Synthetic Vesicles for Sustainable Energy Recycling and Delivery of Building Blocks for Lipid Biosynthesis

ATP is a universal energy currency that is essential for life. l-Arginine degradation via deamination is an elegant way to generate ATP in synthetic cells, which is currently limited by a slow l-arginine/l-ornithine exchange. We are now implementing a new antiporter with better kinetics to obtain faster ATP recycling. We use l-arginine-dependent ATP formation for the continuous synthesis and export of glycerol 3-phosphate by including glycerol kinase and the glycerol 3-phosphate/Pi antiporter. Exported glycerol 3-phosphate serves as a precursor for the biosynthesis of phospholipids in a second set of vesicles, which forms the basis for the expansion of the cell membrane. We have therefore developed an out-of-equilibrium metabolic network for ATP recycling, which has been coupled to lipid synthesis. This feeder–utilizer system serves as a proof-of-principle for the systematic buildup of synthetic cells, but the vesicles can also be used to study the individual reaction networks in confinement.

Table S2.Plasmids used in this study.
Table S3.Primers used in this study.
Table S4.Kinetic parameters of ArcD and ArcE proteins based on in vivo L-arginine uptake.
Table S6.Mass-to-charge ratio of lipid species used in this study.

ArcE homologues, matrix with percentage identity
† The author deceased before publication Keywords: ATP recycling | glycerol 3-P/Pi antiporter | out-of-equilibrium metabolic network | phospholipid biosynthesis | building block delivery | synthetic cells Supplementary Figures and Tables:

Figure S1 .
Figure S1.Amino acid sequence alignment and identity of ArcD and ArcE proteins.

Figure S4 .
Figure S4.In vivo L-arginine uptake by ArcD and ArcE transport proteins.

Figure S5 .
Figure S5.SDS-PAA gel images of purified membrane proteins used in this study.

FigureFigure S7 .Figure S8 .
Figure S6.L-arginine-mediated ATP formation in the presence of L. sakei ArcD or L. lactis ArcD2.Figure S7.Activity of the feeder LUVs measured by online ATP/ADP readout with PercevalHR. Figure S8.Diffusion of phospholipid precursors through polycarbonate dialysis filters.

Figure S2 .
Figure S2.Predicted topology of ArcD and ArcE proteins.The predicted topology 3 of ArcD and ArcE proteins is shown for internal (green) and external (purple) loop regions and transmembrane (grey) segments.

Figure S4 .
Figure S4.In vivo L-arginine uptake by ArcD and ArcE transport proteins.Single colonies of L. lactis JP9000 ΔarcD1ΔarcD2 carrying pNZarcD/E were resuspended in 10 mL of M17 medium supplemented with 1% glucose plus 5 µg/mL of chloramphenicol (GM17Cm).This suspension was used to prepare eight precultures by 1:3 v/v or 1:12 v/v serial dilution into 3 mL fresh GM17Cm that were incubated overnight at 30 ºC without shaking.The following day, 50 mL cultures were prepared by diluting 1:3 v/v the overnight pre-cultures, which were in the late exponential phase of growth (0.7 < OD600 < 1.5); cultures in the stationary phase (OD600 > 2) were diluted 1:100 v/v.These cultures were incubated at 30 ºC, 100 rpm, and over-expression was induced at an OD600 of 0.4-0.5 with 0.5 ng/mL nisin A. After 1 hour induction, cells were harvested by centrifugation (10 min, 4.000 g, 4 ºC), washed with 50 mM K-HEPES pH 7.0 and resuspended to a final OD600 of 50 in 50 mM K-HEPES pH 7.0 supplemented with 1 mM L-ornithine.The resuspended cells were incubated in the presence of 1 mM L-ornithine for 1 hour at RT, or overnight at 4 ºC.Next, a 1:20 v/v dilution was prepared by pipetting 5 µL cells into 90 µL of 50 mM K-HEPES pH 7.0, followed by a 30 sec incubation at 20 ºC with stirring.L-arginine uptake was initiated by adding 5 µL of radiolabeled L-arginine from appropriate 20x stocks (final concentration 0-50 µM).Cells were incubated at 20 ºC with shaking, and L-arginine uptake was quenched at desired time intervals (10-20 seconds) by addition of 2 mL ice-cold 100 mM LiCl.An initial timepoint was prepared by adding the LiCl solution before L-arginine.The cells were poured onto 0.45 µm nitrocellulose filters (Cytiva), and washed with additional 2 mL of ice-cold 100 mM LiCl.Filters were collected in 2 mL tubes (Eppendorf) to which 2 mL Ultima Gold MW scintillation fluid (PerkinElmer) was added.Total counts were obtained by adding 5 µL L-arginine solution directly onto a filter.Filters were dissolved by vortexing and radioactivity was measured in a Tri-Carb 2800 high-performance liquid scintillation analyzer (Perkin Elmer).

Figure S5 .
Figure S5.SDS-PAA gel images of purified membrane proteins used in this study.The bands of purified membrane proteins were used to quantify the efficiency of (co-)insertion into vesicles.ArcD2 from L. lactis (57 kDa) 1 ; ArcD from L. sakei (54 kDa); GlpT from E. coli (54 kDa) 10 .All proteoliposomes have lipid-to-protein ratios of 400:1 (w/w) for each membrane protein.A) Purified ArcD2 from L. lactis and proteoliposomes.B) Purified ArcD from L. sakei and proteoliposomes.C) Co-reconstitution of ArcD2 from L. lactis and GlpT in proteoliposomes.Efficiency of reconstitution of ArcD2 from L. lactis when co-reconstituted with GlpT.D) Co-reconstitution of ArcD from L. sakei and GlpT in proteoliposomes.Efficiency of reconstitution of ArcD from L. sakei when co-reconstituted with GlpT of.E) Purified GlpT and co-reconstitution of ArcD2 from L. lactis or ArcD from L. sakei and GlpT in proteoliposomes.Bands at ~35 and ~70 kDa represent monomers and dimers of the ArcD proteins, respectively.Efficiency of co-reconstitution of GlpT and either ArcD protein.

Figure
Figure S6.L-arginine-mediated ATP formation in L. sakei ArcD or L. lactis ArcD2 proteo-LUVs.The ATP/ADP ratio was determined online by acquiring PercevalHR excitation spectra over time from the same proteoliposomes used for ATP quantification with the luciferase assay; traces are representative of at least 3 independent co-reconstitutions.The fluorescence data is in agreement with the chemiluminescence results (see Main text, Figure3b) and indicates that ATP formation is ~2.5xfaster with ArcD from L. sakei than with ArcD2 from L. lactis.

Figure S7 .
Figure S7.Activity of the feeder vesicles measured by online ATP/ADP readout withPercevalHR.The experiment was carried out as reported in the Main text (see Methods section).Both types of vesicles (with ArcD from L. sakei plus GlpT from E. coli and the negative control with ArcD from L. sakei only) produced ATP upon addition of 20 mM L-arginine, as revealed by the increase in the F500/F430 ratio; traces are representative of at least 3 independent coreconstitutions.

Figure S8 .
Figure S8.Diffusion of phospholipid precursors through polycarbonate filters.A) Schematic of the experimental setup.Utilizer vesicles (0.8 mL, 2.7 mg/mL total lipids, see Main text, Methods section for preparation) were added to a dialysis chamber equipped with a polycarbonate filter (50 nm pore Ø) and pre-equilibrated with 50 mM KPi, pH 7. To determine whether long-chain fatty acids equilibrate through the dialysis filter, oleic acid (0.8 µM) was added directly to the utilizer vesicles (left panel, black), or to the trans compartment (central panel, purple).To determine whether phospholipid biosynthesis is kinetically limited by the slow equilibration of the precursor produced by the feeder vesicles, glycerol 3-phosphate (0.8 µM) was added either directly to the utilizer vesicles (central panel, purple) or trans compartment (left panel, black).A positive control was prepared by adding the utilizer vesicles and the precursors of phospholipid biosynthesis to a test tube (right panel, green).A t=0 a 80 µL sample was taken (control), and the reaction was started by addition of Mg-ATP (2 mM) to the utilizer vesicles.Subsequent time points were taken at t=0.5,1,2.5 and 5 hours and analyzed as detailed in the Main text (see Methods section).B) Normalized oleic acid total ion counts (n=3 independent reconstitutions, error bars are s.e.m.).Oleic acid does not equilibrate much across the polycarbonate dialysis device over the 5 hours.The dialysis chamber can thus be used to avoid possible leakage of the feeder vesicles induced by long-chain fatty acids.C) Normalized LPA total ion counts (n=3 independent reconstitutions, error bars are s.e.m.).Diffusion of glycerol 3phosphate through the polycarbonate dialysis filter is rate-limiting for phospholipid synthesis as can be seen when the green and black lines are compared.

Figure S9 .
Figure S9.Calibration curve for DOPA quantification.A calibration curve was prepared by mixing 2.3-18.8mM DOPA with DOPG, DOPE and DOPC (total 3.5 mM, 25:25:50 mol%) from chloroform stocks.The chloroform was evaporated under gaseous nitrogen flow, the lipids resuspended in 50 µL methanol and analyzed by LC-MS as detailed in the Main text (see Methods section).The DOPA total ion counts were normalized for the internal standard DOPG.The calibration curve (n=3-4, error bars represent s.e.m.) was used for quantification of DOPA (normalized for DOPG) from the feederutilizer LUVs experiment.

Table S1 .
Overview of ArcD and ArcE homologues.The L. lactis IL 1403 ArcD2 1 served as a reference for the screening of homologues by genome database searches.

Amino acid sequence alignment and identity of ArcD and ArcE proteins.
The amino acid sequence alignments and percent identity were obtained with Clustal Omega 2 .L. lactis ArcD2 and L. sakei ArcD sequences are 45% identical.L. lactis ArcD2 and the ArcE homologues have a low sequence identity (<30%).

Table S4 . Kinetic parameters of ArcD and ArcE proteins based on in vivo L-arginine uptake.
In vivo, ArcD from L. sakei transports L-arginine approximately 10-fold faster than ArcD2 from L. lactis at 20 ºC in K-HEPES pH 7.0 (see above).Due to the uncertainty in protein expression level in whole cells, the maximal rates of transport (n=1) were intended as initial screening data and were corroborated further by in vitro characterization of the proteins.

Table S6 . Mass-to-charge ratio of lipid species used in this study.
* Formate adduct [M + CHO2-] -