Since its inception more than thirty years ago, electroformation has become the most commonly used method for growing giant unilamellar vesicles (GUVs). Although the method seems quite straightforward at first, researchers must consider the interplay of a large number of parameters, different lipid compositions, and internal solutions in order to avoid artifactual results or reproducibility problems.
1. Introduction
Artificial vesicles have become an important research tool due to their similarity to biological membranes [1][2][3][4][5][6]. Being lab-created, they enable the study of membrane properties under controlled conditions. When mimicking the biological membrane, we are interested mainly in unilamellar vesicles (only one outer bilayer), but multilamellar (bilayers arranged in concentric circles) and oligolamellar vesicles (containing smaller ones inside) can also be created. Depending on their size, unilamellar vesicles are commonly divided into three groups: small (SUVs, <100 nm), large (LUVs, 100 nm–1 μ m), and giant unilamellar vesicles (GUVs, >1 μ m). SUVs and LUVs are more often studied in the context of drug delivery applications [7][8][9]. GUVs are more useful as artificial cell models for eukaryotic cells due to similarity in size. Additionally, their size enables observation of membrane domain structure using light microscopy.
2. Classic Electroformation Protocol
One of the earliest attempts at forming GUVs was the natural swelling method introduced by Reeves and Dowben in 1969 [3]. According to this method, a lipid solution is deposited on a surface and dried to form a lipid film. The lipid film is then rehydrated and the obtained solution gently stirred to form vesicles. The vesicles are formed mainly due to the osmotic pressure driving the aqueous solution in between the stacked lipid bilayers. Exposing the hydrophobic portion of the bilayer to aqueous solutions is unfavourable and causes them to close up into vesicles. However, the proportion of GUVs that can be generated using this method is small, as most of them are either multilamellar or display other types of defects [10]. In their efforts to devise a protocol that reliably produces a high proportion of cell-sized unilamellar vesicles, Angelova and Dimitrov applied an external electric field during lipid swelling and thus invented the electroformation method [11]. Although the exact mechanism of the method is not yet completely understood, it is believed that the electric field affects lipid swelling through direct electrostatic interactions, redistribution of counterions, changes in membrane surface and line tension, and electroosmotic flow effects [12]. More detailed theoretical discussions on the electroformation mechanism can be found elsewhere [13][14][15].
Although Angelova and Dimitrov used platinum wires for electroformation, indium tin oxide (ITO) electrodes are most commonly used nowadays, so we utilize them in our description of the basic protocol. The only difference between these two protocols is the electroformation chamber layout. The protocol starts with droplets of lipids dissolved in an organic solvent being deposited onto the electrode ( Figure 1a). In addition to lipids, fluorescent dyes are present in the mixture in small quantities to enable the usage of fluorescent microscopy later on.
The solvent is evaporated under vacuum or a stream of inert gas ( Figure 1b). A spacer is attached to the electrode using vacuum grease ( Figure 1c). Another electrode is then attached to the spacer with its conductive side facing inward. Following that, the chamber is filled with an internal solution of choice and connected to a voltage source while maintaining a temperature higher than the phase transition temperature of deposited lipids. Copper tape is often attached to the electrodes to provide better contact with the alternating current function generator ( Figure 1d). The combination of the electric field and lipid film hydration leads to the creation of lipid vesicles which can be observed under a microscope ( Figure 1e).
Figure 1. (a) Deposition of lipid droplets onto the electrode surface. (b) Evaporation of organic solvent under vacuum. (c) Construction of the electroformation chamber. (d) Electroformation chamber filled with an internal solution and connected to an alternating current function generator. (e) An image of fluorescently labeled giant unilamellar vesicles (GUVs) obtained using fluorescence microscopy. The scale bar denotes 50 µm.
At first, direct current was used, but water electrolysis led to formation of bubbles [11], so a transition to alternating currents was made [11][12]. Alternating currents also introduce electroosmotic motion of the fluid, which facilitates the destabilization of the lipid film, thus promoting the formation of vesicles [12][16]. Over the years, the initial protocol has been modified in order to increase the yield, homogeneity, and compositional uniformity of the vesicles or to enable preparation of GUVs from previously incompatible lipids and buffers.
3. Electrode Materials and Cleaning
Alongside platinum electrodes, ITO-coated glass slides are most commonly used nowadays. In comparison to platinum electrodes, ITO electrodes provide a larger and flatter surface, so a higher GUV yield can be obtained. Also, owing to their transparency, microscopic techniques are easier to apply. It is recommended to periodically replace ITO electrodes since they were proven to have limited reusability. Using the same ITO glass more than three times has been shown to decrease the average GUV diameter and their quality. The effect seems to be less pronounced for zwitterionic and negatively charged lipids. Furthermore, the degradation seems to be reversible through annealing at high temperatures
[17].
Compared with ITO electrodes, platinum electrode aging seems to affect the average diameter of vesicles less, but the proportion of GUVs (i.e., the proportion of defect-free unilamellar vesicles) also drops. Electrode replacement or annealing is recommended after approximately five experiments
[18]. Steel syringes and copper electrodes have also been suggested as cost-friendly alternatives
[19][20]. Furthermore, titanium electrodes have been advocated by some groups since their usage seems to decrease lipid peroxidation when compared to ITO electrodes
[21][22][23][24].
In addition to changing the electrode material completely, various modifications to the existing electrode layouts were tested. Okumura and Sugiyama confirmed that GUVs can form on non-electroconductive materials, such as polymer meshes placed onto ITO electrodes and filled with lipid solutions
[25][26]. Lefrançois et al. experimented with an asymmetrical ITO electrode layout (top electrode with a smaller surface area than the bottom one)
[27] (
Figure 2a). This layout proved to be more efficient when physiological salt concentrations were combined with interelectrode separations smaller than 2 mm. Small interelectrode separation did not cause problems when the same parameters were set, but deionized water was used as an internal solution. The effect appears because alternating-current electroosmosis assisted flow does not occur in the symmetrical configuration. However, since both symmetrical and asymmetrical layouts were tested under the same electrical parameters, it is not clear whether GUVs could be grown under physiological conditions for smaller electrode separations just by altering the electrical parameters and retaining the symmetrical ITO chamber layout. Bi et al. abandoned the principle of parallel opposing electrodes altogether and have shown that GUVs can be grown using coplanar interdigitated ITO electrodes
[28] (
Figure 2b).
Figure 2. (a) Asymmetrical electrodes layout. The top indium tin oxide (ITO) electrode has a smaller surface than the bottom one, so it has to be surrounded by glass coverslips in order to close off the chamber. (b) Electroformation chamber with a coplanar interdigitated ITO electrode.
In order to remove contaminants from the electrode surface, researchers use a variety of cleaning protocols. In general, these consist of cleaning the electrodes using organic solvents and then drying them. Of course, there are many variations of these general steps, and some articles do not even include the cleaning protocol
[29][27][30]. Variations include sonication in conjunction with organic solvents
[15][28][31][32], repeated rinsing with organic solvents
[33], and swabbing electrodes using lint-free wipes
[34][35][36][17]. Electrodes are either dried under a stream of inert gas
[15][17][28][32] or just wiped and air dried
[36]. Plasma cleaning has also been tested on ITO glass
[15][28] and has proved to be very effective, since it both cleans the surface and makes it more hydrophilic. Moreover, plasma cleaning could aid hydration of the solid lipid film and the formation of lipid bilayers
[15].
4. Deposition of the Lipid Solution and Removal of the Solvent
In the early days of electroformation, the lipids dissolved in chloroform or other organic solvents were deposited onto the electrodes simply by dropping the solution and evaporating the solvent later
[11][37] (
Figure 3a). In order to efficiently produce GUVs, the lipid film is suggested to be around 5–10 bilayers thick (around 30–60 nm)
[11][32]. When using drop deposition, regions of different thicknesses will be formed. Although this means that at least some portion of the deposited film will be suitable for electroformation at given parameters, other regions will inevitably be too thin or too thick. These nonuniformities lead to lack of reproducibility.
Figure 3. (a) Droplet deposition. (b) Droplet deposition with smearing afterward to better spread the lipid film. (c) Deposition of lipids by pressing a patterned silicon stamp on the electrode surface. (d) Spin-coating of lipid solution by fast rotation of the electrode immediately after the deposition.
In order to make the deposition process more efficient and reproducible, different approaches were investigated as well. Some groups used needles or thin rods to smear the solution after dropping it in order to increase the homogeneity of the film
[34][15][38][39] (
Figure 3b). Others used a Hamilton syringe to deposit non-overlapping snakelike patterns of lipids
[17]. A more systematic approach was tested by stamping the lipid solution onto the ITO electrode using a customized polydimethylsiloxane (PDMS) stamp
[40] (
Figure 3c). This approach resulted in GUVs of a size similar to the width of the gaps in the PDMS stamp. However, the authors reported that the thickness of the lipid film was not uniform over the lipid patches in the gaps. Dip coating (immersing the ITO electrode in the lipid solution and holding it vertically to dry) was explored as well, but due to different rates of solvent evaporation across the glass, the film proved to be inhomogeneous
[32].
Reproducibly controllable lipid film thickness was achieved by Estes and Mayer using the spin-coating method. The lipid solution is dropped onto the flat electrode surface, which is subsequently rotated very fast in order to achieve a homogenous film
[32] (
Figure 3d). The uniformity of films and method reproducibility were confirmed using ellipsometry and atomic force microscopy techniques. The method was proven to be effective by several groups using a wide range of lipid compositions
[36][32][33]. There are, however, disadvantages as well. First, it requires a much larger amount of lipids compared with traditional methods, since much of the solution is washed away during electrode rotation
[32]. Second, it involves a spin-coater, which is not a standard device in labs investigating biological membranes.
Another attempt at achieving uniform lipid films was made by Le Berre et al.
[41]. They drag-dropped an organic solution of lipids on a solid substrate. Constant substrate velocity and temperature were maintained while simultaneously controlling the vapor aspiration. Reproducible results with variations of ±5 nm were obtained, but the method has not been widely adopted, probably due to the device used being relatively complicated.
Before chamber construction, the organic solvent needs to be removed from the deposited lipid solution. This is most commonly achieved by placing the electrodes under a vacuum
[4][34][36][15][17][32][38]. Depending on the research group, the vacuum duration can range from just 5 min
[38] to 2 h
[15]. Alternatively, drying can be performed by placing the electrode with the lipid solution under a stream of an inert gas, such as nitrogen
[11][33][37] or argon
[22]. More rarely encountered, lyophilization can also be used to eliminate traces of the organic solvent
[42][43].
Some researchers went in another direction and instead of trying to improve the existing approach, replaced the organic solution of lipids with aqueous liposome suspensions (SUVs, LUVs, or multilamellar vesicles)
[4][29][31][44][45][46][47]. Using this approach, Pott et al. concluded that GUV formation was better when using aqueous liposome solutions than when using organic lipid solutions
[4]. They attribute this to the ability of such dispersions to produce well-oriented membrane stacks immediately after the evaporation of water. Although any excess water was completely removed from the deposits, an alternative approach was suggested in which the deposits would be only partially dehydrated prior to internal solution addition. The effect of using damp lipid films on GUV properties was explored in more detail by Baykal-Caglar et al.
[29]. By measuring their miscibility transition temperature, they have shown that such an approach produces more compositionally uniform populations of GUVs. This result is explained by a reduction in lipid demixing—a common artifact appearing when drying the lipid solution completely. The artifact is especially pronounced when using mixtures with a cholesterol content near to or above the maximum solubility threshold for that mixture
[42].
Furthermore, avoiding the dry phase and use of organic solvents benefits protocols aimed at protein incorporation into GUVs, since these steps damage the protein structure. The first proof of successful protein reconstitution using this approach came from Girard et al.
[47] and the most recent protocol is described by Witkowska et al.
[31].
Since the deposition of aqueous liposome suspensions was a significant deviation from the classic protocol by itself, until recently, no one addressed the still existing problem of nonuniform film thickness. In 2019, Oropeza-Guzman et al. suggested using the coffee ring effect to solve this issue
[45]. The effect describes a phenomenon in which a drying droplet deposits most of its material on the periphery, forming a ring-like stain. The group used the effect to their advantage by consecutively depositing progressively larger droplets on top of one another. Since a larger droplet has more material, it will create a larger diameter ring and, in the process, will smear and flatten the ring from the previous droplet, thus leaving an area of uniform lipid thickness inside (
Figure 4). Although the mean diameter of vesicle populations hasn’t significantly changed when compared to the single droplet deposition, the multi-droplet preparations displayed a much lower percentage of nonunilamellar vesicles. These results combined with the low lipid mass used and a relatively simple experimental setup make the method a promising option for reproducible and uniform lipid deposition.
Figure 4. Deposition of lipids utilizing the coffee ring effect. After drying, most of the material is carried away toward the periphery and a ring-like stain is formed. This is known as the coffee-ring effect. By depositing progressively larger droplets, the ring from the previous droplets gets smeared and flattened, thus leaving behind an area of uniform lipid film thickness.
5. Temperature and Electroformation Duration
Temperature is an important parameter in electroformation protocols since it governs the bilayer gel to liquid phase transition (melt/transition temperature). Moreover, a continuous increase in temperature (as opposed to a thresholding effect of the transition temperature) seems to be accompanied by an increase in the final diameter of produced GUVs
[15][28][38]. A probable explanation is that the temperature enhances the permeability of the membrane to the solvent through an increase in membrane fluidity
[48]. Although increasing the GUV diameter is a desirable property, there is a disadvantage to increasing the temperature too much since prolonged exposure to high temperatures leads to increased lipid breakdown
[34].
In order for the bilayer to be in the liquid phase and adequate mixing of components to be achieved, the electroformation temperature is usually kept above the transition temperature of the lipid with the highest transition temperature in the mixture. In some cases, the miscibility transition temperature can even exceed this highest transition temperature, so an even higher value is sometimes needed
[34]. However, high temperatures can be harmful in scenarios such as those involving protein reconstitution. Regarding this issue, a study using a DOPC (1,2-dioleoyl-
sn-glycero-3-phosphocholine)/sphingomyelin/Chol (2:2:1 molar mixture) compared the properties of GUVs grown at room temperature and 65 °C
[49]. The proportion of liquid ordered–liquid disordered (Lo + Ld) phase separated vesicles was higher in the higher temperature batch. Nevertheless, the physical properties of vesicles that were phase separated were similar for both temperatures.
The rate of sample cooling after electroformation plays a significant role in the phase separation of membrane domains. First, sudden changes in temperature can break some vesicles
[34]. Second, if cooled too quickly from a fluid one-phase region to the gel–fluid phase, the vesicles seem to be caught in a nonequilibrium state where phase separation has not yet been achieved
[50][51]. On the other hand, in situations where the Ld phase is present in small amounts, cooling too slowly can lead to an artifactual decrease in the amount of Lo + Ld phase due to the Ld phase pinching off the parent GUV
[21].
The duration of electroformation is variable between research groups as well and usually ranges from 0.5 to 7 h, though it can last up to 24 h in some cases
[18]. The effect of prolonged electroformation duration (4–24 h) has been tested using fluorescence microscopy, flow cytometry, spectrofluorimetry, and colorimetric analysis to measure the diameter of vesicles, the proportion of oligolamellar vesicles, and the amount of lipid molecules in the GUV suspension after electroformation
[18]. Prolonged electroformation duration did not significantly impact any of these parameters. However, the study only used a monounsaturated phospholipid, which is poorly oxidizable. Another study involving three lipids with different degrees of unsaturation was also conducted
[52]. Using mass spectrometry and flow cytometry, a longer electroformation duration (5 vs. 20 h) was shown to induce lipid oxidation in more oxidizable lipids (two double bonds or more).
6. Vesicles Population Count and Size
Light scattering methods are a popular tool for determining the average diameter of particles suspended in a solution. For particles up to a diameter of around 1 μ m, methods such as dynamic light scattering can be used to easily obtain the average diameter value [53]. Since GUVs fall out of this size category, the most commonly used light scattering method is flow cytometry [18][52][54][55][56][57][58]. Studies have been performed to confirm the correlation between the forward scatter (FSC) and side scatter (SSC) signal intensity and particle size. For example, the correlation for smaller particles ( ~ 1 µm) has been confirmed using dynamic light scattering measurements [53] and for larger ones ( ~ 10 µm) by observing liposomes under a microscope after fluorescence-activated cell sorting [55]. Consequently, to obtain the average liposome diameter, sample FSC and/or SSC signals are measured and compared with those of calibration beads (usually polystyrene or silica) of known size.
The problem with this approach is that the FSC and SSC signals depend on the internal structure of the particles as well as their refractive indices, giving rise to a certain amount of error in obtained diameters [53][59]. Usually, researchers are content with obtained relative sizing but comment on the possibility of error due to the refractive index mismatch. Possible solutions have been offered to account for the difference. One study first obtained the average refractive index of particles of interest and then corrected the calibration beads measurements using the Mie scattering theory [59]. Another study took a more direct approach and proposed artificial liposomes as calibrators instead of polystyrene beads [54]. An additional detail which is easily missed is that all of the articles above use FSC and/or SSC to extract the dimensions of the particles. However, when particles are approximately the same size as the height of the laser beam (as is the case for most GUVs), pulse width sizing is probably a better approach for determining the size of vesicles as it does not depend on many of the factors plaguing light scattering approaches for size measurements [60][61].
Recently, imaging flow cytometry devices started being used for GUV quantification [57]. This unites the best of both worlds by combining the rapid analysis of thousands of particles using flow cytometry with the capability of image-based approaches to identify the results.
Coulter counters detect particles suspended in an electrolyte based on the change of impedance due to the passing of a particle through an aperture between the two electrodes [62]. In terms of potential for analysis of GUVs, the method has an advantage of being able to detect the size of particles as well as their number.
This entry is adapted from the peer-reviewed paper 10.3390/membranes11110860