Nanoelectrospray based synthesis of  large, transportable membranes with integrated membrane proteins

doi:10.21203/rs.3.rs-4725155/v1

Download PDF

Article

Nanoelectrospray based synthesis of large, transportable membranes with integrated membrane proteins

https://doi.org/10.21203/rs.3.rs-4725155/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Membrane proteins tend to be difficult to study since they need to be integrated into a lipid bilayer membrane to function properly. Here I am presenting a method to synthesize a macroscopically large and freely transportable membrane with integrated membrane proteins. This is useful for studying membrane proteins and protein complexes in isolation. The method could serve as a blueprint for the production of larger quantities of functionalised membranes for integration into technical devices similar to the MinION DNA sequencer.

It is possible to self-assemble larger biological membranes on solid surfaces. However, they cannot be removed from their solid support without destroying them. In transportable form, self-assembled membranes are limited to sizes of about 17 nm in nanodiscs. Here we electrospray a series of molecular layers onto the liquid surface of a buffer solution. This creates a flat, liquid environment on the surface that directs the self-assembly of the membrane. The composition of the membrane, its lipid and protein content, is experimentally controlled. We have succeeded in producing large membranes with integrated OmpG, a transmembrane pore protein. The technique appears to be compatible with the assembly of membrane based protein complexes.

Biological sciences/Biophysics/Membrane biophysics

Biological sciences/Biophysics/Membrane structure and assembly

Biological sciences/Biophysics/Molecular biophysics

Biological sciences/Biophysics/Nanoscale biophysics

Biological sciences/Biotechnology/Biomaterials/Biomaterials proteins

Physical sciences/Physics/Condensed matter physics/Surfaces interfaces and thin films

The aim of this study is to make membranes with integrated proteins accessible for functional and structural studies. We are using a nanoelectrospray device to synthesize an environment from gas phase in which the membranes can self-assemble. The molecular composition of the membrane, its lipid and proteinaceous components, is controlled by the experimental design.

We use a nanoelectrospray ion source to create a planar environment for self-assembly of large lipid bilayers containing membrane proteins. In order to maintain the proteins’ three-dimensional shape, the technique was conducted under atmospheric conditions and ambient temperatures (see Fig. 1)¹.

In the meantime the same technique is used to prepare soluble proteins in a thin layer of water for a cryo-EM analysis². The scientists observe similar characteristics for the preparation which were found here empirically to be required for membrane self-assembly - most noteworthy the hydrophilic enclosure of the proteins.

Larger membranes have thus far only been assembled on solid supports³. The drawback of this method is that once formed, the membranes cannot be removed from the surface without damaging their integrity. In this study, the membrane is assembled within a thin glycerol layer on a liquid buffer surface. Once assembled, the membrane can be easily moved, for instance to a transmission electron microscope for inspection.

Completely unassisted self assembly of membranes in solution is difficult to achieve, is limited in size and can easily lead to the formation of membrane stacks, roles or vesicles⁴. Transportable membranes can otherwise be assembled in the form of nanodiscs⁵. However, these nanodiscs are limited to a diameter of approximately 17 nm, as they become unstable and collapse beyond that point. Here, the self-assembly takes place in a thin layer which prevents the membrane from forming stacks, spheres, or rolls.

The nanoelectrospray method utilized in this study was established over 30 years ago. Relevant experiments have been included as they provided the base of these experiments and had not been published before.

2.1 Nanoelectrospay as a Molecular Beam Device

2.1.1 Nanoelectrospray Based Surface Preparation

Electrospray is routinely used to prepare thin films⁶ even though molecular-level design is not achieved. We used a home-built electrospray device to distribute a dissolved component uniformly over a metal target (figure 1). Figure 2 panel a, b show the surface prepared by spraying a rhodamine solution onto a flat target with a conventional instrument built in the late eighties of the twenties century at the Physics Department of the University of Münster, Germany⁷. The electrospray was operated at flow rates of 10 - 20 μL/min. The purpose was to distribute an analyte uniformly over a surface for subsequent analysis by a static secondary ion mass spectrometer (SIMS). Static SIMS is a sub-monolayer sensitive surface analysis technique. Such an electrospray preparation was state of the art at the time⁶, but far from monolayer characteristics. Subsequent theoretical and experimental studies of the electrospray process itself showed a way to improve the homogeneity of the preparation⁷. A stable liquid Taylor cone ejects droplets exclusively from its tip. The radius at the tip of the Taylor cone for droplet emission depends on the flow rate. The central equation describing the radius of the emission region from the tip of a Taylor cone is⁷:

Equation 1

The central equation linking the radius r_e of the droplet emission region at the tip of a Taylor cone with the flow rate V'. The lower the flow rate the smaller are the initially produced droplets.

Reducing the flow rate reduces the size of the initially produced droplets (figure 2 , panel b). The equation suggests that when spraying an aqueous solution of 1 pmol/μL at a flow rate of 20 nL/min, droplets with the size of the emission radius would contain, on average, only one analyte molecule.

Acetone has a high vapor pressure. When sprayed its droplets start to evaporate in flight and break up into smaller second generation droplets before they dry down and the non volatile rhodamine residuals reach the surface⁸. The droplet emission radius is only a description for the largest droplet generated by the spray, not the most common one. By replacing the metallic nozzle with a gold-coated, pulled glass capillary, the acetone flow rate was reduced to about 200 nL/min. This was the first - later called - nanoelectrospray instrument. 200 nL/min was the lowest flow rate achievable at the time with the first generation instrument. With this emitter, the metallic aluminum surface did not visibly change when the rhodamine solution was sprayed - even under prolonged preparation of 1 h or more. The obtained surfaces were investigated with a raster-electron microscope and a static SIMS mass spectrometer (figure 2, panel c, d). The raster electron microscopic picture showed no discernible particles even though rhodamine covered the surface of the target.

The following experiments were designed to determine whether the nanoelectrospray produced surfaces that behaved like monolayers under static SIMS examination. Such a preparation would allow the hypothesis that the final droplets were so small that they contained on average only one molecule of rhodamine. The experimental results obtained at this stage were not completely conclusive, but they provided enough evidence to continue this line of experimentation.

Figure 2, panel e shows the slow increase of the static SIMS signal with preparation time and a decrease when the rhodamine coverage becomes too thick. This behavior is similar to monolayer coverage on clean surfaces in vacuum by molecular beam deposition⁹. The question was whether the surface is covered by an increasing number of particles, as in panel a, or whether the coverage is monolayer-like. To answer this question, a series of experiments were performed to determine the time required to reach the maximum SIMS response signal t_SIMSmax as a function of rhodamine concentration. When small rhodamine particles cover the target, t_SIMSmaxdepends on their cross section and thus varies with (c_Rh)^-2/3. However, if the coverage is monolayer-like, the static SIMS signal will respond to the total amount of rhodamine sprayed on the surface and will depend on (c_Rh)^-1. Figure 2 panel f shows how the preparation time to reach the maximum SIMS signal t is indirectly proportional to the rhodamine concentration. The linearity indicates that the SIMS signal characteristic changes in proportion to the total amount of rhodamine sprayed onto the target. This result indicates that the surfaces generated by the nanoelectrospray preparation behave like molecular monolayers under static SIMS examination.

2.1.2 Molecular Beam Hypothesis

The surface preparation experiments concluded that when an acetone solution is electrosprayed from a stable Taylor cone using a metal-coated drawn glass capillary at a flow rate of the order of 200 nL/min, the compartmentalization of individual analyte molecules into separate droplets is complete or close to complete. Two hypotheses emerged from these experiments.
First, if such a device is pointed at a vacuum system, the volatile solvent would evaporate, releasing the non-volatile molecules into the vacuum. Thus, a nanoelectrospray could be an ion source for large molecules of virtually unlimited mass - which is what electrospray ion sources are today. And second, it should be possible to use a nanoelectrospray to synthesize molecularly designed layers of large organic compounds¹⁰.

Prof. Fenn et al. had published the first mass spectra of protein ions generated with a conventional electrospray ion source in 1988^11,12. They had no experimental evidence suggesting that the electrospray can produce droplets so small that they contain only one protein molecule. To explain the generation of these ions from larger droplets they proposed the ion evaporation mechanism^12,13.

The first project, the implementation of the nanoelectrospray as an ion source at a mass spectrometer, was realized in 1992 at the European Molecular Biology Laboratory (EMBL), Heidelberg, Germany¹⁴. The direct coupling of the nanoelectrospray ion source to the vacuum system of a mass spectrometer, as shown in figure 3, demonstrated that the compartmentalization of analyte molecules is indeed complete. Under standard operational conditions no clustering of proteins was observed. The nanoelectrospray ion source is a very bright ion source, approximately 100 times more efficient than any other electrospray ion source at the time. The instrument provided the technical basis for the development of low-level protein identification by mass spectrometry. Over the years, it has been used to analyze hundreds of peptide and protein samples.

2.2 Self Assembly of Large Protein Containing Membranes

Can a nanoelectrospray ion source be used to synthesize membranes with incorporated membrane proteins from gas phase? Because purified membrane proteins are only stable when solubilized with detergent molecules, it is not possible to synthesize such membranes directly^15,16. The question was whether nanoelectrosprayed layers could provide an environment in which membranes can self-assemble. The experimental basis for this was the ability of the nanoelectrospray to generate monomolecular films. The question to be addressed was: What are the requirements for self-assembly of protein-filled lipid bilayers when synthesized in a layered fashion? The course of the experiments suggested the answer: self assembly starts with a lipid monolayer as template and hydrostatic stabilization of the membrane proteins on both sides of the lipid bilayer. Although a lipid bilayer can form on the surface of a liquid, protein incorporation will only occur if a hydrostatic layer thick enough to enclose the proteins surrounds the lipid layer. This experimental finding is consistent with the observation that membrane proteins in a supported bilayer can denature if the distance between the membrane and the underlying support is too small³.

Recently, Yang et al. have confirmed this approach of using nanoelectrospray or low-flow electrospray to design flat films at the molecular level. They were able to prepare a thin aqueous film containing soluble proteins for cryo-EM studies². Part of the prepared EM grid contained a layer thin enough to reconstruct the protein structure from the images. The researchers made the same observation as reported here, that in protein structure-preserving monolayers, the proteins are surrounded by a hydrophilic shell.

The first experimental result was the observation that lipid mono- and bilayers can assemble on a liquid surface when the lipids are sprayed onto it. The substrate chosen for layer formation was the liquid meniscus of a small round container with a diameter of 3 mm, the size of a standard electron microscope grid. This corresponds to a surface area of 7.07 mm². With 66 pmol phosphatidylcholine evenly distributed over the surface the viscosity visibly changed and folds appeared in the surface after overnight incubation (see Supplementary Fig. 1). The viscosity change and folding of the surface did not occur when the layer contained only 33 pmol. This was taken as experimental evidence that 66 of pmol phosphatidylcholine was sufficient to form a lipid bilayer. This conclusion is consistent with the calculated surface occupancy of the phosphatidylcholine molecules. With 33 pmol phosphatidylcholine evenly distributed over the surface of the meniscus, the lipid density is 36 Å² per molecule. 36 Å² lies between the surface occupancy of 60 Å² - 70 Å² measured for phosphatidylcholine molecules in vesicles¹⁷ and approximately 20 Å² for a saturated fatty acid molecule in a regular Langmuir-Blodgett film¹⁸. Considering the linear structure of saturated fatty acids and their dense packing in a regular array, 20 Å² is certainly too small for the surface occupancy of a phosphatidylcholine molecule in a closed, flat phosphatidylcholine film. With the globular shape of vesicles 60 Å² for a single phosphatidylcholine molecule is more than this molecule would occupy in a flat layer. Hence, 36 Å² as calculated from the experimentally found conditions for the preparation of a monolayer is within in the expected range.

The behavior of Escherichia coli outer membrane protein G (OmpG) on surfaces generated by spraying 66 pmol and 33 pmol underscores their different nature. OmpG is a relatively robust 35 kDa monomeric channel protein. It is a membrane-spanning protein that denatures in an aqueous environment lacking detergent to cover its hydrophobic regions. Its crystal structure shows that its pore has an inner diameter of 2 nm - 2.5 nm^15,16. It was observed that OmpG sprayed onto a surface prepared with 66 pmol phosphatidylcholine denatures upon removal of the detergent. When sprayed onto a surface with only 33 pmol phosphatidylcholine, it remains intact.

The preparation protocol to generate large membrane layers in a reproducible and robust manner (figure 4) has several steps. The first step consists of spraying 66 pmol of bipolar lipid onto a buffer containing SM-2 biobeads. The vessel was incubated overnight at room temperature. This creates a lipid bilayer that retains subsequent molecules at the surface. A thin layer of glycerol lays the foundation for membrane assembly in a hydrophilic environment by spraying a 1:2 glycerol/ethanol solution. The equivalent of a lipid monolayer is added (33 pmol phosphatidylcholine) followed by the detergent solution of OmpG and enough lipid to form a closed bilayer (20 pmol phosphatidylcholine). A final glycerol layer completes the hydrophilic environment.

The SM-2 beads in the buffer extracted the detergent from the surface during a six-days incubation period. After transfer to an EM grid and washing with distilled water, the grid was exposed to a platinum beam in vacuo. Figure 4 shows a series of images generated from such EM grids. A single, extended, intact OmpG-containing layer has formed. No denatured protein was visible. OmpG denatures in an aqueous environment unless stabilized by detergents or lipids. The presence of the observed structures depends on the addition of OmpG to the surface. The features they are virtually identical in size and shape to a projection of the space-filling model of OmpG calculated from crystallization data¹⁶ (see Supplementary Movie 1).

2.3 Protein Complex Formation in Nanoelectrospray Generated Layers

The assembly of a large protein-containing membrane occurs in a thin layer of glycerol, which is probably too thin to promote the formation of membrane stacks or tubes. The question is whether this highly artificial environment allows the assembly of membrane-based protein complexes. Although this has to be decided on a case-by-case basis, we have tested this using the proteins listeriolysin O and pneumolysin. Listeriolysin O and pneumolysin are both members of a group of pore-forming toxins. They are soluble 53 kDa and 60 kDa proteins, respectively. Up to 50 monomers form a non-covalent ring-shaped complex on membranes. The rings insert themselves and create large pores of about 30 nm in diameter^19,20. Listeriolysin O assembles best at a slightly acidic pH of 5.5. Both proteins require phospholipids in the membrane for docking and complex formation.

2.3.1 Listeriolysin O Assembly on a Lipid Bilayer

The buffer chosen for the preparation of listeriolysin O layers allows the assembly of its complexes on membranes²¹. However, in these experiments, the local environment for the protein is a thin layer of glycerol, not the buffer itself. In thin layers, ion concentrations can be very different from bulk solution. In an experiment to test complex formation, the nanoelectrospray first generated a lipid bilayer film, then a film containing listeriolysin O, followed by a glycerol layer to complete the hydrophilic environment. Figure 5 shows the result. The overall appearance is consistent with published listeriolysin O complexes on membranes²¹. The assembly of the complex is highly efficient. The electron dense centers of many of the completed rings are probably due to residues of glycerol and salt from the buffer solution that were not successfully removed by the washing procedure.

2.3.2 Listeriolysin O and Pneumolysin Assembly on a Lipid Monolayer

The physiological mechanism of listeriolysin O and pneumolysin is to form a ring-shaped complex on a bilayer membrane. The ring inserts itself and forms a pore^21,22. What would happen if the protein were prepared in the same way as an intramembrane protein, by spraying it onto a lipid monolayer ? Figure 6 shows the result.

Some ring-like structures of the expected size are visible, but the efficiency is dramatically reduced compared to the preparation on a lipid bilayer. This is consistent with the mechanism of formation of listeriolysin and pneumolysin pores. In analogy to the experiments with OmpG, which denatures when sprayed onto a lipid bilayer upon detergent removal, the experiment shows that nanoelectrospray is indeed capable of preparing lipid mono- and bilayer structures that behave radically different when interacting with membrane proteins or protein complexes.

2.3.3 Direct Observation of Protein Complex Formation

All observations reported so far were systematic results, repeated several times. Here, we present a sporadic observation that supports the interpretation that the electron dense centers of the ring-like structures are glycerol-filled pores and highlights the potential of nanoelectrospray-based preparation methods: the direct observation of the pore formation process of pneumolysine, from monomer to finished pore.

The images in figure 7 are from a preparation of pneumolysin on lipid monolayers. Initially, there should have been very few pores (see figure 6). However, in this case the spray was not operated at the desired low flow rate. Larger droplets formed that landed on the surface membrane and the water evaporated, leaving crystalline salt residues on the surface. Nanoelectrospray ion sources on mass spectrometers operate with volatile buffers, but for these experiments a more conservative approach was taken and the original solution in which the proteins were kept was not altered. After the preparation was complete, enough lipid was sprayed to form bilayers on the surface. The images in figure 7 suggest that during the incubation period, pneumolysin monomers slowly dissolve from the crystalline residue and diffuse into the glycerol layer. As they encounter a lipid bilayer, they begin to assemble and form pores. This step-wise pore forming process corresponds to the pore forming mechanism proposed by computer simulations of related proteins like listeriolysin O²³. The whole process is visible and suggests that it is slow enough to be extended over the diffusion range. Glycerol has a much higher viscosity than water. And diffusion in thin layers is slower than in bulk solution. All this suggests that the formation of protein complexes in thin layers of glycerol may be directly observable, especially if the temperature of the layer is controlled.

Pore formation in a supported bilayer has been directly observed by atomic force microscopy for the membrane attack complex (MAC)²⁴. The pore is completed in approximately 100 seconds.

The second of the two original hypotheses, that nanoelectrospray should allow the composition of surfaces at the molecular level, seems to hold. The fact that it was possible to self-assemble a large OmpG-containing membrane with a lipid monolayer as base, but not with a lipid bilayer, confirms this hypothesis. The same is true for the formation of listeriolysin O and pneumolysin pores. Only when sprayed on a lipid bilayer does pore formation proceed effectively. These experiments describe a way to generate large, protein-containing membranes that are freely transportable. So far, the experiments have not aimed at anisotropic orientation of the proteins. However, it is likely that the proteins will often be oriented when the membrane assembly process proceeds in a static electric field. Further experiments are needed to show that membranes can be synthesized with other membrane proteins and that the proteins can be incorporated anisotropically when the membrane is assembled in an external electric field.

These experiments don't show that such membranes have the protein-specific biochemical activity. But I think they lay the foundation for further experiments that lead to bio-similar materials that are useful in technical devices like cell-free bioreactors and protein based sensors.

4.1 Protein Expression & Purification

Yildiz et al. describe the expression and purification of the outer membrane porin G (OmpG)¹⁵. Briefly, E. coli expresses the amino acids 22 - 301 plus an N-terminal methionine of the OmpG gene having cloning it into a pET26B plasmid vector. After breaking the cells, solubilizing them in 8 M urea, 1% Triton-X 100 and 25 mM Tris-HCl, unfolded OmpG is separated from impurities via an ion exchange column separation. The protein refolds in a 1% (w/v) n-octyl-β-D-glucopyranoside (OG), 3 M urea solution. Afterward, a final ion-exchange separation process eliminates unfolded or partially folded protein from the solution. By performing ultrafiltration or dialysis with membranes having a 12 kDa weight cutoff against a buffer with 20% PEG-35000, the protein concentration increases to approximately 50 mg/mL. A solution comprised of 10 mM Tris, 1% octylglucoside, and 20 pmol/μL phosphatidylcholine, with a concentration of 32 pmol/μL OmpG, acts as the base for preparing OmpG layers. To shield the protein from the high charges of the electrospray droplets, an extra 25% glycerol is added to the final spraying solution.

Listeriolysin O (LLO) was obtained from two sources. The expression and purification procedure of LLO was described in¹⁹. Briefly, for E. coli BL21 (DE3) expression, the gene segment encoding residues 25-529 of LLO sequence, without N-terminal secretion signal, was cloned into the plasmid pET15b, along with an N-terminal His6 tag. Once the cellular density A600 reached 1.2, the temperature was reduced from 37°C to 30°C to curtail cell proliferation. Introducing 1 mM isopropyl β- D-1-thiogalactopyranoside inhibited the plasmid's Lac-repressor, initiating protein expression. After 4 hours, the cells were harvested, suspended in 50 mM Tris pH 7.7, 150 mM NaCl, and disrupted using a microfluidizer (M-110L, Microfluidics Corp., MA). After centrifugation at 12 000g for 1 hour, the supernatant was applied to an Ni-column to extract the His6-tagged protein. The protein was cleaved from the column with thrombin, concentrated on a 30 kDa cut-off membrane, and purified on a gel filtration column (Superdex200) using 25 mM Tris (pH 7.7), 150 mM NaCl as running buffer.

Alternatively, recombinant LLO from Abcam (Cambridge, UK) was used (residues 60-529, product code ab83345).

A similar method was used to obtain pneumolysin (PLY)²². An Escherichia coli colony expressed the His6 tagged version of the protein by cloning it into the plasmid pET15b. The cell culture was grown overnight at 37°C. Transfer to selective medium containing 50 μg/ml ampicillin ensured that only plasmid-containing cells survived. When an A600 between 1 and 1.5 was reached, the temperature was lowered to 30°C to reduce cell proliferation. The addition of 0.5 mM isopropyl-β-D-1-thiogalactopyranoside induced protein expression. After overnight expression, cells were harvested, suspended in lysis buffer and disrupted using a microfluidizer. The cells were then centrifuged at 185 000g for 1 hour and the supernatant was applied to a HisTrap FF column. Using a lysis buffer containing 500 mM imidazole to displace the histidine tag from its Ni+ binding sites, the protein was eluted from the column. The protein fractions were diluted with a 50 mM Tris pH 7.0, 5 mM β-mercaptoethanol solution. Overnight incubation with 100 units of thrombin at 4°C removed the His tag. Passing the protein solution through a HiTrap Q FF using a salt gradient of 0 to 1 M NaCl in a 50 mM Tris, pH 7.0, 5 mM beta-mercaptoethanol buffer resulted in a purified sample. The protein solution was not frozen but kept at 4°C to avoid precipitation.

4.2 Static Secondary Ion Mass Spectrometry (Static SIMS)

Electrospray prepared targets were analyzed using time-of-flight secondary ion mass spectrometry (TOF-SIMS) I and II instruments^25–27. The TOF-SIMS I instrument had a pulsed Ar⁺ ion source. The ions struck a grounded metal target. 8 keV Ar⁺ ions were focused to a 1 mm spot with a primary ion density of 10¹⁰ Ar⁺ ions per (cm² • second) (static SIMS conditions). Secondary ions were extracted from the grounded target and allowed to drift into two time-of-flight regions connected by a 163° angular and energy focusing toroidal Poschenrieder condenser lens. A single ion counter detector was used. Analytical ions were post-accelerated before being converted to electrons in a channel plate, which excited a scintillator. A photomultiplier counted the photon pulses. For the TOF SIMS II instrument, the Poschenrieder lens was replaced by an energy-compensating reflectron, which increased the mass resolution to more than 10 000. Improvements in the primary Ar⁺ ion source were important for this study. The ion current was more stable than in the TOF-SIMS I instrument, and the diameter of the ion beam on the target was 0.1 mm. The - for the time - small focus allowed spatially resolved scans of the target.

4.3 Electron Microscopy

Carbon-coated grids for electron microscopy were briefly exposed to gas discharge to render them hydrophilic. The on the liquid surface prepared layers were transferred to the electron microscopy grids by carefully lowering them onto the liquid meniscus. To extensively wash the specimens, they were exchanged several times with double distilled water. Then, the grids were negatively stained using 1% wt/vol uranyl acetate in preparation for electron microscopic inspection. Alternatively, the surface was covered with a 1.5 nm thick platinum/carbon layer through the use of an atomic beam angled at 15°. Inspection was conducted using an FEI Tecnai Spirit BioTWIN transmission electron microscope with varying magnifications of up to 150 000x, operating at 120 kV.

4.4 Electrospray based Surface Preparation

4.4.1 Electrospray Apparatus

An electrospray apparatus was constructed to uniformly dispense dissolved material onto a level surface ⁷ (see figure 1 for the second generation instrument). The first generation instrument contained a metallic nozzle with a diameter of 250 μm. The flow rate was determined by the equilibrium between the electrostatic traction on the liquid and the flow resistance within the device. In the initial instrument the height of an accurately fitted metal cylinder in the outlet was modified to regulate its flow resistance. Like this, flow rates ranging from 8 to 15 μL/min were attainable. After the initial experiments, the nozzle was substituted with a hand-pulled glass capillary coated with gold, which lowered the stable flow rate to below 500 nanoliters/min⁷. The liquid sample was sprayed under atmospheric conditions onto a flat metal target. By applying a high voltage, the liquid meniscus at the tip of the nozzle forms a Taylor Cone and emits a stream of highly charged, rapidly spreading small droplets from its tip⁸. The volatile solvent evaporates in flight, and the dried residues cover the target area in a circle approximately 4 mm in diameter. The distance between the nozzle and the target is adjustable.

For a second generation of the apparatus, the glass capillaries were made using a commercial capillary puller (model P-87 Puller, Sutter Instruments Company, Novato, CA, USA) and mounted in a pressurized container (figure 1). Prior to injection, the narrow glass tip was broken to provide an orifice approximately 1 μm in diameter²⁸. With this more mature device, the electrospray operates at stable and constant flow rates between 20 and 100 nL/min.

5.4.2 Surface Preparation

For the static secondary ion time-of-flight mass spectrometer (TOF-SIMS) study, a rhodamine/acetone solution was sprayed onto an aluminum target.

For the membrane preparation the solution was sprayed onto the liquid meniscus of a small 3 mm diameter cylindrical container filled with 10 mM Tris solution, pH 7.5, containing some SM-2 Biobeads (Bio-Rad, CA, USA). The container was connected to a larger buffer reservoir to compensate for evaporation losses. Like this the liquid meniscus was always high enough so that molecules, assembling a membrane on its surface, did not compete for space but could overflow.

Egg-phosphatidylcholine was purchased from Sigma-Aldrich.

The final membrane synthesis protocol for the incorporation of membrane proteins (like OmpG) consisted of five steps. The first material sprayed onto the buffer is the equivalent amount of a lipid bilayer of bipolar lipids dissolved in ethanol: 1 μL 66 pmol/μL lipid with 35% cholesterol, 65% phosphatidylcholine. This preparation is incubated overnight at room temperature to allow for bilayer assembly. In a second step, a thin layer of glycerol is added by spraying a 1:2 solution of glycerol/ethanol onto the meniscus for about 10 minutes. The third step is to spray enough lipid to form a monolayer: 1 μL 33 pmol/μL lipid with 35% cholesterol, 65% phosphatidylcholine in ethanol. In the fourth step, the sprayed membrane protein and lipid complete the molecular mixture to allow self-assembly of an intact, protein-containing lipid bilayer. The sprayed solution consisted of 25% glycerol to protect the protein from the high droplet charges when the water evaporates in flight: 1 μL 32 pmol/μL OmpG, 20 pmol/μL lipid 35% cholesterol, 65% phosphatidylcholine, 1% octylglucoside in 10 mM Tris, 25% glycerol. As a final step, another layer of glycerol is added to complete the hydrophilic environment: 1:2 glycerol/ethanol solution for approximately 10 minutes. The entire assembly is incubated for 6 days at room temperature to allow the SM-2 beads to extract the octylglucoside detergent from the surface layers.

The buffer solution used for the preparation of pneumolysin and listeriolysin O, proteins that are soluble in water by themselves but form a non-covalent intra-membraneous protein complex, contained 50 mM Tris and 150 mM NaCl. The pH value for listeriolysin O solution was adjusted to 5.5, for the pneumolysin solution to 7.0. The initial step in forming a protein membrane layer with initially soluble proteins involves spraying 1 μL of a 66 pmol/μL lipid solution, containing 35% cholesterol and 65% phosphatidylcholine in ethanol, to allow for the formation of a closed lipid bilayer through overnight incubation. Spraying a 1:2 solution of glycerol/ethanol for about 10 minutes generates a thin layer of glycerol, creating a hydrophilic base for the subsequent assembly of the protein complexes. The following step involves the addition of either a lipid monolayer or a lipid bilayer. The bilayer was incubated overnight to complete its formation. At this point, the surface is ready for the addition of the protein-containing film. The pneumolysin solution had 26 pmol/μL of protein in 50 mM Tris, 150 mM NaCl, at a neutral pH. The listeriolysin O solution had a concentration of 20 pmol/μL, dissolved in 50 mM Tris, 150 mM NaCl, at pH 5.5. The solution that was sprayed had an additional component of 25% glycerol. When the surface consisted solely of a lipid monolayer, the protein solution provided enough lipid to complete the bilayer: 20 pmol/μL lipid, with 35% cholesterol and 65% phosphatidylcholine. About 1 microliter of the final solution distributed over the 3 mm liquid meniscus generates the protein-containing layer. Finalizing the hydrophilic environment involves spraying a 1:2 glycerol/ethanol solution for 10 minutes. The self-assembly process continues during the incubation period of several days at room temperature.

4.5 Use of Artificial Intelligence

DeepL was used to bring the writing style of the manuscript closer to that of a native speaker (https://www.deepl.com/write).

OmpG, outer membrane protein G; E. coli, Escherichia coli; TOF-SIMS, Time of Flight - Secondary Ion Mass Spectrometry.

Acknowledgment

The first experiments leading to the electrospray theory and the construction of the first nanoelectrospray were carried out in the laboratory of Prof. A. Benninghoven at the University of Münster, Germany.

The experimental confirmation that the nanoelectrospray works as a molecular beam device for molecules of almost size was obtained in the laboratory of Matthias Mann at the European Molecular Biology Laboratory (EMBL) in Heidelberg, Germany.

The membrane protein experiments were carried out in the laboratory of Prof. W. Kühlbrandt at the Max Planck Institute of Biophysics, Frankfurt, Germany. I am grateful to Prof. Kühlbrandt for providing the necessary support to carry out all these experiments. I thank Dr. Oezkan Yildiz and Dr. Katharina von Pee for their intellectual and material support. Without the help of Deryck Mills with the electron microscope and with technical questions, the project would never have been completed.

The work of Matthias Wilm was supported by the Science Foundation Ireland, SFI grant 07/SK/B1184c.

Authors' contributions

The author did all the experiments and cited calculations relating to the electrospray theory himself.

Competing Interests Statement

The author holds an European patent for the preparation method for bioactive membranes, EP3711848.

Animals and Humans

No experiments on animals had been conducted. No humans were part of the experiments.

Funding:

The work was supported by Science Foundation Ireland, SFI grant 07/SK/B1184c

Data Availability

All relevant electron microscopic images are part of the article or its supplemental information. The overlap of the OmpG structure model derived from its crystal structure with the electron microscopic image has been made available via figshare (https://doi.org/10.6084/m9.figshare.25908022) and the movie is part of the supplemental files.

Wilm, M. Synthesis of Extended, Self-Assembled Biological Membranes containing Membrane Proteins from Gas Phase. bioRxiv 23, 661215 (2019). Preprint at https://www.biorxiv.org/content/10.1101/661215v2
Yang, Z. et al. Electrospray-assisted cryo-EM sample preparation to mitigate interfacial effects. Nat Methods (2024). https://doi.org/10.1038/s41592-024-02247-0
Giess, F., Friedrich, M. G., Heberle, J., Naumann, R. L. & Knoll, W. The Protein-Tethered Lipid Bilayer: A Novel Mimic of the Biological Membrane. Biophysical journal 87, 3213-3220 (2004).
De Zorzi, R., Nicholson, W. V., Guigner, J.-M., Erne-Brand, F. & Vénien-Bryan, C. Growth of large and highly ordered 2D crystals of a K⁺ channel, structural role of lipidic environment. Biophysical journal 105, 398-408 (2013).
Denisov, I. G. & Sligar, S. G. Nanodiscs in Membrane Biochemistry and Biophysics. Chemical reviews 117, 4669-4713 (2017).
Jaworek, A. Electrospray droplet sources for thin film deposition. Journal of Materials Science 42, 266-297 (2007).
Wilm, M. & Mann, M. Electrospray and Taylor-Cone theory, Dole’s beam of macromolecules at last? International Journal of Mass Spectrometry and Ion Processes 136, 167-180 (1994).
Wilm, M. Principles of electrospray ionization. Molecular & cellular proteomics : MCP 10, M111.009407 (2011).
Benninghoven, A., Hagenhoff, B. & Niehuis, E. Surface MS: Probing Real-World Samples. Analytical chemistry 65, 630A-640A (1993).
Wilm, M. & ter Meer-Müller-Steinmeister and Partner, E. P. A., Artur-Ladebeck-Str. 51, D-33617 Bielefeld. Das Micro-Elektrospray, (1991).
CKMeng, Mann, M. & Fenn, J. B. Of protons or proteins. Zeitschrift für Physik D Atoms, Molecules and Clusters 10, 361-368 (1988).
Fenn, J. B., Mann, M., Meng, C. K., Wong, S. F. & Whitehouse, C. M. Electrospray ionization for mass spectrometry of large biomolecules. Science (New York, NY) 246, 64-71 (1989).
Fenn, J. Ion formation from charged droplets: roles of geometry, energy, and time. Journal of the American Society for Mass Spectrometry (1993).
Mann, M. & Wilm, M. Electrospray mass spectrometry for protein characterization. Trends in biochemical sciences 20, 219-224 (1995).
Yildiz, O., Vinothkumar, K. R., Goswami, P. & Kühlbrandt, W. Structure of the monomeric outer-membrane porin OmpG in the open and closed conformation. The EMBO journal 25, 3702-3713 (2006).
Subbarao, G. V. & van den Berg, B. Crystal Structure of the Monomeric Porin OmpG. Journal of molecular biology 360, 750-759 (2006).
Huang, C. & Mason, J. T. Geometric packing constraints in egg phosphatidylcholine vesicles. Proceedings of the National Academy of Sciences of the United States of America 75, 308-310 (1978).
Zasadzinski, J. A., Viswanathan, R., Madsen, L., Garnaes, J. & Schwartz, D. K. Langmuir-Blodgett films. Science (New York, NY) 263, 1726-1733 (1994).
Köster, S. et al. Crystal structure of listeriolysin O reveals molecular details of oligomerization and pore formation. Nature communications 5, 3690 (2014).
Sonnen, A. F.-P., Plitzko, J. M. & Gilbert, R. J. C. Incomplete pneumolysin oligomers form membrane pores. Open biology 4, 140044-140044 (2014).
Mulvihill, E., van Pee, K., Mari, S. A., Muller, D. J. & Yildiz, O. Directly Observing the Lipid-Dependent Self-Assembly and Pore-Forming Mechanism of the Cytolytic Toxin Listeriolysin O. Nano letters 15, 6965-6973 (2015).
van Pee, K., Mulvihill, E., Muller, D. J. & Yildiz, O. Unraveling the Pore-Forming Steps of Pneumolysin from Streptococcus pneumoniae. Nano letters 16, 7915-7924 (2016).
Ilangumaran Ponmalar, I., Sarangi, N. K., Basu, J. K. & Ayappa, K. G. Pore Forming Protein Induced Biomembrane Reorganization and Dynamics: A Focused Review. Front Mol Biosci 8, 737561 (2021).
Parsons, E. S. et al. Single-molecule kinetics of pore assembly by the membrane attack complex. Nat Commun 10, 2066 (2019).
Niehuis, E., Heller, T., Feld, H. & Benninghoven, A. Design and performance of a reflectron based time‐of‐flight secondary ion mass spectrometer with electrodynamic primary ion mass separation. Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films 5, 1243-1246 (1987).
Benninghoven, A., Vickerman, J. C. & Briggs, D. (eds) TOF-SIMS: Materials Analysis by Mass Spectrometry. Chichester, Manchester,
Benninghoven, A. 64-89 (Springer, Berlin, Heidelberg, Berlin, Heidelberg, 1983).
Wilm, M. & Mann, M. Analytical properties of the nanoelectrospray ion source. Analytical chemistry 68, 1-8 (1996).
Yilmaz, N. & Kobayashi, T. Visualization of Lipid Membrane Reorganization Induced by a Pore-Forming Toxin Using High-Speed Atomic Force Microscopy. ACS nano 9, 7960-7967 (2015).

Competing interest reported. The author holds an European patent for the preparation method for bioactive membranes, EP3711848.

Download PDF

Editorial decision: Revision requested
02 Sep, 2024
Reviews received at journal
12 Aug, 2024
Reviewers agreed at journal
31 Jul, 2024
Reviews received at journal
30 Jul, 2024
Reviewers agreed at journal
29 Jul, 2024
Reviewers invited by journal
25 Jul, 2024
Editor assigned by journal
25 Jul, 2024
Editor invited by journal
22 Jul, 2024
Submission checks completed at journal
19 Jul, 2024
First submitted to journal
11 Jul, 2024

You are reading this latest preprint version

Nanoelectrospray based synthesis of large, transportable membranes with integrated membrane proteins

Status:

Version 1

Abstract

Figures

1 Introduction

2 Results