Phospholipid headgroup composition modulates the molecular interactions and antimicrobial effects of sulfobetaine zwitterionic detergents against the “ESKAPE” pathogen Pseudomonas

We determine the efficacy for three known structurally related, membrane active detergents against multidrug resistant and wild type strains of Pseudomonas aeruginosa. Accessible solution state NMR experiments are used to quantify phospholipid headgroup composition of the microbial membranes and to gain molecular level insight into antimicrobial mode of action.


Contents
Section 2: Experimental General remarks: The NMR spectra were obtained using an Avance III 600 Hz spectrometer and processed using processed MestreNova software.Compounds 1 -3 were purchased from commercial sources with a purity level of ≥ 95 % confirmed by HPLC analysis.

DLS potential studies:
All vials used for preparing samples were clean and dry.Milli-Q® water was filtered to remove any particulates that may interfere with the results obtained.Samples were heated to ~50 °C and then filtered also to remove any potential solid beads from synthesis.A of 9 or 10 runs were recorded at 25 °C, allowing 10 mins for sample equilibration before the first measurement was recorded.These data were then averaged.
Phospholipid extraction: Bacterial cells were grown to stationary phase in 500mL tryptic soy broth (TSB), pelleted through centrifugation and resuspended in small volumes of TSB.The cells were then boiled at 98°C for 30 minutes and then snap frozen and stored at -80 to result in non-viable cells.The preparations were resuspended in Lysis buffer (300 mM NaCl, 50 mM NaH2PO4), then sonicated for a total time of 10 minutes, alternated on/off 30 s at a time.The preparations were centrifuged at 10,000 rpm at 4 °C for 15 minutes.The resultant supernatant was then centrifuged at 40,000 rpm at 4 °C for 1 hour.Membrane pellet resuspended in 1 mL Lysis buffer and extracted using the Folch method. 1 Briefly, methanol (7.5 mL) was added to the suspension and mixed, chloroform (15 mL) added and incubated for 1 hour at 25 °C.Phase separation induced by the addition of water (6.25 mL), after 1 hour at 25 °C, the lower layer was collected and dried.
Nanodisc preparation: Lipid films were resuspended in Buffer A (20 mM NaCl, 20 mM NaH2PO4, pH 7.4) and sonicated for 1 hour, at a 5:1 ratio, SMA was added and incubated at 37 °C for 1 hour.Dialysed overnight in 5 L of the same buffer in 10 KDa cut-off dialysis tubing, concentrated, and underwent gel filtration size exclusion chromatography on a Superdex 200 10/300 GL column (GE Healthcare) in the same buffer, whilst monitoring absorbance at 260 nm.Nanodisc quantification carried out via monitoring of absorbance at 260 nm using a calibration curve. 2 Nanodiscs to be stored for future use were aliquoted into 50 µM in 100 µL portions and stored at -80 °C.

H NMR CPMG:
1 H NMR CPMG 3,4 spectra were obtained with a Bruker Avance III 600 MHz spectrometer equipped with a QCI-P cryoprobe at 298 K. Samples were supplemented with 5 % D2O for locking and DSS (0.02 mM) as a chemical shift standard.The standard zgprcpmg pulse sequence from the Bruker library was modified with a watergate element to allow suppression of the water signal and was used for all experiments.This sequence is available on request from the authors.Presaturation was applied between acquisitions for 100 ms at a power level of 1.78x10 -5 W, further water suppression was achieved with a 3-9-19 watergate sequence using 1 ms gradient pulses with a smoothed chirp shape and a peak Z gradient field strength of 21.2 G.The CPMG element had a length of 300 ms with delays between 180 o pulses of 1 ms.Repetition times were chosen to achieve suitable water suppression (RG < 256) with the presat and watergate sequences used with the 95 % H2O : 5 % D2O solvent.The length of the CPMG element was chosen to achieve good differentiation between unbound and strongly bound ligands.Data was collected with 16,384 points and a spectral width of 16.0242 ppm, the receiver gain was set to 256, with 512 scans, 8 dummy scans were acquired with an interscan delay.Data was processed using MestreNova version 14.2.2.All spectra were automatically phased and baselined corrected and chemical shifts were calibrated to the centre of the DSS peak.

Processing titration data:
The NMR resonances of interest were integrated to give the 'absolute area'.These data were then data was normalised and converted to percentage coordination.These data were then fitted to Hill Plot Kinetics: growth/sigmoidal category, Hill function, Levenberg Marquardt iteration algorithm.Vmax was fixed to 100 % as this was the greatest proportion of 1 -3 that could be coordinated to the nanodiscs.Normalising -All absolute integration values was divided by the initial titration point (0 mM of nanodisc) to give values ranging from 0 -1.
2D NMR Experiments: 1 H-31 P HSQC NMR spectra were acquired on a Bruker Avance III 600 MHz spectrometer using a QCIP cryoprobe with a standard 31 P pre-amplifier without enhanced sensitivity from cryogenic cooling.Initial 1D 1 H experiment were measured for quality control.Spectra were acquired using the pulse sequence na_hsqcetf3gpxy. 5Time domain data of 2048 and 128 complex points were used in the direct and indirect domains respectively.The sweep widths used in the 1 H and 31 P dimensions were 14423 and 2915 Hz respectively.The inter scan delay was set to 1 s and 256 scans were acquired per increment with 16 dummy scans at the start of each experiment.Offsets of 4.70 and -1.00 ppm were used for the 1 H and 31 P channels. 1H decoupling was achieved during acquisition using a GARP4 sequence and the CPMG sequence during the 1 H-31 P/ 31 P-1 H transfers used 256 loops, and delays of 215 µs.The gain was limited to a maximum of 256.2D spectra were processed with MestReNova version 14.2.2.All 2D NMR spectra were automatically phased and chemical shifts were calibrated to the centre of the TMP peak ( 1 H = 3.78 ppm, 31 P = 2.07 ppm).
Lipid quantification: Quantification of lipids was carried out using the 2D HSQC 1 H-31 P NMR spectra.The relevant 2D peak was picked and the percentage of total phospholipid in each sample is calculated using the ratio of each phospholipid intensity over the sum of the intensities of all phospholipids.
Membrane fluidity assay: Black bottom 96-well plates were prepared by serially diluting desired compound in H2O across the plate, the appropriate DPH labelled vesicles (100 µL, 30 µM) were added to each well to give a total well volume of 200 µL.FP measurements were taken at 25 °C using a 355 nm filter for excitation and a 430 m for emission.The DPH labelled vesicles were set to a FP value of 100 mP.Data were acquired in endpoint mode.All experiments were repeated in triplicate to ensure experimental reproducibility.

Section 3: Microbial materials and methods
General remarks: Bacterial strains were maintained on Tryptic Soy Agar and all assays were carried out in cation-adjusted Mueller Hinton Broth (MHB) unless otherwise stated.

Minimum Inhibitory Concentrations (MICs):
MICs, defined as the lowest concentration of compound need to inhibit 100 % of visible bacterial growth.MIC50 defined as the lowest concentration of compound needed to inhibit 50 % of visible bacterial growth.Both values were determined experimentally using broth microdilution methods.Briefly, compounds were added to the first column of a 96 well plate and serially diluted down the plate in MH.Bacteria were added to the plate at a final concentration of 5 x 10 6 CFU/mL.Media only controls and untreated bacterial controls were included.Plates were incubated at 37 °C for 20 hours in a CLARIOstar plate reader (BMG Labtech) and the OD600 was read every hour.
Scanning electron microscopy: Bacteria were treated with compounds at 1 mM and 32 mM for 24 hours before being fixed using formaldehyde.Bacteria were then immobilised on poly-l-lysine coated 10mm diameter glass coverslips overnight in a humid chamber at room temperature.Samples on coverslips were secondarily fixed in 2 % osmium tetroxide for 1 hour at room temperature.Coverslips were then dehydrated through a graded ethanol series at room temperature, washed twice in hexamethyldisilazane (HDMS), then air dried.Coverslips were attached to SEM stubs using adhesive carbon disc and gold coated using Atom Tech Ultra Fine Grain Sputter Coater (Z705) and examined using a Zeiss Sigma 300VP SEM.
NPN assay: NPN assay: 1-N-phenylnaphtylamine (NPN) is a fluorescent probe which is excluded from the outer membrane due to its hydrophobic nature.When the outer membrane is damaged, NPN can enter into the phospholipid layer, resulting in prominent fluorescence. 6Bacteria in mid-logarithmic growth were washed twice in buffer A (5 mM HEPES buffer, 5 mM glucose, pH 7.2) and added to a black 96 well plate containing compounds dilutions and the positive control polymyxin b at 10 µg/mL.NPN was added at a final concentration of 10 µM immediately before fluorescence was measured (350ex, 420em) using a CLARIOstar plate reader.The outer membrane permeabilization was calculated as follows: Outer membrane permeabilization (%) = (F0bs -F0) / (F100 -F0) x 100%, where F0bs is the observed fluorescence of the sample, F0 is the initial fluorescence of NPN in bacteria in the absence of compound, and F100 is the fluorescence of NPN with bacteria upon the addition of 100 mM CHAPs.
Section 4: Summary data Tables

Section 1 :
Figure S1 -Chemical structure of 1 (CHAPS), 2 (ASB-14) and 3 (SB 3-14), with the proton environments for 1 H NMR resonances labelled in orange.The same letters have been used for different proton environments where the 1 H NMR resonances have been found to overlap.

Figure S24 -
Figure S24 -The relative normalised area of a downfield aromatic resonance of 1 upon titration with PAO1 nanodiscs.Dark blue = a, orange = b, grey = c, yellow = d, light blue = e, green = f.

Figure S25 -
Figure S25 -The relative normalised area of a downfield aromatic resonance of 1 upon titration with PAO1 nanodiscs, n=3, error = standard deviation of the mean.Dark blue = a, orange = b, grey = c, yellow = d, light blue = e, green = f.

Figure S26 -
Figure S26 -The relative normalised area of a downfield aromatic resonance of 2 upon titration with PAO1 nanodiscs.Dark blue = a, orange = b, grey = c, yellow = d.

Figure S27 -
Figure S27 -The relative normalised area of a downfield aromatic resonance of 3 upon titration with PAO1 nanodiscs.Dark blue = a, orange = b, grey = c, yellow = d.

Figure S28 -
Figure S28 -The relative normalised area of a downfield aromatic resonance of 1 upon titration with NCTC 13437 nanodiscs.Dark blue = a, orange = b, grey = c, yellow = d, light blue = e, green = f.

Figure S29 -
Figure S29 -The relative normalised area of a downfield aromatic resonance of 1 upon titration with NCTC 13437 nanodiscs, n=3, error = standard error of the mean.Dark blue = a, orange = b, grey = c, yellow = d, light blue = e, green = f.

Figure S30 -
Figure S30 -The relative normalised area of a downfield aromatic resonance of 2 upon titration with NCTC 13437 nanodiscs.Dark blue = a, orange = b, grey = c, yellow = d.

Figure S31 -
Figure S31 -The relative normalised area of a downfield aromatic resonance of 3 upon titration with NCTC 13437 nanodiscs.Dark blue = a, orange = b, grey = c, yellow = d.

Figure S33 -
Figure S33 -Graph showing the percentage of 1, peak 2 coordinated to PAO1 nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S34 -
Figure S34 -Graph showing the percentage of 1, peak d/4 coordinated to PAO1 nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S35 -
Figure S35 -Graph showing the percentage of 1, peak e/5 coordinated to PAO1 nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S36 -
Figure S36 -Graph showing the percentage of 1, peak f/6 coordinated to PAO1 nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S37 -
Figure S37 -Graph showing the percentage of 1 coordinated to NCTC 13437 nanodiscs, with respect to increasing nanodisc concentration.

Figure S38 -
Figure S38 -Graph showing the percentage of 1, peak d/4 coordinated to NCTC 13437 nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S39 -
Figure S39 -Graph showing the percentage of 1, peak e/5 coordinated to NCTC nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S40 -
Figure S40 -Graph showing the percentage of 1, peak f/6 coordinated to NCTC nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure
Figure S41 -EC50 (µM) values obtained from the fitting of peaks af of 1 titration data to Hill Plot kinetics using Origin 2018 software, with Vmax fixed to 100 % of 1 bound to the nanodisc.Purple = results from PAO1 nanodiscs, orange = results from NCTC 13437 nanodiscs.

Figure S42 -
Figure S42 -Graph showing the percentage of 2 coordinated to PAO1 nanodiscs, with respect to increasing nanodisc concentration.

Figure S44 -
Figure S44 -Graph showing the percentage of 2, peak b/2 coordinated to PAO1 nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S45 -
Figure S45 -Graph showing the percentage of 2, peak c/3 coordinated to PAO1 nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S46 -
Figure S46 -Graph showing the percentage of 2, peak d/4 coordinated to PAO1 nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S47 -
Figure S47 -Graph showing the percentage of 2 coordinated to NCTC 13437 nanodiscs, with respect to increasing nanodisc concentration.

Figure S48 -
Figure S48 -Graph showing the percentage of 2, peak a/1 coordinated to NCTC 13437 nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S49 -
Figure S49 -Graph showing the percentage of 2, peak b/2 coordinated to NCTC nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S50 -
Figure S50 -Graph showing the percentage of 2, peak c/3 coordinated to NCTC nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S51 -
Figure S51 -Graph showing the percentage of 2, peak d/4 coordinated to NCTC 13437 nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S52 -
Figure S52 -EC50 (µM) values obtained from the fitting of peaks a -d of 2 titration data to Hill Plot kinetics using Origin 2018 software, with Vmax fixed to 100 % of 2 bound to the nanodisc.Purple = results from PAO1 nanodiscs, orange = results from NCTC 13437 nanodiscs.

Figure S53 -
Figure S53 -Graph showing the percentage of 3 coordinated to PAO1 nanodiscs, with respect to increasing nanodisc concentration.

Figure S54 -
Figure S54 -Graph showing the percentage of 3, peak a/1 coordinated to PAO1 nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S55 -
Figure S55 -Graph showing the percentage of 3, peak b/2 coordinated to PAO1 nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S56 -
Figure S56 -Graph showing the percentage of 3, peak c/3 coordinated to PAO1 nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S57 -
Figure S57 -Graph showing the percentage of 3, peak d/4 coordinated to PAO1 nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S58 -
Figure S58 -Graph showing the percentage of 3 coordinated to NCTC 13437 nanodiscs, with respect to increasing nanodisc concentration.

Figure S59 -
Figure S59 -Graph showing the percentage of 3, peak a/1 coordinated to NCTC nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S60 -
Figure S60 -Graph showing the percentage of 3, peak b/2 coordinated to NCTC nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S61 -
Figure S61 -Graph showing the percentage of 3, peak c/3 coordinated to NCTC nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure S62 -
Figure S62 -Graph showing the percentage of 3, peak d/4 coordinated to NCTC nanodiscs, with respect to increasing nanodisc concentration.Data was then fitted to the Hill Plot model with Vmax fixed at 100 %.

Figure
Figure S63 -EC50 (µM) values obtained from the fitting of peaks ad of 3 titration data to Hill Plot kinetics using Origin 2018 software, with Vmax fixed to 100 % of 3 bound to the nanodisc.Purple = results from PAO1 nanodiscs, orange = results from NCTC 13437 nanodiscs.

Figure S64 -
Figure S64 -Effect of 1 on FP measured in DPH labelled PAO1 vesicles at 25 o C. A target FP value of 100 mP was set to the DPH labelled vesicles.

Figure S65 -
Figure S65 -Effect of 2 on FP measured in DPH labelled PAO1 vesicles at 25 o C. A target FP value of 100 mP was set to the DPH labelled vesicles.

Figure S66 -
Figure S66 -Effect of 3 on FP measured in DPH labelled PAO1 vesicles at 25 o C. A target FP value of 100 mP was set to the DPH labelled vesicles.

Figure S67 -
Figure S67 -Effect of 1 on FP measured in DPH labelled NCTC 13437 vesicles at 25 o C. A target FP value of 100 mP was set to the DPH labelled vesicles.

Figure S68 -
Figure S68 -Effect of 2 on FP measured in DPH labelled NCTC 13437 vesicles at 25 o C. A target FP value of 100 mP was set to the DPH labelled vesicles.

Figure S69 -
Figure S69 -Effect of 3 on FP measured in DPH labelled NCTC 13437 vesicles at 25 o C. A target FP value of 100 mP was set to the DPH labelled vesicles.

Table S1 -
Summary table of physicochemical studies from literature and experimentally derived.Size calculated using dynamic light scattering (DLS).Experimental data carried out at 10 mM.

Table S2 -
R 2 , Hill coefficient and EC50 (µM) values obtained from the fitting of nanodisc titration data to Hill Plot kinetics using Origin 2022 software, with Vmax fixed to 100 % of 1 bound to the nanodiscs.Data is ranked, with 1 = 1 H resonance with lowest EC50 value.

Table S3 -
R 2 , Hill coefficient and EC50 (µM) values obtained from the fitting of nanodisc titration data to Hill Plot kinetics using Origin 2022 software, with Vmax fixed to 100 % of 2 bound to the nanodiscs.Data is ranked, with 1 = 1 H resonance with lowest EC50 value.

Table S4 -
R 2 , Hill coefficient and EC50 (µM) values obtained from the fitting of nanodisc titration data to Hill Plot kinetics using Origin 2022 software, with Vmax fixed to 100 % of 3 bound to the nanodiscs.Data is ranked, with 1 = 1 H resonance with lowest EC50 value.