1. Introduction
Vesicles are membrane models consisting of one or more lipid bilayers filled with aqueous solution. They are commonly used to investigate membrane properties under controlled conditions [
1,
2]. Depending on their structure, lipid vesicles can be unilamellar, multilamellar and oligolamellar. Unilamellar vesicles have only a single lipid bilayer, multilamellar vesicles (MLVs) contain multiple lipid bilayers arranged in concentric circles and oligolamellar vesicles contain smaller vesicles within the outer bilayer. With respect to their size, unilamellar vesicles are usually divided into three groups: small (SUVs) (< 100 nm), large (LUVs) (100 nm – 1 μm) and giant (GUVs) (> 1 μm). SUVs and LUVs are the most commonly produced by extrusion and often used for drug delivery studies or in protocols for formation of GUVs [
3] or supported lipid bilayers [
4]. Amongst these three groups, researchers studying membrane properties and organization are most interested in GUVs due their size similar to eukaryotic cells.
The first attempt to form GUVs was performed by Reeves and Dowben [
2]. In their approach, the lipid mixture is deposited on to the electrode and dried to form a dry lipid film that is rehydrated and the aqueous solution is driven between the lipid stacks due to osmotic pressure (gentle hydration). Because of the amphiphilic nature of lipids, it is unfavorable for the hydrophobic acyl chains of lipids to be exposed to the aqueous solution, so the lipid bilayers enclose into vesicles. Even though this method is straightforward and simple, it results in a low GUVs yield with many defects [
5].
One of the most commonly used methods for the production of GUVs today is electroformation, which was developed by Angelova and Dimitrov in 1986 [
6]. This method improved the previous approach by applying an external electric field to the lipid film in addition to hydration. In their protocol, lipids are dissolved in an organic solvent and deposited on an electrode. The organic solvent is then evaporated using a stream of inert gas and vacuum. The electrode with the lipid film is used to construct an electroformation chamber, which is filled with a desired aqueous solution and connected to an external alternating electric field. The electric field and osmotic pressure promote the detachment of the lipids from the electrode, leading to the formation of GUVs [
7]. Compared to samples obtained using the gentle hydration method, the vesicles formed by electroformation have a higher yield, a lower number of defects and a higher proportion of unilamellar vesicles [
5].
The electroformation method has evolved significantly since its inception, with various modifications to the protocols tested [
8,
9,
10] in order to improve the method’s reproducibility and enable compatibility with a wider range of lipid mixtures. Some groups attempt to improve the method by replacing the lipids dissolved in an organic solvent with an aqueous solution of SUVs or LUVs [
11]. It was concluded that the use of an unilamellar vesicles aqueous solution improved the efficiency of GUVs formation compared to deposition of lipids dissolved in an organic solvent. It was also shown that the formation of GUVs with buffers as internal solutions is easier when using buffer loaded SUVs or LUVs compared to the previous approach where the buffer would be applied to a dry lipid film [
11]. Another advantage of this modification is the possibility of improved proteoliposomes formation, due to reduced protein denaturation when the organic solvent is removed from the protocol [
11,
12,
13].
Another issue with the traditional protocol was the use of the drop-deposition method for lipid film deposition [
6,
7,
9]. The problem with this approach is formation of lipid films of nonuniform thickness resulting in high vesicle heterogeneity in the sample and low experimental reproducibility. Various alternative approaches have been tested to address this issue [
14,
15,
16]. The one we found to be optimal in terms of ease of use and quality of the final result is the spin-coating technique in which uniform lipid films are obtained by depositing the lipid solution onto a flat surface and spinning it at a high angular velocity [
15,
17,
18,
19,
20].
The initial electroformation protocol also turned out to be incompatible with lipid mixtures containing a high cholesterol (Chol) concentration. This is due to precipitation of Chol into anhydrous Chol crystals during the lipid film drying phase [
21]. When the film is rehydrated, these anhydrous Chol crystals do not participate in the bilayer formation, resulting in a lower Chol concentration in the vesicle bilayers compared to the Chol concentration in the initial lipid solution. Chol demixing can be avoided by using the rapid solvent exchange (RSE) method [
22,
23]. In this method, lipids dissolved in an organic solvent are first mixed with an aqueous solution. By suddenly decreasing the pressure, the organic solvent is evaporated, leaving behind an aqueous solution of MLVs. In order to avoid Chol demixing artifact during GUVs preparation, Baykal-Caglar et al. tested electroformation from a damp lipid film formed by depositing MLVs, produced by RSE method, onto the electrode [
3]. The final results were positive, indicating a higher Chol content in formed GUVs, compared to GUVs obtained using the original protocol. However, the described protocol is very long due to the long lipid film drying (22 - 25 h), and the MLVs were deposited using the drop-deposition method, inevitably resulting in a non-uniform film.
In addition to the deposition techniques and the properties of the lipid mixtures, the cleanliness of the electrodes has also been shown to have an influence on the electroformation results. Pretreatment of the electrodes with plasma improved the efficiency of the formation of GUVs containing buffers with physiological charge levels. This is probably due to the fact that the plasma makes the electrode surface more hydrophilic, facilitating the hydration of the lipid film and the subsequent formation of the lipid bilayers [
24].
In our previous study, we adapted the traditional electroformation protocol to incorporate all of the aforementioned improvements, allowing us to bypass the dry lipid film phase and produce a uniform damp lipid film (
Figure 1) [
25]. Inspired by the vesicle fusion method for the preparation of supported lipid bilayers [
26,
27,
28], the adapted protocol uses spin-coating to deposit an aqueous solution of LUVs onto a plasma treated electrode. The hydrophilized surface induces LUV rupturing, and the formation of lipid bilayers on the electrode surface. The bilayers later detach and form GUVs under the influence of osmotic pressure and an alternating electric field. With respect to the approach of Baykal-Caglar et al., these modifications significantly shorten the duration of the protocol while also increasing the experiment reproducibility. In order to make the protocol compatible with high Chol concentrations, the deposited LUVs were obtained by extruding a solution of MLVs produced by RSE method, and the lipid film was not dry after spin-coating.
The efficiency of the protocol was tested using 1/1 Chol/1-palmitoyl-2-oleoyl-glycero-3-phosphocholine (POPC) and 1/1/1 Chol/POPC/sphingomyelin lipid mixtures. Compared to the protocol involving a drying step, the new protocol resulted in a similar or better GUV’s yield, while having the potential to significantly reduce the Chol demixing artifact [
25]. In order to further increase the reproducibility and yield of the obtained samples, this study will focus on further optimization of the new protocol.
Special focus is placed on GUVs with a high Chol content for which the Chol demixing issue is more pronounced. Within the lipid bilayer Chol interacts with phospholipids and forms different domains depending on its membrane concentration. For vesicles with a Chol content lower than ~ 15 mol%, the lipid bilayer is in a liquid disorded phase. At 30 – 50 mol% it is in a liquid ordered phase. Cholesterol bilayer domains (CBDs) containing only Chol molecules with no phospholipids start forming when the Chol content is greater than 50 mol%. After reaching the Chol solubility limit at ~66 mol%, excess Chol precipitates in the form of Chol crystals [
29].
Modeling membranes with such a high Chol concentration is important for groups such as ours that study the properties of the eye lens plasma membranes [
21,
30,
31] or for groups performing atherosclerosis research [
32]. The protocol may also be beneficial for the formation of proteoliposomes by reducing the denaturation of proteins that occurs during the preparation of GUVs from lipids dissolved in an organic solvent or during film drying.
2.1. Materials
1-palmitoyl-2-oleoyl-glycero-3-phosphocholine (POPC) and Chol were purchased from Avanti Polar Lipids Inc. (Alabaster, AL, USA), while the fluorescent dye 1,1-dioctadecyl-3,3,3,3-tetramethylindocarbocyanine Perchlorate (DiIC
18(3)) was acquired from Invitrogen, Thermo Fisher Scientific (Waltham, MA, USA). Lipids were stored at −20 °C when not in use. ITO-coated glass (ICG-90 INS 115, resistance 70–100 Ω) measuring 25 × 75 × 1.1 mm was procured from Delta Technologies (Loveland, LO, USA). Fresh ITO-coated glass was employed for each preparation to ensure efficient formation of GUVs [
33]. Mili-Q deionized water (Merck, Rahway, NJ, USA), preheated to 60 °C, served as the internal chamber solution.
2.2. Preparation of Multilamellar Vesicels Using the Rapid Solvent Exchange Method
Initially, multilamellar vesicles (MLVs) were generated using a home built RSE device to circumvent the cholesterol demixing issue. A lipid mixture dissolved in chloroform was prepared by blending 25 mg/ml of POPC, 20 mg/ml of cholesterol (Chol), and 1 mg/ml of DiIC18(3) in appropriate proportions. The Chol/POPC mixing ratios were maintained between 0 and 2.5 (Chol mixing concentration of 0% to 71.4%), while the DiIC18(3)/POPC mixing ratio was fixed at 1/500. The total lipid mass amounted to 2.1 mg. Subsequently, 400 µL of Mili-Q water was introduced into the solution, and the mixture was vortexed (Vortex IR, Star Lab, Blakelands, UK) at an angular velocity of 2200 rpm. Throughout the vortexing process, the pressure was regulated to approximately 0.05 bar using a vacuum pump (HiScroll 6, Pfeiffer Vacuum, Asslar, Germany). The sample was maintained at the specified pressure for 90 seconds to ensure complete removal of residual organic solvent.
2.3. Preparation of Large Unilamellar Vesicles
The MLVs solution was extruded utilizing an Avanti Mini Extruder (Avanti Polar Lipids, Inc, Alabaster, AL.), passing through either a 50 or a 100 nm diameter pores of polycarbonate filters (Nuclepore Track-Etch Membrane, Whatman, UK) 15 times to achieve a uniform LUV suspension. To prevent loss of the lipid suspension, filters and membranes were pre-wetted with Mili-Q water. Additional Mili-Q water was added to the LUV solution to adjust a lipid concentration of 3.5 mg/ml.
2.4. Preparation of the Damp Lipid Film
Prior to conducting the experiments, the ITO-coated glass was submerged in Mili-Q deionized water. Subsequently, the glass was wiped using lint-free cloths saturated with 70% ethanol. Following this, the glass underwent treatment with oxygen plasma for a duration of 1 minute utilizing a plasma cleaner (PDC-002- HPCE with the PLASMAFLO PDC-FMG -2 attachment, Harrick Plasma, Ithaca, NY, USA) connected to a vacuum pump (HiScroll 6, Pfeiffer Vacuum, Assler, Germany).
Unless specified otherwise, a volume of 550 µl of LUV suspension was applied onto the hydrophilic plasma-treated ITO-coated glass electrode and promptly spin-coated using a spin-coater (SM -150, Sawatec, Sax, Switzerland). The electrode was spun at 600 rpm, achieving the final speed within 1 second. The duration of spin-coating, unless specified otherwise, was maintained at 30 seconds to guarantee the formation of a damp, uniform lipid film. To prevent unintended evaporation, the coated ITO-coated glass was transferred into a Petri dish and promptly utilized to assemble an electroformation chamber.
2.5. Electroformation Protocol
The electroformation chamber was constructed by sandwiching two 25 x 37.5 mm ITO-coated glasses with a 1.6 mm thick Teflon spacer in between. Electrodes were made by slicing a 25 x 75 mm ITO-coated glass using a diamond pen cutter. To form the electroformation chamber, the lipid-coated glass was sealed to the Teflon spacer with vacuum grease. Following the addition of 280 µl of Mili-Q water, the stopper was sealed using vacuum grease, ensuring no contact between grease and water to prevent any detrimental effects of grease contamination on GUV formation [
9]. The chamber was secured with clamps at three points along the electrodes, two adjacent to the stopper and one opposite. Subsequently, the chamber was connected to a pulse generator (UTG9005C, UNI - T, Dongguan City, China or PSG 9080, Joy- IT, Neukirchen-Vluyn, Germany) and placed in an incubator set to a temperature of 60 °C. Copper tape was applied to the outer edges of the electrode to enhance wire-electrode contact. Consistent with previous experiments [
17,
18,
25], a voltage of 2 V and a frequency of 10 Hz were maintained. The pulse generator after 2 hours was turned off, and the chamber remained in the incubator for an additional hour.
2.6. Dynamic light scattering
Dynamic light scattering (DLS) was utilized to determine the hydrodynamic diameter and polydispersity index of liposome suspensions (Litesizer 500, Anton Paar, Graz, Austria).
2.7. Fluorescence Imaging and Data Analysis
To cover the entire volume of the chamber, images were captured from 16 different regions of the sample. One hundred vesicles were then randomly selected from these images. In cases where the images did not encompass 100 vesicles, all observed vesicles were included in the analysis. Imaging was conducted utilizing a fluorescence microscope (Olympus BX51, Olympus, Tokyo, Japan), and vesicle diameters were measured using the line tool within the Fiji software [
34].
Unless specifically indicated otherwise, numerical results are presented as mean ± standard deviation. All data analysis and visualization were carried out using the R programming language [
35].
5. Conclusions
In this study, we optimized an improved electroformation protocol that bypasses the dry lipid film phase of the traditional electroformation method and combines the RSE method, the plasma cleaning, and spin-coating techniques to obtain a damp lipid film. To optimize the protocol further, we conducted additional experiments to test the effect of using different vesicle types during lipid film deposition, the effect of spin-coating duration, and the effect of different Chol concentrations. In order to confirm the utility of the optimized protocol, our samples were also compared to those obtained from fully dried lipid films.
Spin-coating duration of 30 s was found to be optimal in terms of balance between electroformation successfulness and lipid film dampness. A longer duration of spin-coating would dry out the lipid film and exacerbate the Chol demixing artifact. Compared to previous protocols that used damp lipid films, such as that of Baykal-Caglar et al, our method significantly shortened the preparation time by eliminating the 22-25 hour high-humidity drying phase and replacing it with 30 seconds of spin-coating [
3].
In terms of vesicle type used for film deposition, the LUV 100 vesicles gave the best results in terms of reproducibility, even though the average diameter was smaller compared to the MLVs and LUV 50 vesicles. The advantage of depositing MLVs instead of LUVs is the shorter protocol duration. However, samples made from MLV deposits displayed lower reproducibility and a more heterogeneous GUV size distribution. This is probably due to a larger average diameter and polydispersity index of MLVs compared to LUVs, as smaller vesicles rupture more easily compared to larger ones. The average diameter of GUVs was largest when LUV 50 vesicles were used, but the reproducibility of the samples was lowest. This is probably a side effect of additional extrusion performed compared to LUV 100 vesicles. Due to high flow resistance through a 50 nm filter, we could not extrude MLVs directly. We first produced LUV 100 vesicles and then passed them through a 50 nm filter. The additional extrusion step extends the duration of the protocol even further and causes a greater loss of final suspension volume, as part of the lipid suspension is always lost through liquid leakage during the extrusion process. This problem was even greater in our case as the resistance increased with increasing Chol/POPC mixing ratio.
Finally, our samples were also compared with those obtained from fully dried lipid films. Drying increased Chol demixing artifact, where higher amount of Chol crystals were found, which proves that the lipid film should never go through a drying phase when using mixtures with a high Chol content.
This protocol allows us to successfully prepare GUVs and study the physical properties, lateral organization and domain function of lipid membranes with a very high Chol content, such as the membranes that resemble composition of the fiber cell plasma membrane of the eye lens. Furthermore, the avoidance of organic solutions and plasma cleaning of the electrode have been shown to be advantageous for the preparation of GUVs containing charged lipids and buffer solutions. Consequently, we believe that our protocol might prove successful in those cases as well. Additionally, the protocol could also be adapted for protein-membrane interaction studies because protein denaturation is reduced by avoiding the dry film phase and the absence of organic solvents [
11,
12,
13].
Author Contributions
Conceptualization, I.M., M.R.; methodology, I.M., Z.B.; software, I.M, Z.B.; validation, I.M., Z.B.; formal analysis, I.M., Z.B.; investigation, I.M., Z.B.; resources, M.R.; data curation, I.M., Z.B.; writing—original draft preparation, I.M.; writing—review and editing, I.M., Z.B., M.R.; visualization, I.M., Z.B.; supervision, M.R.; project administration, M.R.; funding acquisition, M.R. All authors have read and agreed to the published version of the manuscript.