A Metabolite of Pseudomonas Triggers Prophage-Selective Lysogenic to Lytic Conversion in Staphylococcus aureus

Bacteriophages have major impact on their microbial hosts and shape entire microbial communities. The majority of these phages are latent and reside as prophages integrated in the genomes of their microbial hosts. A variety of intricate regulatory systems determine the switch from a lysogenic to lytic life style, but so far strategies are lacking to selectively control prophage induction by small molecules. Here we show that Pseudomonas aeruginosa deploys a trigger factor to hijack the lysogenic to lytic switch of a polylysogenic Staphylococcus aureus strain causing the selective production of only one of its prophages. Fractionating extracts of P. aeruginosa identified the phenazine pyocyanin as a highly potent prophage inducer of S. aureus that, in contrast to mitomycin C, displayed prophage selectivity. Mutagenesis and biochemical investigations confirm the existence of a noncanonical mechanism beyond SOS-response that is controlled by the intracellular oxidation level and is prophage-selective. Our results demonstrate that human pathogens can produce metabolites triggering lysogenic to lytic conversion in a prophage-selective manner. We anticipate our discovery to be the starting point of unveiling metabolite-mediated microbe–prophage interactions and laying the foundations for a selective small molecule controlled manipulation of prophage activity. These could be for example applied to control microbial communities by their built-in destruction mechanism in a novel form of phage therapy or for the construction of small molecule-inducible switches in synthetic biology.

of pyocyanin was obtained with Bruker Avance Neo spectrometer at 800.30 MHz frequency (1H) and a frequency of 201.25 MHz (13C). Chemical shifts (δ) are given in parts per million (ppm) relative to the solvent residual signal of MeOD-d4 = 3.31 ppm. The measured data was processed and analysed with MestreNova 12.0.4 software. Mass spectrometry data were obtained on an ESI-Orbitrap (Thermo Scientific, LTQ Orbitrap Velos) by direct injection and analyzed with Xcalibur (Thermo Scientific) software.

Phage induction and propagation
Overnight cultures of phage host strains were prepared from a glycerol cryo stocks in 5 mL LB. After overnight incubation at 37˚C, 180 rpm, the cultures were diluted 1:100. At exponential phase (OD600=0.8), the culture was split in parts and 10 µL bacterial extracts (in DMSO) or different concentrations of the commercial compounds were added in a final volume of 2 mL. The culture was further incubated at 37˚C, 180 rpm. After 4 h, 1 mL was centrifuged at 3000 rpm at 4˚C for 10 min. The supernatant was collected, sterile filtered through 0.2 µm pore diameter membrane filter and stored at 4˚C until use. For the S. aureus plaque assays, 100 µL of the phage supernatant and 300 µL of an overnight culture of the indicator strain RN4220 were mixed in a glass tube containing 3 mL top agar (0.6% agar) and poured onto agar plates supplemented with 10 mM CaCl2. The plates were incubated overnight at 37˚C to form plaques on the lawns. In the plaque assay experiments with E. coli and P. aeruginosa, same procedure was applied except for using 100 µL of the indicator strains E. coli 3925 and P. aeruginosa PA14.

MIC assay and colony forming units (CFU) counts
From overnight culture of the different strains and the generated mutants, 1:1000 dilution in LB and CCY medium were made. In a 96-well plate, 1 μL of pyocyanin diluted in DMSO with 99 μL of the diluted bacteria were added. The plate was incubated at 37°C at 180 rpm. The plates were read out the next day. An inhibition was considered positive if no haze of the medium and no cell pellet was visible. MIC values were determined in triplicates. S. aureus ATCC 6341 and RN4220 samples treated with different pyocyanin concentrations were diluted in CCY medium and 10 µL of the corresponding dilutions were spread on LB agar plates. The plates were then incubated overnight at 37˚C and the next day CFU were counted for each. Results are expressed as survival percentage from a non-treated DMSO control.
6. Phage precipitation, DNA extraction and TEM microscopy 5 mL of SM buffer (50 mM Tris-HCl, pH 7.5, 100 mM NaCl, 10 mM MgSO4, 0.01% gelatin) were added to plates with developed plaques and the plates were kept at 4˚C overnight. The next day, 4 mL of phage solution was collected from the plates, centrifuged at 10.000 rpm for 10 min and sterile filtered. To the supernatant, an equal amount of 20% PEG 8000/2.5 M NaCl solution was added. Afterwards, phages were allowed to precipitate on ice for 30 min. The phage solution was then centrifuged at 10.000 rpm for 10 min, the supernatant was discarded, and the precipitate dissolved in 2 mL TMN buffer (10 mM Tris-HCl, pH 7.5, 50 mM NaCl, 5 mM MgCl2). DNase and RNase were added in a final concentration of 2.5 µg/mL and the phage solution was incubated for 30 min at 37˚C. The PEG/NaCl precipitation procedure was repeated once more and the supernatant was thoughtfully discarded by two consecutive centrifugations at 10.000 rpm for 10 min and 5.000 rpm for 2 min, respectively. The visible precipitated phage pellet was collected with 200 µL TMN buffer or sterile H2O (for TEM experiments) and the solution was stored at 4˚C until use (2). The phage DNA was extracted using the Phage DNA Isolation Kit (Novagen, Canada). The NGS sequencing was performed by Eurofins GATC Biotech GmbH (Konstanz, Germany). The obtained sequence was searched for homology in the NCBI database using BlastN, BlastP and ORF finder (http://www.ncbi.nlm.nih.gov). For the TEM microscopy experiment on Zeiss EM 912 Omega (Carl Zeiss AG, Oberkochen, Germany), 10 µL of the precipitated phage solutions fixed with 0.5% glutardialdehyde were negatively strained with 2% uranyl acetate and loaded on carbon-coated copper grids. The images were taken with TRS slow scan CCD-camera for TEM (Tröndle Restlichtverstärker Systeme, Moorenweis, Germany) at 20,000-80,000-fold magnification. For the NGS RNA-sequencing experiment, S. aureus ATCC 6341 was grown to OD600 0.8 in LB medium and was then treated with 3 different pyocyanin concentrations (12.5, 25 and 50 µM) with DMSO as control. After 3.5 h of incubation at 37˚C 180 rpm, 1 mL of the cells were pelleted and transferred to DNA/RNA Shield TM Lysis Tubes (Zymo Research). Cells were lysed with FastPrep-24 TM 5G instrument (MP Biomedicals) at 6.0 m/s for 40 s. The lysis cycle was repeated twice with a 300 sec break on ice. RNA isolation, Illumina Next-Seq (10 million reads 1*75) and the differential gene expression analysis was performed at Microsynth AG (Balgach, Switzerland) with using S. aureus ATCC 6341 as a reference strain.

Proteomic analysis
The samples from the different treatments with 25 µM pyocyanin, 1.5 µM mitomycin C and DMSO as control in CCY medium were lysed with FastPrep-24TM 5G instrument (MP Biomedicals) at 6.0 m/s for 40 s. The lysis cycle was repeated twice with a 300 sec break on ice.
All samples were reduced with DTT (30 min, 56°C) and alkylated with chloroacetamide (60 min, RT). Digestions were performed using trypsin (16 h, 30°C). The digests were analysed on a QExactive HF mass spectrometer (ThermoFisher Scientific, Bremen, Germany) interfaced with an Easy-nLC 1200 nanoflow liquid chromatography system (ThermoFisher Scientific, Bremen, Germany). The peptide digests were reconstituted in 0.1 % formic acid and loaded onto the analytical column (75 μm × 15 cm). Peptides were resolved at a flow rate of 300 nL/min using a linear gradient of 5−32% solvent B (0.1% formic acid in acetonitrile) over 45 min. Data-dependent acquisition with full scans in a 350 − 1500 m/z range was carried out at a mass resolution of 120000. The 10 most intense precursor ions were selected for fragmentation. Peptides with charge states 2−7 were selected, and dynamic exclusion was set to 30 sec. Precursor ions were fragmented using higher-energy collision dissociation (HCD) set to 28%. Tandem mass spectra were searched against a suitable protein database using Mascot (Matrix Science) with "Trypsin/P" enzyme cleavage, static cysteine alkylation by chloroacetamide and variable methionine oxidation.

PCR sample preparation and setup
For this experiment, the steps of phage induction and propagation were conducted as described previously, but instead of CCY-Medium, LB-Medium was used to minimize the background of the control. The web-based software Phaster (7,8) was used to map the prophage-like regions (PLRs) in the genome of S. aureus ATCC 6341. For the complete phages (phiMBL2, phiMBL3 and phiMBL4) the major capsid protein genes were selected for amplification, whereas for the pathogenicity islands (SaPImbl1 and SaPImbl6), corresponding SaPI-specific genes were chosen. For PLR V which we tentatively classified as incomplete prophage, a phage scaffold capsid protein gene was amplified. The primers that were designed for each of the PLR genes were synthesized by Metabion AG (Munich, Germany). Genomic DNA was used as positive control and noninduced samples (DMSO) served correspondingly as negative controls in the assay. For each treatment five biological replicates were performed. Initially, 100 μL of the phage supernatant was transferred into a new tube, 0.1 μL of DNase (10 mg/mL) was added and incubated for 15 min at room temperature. In order to inactivate the DNase, the sample was incubated for 5 min at 75°C. The PCR reaction tubes contained: 7 µL milli-Q H2O, 1 µL sample or genomic DNA, 1 µL forward primer, 1 µL reverse primer (final concentration 250 ng/mL) and 10 µL Phusion high-fidelity PCR master mix (ThermoFisher Scientific). The setup for the PCR reaction was as follows: 98°C (30 sec) → 35·[98°C (5 sec) → Tm-2°C (12 sec) → 72°C (20 sec)] → 72°C (10 min) → 7°C (store).
11. Gel electrophoresis Agarose gels for electrophoresis were prepared by heating 0.5 g of agarose in 50 mL 1x TAE buffer (Roth). After cooling, 2.5 μL of peqGREEN (Peqlab, VWR) were added to the gels and mixed well before solidification. On each of the gels, 1 μL of peqGOLD 1 kb DNA ladder (Peqlab, VWR) and 5 μL of the PCR amplicon mixed with 1 μL 6xLoading dye (Peqlab,VWR) were loaded. The gels were run for 50 min at 70V with EasycastTMB1A (ThermoFisher Scientific) electrophoresis system and visualized with the Fusion-FX7 Advanced of Vilber Lourmat (Eberhardzell, Germany).

Reactive oxygen species (ROS) quantification
The production of ROS upon treatment with pyocyanin was quantified using the cell permeable fluorogenic dye 2',7'-dichlorofluorescin diacetate (DCFDA). The bacterial cells were grown to OD600 0.8, washed with PBS and stained with 20 µM DCFDA for 30 min 37˚C with shaking. After the incubation, the cells were again washed with PBS and re-suspended in the corresponding medium. 99 µL were added to black Corning 96-well plates containing different concentrations of pyocyanin, trans-Δ 1 -NQNO or 1 mM H2O2 in a final volume of 100 µL. The plate was incubated for 2 h at 37˚C with shaking (180 rpm) and fluorescence was subsequently measured on a Tecan microplate reader with Ex/Em=485/535 nm.
13. ROS scavenging assay N-acetylcysteine (NAC) was used as ROS scavenger in the assay at different concentrations (3.75, 7.5 and 15 mM). As in most experiments, cells were grown to OD600 0.8 in CCY medium and treated with 25 µM pyocyanin in combination with NAC. The NAC alone served as negative control in each of the indicated concentrations. After incubation and centrifugation, a plaque assay was performed for phage quantification.
14. Mutant generation Overnight culture of S. aureus ATCC 6341 was diluted 1:10k and spread on an LB agar plate. The next day, one colony from the plate was picked and added to 3 mL of fresh LB. After reaching stationary phase, 0.3 µL of the grown culture were added to 3 mL LB tubes containing 12.5, 25, 50 and 100 µM pyocyanin and incubated overnight at 37˚C 180 rpm. From the tube with the highest concentration where bacterial growth was observed a 1:10k dilution was made in fresh LB medium tubes containing the same pyocyanin concentration. This procedure was repeated every day for 2 weeks. Afterwards, the bacteria which grew in the tube with the highest pyocyanin concentration were pelleted by centrifugation, resuspended in PBS and spread on LB plate with 100 µM pyocyanin (two times the MIC). The plate was incubated for 3 days on 37˚C. The colonies that grew after the incubation period were resuspended in 3 mL LB containing 50 µM pyocyanin (one time the MIC) (9). The next day, cryostocks in 15% glycerol were prepared and the MIC value for each of the mutants was determined prior to sequencing. The experiment was performed in three independent biological replicates. Genomic DNA of the mutants was extracted with the Invitrogen TM PureLink TM Microbiome DNA purification kit (ThermoFisher Scientific). The Illumina library preparation, sequencing and mutant analysis was performed at Microsynth AG (Balgach, Switzerland). The gene mutations that were common to all mutants were PCR amplified and confirmed by Sanger sequencing at Eurofins GATC Biotech GmbH (Konstanz, Germany).

Fig. S1.
Preparative HPLC chromatogram from the pyocyanin purification. The elution peak of pyocyanin is indicated with red arrow.         PCR-based detection of phage production. a, Schematic representation of the PCR-based detection method. Cells of S. aureus ATCC 6341 were grown to OD600 0.8 and then incubated with the compounds for 4 h. The collected, sterile-filtered supernatants were DNase treated to digest the genomic DNA, then heated at 75˚C to inactivate the DNase and finally added to the PCR mixture. The intact DNA packed in the phage capsids is released during a heat denaturation step prior to amplification. A prophage region was considered induced when the presence of a DNA band was detected by gel electrophoresis. In the experiments, the major capsid proteins of the prophages, distinctive pathogenicity island proteins and a phage scaffold protein (PLR V) were selected as diagnostic fingerprints. b, Agarose gels of PCR amplified samples from culture supernatants of S. aureus ATCC 6341 after treatment with 1.5 μM mitomycin C and pyocyanin (12.5 μM and 50 μM). DMSO-treated samples served as negative control in the assay and the extracted genome of S. aureus ATCC 6341 as positive control. c, Agarose gels for PCR-based detection of prophage-like elements successfully propagated in S. aureus RN4220. S. aureus ATCC 6341 was treated with 1.5 μM mitomycin C and 25 μM pyocyanin. Subsequently, prophagelike elements in culture supernatants were propagated on a lawn of S. aureus RN4220. After plaques developed on the bacterial lawn, plates were incubated with SM buffer overnight at 4˚C. The phage solution was precipitated with PEG/NaCl and the concentrated phages were collected with TMN buffer. DMSO-treated samples propagated in RN4220 served as negative control in the assay and the extracted genome of S. aureus ATCC 6341 as positive control. b and c, For each treatment five biological replicates were performed and representative results are shown.   S8. Prophage phiMBL3 harbours a truncated repressor phylogentically related to Stl repressors of SaPIs. a, Dotplots displaying genomic sequence identities between different prophage-like regions (PLRs) were generated with the program Genome Pair Rapid Dotter, Gepard 1.40 (http://cube.univie.ac.at/gepard) using a word length of 9 and a window size of 0. b, Sequence alignments of CI-like and Stl-like repressor homologs of the three prophages phiMBL2, phiMBL3, and phiMBL4 with the closely related prophages (phi12 and phiNM2) and a SaPI (SaPI3) performed with PRALINE with the color schemes adapted from CLUSTALX. Secondary structure predictions (orange) with DSSP and PSIPRED are indicated as β-sheets (tubes) and α-helices (arrows). Active site residues are labelled by arrows (red) and the autocleavage site by a triangle (pink). c, Phylogenetic trees of Stl-like repressor homologs of SaPIs and d, CI-like repressor homologs of prophages (pink) and Stl-like repressor homologs of SaPIs (blue). Trees were constructed as maximum likelihood trees with bootstrap test (500 replications) using MEGA-X. Bootstrap values >50 are indicated at the corresponding nodes.