Real-Time Biosynthetic Reaction Monitoring Informs the Mechanism of Action of Antibiotics

The rapid spread of drug-resistant pathogens and the declining discovery of new antibiotics have created a global health crisis and heightened interest in the search for novel antibiotics. Beyond their discovery, elucidating mechanisms of action has necessitated new approaches, especially for antibiotics that interact with lipidic substrates and membrane proteins. Here, we develop a methodology for real-time reaction monitoring of the activities of two bacterial membrane phosphatases, UppP and PgpB. We then show how we can inhibit their activities using existing and newly discovered antibiotics such as bacitracin and teixobactin. Additionally, we found that the UppP dimer is stabilized by phosphatidylethanolamine, which, unexpectedly, enhanced the speed of substrate processing. Overall, our results demonstrate the potential of native mass spectrometry for real-time biosynthetic reaction monitoring of membrane enzymes, as well as their in situ inhibition and cofactor binding, to inform the mode of action of emerging antibiotics.


Bacterial strains and growth conditions
Strains used in this study are listed in Table S1.Bacteria were grown on LB plates or in liquid LB medium (10 g/L tryptone, 5 g/L yeast extract, 10 g/L NaCl for liquid media or 15 g/L for agar plates) at 37°C.Growth of bacteria for protein was in LB or Terrific Broth (24 g/L yeast extract, 20 g Tryptone, 0.4% glycerol, 17 mM KH2PO4, 72 mM K2HPO4), containing antibiotic for selection.Antibiotics were used at the following concentrations: Chloramphenicol (Cam, 25 µg/mL); Kanamycin (Kan, 50 µg/mL), and Ampicillin (Amp, 100 µg/mL).

Construction of E. coli deletion strains
Deletion strains were obtained by moving kan-marked alleles from the Keio E. coli single gene knockout library 1 by P1 phage transduction 2 or by recombineering using the pKD46 and Kan cassette amplified from pKD4 as indicated in refs. 1,3 fterward, the kan cassette was removed by pCP20-encoded Flp recombinase to generate unmarked deletions with a FRT-site scar sequence 3 .The removal of the kan gene was verified by colony PCR.Strains with multiple deletions were generated by sequential P1 transduction or recombineering and kan cassette removal.

Cloning
All template plasmids are listed on Table S2 and oligonucleotide sequences are listed on Table S3.pET28a-ybjG plasmid, encoding for YbjG with an N-terminal His-tag, was amplified by PCR from genomic DNA of E. coli BW25113 using oligonucleotides YbjG_FW_NdeI and YbjG_REV_XhoI, and cloned into pET28a(+) with the appropriate restriction enzymes.pVMH23 was generated by two PCR fragments: an amplified product obtained from pHK226 and the oligonucleotides FhuA2pBB_frag_FW and FhuA2pBB_frag_REV (Table S2), and an amplified product obtained from pBB013 and the oligonucleotides FhuA2pBB_vect_FW and FhuA2pBB_vect_REV.The same volumes of each PCR fragment were mixed, heated to 98˚C and cooled down to room temperature.The DNA mixed was digested with DpnI and transformed into DH5α competent cells.pVMH9 was obtained by the same procedure, using pET28a-uppP and the oligonucleotides hisuppP_frag_FW and hisuppP_frag_REV, and pBB012 and oligonucleotides hisuppP_pBBvect_fw and hisuppP_pBBvect_rev. pVMH10 was obtained by the same procedure using pET28a-pgpB and the oligonucleotides pgpBhis_frag_FW and pgpBhis_frag_REV, and pBB013 and oligonucleotides pgpBhis_pBBvect_fw and pgpBhis_pBBvect_rev. pVMH13 was obtained by the same procedure using BW25113 genomic DNA and the oligonucleotides ybjG_frag_FW and ybjG_frag_REV, and pBB013 and oligonucleotides ybjG _pBBvect_fw and ybjG_pBBvect_rev.pVMH16 was obtained by the same procedure using BW25113 genomic DNA and the oligonucleotides lpxT_frag_FW and lpxT_frag_REV, and pBB013 and oligonucleotides lpxT_pBBvect_fw and lpxT_pBBvect_rev.pVMH11 was obtained by removal of the encodes His-tag in plasmid pVMH9 with primers uppP_deltaHis_FW and uppP _deltaHis_REV using Q5® Site-Directed Mutagenesis Kit (New England Biolabs).A codon-optimized gene block for Bacillus subtilis PgpB (Integrated DNA Technologies) was cloned into pET28a vector between the restriction sites BamHI and XhoI using Infusion cloning kit (Takara) according to the manufacturer's protocol.All constructs were confirmed by DNA sequencing of the genes of interest.
Proteins were finally loaded onto a Superdex S200 Increase column pre-equilibrated with a buffer containing 0.03% DDM, 10% glycerol, 20 mM Tris-HCL, 300 mM NaCl, pH 8.0.Fractions were pooled, concentrated as previously described, aliquoted and snap-frozen in liquid nitrogen, and stored at -80 o C until needed.

C55-PP synthesis
C55-PP synthesis reactions were carried out as described previously 4 with minor modifications.
Reactions (600 µL) contained 100 nmol C15-PP, 1000 nmol C5-PP and 10 nmol of UppS in 50 mM Tris-HCl pH 7.5, 50 mM KCl, and 0.05% DDM, LDAO or Triton X-100 as indicated.After overnight incubation at 37°C, the produced C55-PP was extracted 3 times with 500 µL butanol/pyridine acetate at pH 4.2 (2:1).The organic phase was separated each time, combined and finally aliquoted and dried using a speed-vac concentrator.Completion of the reaction was verified by analysing the product by thin layer chromatography (TLC) and quantification of product concentration by colorimetric phosphate assay after complete digestion with UppP: reactions containing 0.5 µM UppP and ~20 µM C55-PP were incubated for 2 h at 37 °C and stopped by the addition of the acidic reagent in the PiColorLock phosphate detection kit (Expedeon, UK).The amount of phosphate produced was calculated from the absorbance at 620 nm measured using a plate reader, by comparing it with a Pi standard.respectively.The noise level was set at 3. Unless otherwise stated, proteins were activated by applying 100 V in the high-energy collisional dissociation cell without in-source trapping.Data visualised and exported for processing using the Qual browser of Xcalibur 4.1.31.9 (Thermo Scientific).Spectral deconvolution was performed using UniDec. 6Typically the m/z range was set at 2000-10000 without baseline subtraction with the charge range set at 1-20, the mass range was set at 20000-70000 Da, and masses were sampled every 1.0 Da with the peak detection threshold minimum of 5%.Relative binding affinities were obtained from deconvoluted spectra by dividing the intensity of ligand-bound protein peaks by the sum of the intensities of ligand-bound and ligand-free protein peaks.All measurements were performed at least three times and yielded similar results.

Analysis of co-purified lipids
Lipids copurified with UppP were extracted by adding 100 L chloroform: methanol (2:1 vol/vol) mixture to 50 L protein in 200 mM ammonium acetate, pH8, 0.05% LDAO.The mixture was centrifuged at 20,000 g for 10 min to separate the organic and aqueous phases.The organic phase containing copurified lipids was dried under vacuum at 45 o C for 30 min.The resulting lipid film was resuspended in a buffer containing 10 mM ammonium formate, 0.1 %(vol/vol) formic acid, and 60 %(vol/vol) acetonitrile.The sample was analysed in the negative ESI polarity on an Eclipse Tribrid mass spectrometer (Thermo Fisher Scientific) using a CID activation of 175 V for MS 1 and an additional 28-32 V for MS 2 .

Lipopolysaccharide (LPS) isolation, detection and quantification
The protocol for LPS isolation was adapted from Crawford et al. 9      Like the E. coli homologue, this protein copurified with very little phospholipids.37°C for 24 h.The parental strain, BW25113, was included as a control.In brackets, the name of the remaining C55-PP phosphatase is indicated.
pelleted from the lysates by centrifugation at 20,000 g for 20 min.The membrane fraction was separated from the clarified lysates by centrifugation at 100,000 g for 90 min.Membranes were homogenised in a buffer containing 20% glycerol, 20 mmol dm -3 Tris-HCL, 300 mM NaCl, pH 8.0.Aliquots of membrane corresponding to the harvest from 4 L cell culture were solubilised immediately or flash frozen in liquid nitrogen and stored at -80 o C until needed.Membrane fractions were solubilised by adding DDM at a final concentration of 2% and incubation for 2 h.Gentle solubilisation of E. coli PgpB was performed by solubilising the membrane preparation (harvest of 4L cell culture) with 0.5% DDM for 1 hr.Nonsolubilised materials were removed by centrifugation at 20,000 g for 20 min and imidazole was added to the recovered supernatant at a final concentration of 20 mM.The sample was passed over 5-mL HisTrap HP column pre-equilibrated with buffer B20 (20 mM imidazole, 0.03% DDM, 10% glycerol, 20 mM Tris-HCL, 150 mM NaCl, pH 8.0).The column was washed with 100 mL of buffer B50 (50 mM imidazole, 0.03% DDM, 10% glycerol, 20 mM

1
mM IPTG was added to induce expression and cells were incubated at 20°C for 16 h.Cells were pelleted by centrifugation at 7000×g for 10 min and resuspended in 80 mL of buffer A (50 mM Tris-HCl, 1 M NaCl, 10% glycerol, at pH 7.5) supplemented with DNAase I, protease inhibitor cocktail (1:1000 dilution), 2 mM PMSF.Cells were disrupted by sonication on ice and centrifuged (130000×g for 1 h at 4°C) to pellet the membrane fraction.The pellet was resuspended in buffer B (25 mM Tris-HCl, 500 mM NaCl, 1 mM MgCl2, 2% DDM, at pH 7.5) by stirring at 4 o C for 16 h.The extracted membraned was separated from debris by centrifugation (130000×g for 1 h at 4°C) and incubated for 1 h with 4 mL Ni 2+ -NTA beads (Novagen) at 4°C equilibrated in buffer C (25 mM Tris pH 7.5, 500 mM NaCl, 1 mM MgCl2, 0.05% DDM).Beads were washed 10 times with buffer C supplemented with 40 mM imidazole and the protein was eluted with 3 mL buffer C supplemented with 500 mM imidazole.The eluted protein was analysed by SDS-PAGE and the purest fractions were pooled and extensively dialysed against buffer D (25 mM Tris pH 7.5, 300 mM NaCl, 10% glycerol).The protein was finally concentrated using filter concentrators with a 10000 MWCO cut-off, the concentration was measured with a BCA protein concentration kit (Thermo), aliquoted and stored at -80 o C.
POPE and POPG (Avanti) were dissolved in chloroform: methanol (2:1 v/v) mixture and the organic solvents were removed by evaporation.1-Octadecyl lysophosphatidic acid and 1-oleoyl lysophosphatidic (Cayman) was dissolved at 1 mg/mL in mM in 200 mM ammonium acetate (pH 8.0) and 0.05% LDAO by vertexing and then sonication in a water bath for 30 min.C15-PP and C55-PP solution in methanol/ammonia was dried in a SpeedVac and then resuspended in 200 mM ammonium acetate (pH 8.0) and 0.05% LDAO to a final concentration of 1 mM.The stock solution of bacitracin (Sigma-Aldrich) was prepared by dissolving the powder at 1 mM in 200 mM ammonium acetate (pH 8.0) and 0.05% LDAO.Teixobactin was synthesized according to the published procedure,5 dissolved in DMSO at a concentration of 10 mM, and subsequently diluted to 1 mM with 200 mM ammonium acetate (pH 8.0) and 0.05% LDAO.Protein was thawed on ice and buffer-exchanged into 200 mM ammonium acetate (pH 8.0) and 0.05% LDAO using a centrifugal buffer exchange device (Micro Bio-Spin 6, Bio-Rad).Native mass spectrometryAbout 3 μL of a protein aliquot was transferred into a gold-coated borosilicate capillary (Harvard Apparatus).The capillary was mounted on the nano ESI source of a Q-Exactive hybrid quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific, Bremen, Germany).For kinetic experiments, the instrument parameters were pre-optimised via test measurements prior to mixing protein with the desired substrate to minimise the lag times.For measurements longer than 2 minutes, fresh aliquots for the same sample vail were loaded.The instrument settings were 1.2 kV capillary voltage, S-lens RF 200%, argon UHV pressure 3.3 × 10 −10 mbar, capillary temperature 100 °C, and resolution of the instrument was set to 17,500 at a transient time of 64 ms.Voltages of the ion transfer optics -injection flatapole, inter-flatapole lens, bent flatapole, and transfer multipole were set to 5, 3, 2, and 30 V

Fig S2 :
Fig S2: Lipid-stabilized UppP dimer observed in other detergents.A-C.Spectra recorded for UppP in C8E4, G1 (ref. 11) and DM micelles.The mass spectrometer was optimized to maximize the transmission of dimeric species by using an activation voltage of 200V.D,E.Spectra of UppP released from LDAO micelles at 0.5 μM and 20 μM concentrations.F. Relative abundance of UppP species released from LDAO micelles at different protein concentrations.

Fig. S4 :
Fig. S4: Real-time monitoring of UppP interactions with C15-PP or C55-PP and inhibition by EDTA.A,B.(8+) charge state of the mass spectra recorded for 3.5 μM UppP equilibrated with 50 μM C15-PP at indicated times in the absence and presence of 200 μM EDTA.The intensity of peaks assigned to enzyme-substrate complexes remained the same in the absence of divalent cations.C,D.Equivalent dataset for UppP interaction with 20 μM C55-PP (panel C) and inhibition by EDTA (panel D).The reaction against C55-PP is completed before the first 30 s.Only bound products are detected.EDTA inhibited UppP function.The absence of product C55-P molecules in the spectra for reactions in the presence of EDTA indicated that UppP is inactive without divalent cations and that the substrates C15-

Fig. S6 .
Fig. S6.Activity of B. Subtilis PgpB towards C55-PP, DGPP and lysophosphatidic acid (LPA).A. Spectra recorded for 3.5 µM PgpB Ec incubated with 20 µM DGPP (panel A) and 20 µM C55-PP.No residual substrates are detected in the spectra acquired after 30 seconds of mixing protein and the substrate, indicating that Bacillus Subtilis PgpB is very efficient in processing both DGPP and C55-PP.B. Relative intensity of apo and C55-P bound to PgpB Bs for the reaction performed in the absence and in the presence of 200 µM EDTA.PgpB activity is independent of divalent cations.C. Mass spectrum (deconvoluted) for 5 µM PgpB Ec equilibrated with 20 µM LPA.The intensity of the enzyme-substrate complex decreased as a function of time (cf.Fig 2C, main text).Adduct peaks correspond to bound Ni 2+ (+62 Da) and/or sulphate ions (+98 Da) carried over from protein purification, and LPA stock, respectively.

Fig. S7 .
Fig. S7.C55-PP phosphatase activity of YbjG is not sensitive to EDTA.YbjG activity is higher in the presence of Triton X-100 and was not affected by EDTA.TLC analysis of reaction products of C55-PP phosphatase assays with semi-purified YbjG.0.05 mg/mL of the enzyme was incubated with 31.2 µM C55-PP in the presence or absence of 15 mM EDTA.Reactions were performed in buffer 25 mM Tris pH 7.5, 100 mM NaCl, and 0.2% DDM, with or without additional 0.2% LDAO or Triton X-100.Reactions were incubated for 2 h at 25°C.

Fig. S9 .
Fig. S9.Native MS (7+ charge state) for different UppP mutants (at 3.5 μM) incubated with 10 μM C55-PP for 30 min in each case. A. E21A UppP is active, but only a partial activity was observed for S27A UppP.B. R174A UppP and S26A/R174A UppP are inactive.C. S26A/S27A UppP is inactive; C55-PP binds to the inactive double mutant more strongly in the presence of EDTA.Expected masses based on amino-acid sequence and the experimentally observed masses for these mutants are included in Table.S4.

Fig. S10 :
Fig. S10: Effect of EDTA and bacitracin on UppP function.A,B.Spectra for a solution containing UppP (3.5μM) incubated with C55-PP (10 μM) with and without EDTA (500 μM) for 30 min.Only Enzymeproduct complexes are captured for the uninhibited reaction Removal of divalent cations by EDTA resulted in a lack of UppP activity, and therefore only enzyme-substrate complexes were observed.C,D.Equivalent data for UppP activity against DGPP without and with 100 μM bacitracin.The low

Fig. S11 :
Fig. S11: Cells depending on a single phosphatase have different bacitracin sensitivity.Strains with multiple deletions of C55-PP phosphatases were transformed with either pVMH23 plasmid, encoding the plug-less outer membrane transporter FhuA ∆322-355 (panel A), or the empty plasmid pBB012 (panel B).Growth was monitored on LB-Agar plates containing 0, 40 or 120 µg/mL bacitracin and 0 or 0.2 mM IPTG (to induce FhuA ∆322-355 expression), as indicated.Plates were incubated at

Table S1 : Strains used in this work.
. E. coli strains of interest were grown in LB (supplemented with antibiotics when required) at 37°C by orbital shaking up to an optical density at 600 nm of 1.0.Cell growth was stopped by incubating flasks on ice for 10 min.Cell suspensions(2.5 ml) were pelleted by centrifugation (17000×g, 4°C, 5 min) and resuspended in 1 ml of cold PBS.Cells were pelleted and washed in PBS twice more, then pellets were resuspended in 100 µl of 2× SDS-PAGE loading buffer (ThermoFisher) and boiled at 100°C for 10 min.Samples were supplemented with 100 µl of Proteinase K (Sigma-Aldrich, 2 mg/ml stock solution in 10 mM Tris pH 6.8, 20 mM CaCl2, 50% glycerol) and incubated at 65°C for 3 h with vigorous shaking (900 rpm), then boiled again.LPS was detected by loading 10 µl of sample on Tris-Tricine gels containing 18% polyacrylamide and staining with Pro-QTM Emerald 300 Lipopolysaccharide Gel Stain Kit (ThermoFisher), according to the manufacturer's protocol.Briefly, LPS was fixed on polyacrylamide gels by incubating twice in 50% methanol, 5% acetic acid for 45 min, then washed in 3% acetic acid twice.Carbohydrate groups from LPS were oxidised by incubating the gel with periodic acid for 30 min and washing again in 3% acetic acid three times.Gels were stained with Pro-Q Emerald 300 for 2 h in the dark and washed with 3% acetic acid twice more.LPS was visualized in a Gel Doc XR+ System (Bio-Rad) under UV light at 300 nm.