Substrate-induced electrostatic potential varies composition of supported lipid bilayer containing anionic lipid

Ryugo Tero; Natsumi Kobayashi

doi:10.35848/1347-4065/ac3fcc

1. Introduction

The cell membrane is the outermost layer of cells and functions as a reaction field for the transportation of materials, information, and energy into and out of cells.^1–3) Degeneration in these reactions leads to neurological and metabolic diseases, and thus they are important research targets in the fields of medicine and drug discovery. The fundamental structure of the cell membrane is the lipid bilayer, which is a bimolecular sheet of amphiphilic lipid molecules. Natural cell membranes contain various lipid molecules, and their composition determines the physical properties of the lipid bilayer, which affects the transportation reactions at the cell membrane.^4,5)

Artificial lipid bilayer systems, such as free-standing bilayer lipid membranes,^6,7) vesicles,^8,9) and supported lipid bilayers (SLBs),^10–12) have been used to study cell membranes and membrane proteins. SLB is an artificial lipid bilayer existing on a solid substrate, and therefore has a high technical affinity with solid devices and sensors.^13–16) One of the advantages of these artificial experimental systems is that the lipid composition is controllable. One can mix synthetic or natural-derived lipids at an arbitrary ratio.^17–21) Generally, SLBs are made on chemically and biologically inert substrates,²²⁾ such as glass, mica, and the oxide layer on a Si wafer, which are negatively charged because of the hydroxyl groups on their surfaces.^23,24) Approximately 1 nm thick water layer exists between the solid substrate and the lipid bilayer membrane in the SLB system,^25–27) but the chemical and physical properties of the substrate possibly affect the lipid bilayer.^22,28)

The vesicle fusion method is widely used for the formation of SLB. A vesicle is a spherical shell structure of a lipid bilayer. SLB, a planar lipid bilayer membrane on a substrate, is formed through shape transformation processes of vesicles, such as adsorption, rupture, fusion, and spreading. Generally, it is reasonably assumed that the lipid composition in the vesicle is retained in SLB because reasonable phase diagrams are obtained in SLB,^29–31) but not necessarily.³²⁾ Electrostatic interaction between a lipid bilayer and a solid substrate significantly affects the SLB formation process. Electrostatic attraction between charged substrates and lipid vesicles containing oppositely charged lipids promotes the SLB formation from vesicles.^33,34) Inter-lipid electrostatic attraction between anionic and cationic lipids also promotes the lipid transfer and membrane fusion between SLB and vesicles.^35,36)

Recently, we found that the electrostatic interaction between a negatively charged substrate surface and anionic lipid vesicles affected the composition of SLB.³⁷⁾ In this study, we investigated SLB formation from the lipid vesicles containing anionic lipid on a thermally oxidized SiO₂/Si substrate. We aimed to control the ratio of the anionic lipid in the SLB by modulating the substrate-induced electrostatic effect on the lipid bilayer.

2. Experimental methods

The reagents 1,2-dioleoyl-sn-glycero-3-phospho-(1'-rac-glycerol) (DOPG⁽⁻⁾) [Fig. 1(a)] and 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) [Fig. 1(b)] were purchased from Avanti Polar Lipids, Inc. (Alabaster, AL, USA). A fluorescence-tagged lipid, 1,2-Bis-(4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-undecanoyl)-sn-glycero-3-phosphocholine (Bis-BODIPY-PC) [Fig. 1(c)] was purchased from Thermo Fisher Scientific (Waltham, MA, USA). These lipid reagents were used as received, without further purification. Chloroform solutions of DOPG⁽⁻⁾, DPPC, and Bis-BODIPY-PC were mixed at a molar ratio of 40:60:0.2 or 0:100:0.2 in a glass vial. The mixed lipid solution was dried in a nitrogen stream followed by vacuum-drying at least for 6 h. The lipid film was suspended in a buffer solution (120 mM KCl, 10 mM HEPES/ KOH pH 7.2) at 45 °C for 1 h, followed by sonication to prepare a unilamellar vesicle suspension. A thermally oxidized SiO₂/Si substrate was cleaned by boiling in piranha solution (3:1 v/v mixture of conc. H₂SO₄ and 30% H₂O₂ aqueous solution). The SiO₂/Si substrate was incubated in the DOPG⁽⁻⁾ + DPPC vesicle suspension with different Ca²⁺ concentration ([Ca²⁺]) in the range of 0–20 mM at 45 °C for 1 h, to prepare SLBs. The buffer solution for the vesicle suspension (120 mM KCl, 10 mM HEPES/KOH pH 7.2) was used for the condition of [Ca²⁺] = 0 mM, and the buffer solutions including CaCl₂ were used for the conditions of [Ca²⁺] = 0.5–20 mM. The vesicle suspension was exchanged with the buffer solution 15 times, to remove excess vesicles.

The SLBs were observed with an epifluorescence microscope (BX51WI, Olympus, Tokyo, Japan) equipped with a 60× water immersion lens (LUMPlan FL 60×, NA = 1.00) and a CMOS camera (DS-Qi2, Nikon, Tokyo, Japan). The sample temperature was controlled in the range of 15 °C–45 °C during the fluorescence observation and the fluorescence recovery after photobleaching (FRAP) observation. For FRAP observation, fluorescence images were captured every 15 s for 300 s after the photobleaching at each temperature. We performed the FRAP observation increasing the sample temperature by 2 °C–3 °C, to evaluate the phase state of SLB based on its fluidity. The area fraction of lipid domains in the fluorescence images was obtained from at least five images using NIS Elements software (Nikon).

3. Results and discussion

3.1. Results

In this study, SLB was prepared from the lipid vesicle comprising DOPG⁽⁻⁾ and DPPC at the molar ratio 40:60. DOPG^{( − )} [Fig. 1(a)] is an anionic and fluid lipid at the experimental conditions in this study: its headgroup is negatively charged at pH 7.2 because of its low pKa (2.9³⁸⁾); it has low phase transition temperature (T_m) between the gel (solid) and liquid crystalline (L_α) (fluid) phases (−18 °C³⁹⁾), because of the unsaturated acyl chains. DPPC [Fig. 1(b)] is zwitterionic and therefore net neutral at pH 7.2, and its saturated acyl chains give high T_m (41 °C).^39,40) The binary DOPG⁽⁻⁾ + DPPC bilayer mixed at 40:60 causes phase separation between the gel and L_α phases at 25 °C.³⁹⁾

Figures 2(a)–2(c) show fluorescence images of SLB prepared with the DOPG⁽⁻⁾ + DPPC (40:60) vesicle suspension in the absence of Ca²⁺ ([Ca²⁺] = 0 mM) (Ca0mM-SLB) observed at 25 °C. The SLB with uniform fluorescence intensity covered the entire surface [Fig. 2(a)]. Figures 2(b) and 2(c) depict the result of FRAP observation at 25 °C, which detects the lateral molecular diffusion in lipid bilayer membranes.^21,33,41) After a part of the SLB was photobleached [Fig. 2(b)], the shape of the photobleached dark region did not change temporally [Fig. 2(c)]. The lipids did not diffuse laterally between the photobleached region and its surrounding. The FRAP observation revealed that SLB had no fluidity, and thus was in the gel phase at 25 °C. Figure 2(d) shows the FRAP result obtained after the sample was heated to 38 °C. The edge of the photobleached region were blurred over time during the FRAP observation, indicating lateral diffusion across the edge started. We assigned the temperature at which the lateral diffusion in SLB started to T_m. Figures 2(e) and 2(f) show the FRAP result of DPPC-SLB that was prepared from DPPC vesicle suspension without DOPG⁽⁻⁾ at [Ca²⁺] = 0 mM. The photobleached region did not change at 25 °C [Fig. 2(e)], whereas lateral diffusion started at 40 °C, which corresponded to T_m of the DPPC bilayer in the literature (41 °C).^39,40) Therefore, T_m of SLB samples is evaluated by the temperature-controlled FRAP observation.

**Fig. 2** (a)–(c) Fluorescence images during FRAP observation of Ca0mM-SLB that was prepared from the DOPG⁽⁻⁾ + DPPC (40:60) vesicle in the absence of Ca²⁺ observed at 25 °C. (a) Before photobleaching, (b) 0 s after photobleaching, and (c) 227 s after photobleaching. (d) A fluorescence image during FRAP observation of Ca0mM-SLB observed at 38 °C, obtained 213 s after photobleaching. (e), (f) Fluorescence images during FRAP observation of SLB that was prepared from DPPC vesicle without DOPG⁽⁻⁾ in the absence of Ca²⁺ observed (e) at 25 °C, 141 s after photobleaching, and (f) at 40 °C, 234 s after photobleaching. Scale bar = 20 μm.

Download figure:
Standard image High-resolution image

The fluorescence FRAP observations of Ca0mM-SLB [Figs. 2(a)–2(d)] showed the formation of a single-phase lipid bilayer with T_m = 38 °C, which was close to that of DPPC-SLB without DOPG⁽⁻⁾. The result indicates that Ca0mM-SLB was predominantly composed of DPPC, although it was prepared from the binary DOPG⁽⁻⁾ + DPPC vesicle. We hypothesized it is due to the electrostatic repulsion between the negatively charged SiO₂/Si substrate surface and anionic DOPG⁽⁻⁾. To suppress the electrostatic interaction, Ca²⁺ was added during the SLB formation process. Divalent ions effectively shield electrostatic interaction compared to monovalent ions. We avoided increasing KCl concentration to reduce Debye length, because osmotic pressure significantly affects the SLB formation process by the vesicle fusion method.^42,43)

Figure 3 shows fluorescence images of SLBs prepared with the DOPG⁽⁻⁾ + DPPC (40:60) vesicle in the presence of [Ca²⁺] = 0.5 mM (Ca0.5mM-SLB). In the Ca0.5mM-SLB, uniform fluorescence intensity was obtained from the entire sample [Fig. 3(a)], and lateral diffusion started at 35 °C [Fig. 3(b)]. It is a single-phase SLB, and its T_m was 35 °C. Ca0.5mM-SLB mainly consisted of DPPC as Ca0mM-SLB [Fig. 2(a)] but contained more DOPG⁽⁻⁾ than Ca0mM-SLB as indicated by lower T_m. The SLB prepared at [Ca²⁺] = 1.25 mM was also single-phase, and T_m was 25 °C. The values of T_m and its dependence on [Ca²⁺] during the SLB formation are summarized in Table I.

**Fig. 3.** Fluorescence images of Ca0.5mM-SLB that was prepared from the DOPG⁽⁻⁾ + DPPC (40:60) vesicle at [Ca²⁺] = 0.5 mM, (a) observed at 25 °C, and (b) observed at 35 °C, 217 s after photobleaching. Scale bar = 20 μm.

Download figure:
Standard image High-resolution image

Table I. Transition temperature (T_m) of SLB prepared under Ca²⁺ concentration [Ca²⁺].

[Ca²⁺] (mM)	0	0.5	1.25
T_m (°C)	38	35	25

Figure 4 shows fluorescence images of SLBs prepared with the DOPG⁽⁻⁾ + DPPC (40:60) vesicle at [Ca²⁺] = 2.5–20 mM: SLBs prepared at [Ca²⁺] = 2.5 mM [Fig. 4(a)], 5.0 mM, [Fig. 4(b)], 10.0 mM [Fig. 4(c)], and 20.0 mM [Fig. 4(d)], namely Ca2.5mM-SLB, Ca5mM-SLB, Ca10mM-SLB, and Ca20mM-SLB, respectively. Dark regions were observed in Ca2.5 mM-SLB [Fig. 4(a)], and their area fraction (θ) decreased with [Ca²⁺] [Figs. 4(b)−4(d)]: θ = 18.8% in Ca2.5mM-SLB, 14.3% in Ca5mM-SLB, 1.3% Ca10mM-SLB, and 0.7% in Ca20mM-SLB. Two regions with different fluorescence intensities indicate the existence of a binary phase, where the distribution of the fluorescence probe varies. In the FRAP observation, the fluorescence intensity of the photobleached region recovered indicating that the surrounding bright region was in a fluid state [Figs. 4(e) and 4(f)]. The fluorescence probe Bis-BODIPY-PC [Fig. 1(c)] preferably distributes to a lipid bilayer region with disordered acyl chains compared to an ordered region because of its bulky dye-labeled tails.^19,20) Therefore, the bright and dark regions in Ca5mM-SLB—Ca20mM-SLB are assigned to the L_α and gel phases, respectively. The gel phase domains predominantly comprised DPPC (T_m = 41 °C), whereas the L_α region contained a certain amount of DOPG⁽⁻⁾ (T_m = −18 °C). A decrease in θ with increasing [Ca²⁺] indicates the reduction in DPPC in SLB, and thereby increment of the DOPG⁽⁻⁾ composition.

The dependences of T_m and θ on [Ca²⁺] are summarized in Fig. 5. In the range of [Ca²⁺] < 2.5 mM, SLBs were single-phase and in the gel phase at 25 °C. Their T_m decreased with increasing [Ca²⁺]. At [Ca²⁺] = 0 mM, the Ca0mM-SLB had similar T_m (38 °C) to that of DPPC-SLB (40 °C). In the range of [Ca²⁺] = 2.5–20.0 mM, the SLBs separated into the gel and L_α phases, and the occupying area of the L_α region increased with [Ca²⁺]. These results indicate that neutral DPPC in the DOPG⁽⁻⁾ + DPPC (40:60) vesicle was predominantly included in SLB on the SiO₂/Si substrate in the absence of Ca²⁺, and that Ca²⁺ during the SLB formation increased the anionic DOPG⁽⁻⁾ composition in SLB. The substrate-induced electrostatic potential is effectively applied to the vesicles during the SLB formation, even though the Debye length, which is the attenuation distance of electric potential to e⁻¹, in the buffer solution is calculated to be 0.87 nm at 45 °C. Divalent cation attenuated the electrostatic repulsion between DOPG⁽⁻⁾ and the dissociated hydroxyl group (−O⁻) on the SiO₂/Si substrate, and furthermore bridged these monovalent anions.^44,45)

A planar SLB is formed through adsorption, fusion, rupture, and spreading of vesicles. During these processes, a part of lipids in the adsorbed vesicles desorbs from the substrates as reported in previous QCM studies.^46,47) We suggest that the repartition of DOPG⁽⁻⁾ and DPPC in the adsorbed vesicle determine the compositions of SLB remaining on the substrate and desorbing vesicles. When a DOPG⁽⁻⁾ + DPPC vesicle adsorbs on a negatively charged substrate surface, DOPG⁽⁻⁾ is excluded from the vicinal side, where DPPC segregates as a result. The peripheral side of the adsorbed vesicle that is enriched with DOPG⁽⁻⁾ desorb from the substrate. In the presence of Ca²⁺, more DOPG⁽⁻⁾ molecules are retained in the vicinal side at higher [Ca²⁺], and thus in SLB. Continuous change in T_m and θ with increasing [Ca²⁺] indicates that the DOPG⁽⁻⁾:DPPC composition in SLB can be varied with [Ca²⁺]. Himeno et al. reported the phase diagram of DOPG⁽⁻⁾ + DPPC in the giant unilamellar vesicle (GUV) system.³⁹⁾ DOPG⁽⁻⁾ + DPPC (40:60) GUV has the transition temperature between the single fluid phase and the binary gel/L_α phases at approximately 25 °C in 10 mM NaCl aq. Assuming this phase diagram in the present study, the Ca10mM-SLB, in which the gel-phase domain nearly disappeared (θ = 1.3%), contained at ~40 mol% DOPG⁽⁻⁾. This value corresponds to the composition of DOPG⁽⁻⁾ + DPPC vesicle for SLB formation (40:60). We should note that the phase diagram depends on the salt concentration,³⁹⁾ therefore further study is needed for precisely determining the lipid compositions in the DOPG⁽⁻⁾ + DPPC-SLBs. The phase state of DOPG⁽⁻⁾ is rarely affected by Ca²⁺ addition, while phosphatidylserine (PS), which is another anionic phospholipid, segregates in PS + PC mixture bilayer forming domains.⁴⁸⁾

In this study, we did not find a result indicating the asymmetric distribution of lipids between the upper and lower leaflets of the lipid bilayer. An asymmetric gel/L_α bilayer due to Ca²⁺-induced DOPG⁽⁻⁾ accumulation to the lower leaflet would cause half intensity recovery in FRAP observation. A phase-separated SLB with different DOPG⁽⁻⁾:DPPC composition between the upper and lower leaflets would provide three levels in fluorescence intensity; bright L_α/L_α regions, dark gel/gel regions, and gel/L_α in the middle. FRAP results showed no recovery [Figs. 2(b) and 2(c)] or full recovery [Fig 4(f)]. Only two intensity levels were observed in the binary-phase SLBs [Fig 4]. These results indicate that two leaflets had similar lipid compositions and that the lipid domains are coupled between the leaflets. A previous study reported preferential distribution of PS to the lower leaflet of SLB in the presence of Ca²⁺.⁴⁹⁾ PS has several types of binding sites for Ca²⁺.⁵⁰⁾ Therefore we surmise that PG distribution in lipid bilayer membranes are ineffectively affected by Ca²⁺ compared to PS.⁴⁸⁾

4. Conclusions

The electrostatic interaction between the substrate surface and lipid bilayer significantly affected the lipid composition of SLB that were formed from vesicles. The SLB made from the DOPG⁽⁻⁾ + DPPC (40:60) vesicle predominantly comprised neutral DPPC when it was prepared in the buffer solution containing only monovalent ions. The existence of Ca²⁺ during the SLB formation made DOPG⁽⁻⁾ included in SLB, and the DOPG⁽⁻⁾ composition increased with [Ca²⁺]. The results in this study indicate that the lipid composition of charged lipids in SLB is controllable by adjusting the substrate-induced electrostatic interaction as a function of [Ca²⁺]. We expect that these findings are applied for constructing SLB systems on electrodes and biosensors on which electric potentials are applied.

Acknowledgments

This work was supported by JSPS KAKENHI Grant Nos. JP20H02690 and JP 20K21125, JST-A-STEP, and the Nagai Foundation for Science and Technology. We acknowledge support from the Electronics-Inspired Interdisciplinary Research Institute (EIIRIS) Project of Toyohashi University of Technology.

Substrate-induced electrostatic potential varies composition of supported lipid bilayer containing anionic lipid

Article metrics

Permissions

Author e-mails

Author affiliations

ORCID iDs

Dates

Abstract

1. Introduction

2. Experimental methods

3. Results and discussion

3.1. Results

4. Conclusions

Acknowledgments

Substrate-induced electrostatic potential varies composition of supported lipid bilayer containing anionic lipid

Article metrics

Permissions

Share this article

Author e-mails

Author affiliations

ORCID iDs

Dates

Abstract

1. Introduction

2. Experimental methods

3. Results and discussion

3.1. Results

4. Conclusions

Acknowledgments