Peptide discs as precursors of biologically relevant supported lipid bilayers

Abstract

Supported lipid bilayers (SLBs) are commonly used to investigate the structure and dynamics of biological membranes. Vesicle fusion is a widely exploited method to produce SLBs. However, this process becomes less favoured when the vesicles contain complex lipid mixtures, e.g. natural lipid extracts. In these cases, it is often necessary to change experimental parameters, such as temperature, to unphysiological values to trigger the SLB formation. This may induce lipid degradation and is also not compatible with including membrane proteins or other biomolecules into the bilayers. Here, we show that the peptide discs, $~$10 nm discoidal lipid bilayers stabilized in solution by a self-assembled 18A peptide belt, can be used as precursors for SLBs. The characterizations by means of neutron reflectometry and attenuated total reflectance-FTIR spectroscopy show that SLBs were successfully formed both from synthetic lipid mixtures (surface coverage 90–95%) and from natural lipid mixtures (surface coverage $~85\%$). Traces of 18A peptide (below 0.02 M ratio) left at the support surface after the bilayer formation do not affect the SLB structure. Altogether, we demonstrate that peptide disc formation of SLBs is much faster than the SLB formation by vesicle fusion and without the need of altering any experimental variable from physiologically relevant values.

Introduction

A supported lipid bilayer (SLB) is a single lipid bilayer in the proximity of a hydrophilic solid support and separated from it by a thin water layer. Nowadays, SLBs are standard model systems for studying the structure, dynamics and function of biological membranes by a wide range of experimental techniques, such as atomic force microscopy, X-ray and neutron reflectometry, fluorescence microscopy, quartz crystal microbalance with dissipation monitoring and attenuated total reflectance FTIR spectroscopy [1], [2]. On top of their biophysical application, SLBs are also implemented in biosensor design [3], [4] as well as drug testing platforms [5]. The establishment and characterization of versatile SLB preparation protocols, which are compatible with complex lipid compositions and membrane proteins, are required for advancing in the biophysical and physico-chemical investigation of biological membranes as well as for their biotechnological and industrial applications [6].

Several protocols have been developed for SLB preparation including Langmuir-Blodgett (LB)/ Langmuir-Schaefer (LS) depositions [7] and vesicle fusion [8]. While LB/LS deposition requires specialised lab equipment, vesicle fusion simply exploits the spontaneous rupture and fusion of lipid vesicles on a hydrophilic support surface [9]. However, when vesicles are composed of charged lipids (e.g. phosphatidylserine (PS), lipids with high curvature (e.g. phosphatidylethanolamine (PE), sterols (e.g.cholesterol), or even more complex lipid mixtures (e.g. natural lipids extracted from yeast or bacterial cells), the vesicle fusion process becomes less effective compared to pure phosphatidylcholine (PC) lipids [6]. Hence, experimental conditions (e.g. temperature, buffer composition, injection flow rate, lipid concentration) must be optimized according to the specific vesicle lipid composition to promote the vesicle rupture on the support surface [8]. As an example, the exposure to buffers with different ionic strength [10], [11] or heating the natural lipid vesicle suspensions to high temperatures [12], [13] were proven to be efficient strategies to induce vesicle fusion, although they temporarily introduce non-physiological conditions, which may compromise the chemical stability of some lipid species by accelerating the oxidation of the acyl chains double bonds[14], [15], [16]. Lipid chemical degradation can in principle affect the bilayer structure as well as the study of the interaction of SLBs with proteins, peptides or drugs.

Here, we show that peptide discs, $~$10 nm discoidal lipid bilayers stabilized in solution by a self-assembled 18A peptide belt, can be used as precursors for SLBs including both synthetic and natural lipid mixtures. Compared to the traditional nanodiscs [17], where the discoidal lipid bilayer is surrounded by the membrane scaffold protein (MSP), the peptide discs are less stable in solution and have the tendency to quickly rearrange at room temperature into larger particles [18], [19]. Indeed, upon deposition of the peptide discs on a solid support, the 18A peptide belt can be removed by buffer rinsing, leaving a lipid bilayer in the proximity of the support surface. Recently, we exploited protein-loaded peptide discs to produce SLBs with oriented membrane proteins [20].

In the present study, peptide discs were prepared with a reference sample of 1-palmitoyl-2-oleyl-3-sn- glycerophosphatidylcholine (POPC), and mixtures of POPC and either 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-L-serine (POPS) or 1-palmitoyl-2-oleoyl-sn-glycero- 3- phosphoethanolamine (POPE), in both cases with concentration 70/30 mol/mol, respectively. POPC and cholesterol (CHOL, 80/20 mol/mol) mixture was also investigated. PS, PE and PC lipids are all present in the inner leaflet of the eukaryotic plasma membrane [21]. On the other hand, CHOL is present in both membrane leaflets where it regulates membrane fluidity and contributes to regulating the biological function of membrane proteins [22]. Hence, POPC/POPS, POPC/POPE and POPC/CHOL are biologically relevant SLB compositions. We also characterized the peptide discs as precursors of SLBs composed of natural phospholipids extracted from the yeast Pichia Pastoris. This lipid mixture contains phospholipids with acyl chains of different length and number of unsaturations as well as different headgroups, such as PC, PS, PE but also phosphoinositol (PI) and phosphoglycerol (PG) [23], [24]. Most importantly, the ratio between all these lipid species corresponds to Pichia Pastoris cell membranes, and is close to mammalian cell membranes [25]. On top of their biological relevance, these phospholipids can be produced as well in a deuterated version. Indeed, Pichia Pastoris cells can be grown in a deuterated culture medium and used for the production of deuterated phospholipids which are otherwise hard to produce through chemical synthesis[26]. Deuterated natural lipids are particularly valuable for studying biological membranes with neutron scattering techniques, NMR or infrared spectroscopy [27], [12].

The produced SLBs were characterized by means of neutron reflectometry (NR) to assess their structure, composition and surface coverage, which resulted to be between 90 and $95\%$ for synthetic lipids and $~85\%$ for natural lipids. The mechanism through which the peptide discs form a SLB was investigated by time-resolved attenuated total reflectance FTIR (ATR-FTIR). Peptide disc formation of SLBs was markedly faster than SLB formation by vesicle fusion and without the need of altering temperature, buffer composition or any other experimental variables from physiologically relevant values. Indeed, all the experiments, were performed at 25 °C and with a 20 mM Tris-HCl, 100 mM NaCl, 10 mM CaCl₂, pH = 7.4 buffer. Although ATR-FTIR suggests that traces of 18A peptide are left at the support surface, this is below the NR detection limit ($~0.02$ molar ratio), and the very low amount does not affect the SLB structure. Altogether, SLBs were successfully formed with both synthetic and natural lipid mixtures and under the same experimental conditions. Therefore, the peptide disc deposition, and subsequent 18A peptide removal, is a suitable strategy for producing SLBs of variable lipid composition.