Synthetic organelles for energy conservation and delivery of building blocks for lipid biosynthesis

doi:10.21203/rs.3.rs-3385355/v1

Download PDF

Article

Synthetic organelles for energy conservation and delivery of building blocks for lipid biosynthesis

https://doi.org/10.21203/rs.3.rs-3385355/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

ATP is a universal energy currency that is essential for life. ATP generation based on the degradation of L-arginine can be used for the bottom-up construction of synthetic cells. In an earlier design of this pathway, a relatively slow L-arginine/L-ornithine antiporter limited ATP production. After testing homologous proteins, we now implement an antiporter (ArcD from Lactobacillus sakei) with improved transport kinetics, leading to faster ATP recycling in organelle-like vesicles. Furthermore, we exploit the L-arginine-dependent ATP formation for the synthesis and export of glycerol-3-phosphate with complete recycling of waste products, ensuring sustainable production of building blocks. The exported glycerol-3-phosphate from the feeding vesicles serves as a precursor for the biosynthesis of phospholipids in utilizer vesicles, which provides the basis for the expansion of the cell membrane. We thus develop an out-of-equilibrium metabolic network for ATP generation and couple metabolic energy production to lipid synthesis in an organelle-like system, inspired by compartmentalization in eukaryotic cells.

Biological sciences/Biochemistry/Enzymes

Biological sciences/Physiology/Metabolism

ATP recycling

Glycerol 3-P/Pi antiporter

out-of-equilibrium metabolic network

phospholipid biosynthesis

building block delivery

synthetic cells

The bottom-up construction of living cells is one of the major challenges in synthetic biology. While an all-encompassing definition of life has not yet been formulated, it is universally agreed that maintaining an out-of-equilibrium state is a necessity for autonomous growth. Living cells fulfill this requirement by taking up nutrients and excreting waste products, and by coupling of metabolic energy to otherwise unfavorable reactions. Many of these essential processes (e.g., DNA replication, transcription/translation, lipid biosynthesis, uptake of nutrients, etc.) are fueled by the hub metabolite¹ adenosine triphosphate (ATP), which is abundantly present in all living cells. Therefore, any attempt towards the engineering of a synthetic cell should not only focus on the variety of different processes needed, but also on the quantities of metabolites and their turnover rates (k_cat). Sikkema et al.² have calculated that a bottom-up synthetic system with a volume comparable to that of the small, yet free living bacterium Pelagibacter (0.03 fL) would require at least 3.6x10⁸ equivalents of ATP to complete a full cell cycle. In accordance, Breuer et al.³ reported that JCVI-syn3a, a genome-reduced organism of comparable volume (400 nm diameter, 0.034 fL), invests approximately 2.9x10⁸ ATP equivalents per cell cycle⁴. Thus, the reconstitution of a performant, sustainable ATP recycling system is essential for the fueling of an autonomous growing synthetic cell.

ATP synthesis has been successfully achieved in complex photosynthetic systems driven by a proton motive force (PMF) in a light-dependent manner⁵. These artificial chloroplasts have been reconstituted into giant vesicles, mimicking organelles of eukaryotic cells, and used to power actin polymerization⁶ and protein synthesis⁷. In another study, a minimal system for cellular respiration and energy regeneration was obtained by co-reconstituting the mitochondrial complex I with an alternative oxidase and the F_oF₁-ATPase to synthesize ATP in a NADH-dependent fashion⁸. An alternative to ATP synthesis by the complex processes of respiration and photophosphorylation is the exploitation of simple metabolic networks^9–11. The breakdown of L-arginine to ATP requires only four proteins: an L-arginine/L-ornithine antiporter and three metabolic enzymes (Fig. 1a). The strict coupling of L-arginine uptake to the excretion of the end-product L-ornithine forms the basis for maintaining out-of-equilibrium conditions. As the formed CO₂ and NH₃ leave the cell by passive diffusion across the membrane, no side-products remain. ATP formation from L-arginine breakdown has been used to drive solute uptake and volume homeostasis in a reconstituted system¹⁰, and the synthesis of glycerol 3-phosphate, an essential precursor of phospholipids¹¹.

The simplest phospholipid is phosphatidic acid (PA), a central precursor of, e.g., phosphatidylethanolamine (PE), phosphatidylglycerol (PG), phosphatidylserine (PS), phosphatidylcholine (PC) and cardiolipin (CL). PA formation requires the sequential transfer of two acyl chains to glycerol 3-phosphate by dedicated acyltransferases, which in eukaryotic cells occurs on the cytosolic leaflet of the endoplasmic reticulum^12,13. We previously reconstituted the Escherichia coli acyltransferases PlsB and PlsC together with FadD, a soluble enzyme that catalyzes acyl-CoA formation in the presence of ATP plus coenzyme A¹⁴, which led to significant membrane expansion starting from a free fatty acid feed. Additionally, PA formation has been demonstrated from genetically encoded plsB and plsC¹⁵, a fatty acid formation module¹⁶ and from a minigenome encoding the components for headgroup functionalization¹⁷.

We now present a faster, more sustainable version of the L-arginine breakdown pathway and incorporate a glycerol 3-phosphate/Pi antiporter¹¹ to excrete the internally formed glycerol 3-phosphate, thus allowing the recycling of phosphate and maintenance of out-of-equilibrium conditions. Next, we use the energy conservation and glycerol 3-phosphate export to fuel lipid formation in a second set of vesicles. This feeder-utilizer setup operates as functional analog of eukaryotic mitochondria (ATP production) and endoplasmic reticulum (lipid synthesis).

Glycerol 3-phosphate/Pi exchange to overcome phosphate depletion

Glycerol can freely diffuse across the membrane and does not require a membrane transporter²². However, when glycerol 3-phosphate is synthesized inside the vesicles by the glycerol kinase GlpK, the L-arginine breakdown pathway gets depleted of inorganic phosphate¹¹. To overcome this limitation, we expanded the glycerol 3-phosphate synthesis module with a component for glycerol 3-phosphate export and inorganic phosphate import. Hence, we overexpressed, purified and co-reconstituted GlpT from E. coli (Fig. 1b,c).

GlpT is a secondary transporter of the major facilitator superfamily (MFS)²³ that mediates glycerol 3-phosphate/Pi exchange^24,25. GlpT operates bidirectionally and the direction of antiport depends on the sign of the total electrochemical potential. Thus, the transport is independent of the orientation of the protein in the membrane. In our setup, GlpT facilitates the export of the internally formed glycerol 3-phosphate and the import of Pi from the external solution (Fig. 1d). Noteworthy, as the external volume is more than 2-orders of magnitude larger than the cumulative vesicle lumen, we assume a constant external Pi concentration throughout the duration of the experiment. Hence, this expanded metabolic network represents a genuine out-of-equilibrium system where the phospholipid bilayer confines the recycling of ATP and glycerol 3-phosphate formation, while allowing for all substrates to be externally sourced and all products to either be internally recycled, or leave the system.

Selection and activity screening of homologues for faster L-arginine/L-ornithine exchange

The L-arginine/L-ornithine antiporter ArcD2 from Lactococcus lactis has significantly lower k_cat values (0.11 s^− 1) than the water-soluble enzymes (ArcA = 3.3 s^− 1; ArcB, backward reaction = 410 s^− 1; ArcC1 = 200 s^− 1), suggesting that the transport step limits L-arginine breakdown in confinement⁹. To overcome this kinetic bottleneck, we took a functional genomics approach to identify candidate homologues that can replace L. lactis ArcD2. We selected arcD genes with an amino acid identity of 30–60% with respect to the original antiporter (ArcD2 from L. lactis), which allows for significant variation, while the L-arginine/L-ornithine antiport function is likely retained (Figure S1). We also included arcE sequences, encoding putative L-arginine/L-ornithine antiporters that evolved independently from those of the ArcD family^26,27. We selected 10 putative ArcD- and ArcE-like proteins from the genomics approach, further guided by published biochemical data (Table S1, Figure S2, Figure S3).

The arcD and arcE genes were cloned into nisin A-based expression vectors and flanked by a 10x His-tag at either the N- or the C-terminus. The constructs were transformed into L. lactis JP9000 ΔarcD1ΔarcD2²⁸ and the expression was optimized. Immunoblots showed the highest expression levels with ArcD from Clostridium autoethanogenum, Lactobacillus sakei and Pseudomonas aeruginosa, and ArcE from Streptococcus pneumoniae. All these proteins were tested for L-arginine transport activity in vivo (Figure S4). We found that ArcD (with N-terminal His-tag) from L. sakei had the highest rate of uptake. By estimating a similar level of expression as the orginal L. lactis ArcD2, the ArcD from L. sakei would have a 10-fold higher turnover number and was therefore chosen for further in vitro characterization (Table S4).

Purification and characterization of L. sakei ArcD

ArcD from L. sakei was purified and reconstituted in liposomes composed of DOPE:DOPG:DOPC at 25:25:50 mol%. Both transporters had similar reconstitution efficiencies (~ 60%), as estimated by SDS-PAA gel electrophoresis (Fig. 2a-c; Figure S5). The transport activity was measured via uptake of radiolabeled L-arginine (25 µM) in exchange for internal L-ornithine. Accounting for the protein amount, an initial transport rate of 3.5 ± 0.2 µmol.mg protein^− 1.min^− 1 (3.0 ± 0.2 s^− 1) was determined for L. sakei ArcD, a value ~ 5 fold higher than what was found for L. lactis ArcD2 (0.6 ± 0.1 µmol.mg protein^− 1.min^− 1, 0.5 ± 0.1 s^− 1; Fig. 2d,e). By varying the external concentration of L-arginine, the affinity constant (K_M) and turnover number (k_cat) were determined for the L. sakei antiporter (Fig. 2f; Table S5). In agreement with the initial transport rates, the turnover number of L. sakei ArcD was much higher than that of L. lactis ArcD2, while the K_M values were similar⁹.

ATP formation

If L-arginine/L-ornithine exchange would be fully rate-limiting for ATP production, an increase in the rate of transport should lead to a corresponding increase in the rate of ATP synthesis. To test this hypothesis, we equipped vesicles composed of DOPE:DOPG:DOPC (25:25:50 mol%), containing either L. lactis ArcD2 or L. sakei ArcD, with the components of the L-arginine breakdown pathway¹¹ (Fig. 3a) and we quantified the ATP formation rate by chemiluminescence. From the linear increment during the first 5 minutes of L-arginine metabolism, we find that the ATP synthesis is ~ 2.5 fold faster with ArcD from L. sakei than with ArcD2 from L. lactis (Fig. 3b), i.e., under conditions that the difference in the initial L-arginine uptake rate is ~ 5 fold (Fig. 2d). Similar results were obtained when ATP formation was followed with the fluorescent ATP/ADP sensor PercevalHR (Figure S6). We thus conclude that the L-arginine/L-ornithine antiport is not the sole kinetically limiting factor and additional bottlenecks must be present in the pathway. A likely candidate is the reaction catalyzed by the carbamoyl transferase ArcB, which at pH 7.0 has an equilibrium constant of 1.2x10⁵ in the direction of L-citrulline formation⁹ (Fig. 1a). Hence, high concentrations of L-citrulline and inorganic phosphate, and low levels of products are required for a net reaction in the desired direction, i.e., ornithine plus carbamoyl-phosphate formation.

Co-reconstitution of ArcD and GlpT in LUVs

Next, we over-expressed and purified E. coli GlpT³⁰ by affinity and size-exclusion chromatography (Fig. 4a), and co-reconstituted the protein with either L. lactis ArcD2 or L. sakei ArcD in vesicles composed of DOPE:DOPG:DOPC at 25:25:50 mol%. We expected that the rates of glycerol 3-phosphate formation and export would be faster than that of L-arginine breakdown, given the kinetic parameters of the enzymes and transporters. In fact, the synthesis of glycerol 3-phosphate by GlpK depletes the internal pre-formed ATP pool generated by L-arginine breakdown in less than one minute¹¹. In addition, the k_cat of E. coli GlpT is higher than that of the L-arginine/L-ornithine antiporters (Table S5). Thus, we used a 1:1 mol ratio of GlpT:ArcD, at a lipid-to-protein ratio of 400:1 w/w for each protein (Fig. 4b,c). Approximately 60% of GlpT was inserted into the vesicles. However, the reconstitution efficiency of both ArcD variants dropped to ~ 30% in the presence of GlpT (Fig. 4d).

Kinetics and robustness of ATP-driven glycerol 3-phosphate synthesis and export

Glycerol 3-phosphate formation coupled to glycerol 3-phosphate/Pi exchange represents an ideal load to study the performance of the L-arginine breakdown pathway in terms of kinetics (e.g., ATP recycling rates with different ArcD homologues, stability over time) and robustness (e.g., ATP equivalents formed). We encapsulated the L-arginine breakdown pathway in vesicles reconstituted with GlpT plus ArcD2 from L. lactis or ArcD from L. sakei, together with soluble GlpK and the ATP/ADP sensor PercevalHR (Fig. 4e), and we determined the ATP/ADP ratio over time (Fig. 4f,g). We allowed an initial cycle of ATP formation by addition of 5 mM L-arginine. When the ATP/ADP plateau was reached, we initiated glycerol 3-phosphate formation and export by adding glycerol in concentrations ranging from 0.1 to 0.7 mM. At this time, an additional 5 mM L-arginine was added (and hereafter at 3–4 hours intervals) to ensure that ATP recycling would not be limited by L-arginine depletion in the external medium¹⁰. Glycerol rapidly enters the vesicles by passive diffusion²², and glycerol 3-phosphate formation and export are faster than ATP recycling. In agreement, we observed an instantaneous drop in the ATP/ADP ratio upon glycerol addition. The ATP/ADP signal recovers when all glycerol has been converted into glycerol 3-phosphate. The time till recovery is a linear function of the glycerol concentration, while the rate of ATP/ADP recovery upon glycerol depletion is similar in all cases (Fig. 4f,g).

Strikingly, the co-reconstitution of the glycerol 3-phosphate/Pi antiporter allowed ATP/ADP recovery at glycerol concentrations well beyond what is possible on the basis of the levels of inorganic phosphate in the vesicle lumen, demonstrating that the internal Pi pool is replenished efficiently. We calculated the amounts of ATP produced from the amount of glycerol metabolized and the effective volumes of the vesicles and outside medium, which corresponds to 14 rounds of ATP recycling inside the vesicles. These results showcase the robustness of L-arginine breakdown when coupled to a reaction load. The system performs multiple cycles of ATP regeneration that are only limited by the availability of external substrates (here L-arginine, glycerol and Pi) and the elimination of end products (L-ornithine, glycerol 3-phosphate, NH₃ and CO₂), illustrating the sustainable power of out-of-equilibrium metabolic networks.

We then asked how many molecules of ATP are generated on average per vesicle in the time frame of the experiment. A glycerol concentration of 0.7 mM in a 120 µL reaction volume corresponds to 84 nmol, or a total of 5.1x10¹⁶ glycerol equivalents. Hence, the same number of ATP molecules must have been generated by L-arginine breakdown. The total internal vesicle volume in the 120 µL reaction volume is 0.875 µL¹¹ and corresponds to approximately 1.8x10¹⁰ vesicles with an average radius of 226 nm and an average volume of 0.05 fL¹⁰. Thus, on average 2.8x10⁶ ATP equivalents are converted per vesicle over a time period of 3 (L. sakei ArcD) to 5 (L. lactis ArcD2) hours. This value matches the energy needs for DNA replication (10⁶ ATP_eq), transcription (10⁶ ATP_eq) and division (10⁴ ATP_eq), while transport (10⁷ ATP_eq), translation (10⁸ ATP_eq), lipid metabolism (10⁷ ATP_eq) and maintenance (10⁸ ATP_eq) are more costly².

To compare the L-arginine breakdown rate between the two different ArcD variants, we plotted the half recovery time (i.e., the time when half of the ADP pool had been converted back into ATP) of each ATP/ADP trace as a function of the glycerol concentration (Fig. 4h). We observed that the ATP/ADP signal recovers 2- to 3-fold faster in vesicles with ArcD from L. sakei than ArcD2 from L. lactis, which is in line with the 2.5-fold increase in ATP synthesis rate. Taken the data and theoretical considerations collectively, we estimate that a synthetic cell of 400 nm containing the here presented, improved L-arginine breakdown pathway would produce ATP at rate that supports a cell cycle with a doubling time of ~ 10 days.

Design of a feeder-utilizer organelle-like vesicle system

Next, we designed a modular reaction network for the coupling of ATP recycling and building block formation to the synthesis of phospholipids. We used the E. coli acyltransferases PlsB and PlsC in combination with FadD, a soluble enzyme that catalyzes acyl-CoA formation from free fatty acids in the presence of ATP and CoA (Fig. 5a,b)¹⁴. We achieve synthesis of the lipid DOPA by feeding glycerol 3-phosphate from feeder to utilizer organelle-like vesicles with a topology akin lipid biosynthesis in the endoplasmic reticulum of eukaryotic cells.

We spatially segregated L-arginine-driven glycerol 3-phosphate formation and export (the “feeder” module) from DOPA biosynthesis (the “utilizer” module) by separating the two types of vesicles by a polycarbonate dialysis filter with a pore size of ~ 50 nm (Fig. 5c,d). We kept the long-chain free fatty acids away from the feeder module and avoided potential solubilizing effects of the lipid precursors. We exploited our previously developed multi-chamber dynamic dialysis setup¹¹ but did not use the continuous flow needed for dynamic studies. Importantly, small water-soluble metabolites, such as glycerol 3-phosphate, diffuse along their concentration gradient from the feeder to the utilizer module¹¹. Long-chain fatty acids do not cross the dialysis membrane due to their self-assembly into large supramolecular structures in a pH-dependent manner (i.e., vesicles at 7 < pH < 8)³¹.

Coupling of modules for lipid biosynthesis

We filled one compartment of the dialysis chamber with feeder vesicles containing both the L-arginine breakdown pathway and glycerol 3-phosphate synthesis module. We added 5 mM of fructose 1,6-bisphosphate to inhibit external GlpK, in case some enzyme would leak from the vesicles (FBP³²; Fig. 5d, upper panel). ATP from the L-arginine breakdown pathway is used to synthesize glycerol 3-phosphate in the presence of 0.4 mM glycerol. Glycerol 3-phosphate is subsequently exported out of the feeder vesicles and diffuses to the compartment with the utilizer vesicles. This compartment was filled with utilizer vesicles decorated with FadD, PlsB and PlsC, and supplemented with 400 µM oleic acid, 2 mM Mg-ATP plus 0.1 mM CoA (Fig. 5d, lower panel).

Vesicles with and without GlpT synthesized glycerol 3-phosphate, as shown by the ATP/ADP ratio recorded in parallel experiments (Figure S7). Full ATP/ADP recovery after 1.5 hours was observed for vesicles carrying the glycerol 3-phosphate/Pi antiporter, indicating that all glycerol had been converted into glycerol 3-phosphate (Fig. 5e). In agreement, the control vesicles without GlpT failed to recover their ATP/ADP levels because the internal Pi was depleted. The minor ATP/ADP increase observed upon substrate addition to GlpT-deficient vesicles suggests that a sub-population of vesicles lacks GlpK. This is in line with previous observations that a fraction of the vesicles lacks one or more biomolecules due to the stochastic distribution of the components in the encapsulation process¹¹.

Phospholipid biosynthesis was measured by LC-MS analysis of samples collected over time from the compartment with the utilizer vesicles. Little or no DOPA synthesis was observed in the first 2 hours, which is the timeframe required for the diffusion of glycerol 3-phosphate from the feeder (synthesis and export) to the utilizer compartment (Fig. 5f). DOPA was then formed but only when the feeder vesicles were equipped with the glycerol 3-phosphate exporter GlpT; the control without GlpT did not show any DOPA formation. As two fatty acid molecules are needed to form one phospholipid, a maximum yield of 200 µM DOPA would be possible if all the molecules would be available for synthesis. However, at pH 7 most of the fatty acids is present in the form of vesicles from where it is slowly released. After 6 hours, 5.7 µM of DOPA was produced. The diffusion of glycerol 3-phosphate from one compartment to the other is limiting the feeding of building blocks between different vesicles.

A system for continuous, robust energy formation is essential for an autonomous synthetic cell⁴. Here, we present an encapsulated L-arginine-dependent ATP production pathway, which utilizes a novel L-arginine/L-ornithine antiporter with enhanced transport capacity. We equipped the vesicles with a module for glycerol 3-phosphate synthesis and coupled its export to import of inorganic phosphate, thereby avoiding internal depletion of free phosphate. This yielded a fully sustainable synthetic metabolic network where all nutrients are imported, cofactors recycled and end products exported, which is now only limited by the availability of external substrates. Hence, a true out-of-equilibrium, or selectively open, metabolic network, akin how living cells work, but at the same time unique due to its bottom-up construction from components of different (biological) origin.

In view of the global energetic needs of an autonomous synthetic cell, we provide a quantitative assessment of the performance of a simple ATP-generating pathway. Our ATP-producing module would allow a cell to grow with a doubling time of 10 days, which is much slower than that of model organisms grown under optimal laboratory conditions. On the other hand, in nature many microorganisms grow with comparable doubling times, often because they are limited by nutrient availability. In addition, we foresee that the reconstitution of complex processes (e.g., DNA duplication and segregation, protein synthesis, cell division, etc.) in autonomous synthetic cells may initially not require much faster synthesis of ATP. Nevertheless, in the future we aim to optimize ATP synthesis rates by subjecting the L-arginine/L-ornithine antiporter or the entire pathway to evolutionary screening and selection.

To showcase the potential of continuous and sustainable ATP recycling, we coupled L-arginine breakdown-driven glycerol 3-phosphate synthesis to another essential metabolic process, i.e., lipid synthesis. Low phospholipid yields have been previously reported for in vitro transcription-translation based systems reconstituted in giant vesicles, due to the limited solubility of the acyl-chain precursors and impaired ribosomal processivity^17,33,34. In our system, phospholipid formation is limited by the equilibration of glycerol 3-phosphate through the dialysis filter; higher yields are expected in bulk¹⁴, provided that the utilizer vesicles can be stabilized against leakage (e.g., with a peptide or PEG coating). The feeder/utilizer ratio can also be adjusted to boost lipid formation.

The bottom-up construction of complex metabolic systems inspired by eukaryotic cells is gaining momentum¹⁸. For instance, a multi-step enzymatic cascade has been engineered in vesicles, where each reaction step is spatially segregated into a sub-compartment and metabolic intermediates diffuse through membrane pores¹⁹. Living cells have also been exploited as organelle components of metabolic synthetic cells²⁰. Artificial β-cells have been engineered that sense high glucose levels in vitro and in vivo, and they respond by releasing insulin from internal storage vesicles, effectively restoring physiological glucose concentrations²¹.

The feeder-utilizer vesicle system developed in this work is reminiscent of eukaryotic organelles that produce building blocks for other compartments of the cell, namely the mitochondrion for ATP synthesis and the endoplasmic reticulum for lipid formation. Such coupled reaction networks may be exploited as functional containers for the construction of autonomously growing synthetic cells of larger diameter (ideally, bacteria-sized giant unilamellar vesicles⁴). We argue that multi-compartment synthetic cells will have a more favorable surface-to-volume ratio (a parameter important for metabolite exchange and other membrane-related processes³⁵) compared to cells that lack organelles. In addition, the spatial confinement of metabolic modules into different sub-compartments will provide higher control over the construction of complex reaction networks, as each organelle is relatively simple (e.g., in terms of number of components). This will minimize stochastic effects and facilitate modeling the processes⁴. Finally, the physical separation of reaction networks into synthetic organelles will allow the co-existence of mutually exclusive processes that would otherwise be impaired in a single compartment.

In conclusion, we have coupled two essential metabolic processes, i.e., energy conservation and lipid formation, in a two-vesicle system that resembles in function the mitochondria and endoplasmic reticulum of eukaryotes. In the future, this metabolic network can be expanded by directly exporting ATP from the feeder vesicles and by customizing the utilizer vesicles with different enzymes to achieve a bilayer composition akin that of living cells.

Chemicals and media

All chemicals and media were as reported previously¹¹, with the following additions: 1,4-Dithiothreitol (Carl Roth); Ampicillin sodium salt (Formedium); ¹⁴C L-arginine (stock solution 325 mCi/mmol; NovaTec); 2-Mercaptoethanol (Sigma Aldrich); Chloramphenicol (Carl Roth); Coenzyme A, Free Acid (Avanti Polar Lipids); D-(+)-Fructose 1,6-bisphosphate trisodium salt hydrate (Sigma Aldrich); D(+)-Glucose Anhydrous (Formedium); D(+)-Sucrose (Formedium); Ethylenediaminetetraacetic acid disodium salt dihydrate (Sigma Aldrich); HEPES (Carl Roth); Methanol (Sigma); n-Butanol (Sigma); n-Dodecyl-β-D-Maltopyranoside, Anagrade (Anatrace); Oleic acid (Sigma Aldrich); Tris Hydrochloride (ITW Reagents); TritonTM X-100 (Sigma Aldrich).

Strains and plasmids

Genomic DNA from C. autoethanogenum 9 (DSM 10061), P. aeruginosa PAO1 (DSM 22644) and R. denitrificans ATCC 33942 (DSM 7001) were purchased from DSMZ. Genomes from L. fermentum IMDO 130101 and L. sakei ATCC 15521 were kindly donated by Professor Frederic Leroy and Professor Jan Kok, respectively. A codon-optimized synthetic gene was ordered for ArcD from R. fredii HH103 (GenScript). Template vectors encoding L. brevis ATCC 367 arcD and S. pneumoniae D39 arcE were generously provided by Dr. Juke S. Lolkema. Expression vectors encoding arcA, arcB, arcC1 and arcD2¹⁰, percevalHR and glpK¹¹, as well as fadD, plsB and plsC¹⁴ were as previously reported; a plasmid carrying glpT³⁰ was a kind gift of Professor Da-Neng Wang (Table S2).

Cloning of genes for ArcD and ArcE homologues

The genes encoding arcD/E were PCR-amplified from the source DNA templates with Phusion High Fidelity DNA polymerase (Thermo Fisher Scientific) and customized primers (Eurofins Genomics, Table S3). Next, fragment exchange (FX)³⁶ was exploited to clone the genes of interest into pNZ-N(His) or pNZ-C(His) vectors suitable for over-expression in L. lactis strains. According to the procedure, the PCR-amplification products were first cloned into a sequencing vector (pINITIAL). To this end, both the inserts and the backbone (1:5 molar ratio) were digested with 2.5 U of SapI (New England Biolabs) for 1 hour at 37 ºC, followed by a heat-inactivation step at 60 ºC for 20 minutes. Ligation was carried out by adding 5 U of T4 ligase (New England Biolabs) and incubating for 1 hour at RT. The plasmids were transformed by heat-shock into chemically competent E. coli MC1061 cells that were then plated onto lysogeny broth (LB) supplemented with 1% w/w agar and 5 µg/mL chloramphenicol. After overnight incubation at 37 ºC, single colonies were used to prepare 2 mL cultures that were grown overnight at 37 ºC, 200 rpm and used for plasmid isolation, followed by DNA sequencing (Eurofins Genomics, Figure S1). The genes of interest were then transferred from pINITIALarcD/E into the intermediate pREX (1:4 molar ratio) by following the same procedure described above, with the difference that LB-agar plates containing 7% w/w sucrose and 100 µg/mL ampicillin were used for selection. Finally, arcD/E were cloned into the expression plasmids pNZ-N(His) or pNZ-C(His) by vector-backbone exchange (VBEx)³⁶. Briefly, pREXarcD/E and the acceptor vector pERL were mixed (1:3 molar ratio) and digested by incubating with 3 U of SfiI (New England Biolabs) for 2 hours at 50 ºC, followed by enzyme inactivation and ligation for 2 hours at RT. The obtained constructs were transformed by electroporation into electrocompetent L. lactis JP9000 ΔarcD1ΔarcD2 cells²⁸ that were then plated onto M17 medium supplemented with 1% agar, 1% w/w glucose and 5 µg/mL chloramphenicol (GM17Cm-agar) and incubated o/n at 30 ºC.

Over-expression and purification of ArcD homologues

Large-scale cultures were prepared for L. lactis JP9000 ΔarcD1ΔarcD2 carrying pNZarcD2ΔC or pNZarcD-LS in a 10 L bioreactor, followed by cell lysis and membrane vesicle isolation according to a standardized procedure¹⁰. Next, L. lactis ArcD2 and L. sakei ArcD were purified by affinity chromatography as described previously^9–11, with the important difference that 2 mM β-mercaptoethanol plus 10% v/v glycerol were added during membrane vesicle solubilization, and maintained in the buffer (50 mM KPi, 200 mM KCl, 0.5 % w/vDDM, 2 mM β-mercaptoethanol and 10% v/v glycerol) throughout the purification procedure. Elution was performed without glycerol for immediate reconstitution in LUVs (see below), while 10% v/v glycerol was added for storage of the proteins at -80 ºC.

Over-expression and purification of GlpT

Over-expression. Chemically competent E. coli MC1061 cells were transformed with pBADmycHisB-glpT by heat-shock, plated onto LB-agar plates supplemented with 100 µg/mL ampicillin and incubated overnight at 37 ºC. A pre-culture was prepared by inoculating a fresh colony in 175 mL LB-amp, followed by incubation overnight at 37 ºC, 200 rpm. The o/n pre-cultures were diluted 1:40 in 6x 1 L fresh LB-amp, and grown at 37 ºC until an OD₆₀₀ of 0.5 was reached. The temperature was lowered to 25 ºC, and over-expression was induced with 0.01% L-arabinose at an OD₆₀₀ of 0.94. After 2 hours of expression, cells were harvested by centrifugation (15 min, 6000 g, 4°C), washed with 50 mM Tris-HCl pH 8.0 and resuspended to a final concentration of 220 g/L with the same buffer supplemented with 20% w/w glycerol. Cells were flash-frozen with liquid N₂ and stored at -80 ºC for later use.

Membrane vesicles preparation. Membrane vesicles were prepared by thawing 11 g of cells and diluting them to 110 g/L in 100 mL buffer of final composition 50 mM Tris-HCl pH 8.0 plus 400 mM NaCl, to which were added 1 mM PMSF, 1 mM MgCl₂ and 100 µg/mL DNase. Cells were disrupted by a single passage through an HPL6 press (Maximator GmbH) at 20 KPsi, 4°C and supplemented with 5 mM Na-EDTA pH 8.0. Membrane vesicles were purified from the cellular debris by high-speed centrifugation (15 minutes, 15.000 g, 4 ºC) and pelleted by ultracentrifugation (45 minutes, 175.000 g, 4 ºC). The pellets were washed with 25 mL 50 mM Tris-HCl pH 8.0 and the ultracentrifugation step was repeated. Membrane vesicles were resuspended to a final concentration of 21 mg/mL in 10 mL 50 mM Tris-HCl pH 8.0 plus 20% w/w glycerol, flash-frozen with liquid N₂ and stored at -80 ºC for later use.

Purification. A membrane vesicle aliquot containing 10 mg of total membrane proteins was solubilized with 0.5% w/w DDM in solubilization buffer (50 mM Tris-HCl pH 8.0, 20% w/w glycerol, 10 mM imidazole) in a 30 min nutation step at 4 ºC. The debris was removed by ultracentrifugation (12 minutes, 328.000 g, 4 ºC), and the solubilized protein was loaded onto 0.25 mL (1 CV) Ni-Sepharose pre-washed with an excess of milliQ water. Binding was performed for 1 hour at 4 ºC, with nutation. The flow-through was removed, and the protein was washed (20 CVs 50 mM Tris-HCl pH 8.0, 20% w/w glycerol, 0.04% w/w DDM, 50 mM imidazole) and eluted by increasing the imidazole concentration in the washing buffer to 500 mM. The peak fractions from the Ni-Sepharose were pooled and then loaded onto a Superdex 200 (GE Healthcare) column, equilibrated in 50 mM HEPES pH 7.0 plus 150 mM NaCl and 20% w/w glycerol. The elution fractions were concentrated to 1.2 mg/mL with an Amicon Ultra-0.5 Centrifugal Filter Unit (Merck) with a cutoff of 100 kDa, flash-frozen with liquid N₂ and conserved at -80 ºC.

Over-expression and purification of other proteins

ArcA, ArcB, ArcC1, PercevalHR and GlpK, and FadD, PlsB and PlsC were over-expressed and purified as previously reported ^9–11,14.

Formation of proteoliposomes with one or multiple antiporters

Purified L. lactis ArcD2 and L. sakei ArcD were reconstituted separately in liposomes composed of DOPE:DOPG:DOPC at 25:25:50 mol% with a lipid-to-protein ratio of 400:1 w/w. The Triton X-100-mediated vesicle destabilization protocol^9–11 was adapted by adding 2 mM DTT to the reconstitution buffer. To equip vesicles with E. coli GlpT, the antiporter was co-reconstituted together with the L-arginine/L-ornithine antiporter at a lipid-to-protein ratio of 400:1 w/w. The overall lipid-to-protein ratio was 200:1 w/w.

In vitro L-arginine uptake

L-arginine/L-ornithine exchange activity was determined via uptake of radiolabeled L-arginine in vesicles containing L. lactis ArcD2 or L. sakei ArcD at a lipid-to-protein ratio of 400:1 w/w. L-ornithine was encapsulated at a concentration of 10 mM (10 mg, 400 µL) resuspended in 50 mM KPi pH 7.0 plus 2 mM DTT by 5 freeze-thaw cycles. The thawing step was performed in an ice-water bath at 10°C. Proteo-LUVs were extruded 13x through a 400 nm polycarbonate filter pre-equilibrated with 10 mM L-ornithine, 2 mM DTT, 50 mM KPi pH 7.0, diluted to 6 mL with 50 mM KPi pH 7.0, 2 mM DTT without L-ornithine and collected by centrifugation (20 min, 325,000 g, 4°C). Remaining external L-ornithine was diluted further by an additional resuspension of the pellet in 6 mL of 50 mM KPi pH 7.0 plus 2 mM DTT and centrifugation (20 min, 325,000 g, 4°C). Collected proteo-LUVs containing L-ornithine were finally resuspended to a lipid concentration of 200 mg/mL in 50 mM KPi pH 7.0 plus 2 mM DTT.

For the uptake assays, encapsulated proteo-LUVs were diluted 50 times in pre-heated (30°C) reaction buffer containing 25 µM ¹⁴C-L-arginine, 2 mM DTT, 50 mM KPi pH 7.0. A stock of ¹⁴C-L-arginine with a specific activity of 325 mCi/mmol was diluted 125 times with unlabeled L-arginine. A volume of 100 µL was taken from the reaction mixture at 0, 5, 10, 25, 40 and 60 sec, diluted into 2 mL of ice-cold 50 mM KPi pH 7.0 and filtered through a 0.45 µm cellulose nitrate filter. The filter was further washed with 2 mL of ice-cold 50 mM KPi pH 7.0 and dissolved in Ultima Gold MV scintillation liquid (Perkin Elmer). Radioactivity was quantified in a Tri-Carb 2800TR scintillation counter (PerkinElmer). For determination of kinetic parameters of L. sakei ArcD, external L-arginine in the reaction solution was varied between 1–100 µM. Initial rates of L-arginine uptake were plotted versus the external L-arginine concentration and fitted to a Michaelis-Menten equation to obtain the K_M^arg and V_max values.

Encapsulation of soluble components for L-arginine breakdown and glycerol 3-phosphate formation

A previously described procedure¹¹ was followed for the encapsulation in vesicles of the L-arginine breakdown enzymes, glycerol kinase, fluorescent sensors and small molecules, with minor adaptations. Briefly, 10 mM Na-ADP, 10 mM MgCl₂, 0.5 mM L-ornithine, 37.5 µg ArcA, 96 µg ArcB, 72 µg ArcC1, 71 µg PercevalHR plus 35.8 µg GlpK were mixed in 50 mM KPi pH 7.0 plus 2 mM DTT to a final volume of 200 µL. This mixture was added to preformed proteoliposomes bearing either L. lactis ArcD2 or L. sakei ArcD, and E. coli GlpT; control vesicles were prepared without GlpT. For experiments that did not require glycerol 3-phosphate formation, GlpK was omitted from the mixture and ArcD-only proteoliposomes were used. Five freeze-thaw cycles were performed, followed by a 13x extrusion step through a 400 nm polycarbonate filter (Whatman, GE Healthcare). Extruded vesicles were thoroughly washed with 3x 6 mL external buffer (50 mM KPi pH 7.0, 58 mM NaCl, 2 mM DTT) by centrifugation (30 min, 325,000 g, 4°C). The vesicles were resuspended to a stock concentration of 5.55 mg of total lipids/mL and used immediately or stored at 4 ºC up to 48 hours.

ATP quantification by chemiluminescence

Chemiluminescence experiments were carried out similarly as previously reported¹¹. Briefly, vesicles equipped with either ArcD2 from L. lactis or ArcD from L. sakei, plus the soluble components required for L-arginine breakdown were diluted to a concentration of 2.7 mg total lipids/mL with external buffer (50 mM KPi pH 7.0, 58 mM NaCl, 2 mM DTT) and mixed with 1 µM valinomycin plus 1 µM nigericin. A 100 µL sample was taken at t = 0, followed by incubation at 30 ºC and addition of 5 mM L-arginine to initiate ATP formation. Samples were collected over a time course of 2 hours, quenched by perchloric acid treatment and analyzed as described before¹¹.

ATP/ADP determination by fluorescence

ATP/ADP measurements were carried out in parallel with ATP quantification as previously reported¹¹. Briefly, a volume of 120 µL containing 2.7 mg total lipids/mL vesicles (in 50 mM KPi pH 7.0, 58 mM NaCl, 2 mM DTT) equipped with the L-arginine breakdown pathway were mixed with ionophores (1 µM valinomycin plus 1 µM nigericin) in Ultra-Micro-cuvettes 105.252-QS (Hellma Analytics) and incubated at 30 ºC in a FP-8300 or FP-8350 fluorimeter (Jasco). ATP formation was initiated by addition of 5 mM L-arginine, and PercevalHR excitation spectra were acquired over time (excitation = 400–520 nm, bandwidth = 5 nm; emission = 550 nm, bandwidth = 5 nm). The ATP/ADP recovery experiments were carried out analogously, but with vesicles that carried GlpT and GlpK in addition to the components for the L-arginine breakdown pathway. The reaction was started by addition of 5 mM L-arginine, and incubated until a plateau was reached (~ 60 min for L. lactis ArcD2 and ~ 30 min for L. sakei ArcD). Next, glycerol was added at different concentrations (0-700 µM), together with 5 mM fresh L-arginine (5 mM). Additions of 5 mM L-arginine were made at 3–4 hours intervals. PercevalHR excitation spectra were recorded as described.

DOPA biosynthesis in feeder-utilizer vesicles

Feeder proteo-LUVs. Vesicles equipped with L-arginine breakdown (ArcD L. sakei) and glycerol 3-phosphate synthesis plus export were used at a final concentration of 2.7 mg total lipids/mL in 50 mM KPi pH 7.0, 58 mM NaCl plus 2 mM DTT. A 0.8 mL reaction mixture was prepared by adding 1 µM valinomycin, 1 µM nigericin and 5 mM fructose 1,6-bisphosphate (FBP). Control samples were prepared in the same way but with vesicles without GlpT. The vesicles were kept at 4 ºC until the reaction was started.

Utilizer vesicles. Empty phospholipid vesicles composed of DOPE:DOPG:DOPC at 25:25:50 mol% were prepared with 50 mM KPi, 58 mM NaCl plus 2 mM DTT as internal and external buffer. The empty vesicles were used at a final concentration of 2.7 mg total lipids/mL (3.5 mM) in a 0.8 mL reaction volume. To the samples were added 0.4 mM oleic acid, 0.1 mM CoA, 6.2 µg (0.2 µM) FadD, 9.2 µg (0.2 µM) PlsB and 2.8 µg (0.2 µM) PlsC. The vesicles were kept at 4 ºC until the reaction was started.

Dialysis setup and procedure. Four chambers of a previously described dynamic dialysis setup were equipped with a 50 nm polycarbonate filter (Avestin) according to the instructions¹¹. Next, the flow inlet and outlet tubes of each chamber were tightly sealed to create a two-compartment dialysis setup without flow. The four dialysis chambers were connected to a water flow set to 30 ºC, filled with buffer (50 mM KPi pH 7.0, 58 mM NaCl, 2 mM DTT) and allowed to pre-equilibrate overnight with stirring. The following day, the pre-equilibration buffer was removed, and the feeder and utilizer vesicles were transferred to their respective compartments in the dialysis chambers, each with gentle stirring. An 80 µL timepoint was collected for LC-MS analysis from the utilizer vesicle compartment prior to the addition of substrates. Then, ATP recycling and glycerol 3-phosphate formation and export were initiated by addition of 20 mM L-arginine plus 0.4 mM glycerol to the feeder vesicles, while DOPA biosynthesis was started by addition of 2 mM Mg-ATP to the utilizer vesicles. Additional samples were taken over a time course of 8 hours; all samples were immediately quenched by adding 300 µL n-butanol and 5 mM Na-EDTA and vortexing thoroughly. The activity of the feeder vesicles in the presence of 5 mM FBP and upon concomitant addition of 20 mM L-arginine plus 0.4 mM glycerol was followed in parallel by determining the ATP/ADP levels by fluorescence.

LC-MS analysis of phospholipids

Lipid extraction. The n-butanol-quenched reaction samples (total volume 380 µL) were centrifuged at 10.000 rpm, RT for 2 minutes. The lipid-containing organic phase (top layer) was collected and transferred to clean 1.5 mL screw neck vials (VWR), and the extraction process was repeated with 300 µL fresh n-butanol, followed by n-butanol evaporation under gaseous nitrogen flow. The lipid film was then resuspended in 50 µL methanol, transferred into a 0.1 mL micro insert (VWR) and analyzed by LC-MS.

LC-MS analysis. Extracted lipids were analyzed using an established protocol³⁷ using an Accela1250 high-performance liquid chromatography (UHPLC) system coupled to a heated electrospray ionization–mass spectrometry (HESI-MS) Orbitrap Exactive (Thermo Fisher Scientific). A sample of 5 µL was injected into an ACQUITY UPLC CSH C18 1.7 µm Column, 2.1x150 mm (Waters Chromatography Ireland Ltd) operating at 55°C with a flow rate of 300 µL/min. Separation of the compounds was achieved by a changing gradient of eluent A (5 mM ammonium formate in water/acetonitrile 40:60, v/v) and eluent B (5 mM ammonium formate in acetonitrile/1-butanol, 10:90, v/v). The following linear gradient was applied: 45% eluent B for 2.5 min; a gradient from 45–90% eluent B over 19.5 min; holding for 3 min; returning to 45% eluent B in 0.5 min; and holding for 8 min. The column effluent was injected directly into the Exactive ESI-MS Orbitrap operating in negative ion mode. Voltage parameters of 3 kV (spray), − 75 V (capillary), − 190 V (tube lens), and − 46 V (skimmer voltage) were used. Capillary temperature of 300°C, sheath gas flow of 60, and auxiliary gas flow of 5 were maintained during the analysis. Spectral data constituting total ion counts were analyzed using the Thermo Scientific XCalibur processing software (Table S6, Figure S8).

ADP Adenosine diphosphate

ATP Adenosine Triphosphate

CL Cardiolipin

DDM n-Dodecyl-beta-Maltoside

DOPA Dioleoyl Phosphatidic Acid

DOPE Dioleoyl Phosphatidyl Ethanolamine

DOPG Dioleoyl Phosphatidyl Glycerol

DTT Dithiothreitol

EDTA Ethylenediaminetetraacetic Acid

FBP Fructose 1,6-bisphosphate

LC-MS Liquid chromatography–mass spectrometry

LUV Large Unilamellar Vesicle

MFS Major Facilitator Superfamily

PA Phosphatidic acid

PCA Perchloric Acid

PE Phosphatidyl Ethanolamine

PG Phosphatidyl Glycerol

PMF Proton Motive Force

DATA AVAILABILITY STATEMENT

All data are available in the Main Manuscript and Supporting Information; raw data are available upon request.

ASSOCIATED CONTENT

Supporting information. Amino acid sequence alignment and identity of ArcD and ArcE proteins; Predicted topology of ArcD and ArcE proteins; Nucleotide sequence of arcD and arcE genes; In vivo L-arginine uptake by ArcD and ArcE transport proteins; SDS-PAA gel images of purified membrane proteins used in this study; L-arginine-mediated ATP formation in the presence of L. sakei ArcD or L. lactis ArcD2; Activity of the feeder LUVs measured by online ATP/ADP readout with PercevalHR; Calibration curve for DOPA quantification; Overview of candidate ArcD/E homologues; Plasmids used in this study; Primers used in this study; Kinetic parameters of ArcD and ArcE proteins based on in vivo L-arginine uptake; Kinetic parameters of L. sakei ArcD and E. coli GlpT; Mass-to-charge ratio of lipid species used in this study.

CORRESPONDING AUTHOR

Bert Poolman

ORCID ID: 0000-0002-1455-531X

Department of Biochemistry, Groningen Biomolecular Sciences and Biotechnology Institute & Zernike Institute for Advanced Materials, University of Groningen, Nijenborgh 4, 9747 AG, Groningen, The Netherlands

AUTHOR INFORMATION

Eleonora Bailoni

ORCID ID: 0000-0003-1902-1322

Miyer F. Patiño-Ruiz

ORCID ID: 0000-0003-0607-4281

Andreea R. Stan^†

^†The author deceased before publication

Marten Exterkate

ORCID ID: 0000-0003-2204-8008

Department of Membrane Biogenesis and Lipidomics, Institute of Biochemistry, Heinrich-Heine-Universität Düsseldorf, Universitätsstraβe 1, 40225, Düsseldorf, Germany

Gea K. Schuurman-Wolters

Arnold J.M. Driessen

ORCID ID: 0000-0001-9258-9104

Department of Molecular Microbiology Groningen Biomolecular Sciences and Biotechnology Institute & Zernike Institute for Advanced Materials, University of Groningen, Nijenborgh 4, 9747 AG, Groningen, The Netherlands

AUTHOR CONTRIBUTIONS

E.B.: pathway design, protein purification and membrane reconstitution, enzyme and transport assays, ATP quantification, ATP/ADP measurements, phospholipid biosynthesis, data analysis, writing of manuscript. M.F.P.R.: transport assays, transport data analysis, manuscript editing. A.R.S.: database screening, gene cloning, transport assays. G.K.S.W.: protein purification, supervision of experiments in isotope laboratory. M.E.: research design, LC-MS data analysis, project supervision, manuscript editing. A.J.M.D.: research design, project supervision, manuscript editing. B.P.: research design, project supervision, manuscript writing, revision and funding.

NOTES

The authors declare no competing financial interests.

ACKNOWLEDGEMENTS

The authors thank Professor Frederic Leroy and Professor Jan Kok for kindly gifting the L. fermentum IMDO 130101 and L. sakei ATCC 15521 strains, respectively; Dr. Juke Lolkema for providing the DNA templates encoding L. brevis ATCC 367 arcD and S. pneumoniae D39 arcE; Professor Da-Neng Wang for generously donating the over-expression plasmid coding for E. coli glpT; Ronald van der Meulen for preliminary work on ArcD from L. sakei, and Dr. Hjalmar P. Permentier for carrying out the LC-MS analysis. The research was funded by the NWO Gravitation program “Building a synthetic cell” (BaSyC). In memoriam of Andreea R. Stan, who prematurely passed away and she did not see her contribution in the final version of the paper.

Schmidt, S., Sunyaev, S. & Dandekar, T. Metabolites: a helping hand for pathway evolution? Trends Biochem Sci 28, 336–341 (2003).
Sikkema, H. R., Gaastra, B. F., Pols, T. & Poolman, B. Cell Fuelling and Metabolic Energy Conservation in Synthetic Cells. ChemBioChem 20, 2581–2592 (2019).
Breuer, M. et al. Essential metabolism for a minimal cell. Elife 8, e36842 (2019).
Bailoni, E. et al. Minimal Out-of-Equilibrium Metabolism for Synthetic Cells: A Membrane Perspective. ACS Synth Biol 12, 922–946 (2023).
Feng, X., Jia, Y., Cai, P., Fei, J. & Li, J. Coassembly of photosystem II and ATPase as Artificial chloroplast for light-driven ATP synthesis. ACS Nano 10, 556–561 (2016).
Lee, K. Y. et al. Photosynthetic artificial organelles sustain and control ATP-dependent reactions in a protocellular system. Nat Biotechnol 36, 530–535 (2018).
Berhanu, S., Ueda, T. & Kuruma, Y. Artificial photosynthetic cell producing energy for protein synthesis. Nat Commun 10, 1–10 (2019).
Biner, O., Fedor, J. G., Yin, Z. & Hirst, J. Bottom-Up Construction of a Minimal System for Cellular Respiration and Energy Regeneration. ACS Synth Biol 9, 1450–1459 (2020).
Pols, T., Singh, S., Deelman-Driessen, C., Gaastra, B. F. & Poolman, B. Enzymology of the pathway for ATP production by arginine breakdown. FEBS J 288, 293–309 (2021).
Pols, T. et al. A synthetic metabolic network for physicochemical homeostasis. Nat Commun 10, (2019).
Bailoni, E. & Poolman, B. ATP Recycling Fuels Sustainable Glycerol 3-Phosphate Formation in Synthetic Cells Fed by Dynamic Dialysis. ACS Synth Biol 11, 2348–2360 (2022).
Kent, C. Eukaryotic phospholipid biosynthesis. Annu Rev Biochem 64, 315–343 (1995).
Robert Bishop, W. & Bell, R. M. Assembly of Phospholipids into Cellular Membranes: Biosynthesis, Transmembrane Movement and Intracellular Translocation. Annu Rev Cell Biol 4, 579–606 (1988).
Exterkate, M., Caforio, A., Stuart, M. C. A. & Driessen, A. J. M. Growing Membranes in Vitro by Continuous Phospholipid Biosynthesis from Free Fatty Acids. ACS Synth Biol 7, 153–165 (2018).
Kuruma, Y., Stano, P., Ueda, T. & Luisi, P. L. A synthetic biology approach to the construction of membrane proteins in semi-synthetic minimal cells. Biochimica et Biophysica Acta (BBA) - Biomembranes 1788, 567–574 (2009).
Eto, S. et al. Phospholipid synthesis inside phospholipid membrane vesicles. Communications Biology 2022 5:1 5, 1–11 (2022).
Blanken, D., Foschepoth, D., Serrão, A. C. & Danelon, C. Genetically controlled membrane synthesis in liposomes. Nat Commun 11, 1–13 (2020).
Göpfrich, K., Platzman, I. & Spatz, J. P. Mastering Complexity: Towards Bottom-up Construction of Multifunctional Eukaryotic Synthetic Cells. Trends Biotechnol 36, 938–951 (2018).
Elani, Y., Law, R. V. & Ces, O. Vesicle-based artificial cells as chemical microreactors with spatially segregated reaction pathways. Nat Commun 5, 1–5 (2014).
Elani, Y. et al. Constructing vesicle-based artificial cells with embedded living cells as organelle-like modules. Scientific Reports 2018 8:1 8, 1–8 (2018).
Chen, Z. et al. Synthetic beta cells for fusion-mediated dynamic insulin secretion. Nat Chem Biol 14, 86–93 (2018).
Frallicciardi, J., Melcr, J., Siginou, P., Marrink, S. & Poolman, B. Membrane thickness, lipid phase and sterol type are determining factors in the permeability of membranes to small solutes. Nat Commun 13, 1–12 (2022).
Joanne Lemieux, M. et al. Three-dimensional crystallization of the Escherichia coli glycerol-3-phosphate transporter: A member of the major facilitator superfamily. (2003) doi:10.1110/ps.03276603.
Silhavy, T. J., Hartig Beecken, I. & Boos, W. Periplasmic protein related to the sn glycerol 3 phosphate transport system of Escherichia coli. J Bacteriol 126, 951–958 (1976).
Larsons, T. J., Ehrmann, M. & Boos, W. Periplasmic glycerophosphodiester phosphodiesterase of Escherichia coli, a new enzyme of the glp regulon. Journal of Biological Chemistry 258, 5428–5432 (1983).
Majsnerowska, M., Noens, E. E. E., Lolkema, J. S., Henkin, T. M. & Noens, E. E. Arginine and Citrulline Catabolic Pathways Encoded by the arc Gene Cluster of Lactobacillus brevis ATCC 367. J Bacteriol 200, e00182 (2018).
Noens, E. E. E. & Lolkema, J. S. Convergent evolution of the arginine deiminase pathway: the ArcD and ArcE arginine/ornithine exchangers. Microbiologyopen 6, e00412 (2017).
Noens, E. E. E., Kaczmarek, M. B., Zygo, M. & Lolkema, J. S. ArcD1 and ArcD2 arginine/ornithine exchangers encoded in the arginine deiminase pathway gene cluster of Lactococcus lactis. J Bacteriol 197, 3545–3553 (2015).
Law, C. J., Yang, Q., Soudant, C., Maloney, P. C. & Wang, D.-N. Kinetic Evidence Is Consistent with the Rocker-Switch Mechanism of Membrane Transport by GlpT. Biochemistry 46, 12190–12197 (2007).
Auer, M. et al. High-Yield Expression and Functional Analysis of Escherichia coli Glycerol-3-phosphate Transporter. Biochemistry 40, 6628–6635 (2001).
Morigaki, K. & Walde, P. Fatty acid vesicles. Curr Opin Colloid Interface Sci 12, 75–80 (2007).
Ormö, M., Bystrom, C. E. & Remington, S. J. Crystal Structure of a Complex of Escherichia coli Glycerol Kinase and an Allosteric Effector Fructose 1,6-Bisphosphate. Biochemistry 37, 16565–16572 (1998).
van Nies, P. et al. Self-replication of DNA by its encoded proteins in liposome-based synthetic cells. Nat Commun 9, 1–12 (2018).
Doerr, A., Foschepoth, D., Forster, A. C. & Danelon, C. In vitro synthesis of 32 translation-factor proteins from a single template reveals impaired ribosomal processivity. Sci Rep 11, 1–12 (2021).
Śmigiel, W. M., Lefrançois, P. & Poolman, B. Physicochemical considerations for bottom-up synthetic biology. Emerg Top Life Sci 3, 445–458 (2019).
Geertsma, E. R. & Dutzler, R. A Versatile and Efficient High-Throughput Cloning Tool for Structural Biology. Biochemistry 50, 3272–3278 (2011).
de Kok, N. A. W., Exterkate, M., Andringa, R. L. H., Minnaard, A. J. & Driessen, A. J. M. A versatile method to separate complex lipid mixtures using 1-butanol as eluent in a reverse-phase UHPLC-ESI-MS system. Chem Phys Lipids 240, (2021).

There is NO Competing Interest.

20230925supplementary.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Synthetic organelles for energy conservation and delivery of building blocks for lipid biosynthesis

Status:

Version 1

Abstract

Figures

Introduction

Results

Glycerol 3-phosphate/Pi exchange to overcome phosphate depletion

Selection and activity screening of homologues for faster L-arginine/L-ornithine exchange

ATP formation

Co-reconstitution of ArcD and GlpT in LUVs

Kinetics and robustness of ATP-driven glycerol 3-phosphate synthesis and export

Design of a feeder-utilizer organelle-like vesicle system

Coupling of modules for lipid biosynthesis

Discussion

Methods

Chemicals and media

Strains and plasmids

Cloning of genes for ArcD and ArcE homologues

Over-expression and purification of ArcD homologues

Over-expression and purification of GlpT

Over-expression and purification of other proteins

Formation of proteoliposomes with one or multiple antiporters

Encapsulation of soluble components for L-arginine breakdown and glycerol 3-phosphate formation

ATP quantification by chemiluminescence

ATP/ADP determination by fluorescence

DOPA biosynthesis in feeder-utilizer vesicles

LC-MS analysis of phospholipids

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1