Background

The fundamental building blocks for biomembranes are phosphoglycerides. Their low solubility and amphiphilic nature enable these lipids to self-assemble into various structural mesophases including bilayers, the basic barriers that compartmentalize cells. Cells take advantage of membrane bilayers to form internal and external barriers, as well as to carry out transport of materials within and between cells. The latter process involves generation of bilayer vesicles to act as delivery shuttles, by budding from and fusing with different membranes. Such functionality is critically important during periods of cell growth, development, and proliferation for recycling and re-utilization of biomolecules during programmed cell death processes, and for neurotransmission. To gain insights into the physicochemical properties that enable cell membrane functionality, studies of model bilayer membranes produced in vitro, i.e., liposomes, have proven invaluable. The studies have also led to biotechnological advances that include the engineering of liposomes for delivery of drugs, genes, and vaccines, as well as for applications in medical diagnostics, cosmetics, and food industry-related nutraceutics and dietetics (Lichtenberg and Barenholz, 1988; Laouini et al., 2012; Akbarzadeh et al., 2013; Keller et al., 2013; Patil and Jadha, 2014; Nkanga et al., 2019; Maja et al., 2020; Ajeeshkumar et al., 2021).

Over the years, a variety of methodologies have been used to generate bilayer liposomes.Generally, approaches that result in homogeneous populations of single-wall, i.e. unilamellar, bilayer vesicles are preferred to mimic the situation generally encountered in cells, and to optimize trapping efficiency using the least amount of lipid. In biomedical research, approaches involving ultrasonication, reverse-phase evaporation, and extrusion have been widely used. Less widely used are ethanol injection, detergent dialysis/removal, and supercritical fluid approaches. Nearly all of the approaches are either tedious, time-consuming, require specialized equipment, and produce vesicles with limited shelf-life at either room temperature or in cold storage.

We sought a straightforward approach for fast and easy construction of homogeneous unilamellar bilayer vesicles comprised of 1-palmitoyl-2-oleoyl phosphatidylcholine (POPC). We turned to bicelle-based approaches because of insights provided by theoretical and molecular dynamics modeling of mechanistic pathways, involving bicelle disc transformation into unilamellar bilayer vesicles (Fromherz, 1983; Marrink and Mark, 2003; Chng, 2013). These studies suggest that loss of rim-stabilizing short-chain PC (e.g., dihexanoyl PC; DHPC) or detergents from bicelles will lead to their coalescence into larger-diameter discs, followed by the formation of cup-shaped intermediates to reduce long-chain PC hydrocarbon chain exposure to water. With further departure of short-chain DHPC or detergent from the rim areas, the cup-shaped hemi-vesicle intermediates are predicted to undergo rim closure to form unilamellar vesicles. From a thermodynamic standpoint, the loss of DHPC or detergent from the bicelle rims alters the energetic balance between the rim perimeter line tension and the bicelle disc elastic bending energy, with the perimeter line tension becoming more dominant, as DHPC or detergent is lost from the bicelle rim. This situation promotes formation of the cup-shaped hemi-vesicle intermediates that further transition to unilamellar vesicles when nearly complete loss of rim-stabilizing DHPC occurs.

In the majority of previous experimental studies involving bicelles, the long-chain phosphoglyceride component contains saturated acyl chains, e.g., dimyristoyl phosphatidylcholine (DMPC) or DMPC/dimyristoyl phosphatidylglycerol mixtures (DMPG). These lipid species are generally absent or very minor components of eukaryotic biomembranes. Typically, biomembrane phosphoglyceride acyl composition is asymmetric, with the sn1-chain being long and saturated (C ≥ 16), while the sn2-chain is long (C ≥ 18) and contains cis unsaturation, as occurs in POPC (Figure 1). This arrangement has interesting biological consequences, including more efficient membrane vesiculation, and lower membrane permeability compared to phosphoglycerides containing two saturated or polyunsaturated chains (Manni et al., 2018).

Figure 1. Chemical structures of phosphatidylcholines used or discussed in this study.

From previous studies, it is known that the structural stability of DMPC/DHPC bicelles is critically dependent upon the total concentration of long-chain PC+DHPC/detergent (Sanders and Schwonek, 1992; Sanders and Prosser, 1998; Harroun et al., 2005; Lu et al., 2012; Beaugrand et al., 2014). Dilution of bicelles after formation can alter the long-chain PC-to-DHPC/detergent ratio (q-value) and drive mesomorphic structural change, including transformation to liposome bilayer vesicles (Nieh et al., 2004, 2009 and 2011; Yue et al., 2005; Mahabir et al., 2013). The dilution-driven structural transformation is a consequence of the vastly differing aqueous solubilities of the long-chain PC versus the DHPC/detergent lipids forming the bicelle. In the case of DMPC/DHPC mixtures, the aqueous solubilities of DMPC (~6 nM) and DHPC (~16 mM) differ by 2.7 million-fold. Thus, when the bicelle concentration is suddenly decreased and becomes too low (as can occur with aqueous dilution), DHPC moves from the bicelle into aqueous phase. This situation changes the q-value and alters the structural form of the bicelle lipid mixture. This response is accounted for by q*, through provision of the effective DMPC/DHPC molar ratio for bicelle mixtures at various lipid concentrations (Beaugrand et al., 2014). For example, dilution of 1.0 q-value DMPC/DHPC mixtures from 75 to 25 mM results in q increasing from 1.0 up to a q* of 1.8, where isotropic bicelles still prevail. Upon dilution to 16 mM total PC, a 3.2 q*-value results in the prevalence of large bicelles. Further dilution to 2 mM total PC leads to an extremely high q*, where small DMPC liposome vesicles become detectable by ³¹P-NMR.

From a practical standpoint, whether the preceding modeled and experimental behavior that characterizes DMPC/DHPC and DMPC/DMPG/DHPC bicelle lipid mixtures can yield a simple and easy method for producing homogeneous populations of POPC unilamellar bilayer vesicles has remained unclear. Recently, a straightforward approach that streamlines production of POPC vesicles from POPC/DHPC bicelle mixtures has been used in applications for monitoring protein-mediated intervesicular lipid transfer activity (Gao et al., 2020). Characterization has revealed these POPC vesicles to be small (32-36 nm diameter), unilamellar, bilayer vesicles that are surprisingly homogeneous, stable, resistant to freeze-thaw alteration, and usable in assays involving cell cytosolic fractions. The pathway for vesicle formation is illustrated in Figure 2 and production details follow:

Figure 2. Generation of small, homogeneous single-bilayer POPC vesicles by dilution of POPC/DHPC bicelle mixtures. The phosphorylcholine polar head groups are shown as yellow ellipsoids with the long acyl chains of POPC depicted in gray (stick fashion) and the short acyl chains of DHPC depicted in cyan (stick fashion). As noted, dilution (>125-fold) of the 0.5 q-value POPC/DHPC bicelle mixture with aqueous buffer initially leads to larger bicelle formation (not shown), followed by transition to unsealed cup-shaped hemi-vesicles (not shown), that seal to form stable unilamellar vesicles. Illustrations of these short-lived transitional intermediates are depicted in Fromherz (1983), Chng (2013), or Marrink and Mark (2003). The bicelles (left) and vesicle (right) are shown in two-dimensional cross-section.