The AusAB non-ribosomal peptide synthetase of Staphylococcus aureus preferentially generates phevalin in host-mimicking media

Abstract

Non-ribosomal peptide synthetases (NRPSs) are modular multidomain enzymes responsible for the biosynthesis of various secondary metabolites in an mRNA template-independent manner. They are predominantly present in bacteria and fungi, where they synthesize a variety of products, including antibiotics, siderophores, toxins, and signaling molecules. The human pathogen Staphylococcus aureus possesses one single NRPS, AusA, highly conserved in all sequenced S. aureus strains. AusA incorporates the aromatic amino acids (AAAs) phenylalanine or tyrosine, as well as the branched-chain amino acids valine and leucine into three cyclic dipeptides collectively called aureusimines: phevalin, tyrvalin, and leuvalin. By using targeted metabolomics, we found that during growth in the common tissue culture medium RPMI1640, AusA preferentially synthesizes phevalin, despite similar availability for both phenylalanine and tyrosine. Upon cultivation in a chemically defined medium, however, the yields for both products are comparable, albeit with a slight preference for phevalin. Moreover, omission of either "building block" (phenylalanine, tyrosine, or valine) does not abrogate aureusimine biosynthesis, showing that de novo biosynthesis of these amino acids is sufficient to yield aureusimine production. Cultivation of S. aureus in a synthetic medium mimicking human nasal secretions, lacking tyrosine, results in marked phevalin production, despite moderate bacterial growth. Our report on culture medium composition-driven AAA incorporation by a bacterial NRPS provides a useful basis for linking bacterial cell metabolic status to the biosynthesis of secondary metabolites.

Importance: Peptide and protein synthesis are fundamental processes in nature which are largely mediated by the ribosomal machinery. An alternative pathway for peptide synthesis is non-ribosomal mRNA template-independent synthesis, performed by so-called NRPSs. NRPSs are multi-enzyme complexes which serve the simultaneous role of template and biosynthetic machinery. They are mostly found in bacteria and fungi and are responsible for the biosynthesis of many pharmacologically significant products, including antibiotics, anticancer compounds, or immunosuppressants. The human pathogen S. aureus possesses one such NRPS, AusA, which synthesizes three cyclic dipeptides termed "aureusimines" using the aromatic amino acids phenylalanine and tyrosine and the branched-chain amino acids leucine and valine. Although the biological role of aureusimines remains unknown, AusA appears to play a role in the interaction of S. aureus with the host. In addition, owing to its minimal canonical NRPS structure and autonomous function (i.e., most NRPS pathways require the assembly of several NRPS proteins), AusA represents an excellent model system for studying such molecular assembly lines. Our observation is, to our knowledge, the first report of an NRPS showing preferential incorporation of aromatic amino acids, despite their similar availability.

Keywords: Staphylococcus aureus; aureusimines; non-ribosomal peptide synthetases; phevalin; tyrvalin.

Fig 1

Aureusimines are synthesized by different S. aureus strains and by skin-associated staphylococci. (A) Schematic representation of the AusA non-ribosomal peptide synthetase (genomic locus ID: SAUSA300_RS00950). AusA is a ~273 kDa soluble protein with the bimodular architecture A₁-T₁-C-A₂-T₂-R. (B) Aureusimines (phevalin, tyrvalin, and leuvalin) are cyclic dipeptides with monoketopiperazine structure. (C) Phevalin and tyrvalin are synthesized by both cytotoxic (S. aureus JE2) and non-cytotoxic S. aureus (Cowan I), as well as by environmental S. aureus. (D) Phevalin and tyrvalin are detected in Staphylococcus epidermidis and Staphylococcus capitis, two skin-associated coagulase-negative staphylococcal species, but not in Staphylococcus saprophyticus and Streptococcus mutans. Phevalin and tyrvalin were detected by ultra-performance liquid chromatography–mass spectrometry (UPLC-MS) in stationary-phase culture supernatants of bacteria grown in tryptic soy broth (TSB) for 24 h. Commercially available phevalin and tyrvalin serve as standards. Leuvalin is to date not commercially available; therefore, it was not analyzed in the present study. The UPLC spectra shown are representative of three independent biological replicates (n = 3). (A: adenylation; T: thiolation; C: condensation; R: terminal reductase; WT: wild type; env.: environmental; MRSA: methicillin-resistant Staphylococcus aureus; ATCC: American Type Culture Collection).

Fig 2

Aureusimine biosynthesis is dictated by aromatic amino acid availability in S. aureus (A) Phevalin and tyrvalin detection by UPLC-MS in S. aureus JE2 WT supernatants grown in commercial RPMI1640 medium for 4, 8, 12, or 24 h. (B) Dynamics of phevalin and tyrvalin production in commercial RPMI1640 medium. Plotted data represent raw values for area under the curve (arbitrary units) for UPLC-MS chromatograms. Shown data are mean values ± SD from independent biological replicates (n = 3). Same data were used to generate pie charts in panel A. (C) Depletion of Phe, Tyr, and Val during growth of S. aureus JE2 WT in commercial RPMI1640, normalized to input medium (left Y axis). S. aureus JE2 WT growth in RPMI1640 (right Y axis). OD₆₀₀ was measured at 2, 4, 6, 8, 12, and 24 h of growth. Data are shown as mean values ± SD from independent biological replicates (n = 3). (D) Quantification of bacterial pellet-associated Phe, Tyr, and Val for S. aureus JE2 WT cultivated for 24 h in commercial RPMI1640 medium (micromoles amino acid per gram dry pellet). Bar graphs represent mean values ± SD from independent biological replicates (n = 3). Norvaline was used as the internal standard for quantification. Statistical analysis: one-way ANOVA with Tukey’s multiple comparison test (**P < 0.01; ns = not significant). (E) S. aureus JE2 WT growth in chemically defined media (CDMs), lacking Phe, Tyr, Val, Phe/Tyr, or Phe/Tyr/Val. OD₆₀₀ was measured every 18 minutes for 24 or 48 h. Data are shown as mean values ± SD from independent biological replicates (n = 3). Arrows indicate the time points chosen for the aureusimine analysis shown in panel F. (F) Detection of phevalin and tyrvalin by UPLC-MS in S. aureus JE2 WT supernatants grown in CDM lacking Phe, Tyr, Val, and Phe/Tyr or Phe/Tyr/Val for 24 or 48 h (in case of CDM [−Val] and CDM [−Phe/−Tyr/−Val]). (G) Phevalin and tyrvalin detection by UPLC-MS in S. aureus JE2 WT supernatants grown in CDM containing different Phe concentrations. (H) Phe concentration-dependent changes in phevalin production (sigmoidal 4PL fit, where the X axis represents Phe concentrations in micromolar, and the Y axis represents the phevalin:tyrvalin ratio calculated from the area under the curve [AUC] for UPLC-MS chromatograms). (I) Growth of S. aureus JE2 WT and ΔausAB in a synthetic nasal medium. OD₆₀₀ was measured every 18 minutes for 20 h. Data are shown as mean values ± SD from independent biological replicates (n = 3). (J) Phevalin and tyrvalin detection by UPLC-MS in S. aureus JE2 WT supernatants grown in synthetic nasal medium. All pie charts represent the fitted relative area under the curve (arbitrary units) for UPLC-MS chromatograms, where tyrvalin is normalized to phevalin, which is set to 100%. Data used for pie chart generation represent mean values from independent biological replicates (panels A, F, and G, n = 3; panel J, n = 4). Representative total ion chromatograms are shown in panels F and J. (P: phevalin; T: tyrvalin).