Energetics of lipid transport by the ABC transporter MsbA is lipid dependent

The ABC multidrug exporter MsbA mediates the translocation of lipopolysaccharides and phospholipids across the plasma membrane in Gram-negative bacteria. Although MsbA is structurally well characterised, the energetic requirements of lipid transport remain unknown. Here, we report that, similar to the transport of small-molecule antibiotics and cytotoxic agents, the flopping of physiologically relevant long-acyl-chain 1,2-dioleoyl (C18)-phosphatidylethanolamine in proteoliposomes requires the simultaneous input of ATP binding and hydrolysis and the chemical proton gradient as sources of metabolic energy. In contrast, the flopping of the large hexa-acylated (C12-C14) Lipid-A anchor of lipopolysaccharides is only ATP dependent. This study demonstrates that the energetics of lipid transport by MsbA is lipid dependent. As our mutational analyses indicate lipid and drug transport via the central binding chamber in MsbA, the lipid availability in the membrane can affect the drug transport activity and vice versa.

WT and 1% biotin-Lipid-A (b), and with a pHin of 6.8, were loaded with the fluorescent pH indicator BCECF (apparent pKa 6.98) and diluted 100-fold in buffer with pH 8.0 to impose a ΔpH (pHin 6.8/pHout 8.0) (blue trace) or in buffer with pH 6.8 as a control (pHin 6.8/pHout 6.8) (red trace). At the arrows, 2 µM of the ionophore nigericin or 0.025 % (v/v) Triton X-100 was added to increase the proton permeability of the proteoliposomal membrane, allowing protons to diffuse from the acidic interior to the alkaline exterior. Values in the histogram in (a) show fluorescence levels as mean ± s.e.m (n = 3). Asterisks above the square bracket show significant differences following the increase in the proton permeability of the proteoliposomes, indicating the stability of the imposed pH gradient during the biotin-PE and biotin-Lipid-A floppase assays (one-way analysis of variance; ****P＜0.0001).

Supplementary Figure 2. MsbA does not transport free biotin in the biotin-lipid
transport assays. a Proteoliposomes without biotin-PE received 3 µM free biotin before the start of the transport reaction, which yields a very similar fluorescence level as the standard amount of biotin-PE in the proteoliposomes. Reactions were started with the provision of different combinations of metabolic energy: (i) control (pHin 6.8/pHout 6.8), (ii) imposed ΔpH (pHin 6.8/pHout 8.0), (iii) ATP (pHin 6.8/pHout 6.8), and (iv) ATP plus imposed ΔpH (pHin 6.8/pHout 8.0). After the reaction time was completed, the free biotin concentration in the external buffer was measured. No significant differences in biotin concentrations were observed, suggesting that MsbA does not mediate the uptake of free biotin in the proteoliposomes. b Identical transport assays as in (a) were performed with biotin-PE-containing proteoliposomes. The proteoliposomes were then spun down by centrifugation, after which the supernatant was used to measure the concentration of free biotin liberated from the biotin-PE by the end of the transport assay. No significant differences were found for the various incubations, or when compared with proteoliposomes lacking biotin-PE. This result indicates that the biotin-PE is not hydrolysed into free biotin and PE over the time course of the transport experiments.
Values in the histograms represent observations in three experiments (n = 3) with independently prepared batches of proteoliposomes and are expressed as mean ± s.e.m (one-way analysis of variance; not significant).

Chemical synthesis of MsbA inhibitor G907
All reagents were obtained from commercial sources and used without further purification. Tetrahydrofuran was dried over sodium wire and distilled from a mixture of lithium aluminium hydride and calcium hydride with triphenylmethane as the indicator. Diethyl ether was distilled from calcium hydride and lithium aluminium hydride. Acetonitrile, dichloromethane, toluene and methanol were all distilled from calcium hydride. Petroleum ether was distilled before use and refers to the fraction with a boiling point of 40-60 °C.
Yields refer to chromatographically and spectroscopically pure compounds unless otherwise stated. Thin layer chromatography (TLC) was performed on commercially available glass Merck Kieselgel 60 F254 plates, and visualised UV irradiation (254 nm), or by staining with potassium permanganate dip. Rf values are quoted to the nearest 0.01. Flash column chromatography was performed using silica gel (Merck Kieselgel 60 F254, 230-400 mesh) under a positive pressure of nitrogen.
Proton magnetic resonance spectra were recorded using an internal deuterium lock (at 298 K unless stated otherwise) on Bruker DPX (400 MHz; 1H-13C DUL probe), Bruker Avance III HD (400 MHz; Smart probe), Bruker Avance III HD (500 MHz; Smart probe) and Bruker Avance III HD (500 MHz; DCH Cryoprobe) spectrometers. Assignments are supported by 1 H-1 H COSY, 1 H-13 C HSQC or 1 H-13 C HMBC spectra, or by analogy. Chemical shifts (δH) are quoted in ppm to the nearest 0.01 ppm and are referenced to the residual non-deuterated solvent peak: CHCl3 (7.26) or d5-DMSO (2.50). Discernible coupling constants for mutually coupled protons are reported as measured values in Hertz, rounded to the nearest 0.1 Hz. Data are reported as: chemical shift, number of nuclei, multiplicity (br, broad; s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; or a combination thereof), coupling constants and assignment. Two or more possible assignments were given when signals could not be distinguished by any means. Spectra were processed using TopSpin v. 4.0.1 (Bruker).
Carbon magnetic resonance spectra were recorded using an internal deuterium lock (at 298 K unless stated otherwise) on Bruker DPX (101 MHz), Bruker Avance III HD (101 MHz) and Bruker Avance III HD (126 MHz) spectrometers with broadband proton decoupling. Assignments are supported by DEPT editing, 1 H-13 C HSQC or 1 H-13 C HMBC spectra, or by analogy. Chemical shifts (δC) are quoted in ppm to the nearest 0.1 ppm and are referenced to the deuterated solvent peak: CDCl3 (77.2) or d6-DMSO (39.5). Data are reported as: chemical shift and assignment. An aryl (Ar), quaternary (CQ), or two or more possible assignments were given when signals could not be distinguished by any means. Spectra were processed using TopSpin v. 4.0.1 (Bruker).
Infrared spectra were recorded neat on a Perkin Elmer Spectrum One FT-IR spectrometer with internal referencing. Selected absorption maxima (nmax) are quoted in wavenumbers (cm -1 ) and are assigned as: weak (w), medium (m), strong (s), broad (br), or a combination thereof. Melting points were obtained on a Büchi B-545 melting point apparatus and are uncorrected. High resolution mass spectrometry (HRMS) measurements were carried out on a Waters LCT Premier Time of Flight mass spectrometer or a Micromass Quadrupole Time of Flight mass spectrometer.
G907 was prepared following the procedure reported by Ho et al. 2 The overall scheme is outlined below.

2
N-Iodosuccinimide (1.28 g, 5.71 mmol, 1 eq.) was added to a solution of 1 (1.00 g, 5.71 mmol) in acetic acid (12.5 mL). The reaction mixture was stirred at rt for 3 h and then the formed precipitate was filtered, washed with 40-60 petroleum ether (20 mL) and ethyl acetate (20 mL) to afford 2 as a grey solid (1.56 g, 5.18 mmol, 91%) which was carried on without further purification.  Characterisation data is in accordance with the literature. 1

11
Methylmagnesium bromide (4.25 mL of 3 M solution in ether, 12.74 mmol, 2 eq.) was added to an ice-cold solution of 10 (1.15 g, 6.37 mmol) in THF (20 mL) and stirred for 1 h at room temperature. The reaction mixture was quenched with saturated aqueous ammonium chloride solution (20 mL) before being extracted with ethyl acetate (3 x 30 mL). The organic layer was washed with water (30 mL), brine (30 mL), dried over sodium sulfate, and concentrated to afford 11 as a white solid (1.11 g, 5.63 mmol, 88%) which was carried on without further purification.  Characterisation data is in accordance with the literature. 1