This protocol describes the inverted emulsion method used to encapsulate a cell-free expression (CFE) system within a giant unilamellar vesicle (GUV) for the investigation of the synthesis and incorporation of a model membrane protein into the lipid bilayer.
Cell-free expression (CFE) systems are powerful tools in synthetic biology that allow biomimicry of cellular functions like biosensing and energy regeneration in synthetic cells. Reconstruction of a wide range of cellular processes, however, requires successful reconstitution of membrane proteins into the membrane of synthetic cells. While the expression of soluble proteins is usually successful in common CFE systems, the reconstitution of membrane proteins in lipid bilayers of synthetic cells has proven to be challenging. Here, a method for reconstitution of a model membrane protein, bacterial glutamate receptor (GluR0), in giant unilamellar vesicles (GUVs) as model synthetic cells based on encapsulation and incubation of the CFE reaction inside synthetic cells is demonstrated. Utilizing this platform, the effect of substituting the N-terminal signal peptide of GluR0 with proteorhodopsin signal peptide on successful cotranslational translocation of GluR0 into membranes of hybrid GUVs is demonstrated. This method provides a robust procedure that will allow cell-free reconstitution of various membrane proteins in synthetic cells.
Bottom-up synthetic biology has gained increasing interest over the past decade as an emerging field with numerous potential applications in bioengineering, drug delivery, and regenerative medicine1,2. The development of synthetic cells as a cornerstone of bottom-up synthetic biology, in particular, has attracted a wide range of scientific communities due to the promising applications of synthetic cells as well as their cell-like physical and biochemical properties that facilitate in vitro biophysical studies3,4,5,6. Synthetic cells are often engineered in cell-sized giant unilamellar vesicles (GUVs) in which different biological processes are recreated. Reconstitution of cell cytoskeleton7,8, light-dependent energy regeneration9, cellular communication10,11, and biosensing12 are examples of efforts made to reconstruct cell-like behaviors in synthetic cells.
While some cellular processes rely on soluble proteins, many characteristics of natural cells, such as sensing and communication, often utilize membrane proteins, including ion channels, receptors, and transporters. A major challenge in synthetic cell development is the reconstitution of membrane proteins. Although traditional methods of membrane protein reconstitution in lipid bilayers rely on detergent-mediated purification, such methods are laborious, ineffective for proteins that are toxic to the expression host, or are often not suited for membrane protein reconstitution in GUVs13.
An alternative method for protein expression is cell-free expression (CFE) systems. CFE systems have been a powerful tool in synthetic biology that allows in vitro expression of various proteins using either cell lysate or purified transcription-translation machinery14. CFE systems can also be encapsulated in GUVs, thus allowing compartmentalized protein synthesis reactions that can be programmed for various applications, such as the creation of light-harvesting synthetic cells9 or mechanosensitive biosensors15,16. Analogous to recombinant protein expression methods, membrane protein expression is challenging in CFE systems17. Aggregation, misfolding, and lack of post-translational modification in CFE systems are major bottlenecks that hinder successful membrane protein synthesis using CFE systems. The difficulty of bottom-up membrane protein reconstitution using CFE systems is due in part to the absence of a complex membrane protein biogenesis pathway that relies on signal peptides, signal recognition particles, translocons, and chaperoning molecules. However, recently, multiple studies have suggested that the presence of membranous structures such as microsomes or liposomes during translation promotes successful membrane protein expression18,19,20,21. Additionally, Eaglesfield et al. and Steinküher et al. have found that the inclusion of specific hydrophobic domains known as signal peptides in the N-terminus of the membrane protein can significantly improve its expression22,23. Altogether, these studies suggest that the challenge of membrane protein reconstitution in synthetic cells can be overcome if the protein translation occurs in the presence of the GUV membrane and if proper N-terminal signal peptide is utilized.
Here, a protocol for encapsulation of the protein synthesis using recombinant elements (PURE) CFE reactions for membrane protein reconstitution in GUVs is presented. Bacterial glutamate receptor24 (GluR0) is selected as the model membrane protein, and the effect of its N-terminal signal peptide on its membrane reconstitution is studied. The effect of proteorhodopsin signal peptide, which was shown to improve membrane protein reconstitution efficiency by Eaglesfield et al.22, is investigated by constructing a mutated variant of GluR0 denoted as PRSP-GluR0 and its expression and membrane localization with wild-type GluR0 (referred to as WT-GluR0 hereafter) that harbors its native signal peptide is compared. This protocol is based on the inverted emulsion method25 with modifications that make it robust for CFE encapsulation. In the presented method, the CFE reactions are first emulsified using a lipid-in-oil solution that generates micron-sized droplets that contain the CFE system and are stabilized by the lipid monolayer. The emulsion droplets are then layered on top of an oil-water interface that is saturated with another lipid monolayer. The emulsion droplets are then forced to travel across the oil-water interface via centrifugal force. Through this process, the droplets obtain another monolayer, thus generating a bilayer lipid vesicle. The GUVs containing the CFE reaction are then incubated, during which the membrane protein is expressed and incorporated into the GUV membrane. Although this protocol is specified for cell-free expression of GluR0, it can be used for cell-free synthesis of other membrane proteins or different synthetic cell applications such as cytoskeleton reconstitution or membrane fusion studies26.
The reagents and equipment utilized for this study are provided in the Table of Materials.
1. Bulk CFE reactions in the presence of small unilamellar vesicles (SUVs)
2. CFE reactions encapsulated in GUVs
3. Encapsulated CFE reaction incubation and imaging
Prior to encapsulation of the CFE reactions, two variants of GluR0-sfGFP harboring native and proteorhodopsin signal peptides (signal peptide sequences are presented in Supplementary Table 1), and the soluble sfGFP were individually expressed in bulk reactions, and their expression was monitored by detecting the sfGFP signal using a plate reader (Figure 2A). Membrane proteins were expressed in the absence or presence of 100 nm SUVs. Additionally, using a calibration curve that correlates sfGFP signal to its concentration (Supplementary Figure 1), concentrations of synthesized proteins were estimated (Supplementary Table 2). Clearly, soluble sfGFP had the highest expression among all three proteins, which suggests that the expression of membrane proteins imposes a burden on the CFE system, thus slowing down the reaction and lowering its yield. In addition, on average, reactions expressing membrane proteins in the presence of SUVs showed higher sfGFP signal compared to reactions lacking SUVs. This observation aligns with the findings of Steinküher et al., who showed that the expression of membrane proteins reduces the capacity of the CFE systems to produce proteins23. Nevertheless, given the successful demonstration of protein expression in bulk CFE reaction, one can reason that encapsulated CFE will also synthesize proteins inside GUVs.
Next, individual CFE reactions were encapsulated in GUVs using the inverted emulsion method to express variants of GluR0, namely WT-GluR0, PRSP-GluR0, and soluble sfGFP. While WT-GluR0, harboring GluR0 native signal peptide, demonstrated excellent expression and membrane localization (Figure 2B, left panel), its counterpart, PRSP-GluR0, which has proteorhodopsin N-terminal signal peptide, did not show similar strong membrane localization. PRSP-GluR0 was found to be more prone to aggregation and punctate formation (Figure 2B, middle panel). As expected, soluble sfGFP was expressed in GUVs and stayed in the GUV lumen (Figure 2B, right panel; see Supplementary Figure 2 for images of cohorts of GUVs).
Figure 1: Experimental steps of inverted emulsion. (1) Step 2.3.1 through step 2.3.3 of the protocol are visualized to demonstrate the assembly of the lipid monolayer at the interface of the lipid-oil mix and outer buffer solution. (2) Visualization of step 2.3.5 of the protocol is shown here to represent the formation of the lipid monolayer around emulsified droplets encapsulating the inner CFE solution. (3) Step 2.3.6 of the protocol shows the addition of the monolayer GUVs to the microcentrifuge tube with the lipid monolayer at the interface of a lipid-oil mix and outer buffer solution. (4) Step 2.3.7 is depicted here, in which centrifugation leads to the formation of a GUV pellet in the outer solution. (5) Step 2.3.8 is shown here, indicating the process of removing the excess lipid-in-oil mixture and outer solution. (6) Finally, step 2.3.9 is depicted here, where the GUV pellet is resuspended in the outer solution, and the GUVs are ready for incubation, followed by imaging. Please click here to view a larger version of this figure.
Figure 2: Protein expression in bulk CFE reactions and in GUVs encapsulating CFE reactions. (A) Fluorescence readouts of individual bulk CFE reactions expressing WT-GluR0-sfGFP, PRSP-GluR0-sfGFP, and soluble sfGFP. The soluble sfGFP graph represents the signal from a 2.5 µL reaction (standard reaction volume is 20 µL) to avoid oversaturation of the plate reader measurements. Data is presented as mean ± S.D, n = 3. (B) Left: A representative confocal image of a GUV encapsulating CFE reaction expressing WT-GluR0-sfGFP. Middle: A representative confocal image of GUVs encapsulating CFE reaction expressing PRSP-GluR0-sfGFP. Right: A representative confocal image of a GUV encapsulating CFE reaction expressing soluble sfGFP. Scale bars: 10 µm. Please click here to view a larger version of this figure.
Supplementary Figure 1: The sfGFP signal calibration curve and its corresponding linear regression analysis. Please click here to download this File.
Supplementary Figure 2: Representative confocal image of cohorts of GUVs expressing (left) WT-GluR0-sfGFP, (middle) PRSP-GluR0-sfGFP, and (right) soluble sfGFP. Scale bar: 10 µm. Please click here to download this File.
Supplementary Table 1: The amino acid sequence of wild-type and proteorhodopsin signal peptides. Please click here to download this File.
Supplementary Table 2: The concentration of synthesized proteins in bulk CFE reactions. Please click here to download this File.
Virtually any cellular process that depends on the transfer of molecules or information across the cell membrane, like cell signaling or cell excitation, requires membrane proteins. Thus, the reconstitution of membrane proteins has become the main bottleneck in realizing various synthetic cell designs for different applications. Traditional detergent-mediated reconstitution of membrane proteins in biological membranes requires GUV generation methods such as gentle swelling or electroformation. Swelling approaches usually produce small-sized vesicles, and electroformation yield significantly drops when complicated solutions, which is often the case when generating synthetic cells, are encapsulated32. Additionally, detergents solubilize the membrane protein, and their removal during the reconstitution process can cause protein misfolding33,34. On the other hand, the approach presented here relies on the cotranslational incorporation of the membrane protein into the lipid bilayer, which resembles more the natural protein biogenesis pathway in cells22.
From a technical point of view, the presented protocol is advantageous to other common encapsulation methods, such as electroformation and continuous droplet interface crossing encapsulation7,8,35,36 (cDICE), for easier implementation as the only laboratory equipment required for GUV generation is a centrifuge. As opposed to electroformation, the inverted emulsion method allows the encapsulation of different combinations of molecules with various concentrations. Additionally, compared to the original inverted emulsion technique25, this approach generates more stable GUVs that are suitable for encapsulation of CFE lysates or PURE systems. The higher GUV stability is owed to the presence of diblock copolymer in the composition of GUV membrane37 as well as the long incubation of the oil-water interface that allows the interface to be saturated with lipid molecules. Lastly, as opposed to microfluidics approaches, the protocol presented here does not require small channels and tubing. Therefore, the CFE reaction can be encapsulated as soon as it is assembled, and the shorter time of GUV assembly due to lack of flow and possible clogging prevents premature start of the CFE reaction. While the demonstration of membrane protein expression in this protocol is exclusive to PUREfrex reactions, one can extend this method to synthesize proteins using different available CFE systems, such as lysate-based bacterial or mammalian CFE systems.
The presented approach here has limitations that are caused by the oil-dependent nature of the GUV formation process and the intent to have stable GUVs. This approach is typically longer compared to other methods, such as cDICE or microfluidics, due to the long incubation time of the oil-water interface that is required for interface stabilization and high GUV yield. Additionally, lipid composition is primarily limited to POPC with small doses of other lipids or block copolymers, while other methods, such as electroformation, are more suited for the incorporation of lipids with different physical and chemical properties. While the GUV membrane composition in this method is a mixture of POPC and PBD-PEO to maximize CFE yield, possible variations in GUV membrane composition can be tested. However, further optimization of the parameters might be required for other membrane proteins. Since the droplet emulsification occurs through manual pipetting, the GUVs generated via this method are polydisperse and quite heterogeneous in size. Further, the fact that lipids are dissolved in the organic phase may occasionally cause a layer of oil between the two leaflets of the GUV membrane or contaminate the imaging chamber with oil that can be detrimental to image quality. A possible workaround for the challenge of residual oil is to replace mineral oil with a volatile organic solvent, such as diethyl ether, as shown by Tsumoto et al.38, to rely on solvent evaporation along with centrifugation during GUV formation.
While there is no demonstration of channel function in this work, inspired by previous assays used for probing reconstituted mechano- or light-sensitive channel functionality, a fluorescence microscopy-based assay is outlined. The opening of the GluR0 channel is reported to increase the membrane conductivity for K+ ions24. Because CFE reactions already contain a high concentration of K+, typical potassium indicators will not be suitable for assessing channel functionality. However, because potassium influx changes the membrane potential, sensitive membrane potential indicators such as DiBAC4(3)22 or BeRST 139 could report GluR0 activity in the presence of glutamate.
Successful reconstitution of membrane proteins in synthetic cells opens up numerous possibilities for creating synthetic cells with unprecedented abilities that more closely mimic natural cells. A current major disadvantage of synthetic cells is their inability to reproduce and recycle energy. However, with light- and chemical-dependent energy regeneration schemes that rely heavily on membrane proteins, one can envisage long-lasting synthetic cells40. Utilizing CFE systems allows the reconstitution of multiple membrane proteins that can collectively perform certain tasks. For instance, reconstitution of a ligand-gated ion channel similar to GluR0 described here, along with different voltage-gated ion channels, can lead to the construction of an excitable neuron-like synthetic cell.
The authors have nothing to disclose.
APL acknowledges support from the National Science Foundation (EF1935265), the National Institutes of Health (R01-EB030031 and R21-AR080363), and the Army Research Office (80523-BB)
100 nm polycarbonate filter | STERLITECH | 1270193 | |
96 Well Clear Bottom Plate | ThermoFisher Scientific | 165305 | |
BioTek Synergy H1M Hybrid Multi-Mode Reader | Agilent | 11-120-533 | |
Creatine phosphate | Millipore Sigma | 10621714001 | |
CSU-X1 Confocal Scanner Unit | Yokogawa | CSU-X1 | |
Density gradient medium (Optiprep) | Millipore Sigma | D1556 | Optional to switch with sucrose in inner solution |
Filter supports | Avanti | 610014 | |
Fisherbrand microtubes (1.5 mL) | Fisher Scientific | 05-408-129 | |
Folinic acid calcium salt hydrate | Millipore Sigma | F7878 | |
Glucose | Millipore Sigma | 158968 | |
HEPES | Millipore Sigma | H3375 | |
iXon X3 camera | Andor | DU-897E-CS0 | |
L-Glutamic acid potassium salt monohydrate | Millipore Sigma | G1501 | |
Light mineral oil | Millipore Sigma | M5904 | |
Magnesium acetate tetrahydrate | Millipore Sigma | M5661 | |
Mini-extruder kit (including syringe holder and extruder stand) | Avanti | 610020 | |
Olympus IX81 Inverted Microscope | Olympus | IX21 | |
Olympus PlanApo N 60x Oil Microscope Objective | Olympus | 1-U2B933 | |
PEO-b-PBD | Polymer Source | P41745-BdEO | |
pET28b-PRSP-GluR0-sfGFP plasmid DNA | Homemade | N/A | |
pET28b-sfGFP-sfCherry(1-10) plasmid DNA | Homemade | N/A | |
pET28b-WT-GluR0-sfGFP plasmid DNA | Homemade | N/A | |
POPC lipid in chloroform | Avanti | 850457C | |
Potassium chloride | Millipore Sigma | P9541 | |
PUREfrex 2.0 | Cosmo Bio USA | GFK-PF201 | |
Ribonucleotide Solution Set | New England BioLabs | N0450 | |
RNase Inhibitor, Murine | New England BioLabs | M0314S | |
RTS Amino Acid Sampler | Biotechrabbit | BR1401801 | |
Sodium chloride | Millipore Sigma | S9888 | |
Spermidine | Millipore Sigma | S2626 | |
Sucrose | Millipore Sigma | S0389 | |
VAPRO Vapor Pressure Osmometer Model 5600 | ELITechGroup | VAPRO 5600 |