A widespread methylotroph acyl-homoserine lactone synthase produces a new quorum sensing signal that regulates swarming in Methylobacterium fujisawaense

ABSTRACT Pink-pigmented facultative methylotrophs of the genera Methylorubrum and Methylobacterium are metabolically versatile bacteria that colonize many diverse environments on earth. Despite their widespread occurrence, the molecular details of how these organisms interact with each other and their environment remain understudied. We analyzed genes encoding N-acylhomoserine lactone (acyl-HSL) quorum sensing signal synthases in the published genomes of these bacteria and determined that the product of the largest family of signal synthases had not been characterized. We identified 3R-OH-5Z-C12:1-HSL as a novel signal produced by these synthases using inverse stable isotopic labeling and structural characterization by mass spectrometry, nuclear magnetic resonance spectroscopy, and chemical synthesis. We show that in the methylotroph Methylobacterium fujisawaense DSM5686, this signal activates its cognate LuxR-family transcription factor and is produced in a positive feedback loop. We also discover that in this strain, quorum sensing negatively regulates swarming motility by activating the expression of a small protein that binds a predicted transcription factor. These results characterize a previously undescribed, yet widespread quorum sensing system used by pink-pigmented facultative methylotrophs, which helps us understand the chemical ecology of these important bacteria. IMPORTANCE Bacteria known as pink-pigmented facultative methylotrophs colonize many diverse environments on earth, play an important role in the carbon cycle, and in some cases promote plant growth. However, little is known about how these organisms interact with each other and their environment. In this work, we identify one of the chemical signals commonly used by these bacteria and discover that this signal controls swarming motility in the pink-pigmented facultative methylotroph Methylobacterium fujisawaense DSM5686. This work provides new molecular details about interactions between these important bacteria and will help scientists predict these interactions and the group behaviors they regulate from genomic sequencing information.


SUPPLEMENTARY METHODS
Key reagents. 13C-labeled methanol was purchased from Cambridge Isotope Laboratories.3-OH-C12-HSL was purchased from Millipore Sigma.All other acyl-HSLs were purchased from Cayman Chemical.
Inverse stable isotopic labeling (InverSIL) experiments.Inverse labeling and subsequent analysis was performed as previously described (1).Exponentially growing bacterial cultures were pelleted at 16,100 rcf for one minute and resuspended in growth medium with no carbon source.Subsequently, three separate six milliliter cultures were inoculated with the resuspended strain at a starting OD of 0.02.The 12 C-carbon source was added to one culture, the 13 C-carbon source to the second, and the 13 C-carbon source plus 500 nM 12 C-methionine to the last culture.The carbon sources used were 50 mM methanol, 50 mM methylamine, or 50% (v/v) methane.Cultures were grown until reaching stationary phase (OD of approximately 0.8) and then were centrifuged at 4,800 rcf for ten minutes.The resulting supernatant was extracted twice with an equal volume of ethyl acetate containing 0.01% (v/v) acetic acid, and the combined organic extract was evaporated to dryness using a nitrogen stream and stored at -20 ºC until analysis by LC-MS.
LC-MS for acyl-HSL signal detection.Dried culture supernatant extracts were resuspended in 200 microliters of 1:1 water:acetonitrile, and subsequently 65 microliters were injected onto an Agilent 1260 Infinity liquid chromatography system connected to an Agilent 6120 single quadrupole mass spectrometer operating with positive polarity and a mass range of 150-1500 m/z.A Waters Xselect HSS T3 column (2.5 µm particle size, 2.1 mm x 50 mm) held at 30 ºC was used for reverse phase separation with a flow rate of 0.4 mL min -1 .Solvent A: Water + 0.1 % formic acid, Solvent B: Acetonitrile + 0.1% formic acid.Gradient: 0-2 min, 0% B. Inverse stable isotopic labeling (InverSIL) analysis.Inverse labeling analysis was performed as previously described (1).Raw data files in netCDF format were exported using Agilent OpenLab CDS (rev C.01.07).Features were detected using MZmine version 2.53 (2) using the following workflow: 1. Mass detection (centroid, noise level 1.0E3).2. ADAP chromatogram builder (minimum group size 5 scans, group intensity threshold 1.0E3, min highest intensity 5.0E3), m/z tolerance 0.3).3. Chromatogram deconvolution (local minimum search, chromatogram threshold 30%, search minimum 0.1 min, minimum relative height 10%, minimum absolute height 6.0E3, minimum ratio of peak top/edge 2, peak duration 0-2 min).4. Adduct search (RT tolerance 0.1 min, adducts [M+Na] + and [M+NH4] + selected, m/z tolerance 0.2, max relative peak height 200%). 5. Feature list rows filter (remove identified adducts).Subsequently, isotopes were removed from 12 C samples using the Isotopic peaks grouper (m/z tolerance 0.2, retention time tolerance 0.1 min, monotonic shape required, maximum charge 3, representative isotope most intense), and the three feature lists were aligned in the order 13 C-carbon source + 12 C-methionine, 12 C-carbon source, 13 C-carbon source using the Join aligner (m/z tolerance 0.3, weight of m/z 50, retention time tolerance 0.1 min, weight for retention time 50).The alignment was exported in .csvformat with the row retention time as a common element and peak m/z as the data file element.Features containing the desired four m/z unit difference in the 13 C-carbon source and 13 C-carbon source + 12 C-methionine samples were then detected using a custom Python script (available at https://github.com/purilab/inverse).

Plasmid construction.
Plasmids used in this study are listed in Table S5.Primers used in this study are listed in Table S6.All plasmids were constructed using Gibson Assembly (4) and selection was performed with kanamycin (50 µg mL -1 ).
High-resolution tandem mass spectrometry.Mass spectrometry data were collected using a Waters Acquity I-class ultra-high pressure liquid chromatography (UPLC) instrument coupled to a Waters Xevo G2-S quadrupole time-of-flight mass spectrometer.An Acquity UPLC BEH C18 column (2.1 x 50 mm) was used for separation and resolving samples.Solvent A: Water + 0.1 % (v/v) formic acid, Solvent B: Acetonitrile + 0.1% (v/v) formic acid.The sample was eluted from the column using a ten minute linear solvent gradient: 0-0.1 min, 1% B; 0.1 -10 min, 1-100% B. The solvent flow rate was 0.45 mL min -1 .Mass spectra were collected in positive ion mode, with following parameters: 3 kV capillary voltage; 25 V sampling cone voltage; 150 °C source temperature; 500 °C desolvation temperature; nitrogen desolvation at 800 L/hr.The fragmentation spectra were collected using the same parameters with a 10-25 eV collision energy ramp.The lockspray solution was 200 pg/µL leucine enkephalin.The lockspray flow rate was 6 µL/min.Sodium formate was used to calibrate the mass spectrometer.
Marfey's analysis of 3R-OH-5Z-C12:1-HSL.3R-OH-5Z-C12:1-HSL (0.1mg) was suspended in 250 μL of 6M HCl in water overnight at 110 °C to cleave the amide bond.This reaction was then lyophilized and the resulting solid was resuspended in 250 μL of saturated sodium bicarbonate.16 μL of 1% (m/v) of Marfey's reagent (Sigma-Aldrich, 71478) was added and the mixture was allowed to react for 1 hour at 40 °C.The reaction was then quenched with 20 μL of 2M HCl in water and analyzed by LC-MS, where an m/z of 372 was found, corresponding to the [M+H] + of the hydrolyzed HSL with a Marfey's adduct.Retention times were compared to Marfey's derivatized L-and R-HSL standards.
Excess acid was neutralized with sodium bicarbonate, then the solution was filtered through cotton in preparation for GC analysis.
Chiral gas chromatography analysis.To determine the absolute stereochemistry of 5, chiral GC analysis was performed.Retention times of 2, racemic 3-hydroxydodecanoic acid methyl ester (TRC, H939600), and 5 were compared.Separation of the R/S methyl ester enantiomers was performed on an Agilent 6890 GC fitted with an Agilent HP-Chiral column (30m, 0.32mm i.d., Agilent) and a flame ionization detector.1 μL of each sample was injected in split-injection mode (10:1).The instrument was run at 115 °C isocratic for 90 minutes, followed by an increase to 220 °C (5 °C/min).H2 carrier gas (4.0 mL/min) was used.
MmaRDSM5686 reporter assay.An overnight culture of AWP370 reporter strain (AWP348+pAWP492) was subcultured to an optical density of 0.05 in fresh AMS containing kanamycin (50 ug/mL), 0.01% (m/v) yeast extract, and 50mM MeOH in a 50 mL sterile conical tube.0.5 mL of this culture was added to each of the wells of a 96-well deep well plate containing 5 µL of the appropriate acyl-HSL dissolved in acidified ethyl acetate.The plate was then incubated at 30 °C for 24 hours with shaking (200 rpm).100 µL of each well was transferred to a black with clear flat bottom 96-well plate (Corning 3631) and red fluorescence was quantified at 570 nm excitation and 605 nm emission using a SpectraMax i3x plate reader.Absorbance at 600 nm was also quantified for normalization.Each condition was performed in triplicate.Results were analyzed using Graphpad Prism version 8.0.2.
Swarming assay.6 ml cultures of DSM5686-derived strains were inoculated from plates and grown to stationary phase plus an additional two days in R2A media supplemented with 50mM MeOH.For strains that were grown on kanamycin, 50 µg mL -1 was used.Cultures were normalized by OD and 5 μL of the culture was spotted onto the center of a soft agar R2A plate supplemented with 50 mM methanol, and kanamycin when necessary.Soft agar R2A plates had an agar concentration of 0.4% (m/v) and a volume of 25mL.10μM 3-OH-C12-HSL was used for complementing the ΔmmaI strain.After pouring, plates were left to air dry without lids in a biosafety cabinet for ~20 minutes before spotting.Plates were incubated, inverted, at 30 ºC for 4-6 days before photographing.Swarming area was quantified in ImageJ (https://imagej.net/ij/)by tracing the perimeter of the bacterial swarm and measuring the number of pixels contained within that area.This area was then divided by the total area of the plate to give % plate area.
Expression of His6-PraR and MmaP-3XFLAG.10mL terrific broth (TB) overnight cultures of BL21 E. coli expressing MmaP-3XFLAG with or without His6-PraR were used to inoculate 1L of TB.These cultures were grown at 30ºC to an OD of ~0.5, then induced with a final concentration of 0.5mM IPTG.Induced cultures were grown overnight at 17 ºC.The following morning, two 450 mL portions of the culture were pelleted and the pellets were frozen overnight at -80ºC.The following day, the 450 mL culture pellets were resuspended in 15 mL of lysis buffer (50 mM Tris, 20 mM imidazole, 1 mM PMSF, pH 8.0). 1 mg/mL lysozyme was added and the mixture was left on ice for 30 minutes before sonication (2s on, 2s off for 4 min at 70% amplitude).Cell lysate was clarified for 60 minutes at 21,000 rcf.Protein concentrations were determined by absorbance at 280nm and normalized to the lowest concentration by addition of lysis buffer.
His6-PraR pull-down.8 mL of clarified lysate was combined with 1 mL of IMAC resin that had been pre-equilibrated in wash buffer (50mM phosphate buffer, 250mM NaCl, 25 mM imidazole, pH 8.0).This mixture was left to equilibrate in at 4 ºC for one hour.The resin/clarified lysate mixture was loaded onto a gravity column and the flow through was discarded.The resin was washed twice with 10 mL wash buffer.The resin was then washed the following number of times with 1mL of wash buffer with increasing concentrations of imidazole: 2 x 100 mM, 2 x 200 mM, 3 x 300 mM.His6-PraR was then eluted with 8 x 0.5 mL of elution buffer (50mM phosphate buffer, 250mM NaCl, 500 mM imidazole, pH 8.0) and all fractions were analyzed by SDS-PAGE.
Western blot detection of purified MmaP-3XFLAG.Protein samples to be analyzed were denatured in an equal volume of 2X Laemmli SDS sample buffer at 97ºC for six minutes and then separated on a 15% polyacrylamide gel.The gel was then transferred to a 0.2 μm PVDF membrane.The membrane was rinsed twice with PBST and then blocked in a PBST + 5% (m/v) dry milk for 30 minutes at room temperature.The membrane was rinsed twice with PBST, then incubated with the primary antibody overnight at 4ºC (Sigma-Aldrich, F3165 1:3000 in PBST).The following morning, the membrane was washed 3X with PBST and incubated with the secondary antibody (LI-COR, 926-32210, 1:10,000 in PBST) for two hours at room temperature.The membrane was then washed and imaged on the LiCor Odyssey CLx imager at the highest resolution.Gene neighborhood analysis of mmaP in DSM5686 and representative PPFM strains.Analysis was performed using the CAGECAT cblaster and clinker webtools (7).The percent amino acid identities shown are comparing the products of genes to those in strains listed immediately above and below in the figure.Of the strains listed, the minimum percent identities to the three Mma QS genes products in DSM5686 are: Methylobacterium sp.PvP109 (91% identical to MmaR and 95% identical to MmaI) and Methylobacterium sp.Leaf361 (83% identical to MmaP) The cutoff for a link being shown is 30% identity.

Figure S11 .
Figure S11.The Mma QS gene neighborhood is highly homologous in different PPFM genomes.Gene neighborhood analysis of mmaP in DSM5686 and representative PPFM strains.Analysis was performed using the CAGECAT cblaster and clinker webtools(7).The percent amino acid identities shown are comparing the products of genes to those in strains listed immediately above and below in the figure.Of the strains listed, the minimum percent identities to the three Mma QS genes products in DSM5686 are: Methylobacterium sp.PvP109 (91% identical to MmaR and 95% identical to MmaI) and Methylobacterium sp.Leaf361 (83% identical to MmaP) The cutoff for a link being shown is 30% identity.

Figure S12 .
Figure S12.mmaI is cotranscribed with mmaP.PCR amplification of a genomic region that spans the mmaI and mmaP open reading frames.The template for PCR was either gDNA, cDNA, or RNA not treated with reverse transcriptase (RT).Expected product size is 526 bp.Product identity was confirmed by Sanger sequencing.

Figure S13 .
Figure S13.Wild-type DSM5686 and DSM5686ΔmmaP produce the same amount of 3R-OH-5Z-C12:1-HSL.The area under the curve of the LC-MS peak of the 3R-OH-5Z-C12:1-HSL feature was quantified.Data show the mean and standard deviation of five technical replicates and are representative of two independent experiments.n.s., not significant (student's t-test, p < 0.05).

Figure S14 .
Figure S14.Gene neighborhood analysis of praR in Rhizobium leguminosarum bv.viciae 3841, DSM5686 and representative Methylobacterium strains.Analysis was performed using the CAGECAT cblaster and clinker webtools(7).The cutoff for a link being shown is 30% amino acid identity.

Figure S15 .
Figure S15.A C-terminally 3XFLAG-tagged MmaP is still capable of repressing swarming in DSM5686.Data show the mean and standard deviation of three plates and are representative of two independent experiments.Means were compared to WT using a 1-way ANOVA with Tukey's post-hoc test.*, P <0.001; ns, not significant.

Table S4 .
Strains used in this study.

Table S5 .
Plasmids used in this study.

Table S6 .
Primers used in this study.Homology regions used for Gibson Assembly are bolded.