 1.0, at which point cell samples (500 μL) were withdrawn and mixed with equal volumes of Triton X-100 (1%) and 30% hydrogen peroxide (30%). Reactions were mixed and incubated at room temperature. After completion of the reaction (15 min), the height of O2-forming foams in the test tubes were visually compared. The assay was conducted with three biological replicate samples. Experimental Infection. One udder quarter of two primiparous, midlactating dairy cows was experimentally infected with the human SE strain ATCC12228 as previously described.3 Briefly, one udder quarter of each cow was used as an experimental quarter, and another, as a control quarter. The quarters were infused through the teat canal within 30 min after the morning milking using a blunt cannula. The inoculate contained 5.7 × 106 cfu in 7 mL of saline. Milk samples and clinical data were collected at regular intervals over the 2 week study period. Bacterial counts, somatic cell counts (SCC), and milk N-acetyl-β-D-glucosaminidase (NAGase) activity in the milk of the inoculated and control quarters were determined. The cows were examined at every sampling, and local and

Table 1. Comparison of Genomic Features of the BovineSpecific PM221 Strain and Human-Specific ATCC12228 and RP62A Strainsa features

PM221

ATCC12228

RP62A

CDS all RNA tRNA 5S 16S 23S prophages plasmids

2393 77 58 7 6 6 3 4

2452 118 61 6 5 5 1 6

2465 120 59 6 6 6 1 2

a

Prophages were predicted using Prophinder (aclame.ulb.ac.be/ Tools/Prophinder). Other features were predicted using RAST annotation pipeline (rast.nmpdr.org).

organization between the three SE genomes implied that PM221 more closely resembles the commensal-type ATCC12228 strain than the virulent representative human RP62A strain (Figure 1B). A PhyML analysis with 1:1 singlecopy r-protein (ribosomal protein) orthologues from PM221, ATCC12228, RP62A, and evolutionary more distant Staphylococcus species, S. carnosus TM300, confirmed that the bovine strain is evolutionary closer to the commensal-type than to the virulent human strain (Figure 1C). In addition, the fdr (formate dehydrogenase) gene, which was recently reported as a potential marker for discriminating commensal from pathogenic human SE strains,60 was commonly identified only in the 3752

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762

Journal of Proteome Research

Article

genomes of PM221 (SEB_01866, SEB_01877) and ATCC12228 (SE_1853, SE_1854). In addition, PM221 and ATCC12228 were found to be devoid of a CRISPR (clustered regularly interspaced short palindromic repeat) element (Supporting Information Table 3) that is present in RP62A (SERP2462−SERP2455) and which is thought to confer resistance to exogenous genetic elements such as plasmids and phages.61 In the case of PM221, this deficiency might have played some role in the acquisition of three prophage-like elements, amounting to 3.1% of the PM221 genome content (Supporting Information Table 3), as well as high number of specific genes (47 in total) coding for potential phage proteins. One of the identified prophage-like elements was found to be embedded in a 12 kb genomic island that contains two superantigen encoding elements (SEB_01625, SEB_01627) with no counterpart in RP62A or ATCC12228. Pathogenicity islands (PAIs) are typically part of 15−17 kb mobile phagerelated genetic elements delivering superantigen-related diseases and other virulence factors in SA.62 PAIs and phagemediated pathogenesis are less explored in CoNS, with the first enterotoxin-bearing PAI being recently identified from a clinical human SE strain, FRI909.63 To search for functional differences between the bovine and human strains, the predicted protein complements were classified according to the mechanism of cell wall or cell membrane anchoring or the presence of secretion motifs that facilitate protein export out of the cell (Supporting Information Table 3). RP62A was predicted to harbor more sortasecatalyzed surface proteins as well as lipoproteins and proteins secreted into the extracellular milieu, whereas a higher number of transmembrane helices (TMHMM) containing proteins was present in PM221 (Figure 2A). Screening for strain-specific and strain-shared genes indicated the core genome coding for 2130−2151 proteins in bovine and human strains and flexible gene pools comprising 249 (PM221), 198 (ATCC12228), and 302 (RP62A) genes encoding strain-specific proteins (Figure 2B, Supporting Information Table 2). The most distinct differences between the bovine and human strains were associated with gene paralogues; there are close to 600 genes with one or more paralogue in PM221, whereas only 286 and 252 gene paralogues are found in the genomes of ATCC12228 and RP62, respectively. Other specific differences included an arginine catabolic mobile element (ACME), a key virulence factor in some non-biofilm-producing SE strains,64 which was identified only in the PM221 and ATCC12228 (SEB_02215− SEB_02219, SE2214−SE2218) genomes. Although PM221 was found to lack the icaABCD operon genes encoding PIA/PNAG that mediates biofilm formation in RP62A (SERP2292− SERP2296), we have recently demonstrated that this strain is able to promote protein- and DNA-mediated adherent/biofilm growth, suggesting that proteins and DNA play important roles for biofilm formation.65 PM221 was found to share several cellsurface bound and other extracellular (i.e., surfome proteins) virulence factors with the human SE strains, including proteases (HtrA, SspA, SspB, SspC), lipases (Lip, Geh, Geh1/Geh2, LipA), iron and heme transporters (SitADB, HtsABC), toxins/ hemolysins (PSMβ, PSMβ1/2), adhesive matrix proteins (CidA, Aae, AtlE, Aap, EbhAB, Fbp, EbpS, SdrF, SdrH, PGA, SesA,C,E,H), and antigens (SsaA, MCRA) (Supporting Information Table 3). A more detailed comparison demonstrated that PM221 shares more surfome protein encoding genes with ATCC12228 than with RP62A and that it lacks certain sdr and ses family protein encoding genes present in

Figure 2. (A) Diagram representing the number of predicted cell wallanchored proteins and proteins secreted out of the cell in PM221, ATCC12228, and RP62A. TMHMM, proteins with one or more transmembrane helices; lipobox, proteins with an N-terminal lipobox that mediate covalent binding of a conserved Cys residue to a lipid; LysM, proteins with a C-terminal lysine mediating noncovalent attachment to peptidoglycan; secreted, proteins secreted in the extracellular environment; LPxTG, proteins with a C-terminal LPxTG cell wall anchoring signal for covalent attachment to peptidoglycan by sortase; TLxTC, a sortase signature motif containing a catalytic cysteine residue; NPQTN, a C-terminal sorting signal for sortase B in SA; EVPTG, a C-terminal cell wall sorting signal motif; LRR, a 20−30 residues long leucine-rich repeat region involved in protein−ligand and protein−protein interactions (e.g., cell adhesion and signal transduction). (B) Venn diagrams showing the number of all predicted proteins or all predicted surfome proteins with or without counterparts (orthologues/paralogues) in the genomes of PM221, ATCC12228, and RP62A.

RP62A (Figure 2B, Supporting Information Table 3). In conclusion, our findings suggest a close evolutionary relation between the bovine and low-infectious/commensal-type ATCC12228 strains and show that the PM221 genome harbors markers of both commensal and pathogenic lifestyles. Proteome Profiling Demonstrates a Close Relationship between PM221 and ATCC12228

To complement the genome-level comparisons among PM221, ATCC12228, and RP62A, 2D DIGE was next applied to explore proteome dynamics in PM221, ATCC12228, and RP62A cells during the logarithmic and stationary phases of growth (Supporting Information Figure 1). Supporting Information Figure 3 shows representative 2DE image overlays displaying the expression pattern of proteins (in the pI range of 3−10) in each strain at the indicated growth stages. DeCyder analysis resulted in 485 spots showing significant change (oneway ANOVA, p < 0.05) among the six experimental groups (PM221 log , PM221 stat , ATCC12228 log , ATCC12228 stat , RP62Alog, and RP62Astat). The data were further investigated through unsupervised principle component analysis (PCA) and hierarchical clustering analysis (HCA), demonstrating high reproducibility between the independent replicate samples (Figure 3A). The PM221 and ATCC12228 groups (PM221log and ATCC12228log; PM221stat and ATCC12228stat) cluster closer together, whereas those of RP62A form a separate cluster 3753

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762

Journal of Proteome Research

Article

Figure 3. Protein expression patterns in PM221, ATCC12228, and RP62A during logarithmic and stationary phases of growth, as assessed by clustering analyses. (A) PCA analysis showing the clustering of different DIGE spot maps by two principle components and demonstrating high reproducibility between replicate samples within each of six groups (log and stat refer to logarithmic and stationary phase proteomes, respectively). Proteins that participated in PCA analysis were present in at least 80% of the spot maps and passed filtering using a one-way ANOVA (p < 0.05) test. (B) Heat map showing the separation of different experimental groups after the unsupervised HCA, with relative expression values for each protein and a standardized log abundance scale ranging from −1 (green) to +1 (red). Red, green, and black refer to proteins showing increased, decreased, or no significant change in protein abundance, respectively.

Figure 4. (A) Representative overlay image of the fluorescent 2DE gel representing the stationary phase proteomes of PM221 and RP62A. The numbered protein spots were identified from the poststained 2DE gels (Supporting Information Figure 4) and are listed in Table 2 and Supporting Information Table 4. Pink circles refer to proteins that were identified in multiple spots. (B) Representative three-dimensional (3D) image of DIGE spot identified as ClpC (spot 75), which displays decreased abundance in PM221 and ATCC12228 compared to that in RP62A. 3D images and statistics were generated using the BVA module of the DeCyder software.

Expression Patterns of Several Plausible Virulence Factors Were Similar in PM221 and ATCC12228

from the other four experimental groups (Figure 3A). The unsupervised HCA reproduced the same clustering (Figure 3B). Thus, the clustering clearly indicated that the protein expression profiles of PM221 and ATCC12228 are more alike, potentially resulting in similar phenotypic traits for the bovine and ATCC12228 strains, and revealing that the sepsisassociated RP62A human strain differs considerably from the bovine and commensal-type SE strains.

According to DeCyder analyses, 485 protein spots showed a statistically significant change in spot volume abundances (p < 0.05), from which 119 spots displayed at least a 1.3-fold change (p < 0.05) in at least one of the conditions tested (strain and/or growth stage) (Figure 4). These proteins were selected for LC−MS/MS-based identification and are listed in Table 2 and Supporting Information Table 4. Proteins showing differential 3754

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762

Journal of Proteome Research

Article

Table 2. Proteins Showing Strain- and Condition-Specific Changes in PM221, ATCC12228, and RP62A during the Logarithmic and/or Early Stationary Phases of Growtha protein Tricarboxylic Acid Cycle malate: quinone oxidoreductase, Mqo2 2-oxoglutarate oxidoreductase Carbohydrate Metabolism Pyruvate oxidase, CidC/Pyruvate oxidase Nucleotide and Nucleoside Metabolism Ribose-phosphate pyrophosphokinase, RPPK Thymidylate synthase Uracil phosphoribosyl transferase, UPRT Phosphopentomutase Purine Metabolism Phosphoribosylformylglycin amidine synthase, glutamine amidotransferase subunit, PurQ Manganese superoxide dismutase, MnSOD Amino Acid Metabolism Urease beta subunit, UreB

spot no.

strain-specific change (log. to stat. phase)b

1−9 13−14

↑c RP62A, ↓c PM221, ATCC12228 More abundant in RP62A under both conditions

97

↓ PM221, ATCC12228, ↔cRP62A

26−28 29 31 34

↓ PM221, ATCC12228, ↔ RP62A More abundant in RP62A under both conditions ↑ PM221, ATCC12228 ↓ RP62A More abundant in RP62A and ATCC12228, under both conditions

38

More abundant in RP62A under both conditions

78

More abundant in RP62A under both conditions

98

More abundant in RP62A and ATCC12228 under both conditions More abundant in RP62A and ATCC12228 under both conditions

Urease alpha subunit, UreA

99

Catabolite control protein A, CcpA

65

GTP sensing transcriptional pleiotropic repressor, CodY GntR family transcriptional regulator, GntR

66−67 68

ABC transporter EcsA-like protein

71−72

Glycine betaine ABC transport system, ATP-binding protein OpuAA

73

Transcription ↓ in each three strains, more abundant in PM221 and RP62A during log. growth ↓ PM221, ATCC12228, ↔ RP62A Lowest abundance in ATCC12228 during stat. growth

Transport

Stress Response Catalase, KatE

119

Highest abundance in ATCC12228 under both conditions Lowest abundance in ATCC12228 under both conditions

ATP-dependent Clp protease, ATP-binding subunit ClpC/Negative regulator of genetic competence ClpC/MecB Universal stress protein family, UspA

75−76

Highest abundance in ATCC12228 under both conditions ↓ PM221, ATCC12228, ↔ RP62A

77

↑ RP62A, ↓ PM221, ATCC12228

Proline dipeptidase, PepQ Glutamyl endopeptidase precursor SspA

81−82 83−84

↑ RP62A, ↓ PM221, ATCC12228 More abundant in ATCC12228 and RP62A during stat. growth

91

Highest abundance in RP62A under both conditions

94−95

↓ PM221, ATCC12228, ↔ RP62A

Proteolysis

Folate Biosynthesis Dihydrofolate reductase Cell Cycle Chaperone trigger factor, Tf a

Detailed data related to statistical analysis are provided in Supporting Information Table 2. Numbers refer to identified proteins in Figure 4. bLog and stat, logarithmic stages of growth, respectively. c↑ and ↓, protein abundance increased or decreased after switch to stationary phase of growth, respectively. ↔, no change in protein abundances during the logarithmic and stationary phases of growth.

in SA,68 and these proteins were also identified in the present study. From these proteins, metalloenzyme MnSOD displayed a higher abundance in the stationary phase RP62A cells compared to those of PM221 and ATCC12228. MnSOD is required to deal with reactive oxygen species (ROS) and plays an important role in host defense by promoting the optimal antimicrobial activity of neutrophils and other phagocytes.69 Catalase, KatE, is another enzyme dealing with ROS, which was found to be produced at a higher level under both growth conditions in ATCC12228 compared to that in PM221 and RP62A. Increased catalase activity in ATCC12228 was confirmed using a recently described visual approach.59 In this assay, catalase-generated oxygen bubbles are trapped by Triton X-100, forming foam, which was found to be more

abundance included several regulators and potential virulence factors such as Ecs, PepQ, trigger factor, UreAB, GntR, CodY, and OpuCA. CodY is a pleiotropic regulator that responds to nutrient availability in SA and contributes to virulence and stationary phase adaptation by repressing genes involved in nitrogen metabolism and activating the expression of virulence factors.66,67 In our study, CodY was suggested to be downregulated in PM221 (∼2-fold) and ATCC12228 (∼3to 4-fold) after shifting to the stationary phase of growth. On the other hand, the relative abundance of CodY was unaffected by the growth phase in RP62A, in which the regulator displayed 2- to 3-fold higher expression levels in stationary phase cells compared to that in PM221 and ATCC12228. KatA, OpuCA, SspA, PurQ, GntR, and MnSOD are the potential CodY targets 3755

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762

Journal of Proteome Research

Article

multisubunit enzyme is known to function as the major virulence factor during urinary tract infection with CoNS such as Staphylococcus saprophyticus.75 Efficient urease production in ATCC12228 and RP62A could reflect their lifestyles on human skin or on indwelled medical devices, where they encounter urea as a part of body fluids. In our recent exoproteome study, we have shown that RP62A and ATCC12228 are more efficient in hydrolyzing urea compared to that of PM221.76 On the other hand, cow milk is a more alkaline environment compared to the human skin, which could explain the reduced production of the urease enzyme in the bovine strain. Thus, 2D DIGE analyses suggest that PM221 is not as virulent as RP62A and that PM221 is likely to exploit a less-aggressive strategy to establish persistent infection.

abundant in samples containing ATCC12228 cells compared to that of samples containing an equal number of cells of the other two strains (Figure 5). Studies of SA have shown that KatE is

TCA Cycle Activity and Formation of SCVs Are Likely To Be Regulated in a Similar Way in PM221 and ATCC12228

Figure 5. Catalase activity of PM221, ATCC12228, and RP62A using a visual approach. The image shows the height of the foams developed in test tubes after the reaction of catalase with Triton X-100 and hydrogen peroxide. The assay was conducted with three biological replicate samples withdrawn from logarithmically growing cells (OD600 ∼ 1.0). Arrows indicate the differential heights of the generated foams.

The DeCyder analyses indicated a total of 24 protein orthologues appearing with differing pI values on 2DE gels, resulting from changes in amino acid composition or charged modification of these proteins (Figure 4). Some of these proteins were identified as Mqo2 (malate:quinone oxidoreductase), ClpC ATPase (molecular chaperone), CidC (pyruvate oxidase), and 2-oxoglutarate oxidoreductase (Figure 4; Supporting Information Table 4). In ATCC12228 and PM221, these proteins showed decreased abundance or were less produced in stationary phase cells compared to that in RP62A (Supporting Information Table 4). All of these proteins have implications in coordinating TCA cycle activity and late stationary phase survival.77−79 The identification data also suggested that CcpA (catabolite control protein A), a conserved regulator of carbon catabolite repression,78 is less produced in stationary phase cells of ATCC12228 and PM221 compared to that in RP62A. Lack of CidC and ClpC is reported to increase stationary phase survival of SA,73,76 whereas Mqo2 is reported to be increasingly carbonylated in ClpC cells.77 This has been suggested to lead to reduced TCA cycle activity and oxidative damage, which eventually increased stationary phase survival.80 Thus, it could be that the appearance of Mqo2 in several proteins spots is indicative of oxidative carbonylation resulting from reduced synthesis of ClpC in PM221 and ATCC12228. ClpC-deficient SA cells have also been reported to display decreased folate and nucleotide metabolism.79 This is in line with our findings obtained with stationary phase cells of PM221 and ATCC12228; decreased abundance of ClpC is accompanied by reduced levels of dihydrofolate reductase (FolA), thymidylate synthase (ThyA), ribose-phosphate pyrophosphokinase (RppK), phosphopentomutase, and uracil phosphoribosyl-transferase (Upp) (Supporting Information Table 4). Compromised thymidine biosynthesis is closely associated with formation of small colony variants (SCVs), which show unusual colony morphology and superior stationary-phase survival.81,82 This hypothesis was tested on PM221, ATCC12228, and RP62A by plating serially diluted cells onto TSA for different time periods. PM221 and ATCC12228 cells cultured for 24 h produced colonies of different sizes, whereas those generated by RP62A were large and uniform in size (Figure 6). We also observed that the PM221 and ATCC12228 strains cultured for a prolonged time (36 h) resulted in the appearance of proportionally more small-sized colonies, possibly representing SCVs in both strains, whereas uniformly sized colonies were observed for RP62A (Figure 6). Unlike ATCC12228, PM221

important in niche competition on nasal mucosa against pathogens such as Streptococcus pneumoniae and for intracellular survival of SA.70,71 Differences in KatE expression may thus enable the bovine and human strains to adapt to conditions frequently encountered on the human skin or mucous membranes. Other potential virulence factors displaying different abundances included Ecs, PepQ, TF, and UreAB. From these proteins, EcsA, an ATP-binding transporter, was found to be produced in ATCC12228 at levels that were 14−2 -times as much as that in PM221 and RP62A during the logarithmic growth stage (Supporting Information Table 4). In SA, EcsAB is essential for the normal structure and function of the cell wall as well as for virulence and resistance against antimicrobial compounds such as α-defensins.72 PepQ, a potential proline dipeptidase, displayed decreased abundance after the shift from the logarithmic to the stationary phase of growth in PM221 and ATCC12228, whereas in RP62A, this peptidase was increasingly produced after entering the stationary growth phase (Supporting Information Table 4). The significance of intracellular peptidases for infection is not well known, but the inactivation of some peptidase encoding genes such as pepZ (an aminopeptidase Z) is reported to result in a dramatic attenuation of virulence of SA.73 That study also demonstrated that PepZ is required for survival inside human macrophages and that the pepZ expression is highest in the intracellular environment. Chaperone trigger factor TF, representing a ribosomeassociated chaperone, is reported to regulate production of central virulence factors of some Gram-positive pathogens.74 In our study, TF was shown to be constantly produced in RP62A during both growth conditions, whereas it was dramatically downregulated in the other two strains toward the stationary phase of growth (Supporting Information Table 4), implying that downregulation of TF could underpin the reduced virulence in PM221 and ATCC1228. We also identified the alpha and beta subunits of the urease complex, which both were produced more in ATCC12228 and RP62A compared to that in PM221 (Supporting Information Table 4). Urease is encoded by the ureABCDERGD operon, and the resulting 3756

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762

Journal of Proteome Research

Article

ATCC12228 remained persistently infected. Cows infected with ATCC12228 showed no systemic signs such as fever, whereas PM221 induced clinical signs, with half of the cows having fever (>39.5 °C) at one time point.6 In addition, infection with the ATCC12228 strain did not induce any changes in the udder or appearance of the milk. In cows infected with ATCC12228, milk SCC peaked at 27 h PC to 5.0 × 106 cells/mL (4.2−5.7 × 106) and NAGase activity was ∼2.5 pmol 4-MU/min/μL, whereas cows infected with PM221 had milk SCC of 8.8 × 106 cells/mL (range 0.45−10.0 × 106) and milk NAGase activity of ∼4.35 pmol 4-MU/min/μL (Figure 7).

Figure 6. Formation of small colony variants by PM221, ATCC12228, and RP62A assessed by plating the indicated dilutions from PM221, ATCC12228, and RP62A cultures at 24, 36, and 48 hours postinoculation (hpi) onto TSA. Plates were photographed after 24 h of incubation at 37 °C under aerobic conditions.

Figure 7. Experimentally induced IMI in a bovine model using PM221 and ATCC12228. Mean milk somatic cell count (SCC) and NAGase activity in quarters experimentally infected with PM2216 isolated in persistent bovine mastitis (n = 8) and ATCC12228 (n = 2) (representing a low-pathogenic human strain). Symbols: ○, milk SCC for the bovine strain; ●, milk SCC for the human strain; □, milk NAGase activity for the bovine strain; ■, milk NAGase activity for the human strain. SCC and NAGase values of PM221 are from a previous study.6

cells seemed to undergo lysis at 48 h, which may result from the activation of autolysins or phage-related functions. Cell lysis was observed for ATCC12228 after 60 h (data not shown), whereas RP62A cells withdrawn at this time point remained viable, displaying efficient colony forming ability (Figure 6). RP62A has been shown to produce a high number persister cells during the stationary growth phase,14 and, furthermore, oxidative stress may trigger formation of persisters.83 Oxidative stress is maximal in stationary phase cultures,84 which is the likely explanation for the increased abundance of oxidative stress proteins (e.g., MnSOD, universal stress protein A, glutathione reductase) in stationary phase cultures of RP62A. These results suggest that PM221 and ATCC12228 could downregulate the oxidative stress load by decreasing TCA cycle activity and increasing the formation of SCVs. This strategy appears to differ from that of RP62A, which is proposed to coordinate the oxidative damage responses in a way that ensures long-term survival and adaptation toward a persister state.

Although the ATCC12228 strain is also able to induce IMI in a bovine mastitis model, the course of the disease and the inflammatory reaction in the udder were milder than those observed with the bovine PM221 strain. These findings are in line with the previous testing of ATCC12228 in an experimental infection model, which demonstrated that this strain shows a trend toward lower virulence in the foreign body infection mouse model and in the intravascular catheterassociated infection rat model.16 Cell-Surface Shaving Proteomics Suggests Strain- and Condition-Specific Differences between PM221 and ATCC12228

Because the results of the pilot test suggest that ATCC12228 induces bovine IMI with different clinical features compared to those observed with PM221, we next screened for specifically expressed cell-surface proteins with plausible roles in interaction with the bovine host. For this purpose, cell-surface proteins (surfacome) were enzymatically released using trypsin from PM221 and ATCC12228 cells cultured under microaerophilic (M) and aerobic (A) conditions, which represent factors (high- and low-levels of oxygen) that are likely to be encountered during infection. Trypsin digestions with intact alive cells were conducted under conditions that minimize cell lysis and maximize the number of released proteins. Viable counts of SE cells before and after surface shaving demonstrated that the optimized trypsin-digestion conditions did not result in noticeable cell lysis (data not shown). An equal

The Human ATCC12228 Strain Is Able To Induce Persistent Bovine Mastitis

A previous study using an experimental mastitis model with PM221 demonstrated that this strain is able to induce clinical mastitis and to persist in the mammary gland.3 Because both the genome- and proteome-level findings suggest that PM221 is highly similar to the human ATCC12228 strain, we tested the infectious potential of the human strain in a bovine host. Preliminary results of this pilot study involving two cows suggested that ATCC12228 is able to cause infection. The bacterial count in the milk samples was the highest at the second sampling (at 6 h PC): 9.2 log cfu/mL for the ATCC12228 strain, whereas that for PM221 was reported to be 9.3 log cfu/mL.6 One of the two quarters infected with 3757

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762

Journal of Proteome Research

Article

Table 3. Identified Cell-Surface Exposed Proteins with Differing Identification Scores from PM221 and ATCC12228a identified (±)b gene ID

protein

Adhesion, Biofilm Formation, Antigens, Virulence, Stress SE_0175 Aap, accumulation-associated protein SEB_p102526 Aap, accumulation associated protein SE_1169 EbpS elastin binding protein SEB_01191 EbpS, elastin binding protein SE_1578c Ferritin, iron binding protein SE_0383c Iron-binding protein SEB_01954 TpgX, lipoprotein SE_1947 TpgX, lipoprotein SE_0848 PSM β1/2, phenol soluble modulin beta 1/beta 2 SEB_00827 PSM β1/2, phenol soluble modulin beta 1/beta 2 SEB_02475 PSM β1/2, phenol soluble modulin beta 1/beta 2 SE_1634c PSM δ, phenol soluble modulin delta SE_1872 SsaA, secretory antigen A SE_1876 SsaA, secretory antigen precursor SEB_01884 SsaA, secretory antigen precursor SEB_02430c IsaB, immunodominant antigen B SE_p517c Cell-wall anchored protein (LPxTG) SdrE, adhesin of unknown SEB_00253c specificity SE_0331c SdrG, Ser-Asp rich fibrinogenbinding protein SE_2319 Autolysin SE_2231 Autolysin precursor SEB_02314 Autolysin precursor SE_0750 Autolysin AtlE SEB_00726 Autolysin AtlE SEB_02314 Autolysin Aae SE_2319 Autolysin Aae SE_1866c UreG, urease accessory protein SEB_01428 HtrA, serine protease SE_1405 HtrA, serine protease SE_1138c Penicillin-binding protein Potential Moonlighting Proteins SE_0674 ClpB ATPase SEB_00654 ClpB ATPase SEB_00210c ClpC ATPase SE_1349c ClpX ATPase

A

identified (±)b

M

gene ID SE_0551c

++++++

+++++

++++++ +++ ++ ++ − +++ ± +++ +++

++++++ +++ ++ + ++ − +++ +++ +++

+

+

+++



+

+

+++ − +++ + +

+ ++ + + −

+++++

+++

++++



++ ++ + +++++ ++++ + ++ − + + ++++

++ + + +++++ +++ + ++ + − ± +++

− − − +

++++ + ++ ++

SE_1266c SEB_01288 SE_1267 SE_1629 SEB_01643 SE_1630c SE_1268 SEB_01289 SE_1774c SEB_02356 SE_2357 SE_2358c SEB_01260 SE_1240 SE_0311 SEB_00233 SE_1213c SE_0933 SEB_00913 SE_0312 SEB_00234 SE_0561 SEB_00477 SE_1723 SEB_02159 SEB_01734 SE_2156 SE_0557 SEB_00473 SE_1373c

protein ClpP, Clp protease proteolytic subunit DnaJ chaperone protein DnaK chaperone protein DnaK chaperone protein GroEL chaperone protein GroEL chaperone protein GroES chaperone protein GrpE chaperone protein GrpE, chaperone protein Alkaline shock protein 23 AhpC, alkyl hydroperoxide reductase protein C AhpC, alkyl hydroperoxide reductase subunit C AhpF, alkyl hydroperoxide reductase subunit F SodA superoxide dismutase SodA, superoxide dismutase Elongation factor EF-G Elongation factor EF-G Elongation factor EF-P Elongation factor EF-Ts Elongation factor EF-Ts Elongation factor EF-Tu Elongation factor EF-Tu Enolase Enolase Fructose-bisphosphate aldolase Fructose-bisphosphate aldolase class I Fructose-bisphosphate aldolase class II Fructose-bisphosphate aldolase-like protein Gap, glyceraldehyde-3-phosphate dehydrogenase Gap, glyceraldehyde-3-phosphate dehydrogenase Pyk, pyruvate kinase

A

M



+

− ++++ ++++ ++ +++ − ± +++ ++ +++

+ ++++ +++++ ++++ ++ ++ +++ ++ +++ -

+++

+++



±

+ + +++ +++ − +++ +++ ++++ ++++ ++ ++ +++ +++

++++ + + +++ + +++++ +++ +++++ +++ +++ ++++

++

+

++++

+++++

+++

++++

+++

++

+

+++++

a

Strains were cultured under aerobic (A) and mircroaerophilic (M) conditions. b++++++++++, mascot scores [ms] > 20 000; ++++++, [ms] scores > 3000; +++++, [ms] scores ∼ 1500−2500; ++++, [ms] scores ∼ 800−1500; +++, [ms] scores ∼ 250−350; ++, [ms] scores ∼ 100−150; +, [ms] scores ∼ 50−100. −, not identified or identified with [ms] scores < 50. cStrain-specific identification.

associated protein, Aap, that is predicted to be covalently attached to the cell wall via a LPxTG motif and that, according to the identification data, is suggested to be produced at a higher level in PM221 compared to that in ATCC12228. In SE strains that lack PIA/PNAG coding ability, Aap is thought to play the primary role in biofilm formation by promoting intercellular adhesion in the developing biofilm.86 Aap is composed of a 556 amino-acid N-terminal A domain (comprising 10 imperfect repeats of 16 amino acids and a nonrepetitive region) and a B repeat region with a varying number of 128 amino acid repeats known to contribute to adhesion to human corneocytes and to mediate cell-to-cell adhesion via Zn2-dependent dimerization.87−89 Our recent indepth proteome characterization results obtained with PM221

amount of released peptides from two biological replicate cultures were analyzed by LC−MS/MS. The total number of high-quality identifications ([ms] scores > 50 and p < 0.05 and/ or [pg] scores > 1.3 and p < 0.05) were higher with ATCC12228 cultured under both conditions (A: PM221, 89; ATCC12228, 150 and M: PM221, 68; ATCC12228, 245) (Supporting Information Table 5). Although the present study does not provide accurate protein quantification data, higher identification scores and sequence coverage can be used to estimate changes in protein relative abundances between two organisms.85 Using these criteria, commonly identified proteins with clearly different identification scores were next compared, and proteins with likely roles in virulence or adaptation are listed in Table 3. One of these proteins was the accumulation 3758

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762

Journal of Proteome Research

Article

for potential phage-like functions, and plasmid-associated virulence factors were identified in the genome of PM221, which could provide the bovine strain with a higher adaptive potential and an ability to promote adherent growth. Proteomelevel confirmations suggest that PM221 and ATCC12228 regulate late-stationary phase metabolism and generation of SCVs in such a way that could increase the viability and persistence of the bacteria. An additional pilot study indicated that ATCC12228 was able to induce persisting IMI in a bovine model with milder clinical symptoms compared to PM221. Although additional protein identification analyses suggested differences with respect to the expression of certain cell-surface adhesins and adhesive moonlighting proteins, further studies are required to uncover mechanistic details behind PM221- and ATCC12228-mediated IMI. Altogether, the present study strengthens the prevailing hypothesis that SE strains originating from humans are able to induce bovine mastitis and suggests that SE is likely to exploit diverse strategies to achieve successful infection.

and ATCC12228 indicated higher expression level for Aap in PM221.62 In addition, this study revealed that PM221 forms a protein-mediated biofilm on a polystyrene support in vitro.65 Although the individual proteins mediating PM221 biofilm formation remain to be identified, Aap provides an interesting candidate for these further studies. Furthermore, the Aap encoding gene (SEB_p102468) was shown to be located on a plasmid in PM221, whereas in ATCC12228 (SE0175) and RP62 (SERP2398), Aap is predicted be chromosomally encoded. These findings imply that the bovine strain has the possibility to transfer the Aap encoding plasmid to other bacteria, which could make them more pathogenic. Specifically identified PM221 proteins were nucleic acid binding cell-surface protein IsaB (i.e., immunodominant antigen B) and the complement factor H binding cell-surface proteins SdrE,90,91 which suggests that PM221 may differ from ATCC12228 in terms of immunomodulatory functions. Surfome proteins assumed to be more abundantly or specifically produced by ATCC12228 included phenol-soluble modulin delta (PSMδ), which has reported antimicrobial activity,92 a cell wall-anchored protein (LPxTG motif), autolysins Aae and AtlE, penicillin- and two iron-binding proteins, and fibrinogen-binding protein SdrG. In contrast to PM221, the ATCC12228 seems to produce more moonlighting proteins with significantly higher identification scores, suggesting that the production of these multitasking proteins is more efficient in ATCC12228 under the conditions tested. Moonlighting proteins are conserved cytoplasmic proteins involved in metabolic pathways or the cell stress response, which are exposed at the bacterial surface to take on additional activities contributing to, e.g., virulence or bacterial benefit.93,94 The identified moonlighting proteins with adhesive and immunomodulatory properties included elongation factors Tu, Ts, and G (EfTu, EfTs, EfG), glycolytic proteins (GapPDH, Fba, Pyk), enolase, superoxide dismutase (SodA), and chaperones (GroEL, DnaK).94 In addition, the identification data suggests that the production of several surfome proteins is conditiondependent. These included, e.g., the serine-type protease HtrA that could be identified only from aerobically cultured cells of ATCC12228 and PM221. In the case of PM221, conditiondependent identification was obtained with potential lipoproteins TpgX and SdrE, which are suggested to be produced more under microaerophilic and aerobic conditions, respectively. In ATCC12228, the growth conditions are likely to have a great impact on the expression of ferritin-binding proteins, SdrG, and potential moonlighting proteins such as enolase, pyruvate kinase, DnaK, and GroEL. Thus, the surfacome comparisons suggest both strain-specific and condition-dependent variation in cell wall-anchored and moonlighting proteins between the PM221 and ATCC12228 strains. It remains to be seen if these proteins can account for some of the clinical phenotypes of PM221 and ATCC12228 observed in vivo.



ASSOCIATED CONTENT

S Supporting Information *

Figure 1: Cell density of four biological replicate cultures of PM221, ATCC12228, and RP62A in TSB as a function of time. An equal amount of cells were withdrawn for 2D DIGE analyses at the indicated time points. Figure 2: MS/MS spectra and peaklists for single-peptide-identifications. Figure 3: Pseudocolor 2DE images representing protein expression in PM221, ATCC12228, and RP62A strains during logarithmic and stationary phases of growth. Protein samples (400 μg) were labeled using the experimental setup described in Supporting Information Table 1 prior to 2DE involving IEF with 24 cm IPG strips (pH 3 to 11) and 12% SDS-PAGE. Protein spots appearing in purple or green were either more or less abundant in the indicated strains. Protein spots appearing in yellow showed no strain-dependent difference in abundance. Table 1: Staphylococcal strains used in comparative genomics for screening genes that are common and specific to PM221, ATCC12228, and RP62A. Table 2: Setup of a DIGE experiment using four repeats of PM221, ATCC12228 and RP62A protein samples and a dye swap between Cy3 and Cy5. Log and stat refer to protein samples extracted from logarithmic and stationary phases of growth, respectively. Table 3: List of all predicted genes in the genome and the four plasmids of the bovine mastitis-causing S. epidermidis PM221 strain. Relevant information related to protein/gene functions, protein conservation, physicochemical characters, subcellular location, and the presence of cell-membrane or peptidoglycananchoring motifs and domains are described in materials and methods. Prophage elements and potential transposases/ transposable elements are shaded in gray and blue, respectively. Table 4: The PM221, ATCC12228, and RP62A proteomes at two stages of growth (logarithmic vs stationary phase of growth) analyzed by 2D DIGE. Table 5: Cell-surface-exposed proteins identified from PM221 cultured under aerobic and microaerophilic conditions. This material is available free of charge via the Internet at http://pubs.acs.org.



CONCLUSIONS It is currently believed that humans could be a major source of SE-induced IMIs in dairy cattle, but the mechanisms by which SE adapts to infect the bovine is poorly understood. Here, we report the first genome sequence of a bovine IMI-associated SE strain (PM221) and demonstrate by genome- and proteomelevel comparisons that the bovine strain more closely resembles the commensal-type and low-infectious human ATCC12228 strain than the sepsis-associated virulent human RP62A strain. A significantly higher number paralogous genes, genes coding



AUTHOR INFORMATION

Corresponding Authors

*(A.I.) E-mail: antti.iivanainen@helsinki.fi. Tel.: +358 2 941 57034. 3759

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762

Journal of Proteome Research

Article

*(P.A.) E-mail: petri.auvinen@helsinki.fi. Tel.: +358 2 941 58902 *(P.V.) E-mail: pekka.varmanen@helsinki.fi. Tel.: +358 2 941 57057. Fax: +358 2 941 58575.

(14) Shapiro, J. A.; Nguyen, V. L.; Chamberlain, N. R. Evidence for persisters in Staphylococcus epidermidis RP62a planktonic cultures and biofilms. J. Med. Microbiol. 2011, 60, 950−960. (15) Taponen, S.; Koort, J.; Björkroth, J.; Saloniemi, H.; Pyörälä, S. Bovine intramammary infections caused by coagulase-negative staphylococci may persist throughout lactation according to amplified fragment length polymorphism-based analysis. J. Dairy Sci. 2007, 90, 3301−3307. (16) Zhang, Y. Q.; Ren, S. X.; Li, H. L.; Wang, Y. X.; Fu, G.; Yang, J.; Qin, Z. Q.; Miao, Y. G.; Wang, W. Y.; Chen, R. S.; Shen, Y.; Chen, Z.; Yuan, Z. H.; Zhao, G. P.; Qu, D.; Danchin, A.; Wen, Y. M. Genomebased analysis of virulence genes in a non-biofilm-forming Staphylococcus epidermidis strain (ATCC 12228). Mol. Microbiol. 2003, 49, 1577−1593. (17) Gill, S. R.; Fouts, D. E.; Archer, G. L.; Mongodin, E. F.; Deboy, R. T.; Ravel, J.; Paulsen, I. T.; Kolonay, J. F.; Brinkac, L.; Beanan, M.; Dodson, R. J.; Daugherty, S. C.; Madupu, R.; Angiuoli, S. V.; Durkin, A. S.; Haft, D. H.; Vamathevan, J.; Khouri, H.; Utterback, T.; Lee, C.; Dimitrov, G.; Jiang, L.; Qin, H.; Weidman, J.; Tran, K.; Kang, K.; Hance, I. R.; Nelson, K. E.; Fraser, C. M. Insights on evolution of virulence and resistance from the complete genome analysis of an early methicillin-resistant Staphylococcus aureus strain and a biofilmproducing methicillin-resistant Staphylococcus epidermidis strain. J. Bacteriol. 2005, 187, 2426−2438. (18) Proctor, R. A.; von Eiff, C.; Kahl, B. C.; Becker, K.; McNamara, P.; Herrmann, M.; Peters, G. Small colony variants: a pathogenic form of bacteria that facilitates persistent and recurrent infections. Nat. Rev. Microbiol. 2006, 4, 295−305. (19) Margulies, M.; Egholm, M.; Altman, W. E.; Attiya, S.; Bader, J. S.; Bemben, L. A.; Berka, J.; Braverman, M. S.; Chen, Y. J.; Chen, Z.; Dewell, S. B.; Du, L.; Fierro, J. M.; Gomes, X. V.; Godwin, B. C.; He, W.; Helgesen, S.; Ho, C. H.; Irzyk, G. P.; Jando, S. C.; Alenquer, M. L.; Jarvie, T. P.; Jirage, K. B.; Kim, J. B.; Knight, J. R.; Lanza, J. R.; Leamon, J. H.; Lefkowitz, S. M.; Lei, M.; Li, J.; Lohman, K. L.; Lu, H.; Makhijani, V. B.; McDade, K. E.; McKenna, M. P.; Myers, E. W.; Nickerson, E.; Nobile, J. R.; Plant, R.; Puc, B. P.; Ronan, M. T.; Roth, G. T.; Sarkis, G. J.; Simons, J. F.; Simpson, J. W.; Srinivasan, M.; Tartaro, K. R.; Tomasz, A.; Vogt, K. A.; Volkmer, G. A.; Wang, S. H.; Wang, Y.; Weiner, M. P.; Yu, P.; Begley, R. F.; Rothberg, J. M. Genome sequencing in microfabricated high-density picolitre reactors. Nature 2005, 437, 376−380. (20) Delcher, A. L.; Bratke, K. A.; Powers, E. C.; Salzberg, S. L. Identifying bacterial genes and endosymbiont DNA with Glimmer. Bioinformatics 2007, 23, 673−679. (21) Aziz, R. K.; Bartels, D.; Best, A. A.; DeJongh, M.; Disz, T.; Edwards, R. A.; Formsma, K.; Gerdes, S.; Glass, E. M.; Kubal, M.; Meyer, F.; Olsen, G. J.; Olson, R.; Osterman, A. L.; Overbeek, R. A.; McNeil, L. K.; Paarmann, D.; Paczian, T.; Parrello, B.; Pusch, G. D.; Reich, C.; Stevens, R.; Vassieva, O.; Vonstein, V.; Wilke, A.; Zagnitko, O. The RAST server: rapid annotations using subsystems technology. BMC Genomics 2008, 9, 75. (22) Hunter, S.; Apweiler, R.; Attwood, T. K.; Bairoch, A.; Bateman, A.; Binns, D.; Bork, P.; Das, U.; Daugherty, L.; Duquenne, L.; Finn, R. D.; Gough, J.; Haft, D.; Hulo, N.; Kahn, D.; Kelly, E.; Laugraud, A.; Letunic, I.; Lonsdale, D.; Lopez, R.; Madera, M.; Maslen, J.; McAnulla, C.; McDowall, J.; Mistry, J.; Mitchell, A.; Mulder, N.; Natale, D.; Orengo, C.; Quinn, A. F.; Selengut, J. D.; Sigrist, C. J.; Thimma, M.; Thomas, P. D.; Valentin, F.; Wilson, D.; Wu, C. H.; Yeats, C. InterPro: the integrative protein signature database. Nucleic Acids Res. 2009, 37, 211−215. (23) Tatusov, R. L.; Natale, D. A.; Garkavtsev, I. V.; Tatusova, T. A.; Shankavaram, U. T.; Rao, B. S.; Kiryutin, B.; Galperin, M. Y.; Fedorova, N. D.; Koonin, E. V. The COG database: new developments in phylogenetic classification of proteins from complete genomes. Nucleic Acids Res. 2001, 29, 22−28. (24) Moriya, Y.; Itoh, M.; Okuda, S.; Yoshizawa, A. C.; Kanehisa, M. KAAS: an automatic genome annotation and pathway reconstruction server. Nucleic Acids Res. 2007, 35, 182−185.

Author Contributions ○

These authors equally contributed to this work.

Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This study was supported by the Academy of Finland (grant nos. 139296 to P. Varmanen and 135628 and 140950 to T. Nyman), the Ministry of Agriculture and Forestry (grant no. 828/312/2009 to A. Iivanainen and P. Varmanen,) and the Walter Ehrström Foundation to K. Savijoki.



REFERENCES

(1) Taponen, S.; Simojoki, H.; Haveri, M.; Larsen, H. D.; Pyörälä, S. Clinical characteristics and persistence of bovine mastitis caused by different species of coagulase-negative staphylococci identified with API or AFLP. Vet. Microbiol. 2006, 115, 199−207. (2) Pyörälä, S.; Taponen, S. Coagulase-negative staphylococci emerging mastitis pathogens. Vet. Microbiol. 2009, 134, 3−8. (3) Simojoki, H.; Salomäki, T.; Taponen, S.; Iivanainen, A.; Pyörälä, S. Innate immune response in experimentally induced bovine intramammary infection with Staphylococcus simulans and S. epidermidis. Vet. Res. 2011, 42, 49. (4) Simojoki, H.; Hyvönen, P.; Plumed Ferrer, C.; Taponen, S.; Pyörälä, S. Is the biofilm formation and slime producing ability of coagulase-negative staphylococci associated with the persistence and severity of intramammary infection? Vet. Microbiol. 2012, 158, 344− 352. (5) Thorberg, B. M.; Kühn, I.; Aarestrup, F. M.; Brändström, B.; Jonsson, P.; Danielsson-Tham, M. L. Pheno- and genotyping of Staphylococcus epidermidis isolated from bovine milk and human skin. Vet. Microbiol. 2006, 115, 163−172. (6) Jaglic, Z.; Michu, E.; Holasova, M.; Vlkova, H.; Babak, V.; Kolar, M.; Bardon, J.; Schlegelova, J. Epidemiology and characterization of Staphylococcus epidermidis isolates from humans, raw bovine milk and a dairy plant. Epidemiol. Infec. 2010, 138, 772−782. (7) Piessens, V.; Van Coillie, E.; Verbist, B.; Supré, K.; Braem, G.; Van Nuffel, A.; De Vuyst, L.; Heyndrickx, M.; De Vliegher, S. Distribution of coagulase-negative Staphylococcus species from milk and environment of dairy cows differs between herds. J. Dairy Sci. 2011, 94, 2933−2944. (8) Schoenfelder, S. M.; Lange, C.; Eckart, M.; Hennig, S.; Kozytska, S.; Ziebuhr, W. Success through diversity − how Staphylococcus epidermidis establishes as a nosocomial pathogen. Int. J. Med. Microbiol. 2010, 300, 380−386. (9) Iwase, T.; Uehera, Y.; Shinji, H.; Tajima, A.; Seo, H.; Takada, K.; Agata, T.; Mizunoe, Y. Staphylococcus epidermidis Esp inhibits Staphylococcus aureus biofilm formation and nasal colonization. Nature 2010, 465, 346−349. (10) Otto, M. Staphylococcus epidermidis − the ‘accidental’ pathogen. Nat. Rev. Microbiol. 2009, 7, 555−567. (11) Otto, M. Molecular basis of Staphylococcus epidermidis infections. Semin. Immunopathol. 2012, 34, 201−214. (12) Fey, P. D.; Olson, M. E. Current concepts in biofilm formation of Staphylococcus epidermidis. Future Microbiol. 2010, 5, 917−933. (13) Vuong, C.; Dürr, M.; Carmody, A. B.; Peschel, A.; Klebanoff, S. J.; Otto, M. Regulated expression of pathogen-associated molecular pattern molecules in Staphylococcus epidermidis: quorum-sensing determines pro-inflammatory capacity and production of phenolsoluble modulins. Cell. Microbiol. 2004, 6, 753−759. 3760

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762

Journal of Proteome Research

Article

(25) Lima-Mendez, G.; Van Helden, J.; Toussaint, A.; Leplae, R. Prophinder: a computational tool for prophage prediction in prokaryotic genomes. Bioinformatics 2008, 24, 863−865. (26) Carver, T. J.; Rutherford, K. M.; Berriman, M.; Rajandream, M. A.; Barrell, B. G.; Parkhill, J. ACT: the artemis comparison tool. Bioinformatics 2005, 16, 3422−3423. (27) Carver, T.; Thomson, N.; Bleasby, A.; Berriman, M.; Parkhill, J. DNAPlotter: circular and linear interactive genome visualization. Bioinformatics 2009, 1, 119−120. (28) Siguier, P.; Perochon, J.; Lestrade, L.; Mahillon, J.; Chandler, M. ISfinder: the reference centre for bacterial insertion sequences. Nucleic Acids Res. 2006, 1, D32−D36. (29) Altschul, S. F.; Wootton, J. C.; Gertz, E. M.; Agarwala, R.; Morgulis, A.; Schäffer, A. A.; Yu, Y.-K. Protein database searches using compositionally adjusted substitution matrices. FEBS J. 2005, 272, 5101−5109. (30) Li, L.; Stoeckert, C. J.; Roos, D. S., Jr. OrthoMCL: identification of ortholog groups for eukaryotic genomes. Genome Res. 2003, 13, 2178−2189. (31) Niskanen, S.; Ö stergård, P. R. J. Cliquer User’s Guide, Version 1.0; Technical Report T48 edn, Communications Laboratory; Helsinki University of Technology: Espoo, Finland, 2003. (32) Yutin, N.; Puigbò, P.; Koonin, E. V.; Wolf, Y. I. Phylogenomics of prokaryotic ribosomal proteins. PLoS One 2012, 7, e36972. (33) Edgar, R. C. MUSCLE: a multiple sequence alignment method with reduced time and space complexity. BMC Bioinf. 2004, 5, 113. (34) Castresana, J. Selection of conserved blocks from multiple alignments for their use in phylogenetic analysis. Mol. Biol. Evol. 2000, 17, 540−552. (35) Guindon, S.; Gascuel, O. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst. Biol. 2003, 52, 696−704. (36) Halligan, B. D.; Ruotti, V.; Jin, W.; Laffoon, S.; Twigger, S. N.; Dratz, E. A. ProMoST (protein modification screening tool): a webbased tool for mapping protein modifications on two-dimensional gels. Nucleic Acids Res. 2004, 32, 638−644. (37) Kyte, J.; Doolittle, R. F. A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 1982, 157, 105−132. (38) Yu, N. Y.; Wagner, J. R.; Laird, M. R.; Melli, G.; Rey, S.; Lo, R.; Dao, P.; Sahinalp, S. C.; Ester, M.; Foster, L. J.; Brinkman, F. S. PSORTb 3.0: improved protein subcellular localization prediction with refined localization subcategories and predictive capabilities for all prokaryotes. Bioinformatics 2010, 26, 1608−1615. (39) Petersen, T. N.; Brunak, S.; von Heijne, G.; Nielsen, H. SignalP 4.0: discriminating signal peptides from transmembrane regions. Nat. Methods 2011, 8, 785−786. (40) Bendtsen, J. D.; Kiemer, L.; Fausboll, A.; Brunak, S. Nonclassical protein secretion in bacteria. BMC Microbiol. 2005, 5, 1−13. (41) Krogh, A.; Larsson, B.; von Heijne, G.; Sonnhammer, E. L. Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J. Mol. Biol. 2001, 305, 567− 580. (42) Bagos, P. G.; Tsirigos, K. D.; Liakopoulos, T. D.; Hamodrakas, S. T. Prediction of lipoprotein signal peptides in Gram-positive bacteria with a hidden Markov model. J. Proteome Res. 2008, 7, 5082− 5093. (43) Sutcliffe, I. C.; Harrington, D. J. Pattern searches for the identification of putative lipoprotein genes in Gram-positive bacterial genomes. Microbiology 2002, 148, 2065−2077. (44) de Castro, E.; Sigrist, C. J.; Gattiker, A.; Bulliard, V.; LangendijkGenevaux, P. S.; Gasteiger, E.; Bairoch, A.; Hulo, N. ScanProsite: detection of PROSITE signature matches and ProRule-associated functional and structural residues in proteins. Nucleic Acids Res. 2006, 34, 362−365. (45) Matsushima, N.; Miyashita, H.; Mikami, T.; Kuroki, Y. A nested leucine rich repeat (LRR) domain: the precursor of LRRs is a ten or eleven residue motif. BMC Microbiol. 2010, 10, 235.

(46) Scott, J. R.; Barnett, T. C. Surface proteins of Gram-positive bacteria and how they get there. Annu. Rev. Microbiol. 2006, 60, 397− 423. (47) Heilmann, C.; Hartleib, J.; Hussain, M. S.; Peters, G. The multifunctional Staphylococcus aureus autolysin aaa mediates adherence to immobilized fibrinogen and fibronectin. Infect. Immun. 2005, 73, 4793−4802. (48) Desvaux, M.; Dumas, E.; Chafsey, I.; Hebraud, M. Protein cell surface display in Gram-positive bacteria: from single protein to macromolecular protein structure. FEMS Microbiol. Lett. 2006, 256, 1− 15. (49) Mazmanian, S. K.; Ton-That, H.; Su, K.; Schneewind, O. An iron-regulated sortase anchors a class of surface protein during Staphylococcus aureus pathogenesis. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 2293−2298. (50) Scott, J. R.; Zahner, D. Pili with strong attachments: Grampositive bacteria do it differently. Mol. Microbiol. 2006, 62, 320−330. (51) Koskenniemi, K.; Koponen, J.; Kankainen, M.; Savijoki, K.; Tynkkynen, S.; de Vos, W. M.; Kalkkinen, N.; Varmanen, P. Proteome analysis of Lactobacillus rhamnosus GG using 2-D DIGE and mass spectrometry shows differential protein production in laboratory and industrial-type growth media. J. Proteome Res. 2009, 8, 4993−5007. (52) O’Connell, K. L.; Stults, J. T. Identification of mouse liver proteins on two-dimensional electrophoresis gels by matrix-assisted laser desorption/ionization mass spectrometry of in situ enzymatic digests. Electrophoresis 1997, 18, 349−359. (53) Shevchenko, A.; Wilm, M.; Vorm, O.; Mann, M. Mass spectrometric sequencing of proteins silver-stained polyacrylamide gels. Anal. Chem. 1996, 68, 850−858. (54) Nyman, T. A.; Rosengren, A.; Syyrakki, S.; Pellinen, T. P.; Rautajoki, K.; Lahesmaa, R. A proteome database of human primary T helper cells. Electrophoresis 2001, 22, 4375−4382. (55) Rodriguez-Ortega, M. J.; Norais, N.; Bensi, G.; Liberatori, S.; Capo, S.; et al. Characterization and identification of vaccine candidate proteins through analysis of the group A Streptococcus surface proteome. Nat. Biotechnol. 2006, 24, 191−197. (56) Lietzén, N.; Natri, L.; Nevalainen, O. S.; Salmi, J.; Nyman, T. A. Compid: a new software tool to integrate and compare MS/MS based protein identification results from Mascot and Paragon. J. Proteome Res. 2010, 9, 6795−6800. (57) Vizcaino, J. A.; Cote, R. G.; Csordas, A.; Dianes, J. A.; Fabregat, A.; Foster, J. M.; Griss, J.; Alpi, E.; Birim, M.; Contell, J.; O’Kelly, G.; Schoenegger, A.; Ovelleiro, D.; Perez-Riverol, Y.; Reisinger, F.; Rios, D.; Wang, R.; Hermjakob, H. The PRoteomics IDEntifications (PRIDE) database and associated tools: status in 2013. Nucleic Acids Res. 2013, 41, D1063−D1069. (58) Elias, J. E.; Gygi, S. P. Target-decoy search strategy for increased confidence in large-scale protein identifications by mass spectrometry. Nat. Methods 2007, 4, 207−214. (59) Iwase, T.; Tajima, A.; Sugimoto, S.; Okuda, K.; Hironaka, I.; Kamata, Y.; Takada, K.; Mizunoe, Y. A simple assay for measuring catalase activity: a visual approach. Sci. Rep. 2013, 3, 3081. (60) Conlan, S.; Mijares, L. A.; NISC Comparative Sequencing Program; Becker, J.; Blakesley, R. W.; Bouffard, G. G.; Brooks, S.; Coleman, H.; Gupta, J.; Gurson, N.; Park, M.; Schmidt, B.; Thomas, P. J.; Otto, M.; Kong, H. H.; Murray, P. R.; Segre, J. A.; Staphylococcus epidermidis pan-genome sequence analysis reveals diversity of skin commensal and hospital infection-associated isolates. Genome Biol. 2012, 13, R64. (61) Barrangou, R.; Fremaux, C.; Deveau, H.; Richards, M.; Boyaval, P.; Moineau, S.; Romero, D. A.; Horvath, P. CRISPR provides acquired resistance against viruses in prokaryotes. Science 2007, 315, 1709−1712. (62) Novick, R.; Subedi, A. The SaPIs: mobile pathogenicity islands of Staphylococcus. In Superantigens and Superallergens; Marone, G, Ed.; Karger: Basel, Switzerland, 2007; Vol. 93, pp 42−57. (63) Madhusoodanan, J.; Seo, K. S.; Remortel, B.; Park, J. Y.; Hwang, S. Y.; Fox, L. K.; Park, Y. H.; Deobald, C. F.; Wang, D.; Liu, S.; Daugherty, S. C.; Gill, A. L.; Bohach, G. A.; Gill, S. R. An enterotoxin3761

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762

Journal of Proteome Research

Article

bearing pathogenicity island in Staphylococcus epidermidis. J. Bacteriol. 2011, 193, 1854−1862. (64) Diep, B. A.; Stone, G. G.; Basuino, L.; Graber, C. J.; Miller, A.; des Etages, S. A.; Jones, A.; Palazzolo-Ballance, A. M.; PerdreauRemington, F.; Sensabaugh, G. F.; DeLeo, F. R.; Chambers, H. F. The arginine catabolic mobile element and staphylococcal chromosomal cassette mec linkage: convergence of virulence and resistance in the USA300 clone of methicillin-resistant Staphylococcus aureus. J. Infect. Dis. 2008, 197, 1523−1530. (65) Siljamäki, P.; Varmanen, P.; Kankainen, M.; Pyörälä, S.; Karonen, T.; Iivanainen, A.; Auvinen, P.; Paulin, L.; Laine, P. K.; Taponen, S.; Simojoki, H.; Sukura, A.; Nyman, T. A.; Savijoki, K. Comparative proteome profiling of bovine and human Staphylococcus epidermidis strains for screening specifically expressed virulence and adaptation proteins. Proteomics 2014, DOI: 10.1002/pmic.201300275. (66) Pohl, K.; Francois, P.; Stenz, L.; Schlink, F.; Geiger, T.; Herbert, S.; Goerke, C.; Schrenzel, J.; Wolz, C. CodY in Staphylococcus aureus: a regulatory link between metabolism and virulence gene expression. J. Bacteriol. 2009, 191, 2953−2963. (67) Sonenshein, A. L. CodY, a global regulator of stationary phase and virulence in Gram-positive bacteria. Curr. Opin. Microbiol. 2005, 8, 203−207. (68) Majerczyk, C. D.; Dunman, P. M.; Luong, T. T.; Lee, C. Y.; Sadykov, M. R.; Somerville, G. A.; Bodi, K.; Sonenshein, A. L. Direct targets of CodY in Staphylococcus aureus. J. Bacteriol. 2010, 192, 2861− 2877. (69) Miller, R. A.; Britigan, B. E. Role of oxidants in microbial pathophysiology. Clin. Microbiol. Rev. 1997, 10, 1−18. (70) Park, B.; Nizet, V.; Liu, G. Y. Role of Staphylococcus aureus catalase in niche competition against Streptococcus pneumonia. J. Bacteriol. 2008, 190, 2275−2278. (71) Das, D.; Bishayi, B. Staphylococcal catalase protects intracellularly survived bacteria by destroying H2O2 produced by the murine peritoneal macrophages. Microb. Pathog. 2009, 47, 57−67. (72) Jonsson, I. M.; Juuti, J. T.; François, P.; AlMajidi, R.; Pietiäinen, M.; Girard, M.; Lindholm, C.; Saller, M. J.; Driessen, A. J.; Kuusela, P.; Bokarewa, M.; Schrenzel, J.; Kontinen, V. P. Inactivation of the Ecs ABC transporter of Staphylococcus aureus attenuates virulence by altering composition and function of bacterial wall. PLoS One 2010, 5, e14209. (73) Carroll, R. K.; Robison, T. M.; Rivera, F. E.; Davenport, J. E.; Jonsson, I. M.; Florczyk, D.; Tarkowski, A.; Potempa, J.; Koziel, J.; Shaw, L. N. Identification of an intracellular M17 family leucine aminopeptidase that is required for virulence in Staphylococcus aureus. Microbes Infect. 2012, 14, 989−999. (74) Wu, T.; Zhao, Z.; Zhang, L.; Ma, H.; Lu, K.; Ren, W.; Liu, Z.; Chang, H.; Bei, W.; Qiu, Y.; Chen, H. Trigger factor of Streptococcus suis is involved in stress tolerance and virulence. Microb. Pathog. 2011, 51, 69−76. (75) Gatermann, S.; Crossley, K. B. Urinary tract infections. In Staphylococci in Human Disease; Crossley, K. B., Jeffersson, K. K., Archer, G. L., Fowler, V. G., Jr., Eds.; Wiley-Blackwell: West Sussex, UK, 2009; pp 455−469. (76) Siljamäki, P.; Varmanen, P.; Kankainen, M.; Sukura, A.; Savijoki, K.; Nyman, T. A. Comparative exoprotein profiling of different Staphylococcus epidermidis strains reveals potential link between nonclassical protein export and virulence. J. Proteome Res. 2014, 13, 3249− 3261. (77) Patton, T. G. The Staphylococcus aureus cidC gene encodes a pyruvate oxidase that affects acetate metabolism and cell death in stationary phase. Mol. Microbiol. 2005, 56, 1664−1674. (78) Sadykov, M. R.; Hartmann, T.; Mattes, T. A.; Hiatt, M.; Jann, N. J.; Zhu, Y.; Ledala, N.; Landmann, R.; Herrmann, M.; Rohde, H.; Bischoff, M.; Somerville, G. A. CcpA coordinates central metabolism and biofilm formation in Staphylococcus epidermidis. Microbiology 2011, 157, 3458−3468. (79) Chatterjee, I.; Schmitt, S.; Batzilla, C. F.; Engelmann, S.; Keller, A.; Ring, M. W.; Kautenburger, R.; Ziebuhr, W.; Hecker, M.; Preissner, K. T.; Bischoff, M.; Proctor, R. A.; Beck, H. P.; Lenhof, H. P.;

Somerville, G. A.; Herrmann, M. Staphylococcus aureus ClpC ATPase is a late growth phase effector of metabolism and persistence. Proteomics 2009, 9, 1152−1176. (80) Chatterjee, I.; Maisonneuve, E.; Ezraty, B.; Herrmann, M.; Dukan, S. Staphylococcus aureus ClpC is involved in protection of carbon-metabolizing enzymes from carbonylation during stationary growth phase. Int. J. Med. Microbiol. 2011, 301, 341−346. (81) Chatterjee, I.; Herrmann, M.; Proctor, R. A.; Peters, G.; Kahl, B. C. Enhanced post-stationary-phase survival of a clinical thymidinedependent small-colony variant of Staphylococcus aureus results from lack of a functional tricarboxylic acid cycle. J. Bacteriol. 2007, 189, 2936−2940. (82) Chatterjee, I.; Kriegeskorte, A.; Fischer, A.; Deiwick, S.; Theimann, N.; Proctor, R. A.; Peters, G.; Herrmann, M.; Kahl, B. C. In vivo mutations of thymidylate synthase (encoded by thyA) are responsible for thymidine dependency in clinical small-colony variants of Staphylococcus aureus. J. Bacteriol. 2008, 190, 834−842. (83) Noor, R.; Murata, M.; Yamada, M. Oxidative stress as a trigger for growth phase-specific sigmaE-dependent cell lysis in Escherichia coli. J. Mol. Microbiol. Biotechnol. 2009, 17, 177−187. (84) Nyström, T. Stationary-phase physiology. Annu. Rev. Microbiol. 2004, 58, 161−181. (85) Wang, H.; Chang-Wong, T.; Tang, H.-Y.; Speicher, D. E. Comparison of extensive protein fractionation and repetitive LC−MS/ MS analyses on depth of analysis for complex proteomes. J. Proteome Res. 2010, 9, 1032−1040. (86) Rohde, H.; Burdelski, C.; Bartscht, K.; Hussain, M.; Buck, F.; Horstkotte, M. A.; Knobloch, J. K.; Heilmann, C.; Herrmann, M.; Mack, D. Induction of Staphylococcus epidermidis biofilm formation via proteolytic processing of the accumulation-associated protein by staphylococcal and host proteases. Mol. Microbiol. 2005, 55, 1883− 1895. (87) Conrady, D. G.; Brescia, C. C.; Horii, K.; Weiss, A. A.; Hassett, D. J.; Herr, A. B. A zinc-dependent adhesion module is responsible for intercellular adhesion in staphylococcal biofilms. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 19456−19461. (88) Conrady, D. G.; Wilson, J. J.; Herr, A. B. Structural basis for Zn2+-dependent intercellular adhesion in staphylococcal biofilms. Proc. Natl. Acad. Sci. U.S.A. 2013, 110, 202−211. (89) Macintosh, R. L.; Brittan, J. L.; Bhattacharya, R.; Jenkinson, H. F.; Derrick, J.; Upton, M.; Handley, P. S. The terminal A domain of the fibrillar accumulation-associated protein Aap of Staphylococcus epidermidis mediates adhesion to human corneocytes. J. Bacteriol. 2009, 191, 7007−7016. (90) Sharp, J. A.; Echague, C. G.; Hair, P. S.; Ward, M. D.; Nyalwidhe, J. O.; Geoghegan, J. A.; Foster, T. J.; Cunnion, K. M. Staphylococcus aureus surface protein SdrE binds complement regulator factor H as an immune evasion tactic. PLoS One 2012, 7, e38407. (91) Mackey-Lawrence, N. M.; Potter, D. E.; Cerca, N.; Jefferson, K. K. Staphylococcus aureus immunodominant surface antigen B is a cellsurface associated nucleic acid binding protein. BMC Microbiol. 2009, 9, 61. (92) Cogen, A. L.; Yamasaki, K.; Muto, J.; Sanchez, K. M.; Crotty, A. L.; Tanios, J.; Lai, Y.; Kim, J. E.; Nizet, V.; Gallo, R. L. Staphylococcus epidermidis antimicrobial delta-toxin (phenol-soluble modulin-gamma) cooperates with host antimicrobial peptides to kill group A Streptococcus. PLoS One 2010, 5, e8557. (93) Wang, G.; Xia, Y.; Cui, J.; Gu, Z.; Song, Y.; Chen, Y. Q.; Chen, H.; Zhang, H.; Chen, W. The roles of moonlighting proteins in bacteria. Curr. Issues Mol. Biol. 2013, 16, 15−22. (94) Kainulainen, V.; Korhonen, T. Dancing to another tune − adhesive moonlighting proteins in bacteria. Biology 2014, 3, 178−204.

3762

dx.doi.org/10.1021/pr500322d | J. Proteome Res. 2014, 13, 3748−3762