A modular platform for one-step assembly of multi-component membrane systems by fusion of charged proteoliposomes

Abstract

An important goal in synthetic biology is the assembly of biomimetic cell-like structures, which combine multiple biological components in synthetic lipid vesicles. A key limiting assembly step is the incorporation of membrane proteins into the lipid bilayer of the vesicles. Here we present a simple method for delivery of membrane proteins into a lipid bilayer within 5 min. Fusogenic proteoliposomes, containing charged lipids and membrane proteins, fuse with oppositely charged bilayers, with no requirement for detergent or fusion-promoting proteins, and deliver large, fragile membrane protein complexes into the target bilayers. We demonstrate the feasibility of our method by assembling a minimal electron transport chain capable of adenosine triphosphate (ATP) synthesis, combining Escherichia coli F₁F_o ATP-synthase and the primary proton pump bo₃-oxidase, into synthetic lipid vesicles with sizes ranging from 100 nm to ∼10 μm. This provides a platform for the combination of multiple sets of membrane protein complexes into cell-like artificial structures.

Figure 1. Fusion of lipid vesicles studied with cobalt-calcein liquid content transfer assay.

(a) Schematic representation of the assay. Fusion of non-fluorescent cobalt-calcein loaded positively charged small unilamellar vesicles (SUV⁺) (red) with EDTA-loaded negatively charged vesicles of various sizes (blue) monitored by calcein fluorescence in the fusion product (pink) upon liquid content mixing. (b) Calcein fluorescence versus time upon mixing suspensions of oppositely charged SUV in buffer (1 mM MOPS, pH 7.4) (red) or buffer with 20 (blue) or 200 (black) mM KCl. Increasing fluorescence indicates vesicle fusion. Replacing cationic with neutral vesicles gives no fusion (grey). The inset: % of fusion as a function of KCl concentration. (c) Fusion between SUV⁺ and GUV⁻, in bright-field (left) and epi-fluorescence (right) microscopy. Post-fusion vesicles containing free calcein form fluorescent clumps (bottom), which are absent in controls lacking GUV⁻ (top) or SUV⁺ (middle). (d) A surface-immobilized anionic GUV, following fusion with one or more cationic proteoliposomes (PL⁺) anchored to the surface by ATP-synthase in 20 mM KCl, 1 mM MOPS pH 7.4, in bright-field (left) and fluorescence (right). Fluorescence is due to fluorescent cholesterol originally included in the PL⁺ membranes. Unfused PL⁺ are visible out of focus in the background. Fluorescence of the GUV membrane indicates fusion with the PL⁺. Scale bars are 10 μm in c,d. (e) A schematic of the experiment illustrated in d.

Figure 2. Delivery of functional ATP-synthase into target membranes by vesicle fusion.

(a) Pink: proton pumping driven by ATP hydrolysis in ATP-synthase, delivered into anionic SUV by fusion with cationic proteoliposomes (PL⁺) in 10 mM KCl, 1 mM MgCl₂, 1 mM MOPS pH 7.4 and assayed by ACMA quenching in the same buffer. Protons are pumped into the fusion product in the presence of 0.2 mM ATP, acidifying the interior which quenches ACMA fluorescence. Red: PL⁺ alone show less ACMA quenching without the addition of anionic SUV. Black, blue: neutral or anionic proteoliposomes show more ACMA quenching than PL⁺. Dark blue: addition of anionic SUV to PL⁻ has no effect on ACMA quenching. Schematic representation of this experiment is shown in the right panel. (b,c) Red: proton pumping driven by ATP hydrolysis in ATP-synthase, delivered into anionic LUV (b) or GUV (c) by fusion with PL⁺, assayed by ACMA quenching as described for a. Addition of uncoupler nigericin dissipated the proton gradient. Unfused PL⁺ were removed by low g centrifugation. Black: as above, but with PL⁺ replaced by neutral PL⁰, which do not fuse with LUV and GUV, and thus ATP-synthase is removed by centrifugation leading to negligible ACMA quenching. Blue: as above, but with no proteoliposomes.

Figure 3. Modular assembly of a functional electron transport chain by delivering ATP-synthase and bo₃-oxidase into target membranes.

(a) Schematic representation of the experiment. (top) Binary system (two components): cationic (PL⁺) and anionic proteoliposomes (PL⁻) fused together. (bottom) Ternary system (three components): anionic LUV or GUV fused with two types of PL⁺ containing either ATP-synthase or bo₃-oxidase. The vesicles were fused in 20 mM KCl, 5 mM MOPS pH 7.4, 1 mM MgCl₂ for 5–7 min followed by addition of KCl and MOPS to final concentrations of 100 and 50 mM, respectively, the luciferin-luciferase-ADP cocktail prepared as described in Methods, and oxidized Coenzyme Q₁. In the assembled electron transport chain energization of the membrane is triggered by addition of dithiothreitol (DTT_red), which reduces Q₁ (Q₁H) to make it available to bo₃-oxidase. Oxidation of reduced Q₁ by bo₃-oxidase pumps protons into each post-fusion vesicle. This builds up a PMF which drives ATP synthesis by ATP-synthase in the same vesicle. ATP synthesis is initiated by adding potassium phosphate (KP_i) to energized vesicles 1 min after addition of DTT. The ATP synthesized is detected by the luciferin-luciferase system, where conversion of synthesized ATP into pyrophosphate and AMP by luciferase is followed by light emission registered in a luminometer. (b) ATP synthesis by binary system. Blue trace: ATP synthesis by F₁F_o PL⁻ fused with bo₃ PL⁺; grey: same as blue but with nigericin; pink: same as blue but with DCCD-treated F₁F_o PL⁻; green: same as blue but in 200 mM KCl; black: same as blue, but with bo₃ PL⁰;. Red trace: ATP synthesis by F₁F_o PL⁺ fused with bo₃ PL⁻. (c,d) ATP synthesis by ternary system. Red trace: ATP synthesis by F₁F_o PL⁺ and bo₃ PL⁺ fused with LUV⁻ or GUV; green: same as red but in 200 mM KCl. Black: PL⁻ were used instead of PL⁺ in the fusion reaction.