Modulating Liposome Surface Charge for Maximized ATP Regeneration in Synthetic Nanovesicles

Abstract

In vitro reconstructed minimal respiratory chains are powerful tools to investigate molecular interactions between the different enzyme components and how they are influenced by their environment. One such system is the coreconstitution of the terminal cytochrome bo₃ oxidase and the ATP synthase from Escherichia coli into liposomes, where the ATP synthase activity is driven through a proton motive force (pmf) created by the bo₃ oxidase. The proton pumping activity of the bo₃ oxidase is initiated using the artificial electron mediator short-chain ubiquinone and electron source DTT. Here, we extend this system and use either complex II or NDH-2 and succinate or NADH, respectively, as electron entry points employing the natural long-chain ubiquinone Q₈ or Q₁₀. By testing different lipid compositions, we identify that negatively charged lipids are a prerequisite to allow effective NDH-2 activity. Simultaneously, negatively charged lipids decrease the overall pmf formation and ATP synthesis rates. We find that orientation of the bo₃ oxidase in liposomal membranes is governed by electrostatic interactions between enzyme and membrane surface, where positively charged lipids yield the desired bo₃ oxidase orientation but hinder reduction of the quinone pool by NDH-2. To overcome this conundrum, we exploit ionizable lipids, which are either neutral or positively charged depending on the pH value. We first coreconstituted bo₃ oxidase and ATP synthase into temporarily positively charged liposomes, followed by fusion with negatively charged empty liposomes at low pH. An increase of the pH to physiological values renders these proteoliposomes overall negatively charged, making them compatible with quinone reduction via NDH-2. Using this strategy, we not only succeeded in orienting the bo₃ oxidase essentially unidirectionally into liposomes but also found up to 3-fold increased ATP synthesis rates through the usage of natural, long-chain quinones in combination with the substrate NADH compared to the synthetic electron donor/mediator pair.

Keywords: artificial ATP production; charge-mediated fusion; energy conversion; ionizable lipids; liposomes; membrane protein orientation; synthetic biology.

Figure 1

(A) Bottom-up approaches of artificial respiratory chains usingE. coli enzymes. For approach I, FRD, bo₃ oxidase, and ATP synthase are coreconstituted into liposomes containing electron mediator ubiquinone Q₈. The addition of succinate leads to Q₈ reduction by FRD, followed by the reoxidation of Q₈H₂ and simultaneous proton translocation through bo₃ oxidase. The so-generated pmf is used by F₁F_O-ATP synthase to produce ATP, which in turn is detected via luminescence. In an alternative approach II, Q₈ is reduced upon NADH addition by the peripheral membrane protein NDH-2, which can be added to proteoliposomes during measurements. (B) Approach I (black trace)–ATP synthesis was initiated by adding 1 mM succinate and inhibited by 400 μM of FRD inhibitor malonate. Approach II (red trace)-to start ATP synthesis, 300–500 nM NDH-2 and 200 μM NADH were added, and ATP production was monitored. The reaction was stopped by bo₃ oxidase inhibitor KCN. (C) Coupled ATP synthesis rate initiated either by NDH-2/NADH or Q₁/DTT in DOPC liposomes containing varying amounts of DOPG. For Q₁/DTT-induced ATP synthesis, 20 μM Q₁ and 4 mM DTT were used to start the reaction, while NDH-2/NADH-induced ATP synthesis was measured, as described in (B). Rates were normalized to 100% DOPC (Q₁/DTT induced).

Figure 2

(A) Homology model of E. coli peripheral membrane protein NDH-2 (PDB access: 6BDO, from C. thermarum). Amino acids in the N-terminal domain are colored green. The C-terminus is depicted in blue, while the FAD cofactor is highlighted in orange. Interaction of the C-terminal helices to the negatively charged membrane (gray) is indicated. (B) NADH/quinone oxidoreductase activity measurement of NDH-2. Absorption of NADH is monitored at 340 nm. After reaching a baseline of buffer (20 mM HEPES pH 7.4, 200 mM NaCl, 20 mM KCl) containing 100 μM NADH, 1 mg/mL liposomes, and 100 μM Q₂, NADH oxidation is initiated by the addition of 5-10 nM NDH-2. (C) Lipid-dependent NDH-2 activity. NADH oxidation activity of NDH-2 was measured in the presence or absence (buffer) of different liposomes (1 mg/mL), as described for (B). To adjust for different specific activities of protein preparations, measurements from different NDH-2 batches have been normalized to activity with ECPE (100%). CL: cardiolipin and ECPE: E. coli polar extract.

Figure 3

(A) Structure of bo₃ oxidase (PDB access: 6WTI). Single-cysteine mutants used for orientation determination are depicted in spheres. Cysteines were located at the cytoplasmic side (ID578C or IIIA21C). (B) TCEP-based orientation determination of bo₃ oxidase. Site-specifically DY647P1-labeled bo₃ oxidase mutants are reconstituted in liposomes. To determine the orientation, fluorescence was monitored, and fluorophores located on the outside of liposomes are quenched in a first step by 14 mM TCEP. Full quench was achieved in a second step after adding 0.05% Triton X-100. To calculate the orientation, the first quench was set in relation to the full quench. Liposomes were composed of either only PC or 6:4 PC/PG. (C) bo₃ oxidase orientation in different liposomes. Different liposomes (10 mg/mL) were partially solubilized by 0.4% sodium cholate, and bo₃ oxidase was added. After detergent removal by gel filtration, liposomes were pelleted by ultracentrifugation, and orientation was determined via the TCEP-based assay. Liposomes were composed of either 100% PC, or of 60% PC and either 40% PG, 40% PS, or 40% TAP. (D,E) Orientation of bo₃ oxidase after reconstitution in the presence or absence of salt. DY647P1-labeled bo₃ oxidase-IIIA21C was reconstituted in absence or in the presence of 100 mM/300 mM NaCl either in pure PC liposomes (PC) or in 4:6 PG/PC (PG). Orientation was determined via TCEP-based assay and depicted in bar plots either as a fraction of inside-out orientation (D) or normalized to the orientation in the absence of salt (E). (F) Surface charge distribution of bo₃ oxidase with side view (left) and top and bottom view (right), respectively. Positively charged and negatively charged areas are colored in blue and red, respectively (drawn with PyMOL with PDB access 6WTI). (G) Orientation of bo₃ oxidase after coreconstitution with ATP synthase into liposomes containing TAP lipids. (H) Relative ATP synthesis rates of proteoliposomes of (G) energized with DTT/Q₁.

Figure 4

(A) Strategy to use the ionizable DODAP lipid to temporarily provide a positively charged membrane which becomes negatively charged upon liposome fusion and pH adjustment. Initially neutrally charged liposomes become temporarily positive when applying acidic pH, under which condition also coreconstitution (ATP synthase and bo₃ oxidase) is performed. Subsequent lipid mixing with negatively charged liposomes (e.g., 100% PG, brown) and physiological pH renders the liposome membrane overall negative, thus allowing NDH-2 to interact. Headgroups of TAP and ionizable DAP lipids are depicted. (B) Zeta potential measurements of differently charged liposomes. (C) Comparison of ATP synthesis rates between permanently (DOTAP) and transiently (DODAP) positively charged liposomes at different pH values. ATP production was chemically initiated with Q₁ and DTT. (D) Normalized ATP synthesis efficiency induced chemically (Q₁/DTT) in uncharged (gray), negatively charged (red), permanently positively charged (light blue), or ionizable liposomes (PC/DAP, different reconstitution pH, blue-gray mesh). (E) Normalized ATP synthesis induced by NDH-2/NADH (except first column) using DOPC/DODAP liposomes fused with different negatively charged liposomes (see text for details).