Processing, Export, and Identification of Novel Linear Peptides from Staphylococcus aureus.

Here, we provide evidence indicating that S. aureus secretes small linear peptides into the environment via a novel processing and secretion pathway. The discovery of a specialized pathway for the production of small linear peptides and the identification of these peptides leads to several important questions regarding their role in S. aureus biology, most interestingly, their potential to act as signaling molecules. The observations in this study provide a foundation for further in-depth studies into the biological activity of small linear peptides in S. aureus.

S taphylococcus aureus colonizes approximately a third of the human population and is the most common cause of skin and soft tissue infections in otherwise healthy community individuals (reviewed in references 1 and 2). S. aureus virulence is controlled by numerous regulatory systems, one of which is a quorum sensing system that responds to cyclic peptide signals (reviewed in references 3 and 4). Other Gram-positive bacteria have evolved small linear peptide signaling systems to control a variety of critical processes (reviewed in references 5 and 6). Enterococcus faecalis uses small linear peptides (7 to 8 amino acids in length) that induce a mating response of donor cells carrying cognate conjugative plasmids (7). Interestingly, S. aureus produces the hydrophobic, linear heptapeptide staph-cAM373 (amino acids AIFILAA), which mimics the enterococcal cAM373 peptide (AIFILAS) by mediating aggregation of E. faecalis cells harboring the pAM373 plasmid and subsequent interspecies horizontal gene transfer (HGT) in vitro (8)(9)(10). Interspecies HGT is of special interest, because vancomycin resistance in S. aureus has been proposed to have occurred through HGT from vancomycin-resistant E. faecalis (11).
Small linear peptides often derive from the secretion signal sequences of lipoprotein precursors that contain an N-terminal sequence with a conserved C-terminal lipobox motif. After predominantly Sec-dependent secretion, the lipoprotein is processed and attached to the membrane via the conserved phosphatidylglycerol prolipoprotein diacylglyceryl transferase (Lgt) and the specific signal peptidase II (LspA) (reviewed in references 12 and 13). In addition to the upstream processing events, the intramembrane metalloprotease RseP/Eep and the EcsAB/PptAB transporter were shown to be involved in linear peptide processing and secretion in diverse bacterial species (14)(15)(16)(17). This is especially intriguing because the ecsAB locus is conserved in S. aureus (18) and shows high homology to the enterococcal proteins (60% and 30% amino acid identity for EcsA and EcsB, respectively), while Eep shares 39% amino acid identity with the enterococcal metalloprotease RseP/Eep. However, the substrates for both membrane proteins and the machinery responsible for processing and export of linear peptides in S. aureus are undefined. We aimed to identify the molecular determinants of linear peptide processing and export in the community-acquired methicillin-resistant S. aureus (MRSA) USA300 LAC strain. To identify candidate genes, we employed an E. faecalis aggregation assay to assess the production of the linear peptide staph-cAM373 in S. aureus culture supernatant. Briefly, E. faecalis cells harboring the pAM373 plasmid respond to the presence of staph-cAM373 by formation of cellular aggregates (8). Exogenous addition of synthetic staph-cAM373 peptide (AIFILAA) or LAC wild-type (WT) supernatant resulted in aggregation of E. faecalis pAM373 cells and low culture turbidity ( Fig. 1A to C). In contrast, supernatant of the isogenic camS mutant strain, which lacks the lipoprotein and staph-cAM373, resulted in a significant reduction of E. faecalis aggregation accompanied by higher turbidity (Fig. 1A). The loss of the aggregation phenotype was also observed with supernatants from the lgt and lspA mutants (Fig. 1B). Most interestingly, mutation of eep and ecsAB inhibited the aggregation phenotype of E. faecalis pAM373 to the same degree as that observed with the camS mutant ( Fig. 1B and C). Complementation of the mutated genes on a plasmid restored the aggregation phenotype for each of the five mutant strains to WT levels. To confirm the importance of Eep and EcsAB for staph-cAM373 production, we analyzed supernatant of the LAC WT strain as well as the camS, eep, and ecsAB mutant strains by high-resolution mass spectrometry (MS). As expected, staph-cAM373 was present in the supernatant of the WT but not in that of the camS mutant (Fig. 1D). Importantly, we detected only minimal levels of staph-cAM373 in the supernatant of the eep and ecsAB mutant strains, confirming their importance for staph-cAM373 processing and export. Taken together, we can now model the processing and secretion pathway of the interspecies linear peptide signal staph-cAM373 (Fig. 1E).
To determine whether Eep and EcsAB are components of a global S. aureus linear peptide biosynthetic pathway, we mined the peptidome data obtained from the LAC WT strain for the presence of additional linear peptides derived from the secretion signal sequence of reported staphylococcal lipoproteins (19) with characteristics similar to those of staph-cAM373. Ions with masses matching (mass error Ͻ 5 ppm) those of eight peptides were detected in the supernatant of the LAC WT strain but not in the eep and ecsAB mutant strains ( Fig. 2A; see also Table S1 in the supplemental material). We confirmed the presence of four of the putative linear peptides in the LAC WT supernatant by comparing their m/z values and retention times to those of synthetic standards (Fig. 2B). Subsequently, we also quantified these four linear peptides in the peptidome of the eep and ecsAB mutant strains. All four peptides were either completely absent or present only at very low levels compared to the WT strain ( Fig. 2B and C). These data clearly demonstrate that Eep and EcsAB have global importance for linear peptide signal production in S. aureus.
Taking the results together, we used a combination of genome mining, genetic characterization, and mass spectrometry to model the processing and secretion pathway of the interspecies linear peptide signal staph-cAM373 and to demonstrate that Eep and EcsAB are of global importance for linear peptide signal production in S. aureus. Whereas many of the peptides identified by our approach derive from unchar-acterized proteins, the staph-cAM373 linear peptide is a known interspecies signaling peptide (8)(9)(10), suggesting that Eep and EcsAB are likely to be additional important determinants for HGT. Two of the Eep-dependent and EcsAB-dependent peptides derive from lipoproteins (SAUSA300_0079 and SAUSA300_0618) involved in metal binding (20,21). Thus, Eep and EcsAB might be of importance for survival when S. aureus encounters host-imposed metal stress during infection. With our new knowl-  Table S2 and primers are listed in Table S3.
Reagents and growth conditions. Peptides were custom synthesized by AnaSpec Inc. (Fremont, CA) and resuspended in dimethyl sulfoxide (DMSO) at a concentration of 1 mg/ml. All bacterial strains and plasmids used in this work are listed in Table S2. Escherichia coli was cultured in lysogeny broth (LB), and for most experiments, S. aureus and E. faecalis were cultured in tryptic soy broth (TSB) at 37°C with shaking at 225 rpm. For assessment of aggregation and mass spectrometry analysis, S. aureus strains were grown in chemically defined medium (CDM), prepared as described by Ibberson et al. (22). Antibiotics for S. aureus selection and plasmid maintenance were added to the media at the following concentrations: chloramphenicol (Cm), 10 g/ml; erythromycin (Erm), 5 g/ml. E. coli plasmids were maintained in media supplemented with ampicillin (Amp) (100 g/ml) or chloramphenicol (Cm) (10 g/ml). Tetracycline (Tet) was added at 10 g/ml to E. faecalis cultures when necessary.
S. aureus supernatant preparation for cell aggregation and mass spectrometry. S. aureus overnight cultures were grown in CDM and diluted to an optical density at 600 In the top panel, all peptides are shown at a concentration of 10 ng/ml. All peaks highlighted in the bacterial strains have been normalized to the peak height of the maximum detected peak (1.95 ϫ 10 6 ) to illustrate differences in production within and between strains. Peaks that have not been colored in represent isobars of peptides under study. (C) Production of all four linear peptides is significantly reduced in the eep (Δeep) and ecsAB (ΔecsAB) mutant strains. Shown are the absolute quantifications of all four peptides determined using a calibration curve of the corresponding synthetic peptide standard for each peptide (n ϭ 5 biological replicates, means Ϯ SD). ****, P Յ 0.0001. nm (OD 600 ) of 0.02 in 5 ml fresh CDM and grown at 37°C with shaking (225 rpm) to an OD 600 of 2.8 to 3. Cells were harvested by centrifugation, and the supernatant was filtered through a 0.22-m-pore-size filter (Millipore; Millex polyethersulfone [PES] membrane). Proteinase inhibitor cocktail (SigmaFAST, EDTA free; Sigma) was added to the supernatant, and the samples were stored at 4°C.
Cell aggregation assay. E. faecalis cell aggregation was assayed using a protocol modified from that previously reported by Dunny et al. (7). E. faecalis JH2-2 cells containing peptide-responsive pAM373::Tn918 plasmid were grown overnight in TSB at 37°C with 225 rpm shaking, washed twice, and resuspended to an OD 600 of 0.2 (9.1E ϩ 07 Ϯ 6.2E ϩ 06 CFU per ml) in CDM. S. aureus supernatant (1 ml) was mixed with 1 ml of E. faecalis cells in a round-bottom glass tube (Kimax; Kimble) and incubated at 37°C in a shaking incubator (250 rpm) for 4.5 h. Cell aggregation was quantified by measuring the turbidity of the suspension with a MicroScan turbidity meter (Dade Behring) compared to that of the medium. To control for E. faecalis growth effects, cell aggregates were dissolved by addition of 2 mM EDTA followed by vortex mixing of the suspension for 2 min. Subsequently, turbidity was determined as described above and bacteria were plated on tryptic soy agar (TSA) to determine CFU counts per ml (see Fig. S1A and B in the supplemental material).
Mass spectrometry analysis. All liquid chromatography-mass spectrometry (LC-MS) analyses were conducted using a Waters Acquity ultraperformance LC (UPLC) system (Waters Corporation) coupled to a Thermo Fisher Scientific Q-Exactive Plus mass spectrometer (Thermo Fisher Scientific) with electrospray ionization (ESI). Samples (3 l) were injected into a Waters BEH C 18 column (1.7-m pore size; 2.1 by 50mm) at a flow rate of 0.3 ml/min with a gradient of H 2 O (solvent A) and acetonitrile (solvent B), each spiked with 0.1% formic acid. The gradient began at 5% solvent B for 0.5 min and increased to 40% solvent B using a gradient curve of 5 from 0.5 to 7.5 min. From 7.5 to 8.0 min, the gradient linearly increased to 100% solvent B and was held at 100% solvent B for 0.5 min. The starting conditions were reestablished from 8.5 to 9.0 min and held until 10.0 min.
Analyses were conducted in positive-ion and negative-ion modes with a scan range of m/z 175 to 2,000, spray voltage of 3.7 kV, capillary temperature of 265.63°C, sheath gas flow of 51.25, auxiliary gas flow of 13.13, and S-lens radio frequency (RF) level of 80.00. Peptide concentrations were calculated using calibration curves of synthetic peptides ranging from 0 to 100 ng/ml by 2-fold dilutions (Fig. S2A). Calibration curves were generated with positive-mode data by plotting the peak area versus the concentration, and the linear portion of each curve was determined by log transformation of both axes. Peak areas were selected for m/z values of 783.5330, 805.4812, 773.4774, 785.5125, and 718.4492, corresponding to ALLLLVAA, ALSLFLAA, GSLLILSA, GLVILLSA, and AIFILAA, respectively. Concentrations were calculated using the best-fit line obtained by linear regression of each data set. Ions whose peak areas were below the detection limit were assigned a concentration of 0. Concentrations were normalized to CFU counts (Fig. S2B).
Statistical analysis. Statistical significance was calculated with GraphPad Prism (8.2.0) using a one-way analysis of variance (ANOVA) with a Tukey's or Dunnett's posttest to correct for multiple comparisons.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. TEXT S1, DOCX file, 0.03 MB.