Article pubs.acs.org/JPCB
Phase Transition-Controlled Flip-Flop in Asymmetric Lipid Membranes Yujia Jing, Angelika Kunze, and Sofia Svedhem* Department of Applied Physics, Chalmers University of Technology, SE-412 96 Göteborg, Sweden S Supporting Information *
ABSTRACT: Lipid membrane asymmetry is of fundamental importance for biological systems and also provides an attractive means for molecular control over biomaterial surface properties (including drug carriers). In particular, temperature-dependent changes of surface properties can be achieved by taking advantage of distinct phase transitions in lipid membrane coatings where lipids exchange (flip-flop) between leaflets. In this study, temperature is used to control flip-flop of lipids in asymmetric lipid membranes on planar solid supports, where the two leaflets of the lipid membrane are in different phase states. More specifically, the lower leaflet is prepared from a supported lipid membrane composed of a high Tm lipid mixture of phosphocholine (PC), phosphatidylserine (PS), and a bioactive lipid on TiO2, followed by selective removal of the top leaflet by detergent. Next, at a lower temperature, where the remaining leaflet is in the gel state, a top leaflet of a different lipid composition and in the fluid phase is formed. Phase transition-induced changes in membrane surface properties following upon temperature-activation of the prepared asymmetric membrane are demonstrated by the detection of biotinylated lipids, which were initially located (thus “hidden”) in the lower-gel phase leaflet, at the surface of the top leaflet. These processes were monitored in real-time by the quartz crystal microbalance with dissipation (QCM-D) and the dual polarization interferometry (DPI) techniques, allowing modeling of the mass and the anisotropic property of the lipid structures in different phase states.
1. INTRODUCTION Understanding of the assembly and dynamics of asymmetric membranes is of fundamental importance, both for the engineering of new materials,1 as well as for functions in biological systems.2,3 In particular, the maintenance of lipid asymmetry in the plasma membrane is an energy-dependent process,4 tightly connected to important cell processes, such as, for example, apoptosis5 or platelet activation.6 On a mechanistic level, transmembrane lipid redistribution is believed to be controlled by a range of flippases, floppases, and scramblases,7 but also includes spontaneous flip-flop of lipid molecules between the membrane leaflets.8 Generally, lateral redistributions are thought to occur more rapidly than transmembrane redistributions.9 An important strategy for the study of lipid membrane asymmetry is by the use of surface-supported lipid membranes.10−12 Often, the Langmuir−Blodgett (LB) technique has been used to prepare asymmetric supported lipid membranes.13 With the LB technique, the two leaflets are deposited stepwise on the solid support. A more convenient protocol for the formation of supported lipid membranes is based on the spontaneous rupture of small unilamellar liposomes on a solid support.11,14,15 Using this latter approach, asymmetric membranes have been reported for membranes containing phosphatidylserine (PS) lipids on a TiO2 coated surface.16−18 The properties and interactions of supported membranes can be conveniently studied by surface sensitive analytical techniques. Notably, a combination of an acoustic © 2013 American Chemical Society
and an optical technique has been proven useful to gain, through real-time measurements, information about such systems; typically mass,20 thickness,19 phase transition temperature,19,20 and sometimes birefringence (a measure of the anisotropy).20,21 For the purpose of characterizing temperaturedependent properties of lipid membranes, birefringence studies are of particular interest since lipid packing and arrangement within the membrane changes at the temperature (Tm) for the transition between the liquid and gel phases. In this study, we describe a protocol for the preparation of asymmetric lipid membranes, where the lower leaflet is in the gel (solid) phase (prepared from high Tm liposomes) and the top leaflet is in the fluid phase (prepared from low Tm liposomes) (Figure 1A). In particular, we are interested in lipid compositions yielding a Tm suitable for hyperthermal medical treatments (i.e., local increasing of the tissue temperature to around 42 °C). We suggest that hidden surface functions in such membranes can be activated by increasing the temperature above the effective Tm of the asymmetric membrane, thus inducing lipid flip-flop. As a proof-of-concept, binding of streptavidin to biotinylated lipids transferred from the lower membrane leaflet to the upper leaflet by phase transition-induced flip-flop is used. The different steps of the process were monitored by the two complementary techniques; Received: July 1, 2013 Revised: November 25, 2013 Published: December 23, 2013 2389
dx.doi.org/10.1021/jp406502b | J. Phys. Chem. B 2014, 118, 2389−2395
The Journal of Physical Chemistry B
Article
2.2. Liposome Preparation and Size Characterization. DPPC/DPPS/DPPE-biotin liposomes were prepared by the following procedure: 3.5 mg of DPPC, 1.5 mg of DPPS, and 250 μg of DPPE-biotin dissolved in chloroform were added to a round-bottomed flask. The solvent was evaporated under a gentle nitrogen flow in a fume hood, after which residual chloroform was removed under reduced pressure. The subsequent rehydration and extrusion were performed at 60 °C, that is, well above the Tm for DPPC (Tm = 41 °C) and DPPS (Tm = 54 °C). The lipid film was rehydrated in 1 mL of preheated Tris buffer and the suspension was kept at 60 °C in an oven for at least 1 h with occasional vortexing until lipids were fully dissolved. The solution was then extruded through polycarbonate membranes with pore sizes of 30 nm (Whatman, U.K.) in a temperature-controlled mini-extruder (Avanti Polar Lipids, Inc., U.S.A.). Unilamellar small liposomes were obtained by extruding the suspension 51 times. Other liposomes (POPC, POEPC, DOPEt, Lyso-PC) were prepared following the same protocol but at ambient temperature. The liposomes were characterized by nanoparticle tracking analysis (NTA) (Nanosight, U.S.A.) with respect to size. Measurements were performed at room temperature using liposome solutions that were diluted in Tris buffer to 0.0005 mg/mL. The typical size of liposomes prepared by the protocol described above is 69 ± 1 nm, with a diameter distribution ranging from 30 nm to 180 nm. 2.3. Quartz Crystal Microbalance with Dissipation (QCM-D) Measurements. The preparation of asymmetric lipid membranes and the subsequent transmembrane lipid exchange (flip-flop) were monitored by QCM-D and repeated several times for each condition. Prior to mounting, the QCMD sensors (TiO2-coated sensors obtained from Q-Sense AB, Sweden) were rinsed with 10 mM SDS solution, water, and further cleaned in a UV-ozone cleaner for 30 min. QCM-D measurements were performed at several harmonics (3, 5, 7, 9, 11, and 13) using a QCM-D E4 system (Q-Sense AB, Sweden). The frequency and dissipation changes were plotted and analyzed with the Q-tools software (Q-Sense AB, Sweden). The presented frequency shifts obtained at different harmonics were normalized by dividing each frequency shift by the corresponding harmonic number. A typical experiment was performed as follows: DPPC/ DPPS/DPPE-biotin liposomes (diluted just before injection to a total lipid concentration of 0.2 mg/mL in Tris-CaCl2) were adsorbed onto TiO2-coated QCM-D sensors at 55 °C and formed supported lipid membranes. Excess intact liposomes attached to the surface of the formed supported membranes were removed by washing with Tris-EDTA for 5 min. After stabilization of the QCM-D signals at Δf = −25 Hz and ΔD < 0.5 × 10−6 (characteristic values of a supported lipid membrane14,19), a DPPC/DPPS/DPPE-biotin lipid monolayer was obtained by rinsing (shortly) with Tris-SDS followed by Tris. The temperature was then decreased to 22 °C, and POPC, Lyso-PC, POEPC, or DOPEt liposomes were introduced to form different top leaflets. The temperature was again increased to 55 °C and maintained for 1 h to allow the lipids to redistribute between the two leaflets after which the temperature was again decreased to 22 °C for further tests. In these further tests, a BSA solution was injected before or after the heat treatment to block the unspecific adsorption of streptavidin. The streptavidin solution was introduced directly after the BSA injection to detect the biotinylated lipids exposed at the membrane surface.
Figure 1. (A) Schematic representation of the three-step-formation of an asymmetric membrane with the bottom leaflet being in gel phase and the top leaflet being in fluid phase. Blue head and green tails, DPPC; yellow head and tails, DPPS; pink head and tails, DPPE-biotin lipids; light brown head and tails, POEPC. (B) QCM-D data recorded during the formation of an asymmetric lipid membrane, where a high Tm membrane (DPPC/DPPS/DPPE-biotin) and monolayer were first formed at 55 °C and where a different lower Tm top leaflet (POEPC) was formed after lowering the temperature to 22 °C. The presented data have been corrected with respect to the inherent temperature effect on the QCM-D signal.27
the quartz crystal microbalance with dissipation (QCM-D) and the dual polarization interferometry (DPI).
2. MATERIALS AND METHODS 2.1. Materials. Unless otherwise stated, chemicals were purchased from commercial suppliers and used without further purification. The compounds 1,2-dipalmitoyl-sn-glycero-3phosphatidylcholine (DPPC), 1, 2-dipalmitoyl-sn-glycero-3phospho-L-serine (DPPS), 1, 2-dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-(biotinyl) (DPPE-biotin), 1-palmitoyl-2oleoyl-sn-glycero-3-phosphocholine (POPC), 1-palmitoyl-2hydroxy-sn-glycero-3-phosphocholinelipids (Lyso-PC), 1-palmitoyl-2-oleoyl-sn-glycero-3-ethylphosphocholine (POEPC), and 1, 2-dioleoyl-sn-glycero-3-phosphoethanol (DOPEt) were purchased from Avanti Polar Lipids Inc., U.S.A. Ethylenediaminetetraacetic acid (EDTA), sodium dodecyl sulfate (SDS), bovine serum albumin (BSA), and streptavidin were purchased from Sigma. Water was purified using a Milli-Q water purification system (Millipore, France) to a minimum resistivity of 18.2 MΩ·cm. Tris-buffered saline (Tris) contained 10 mM Tris and 10 mM sodium chloride (NaCl) in water. The prepared buffer was adjusted to pH 8.0 by addition of 0.1 M HCl solution. Different Tris buffers were prepared by including additionally CaCl2, EDTA, or SDS: Tris-CaCl2 (2 mM CaCl2, pH 8.0), Tris-EDTA (2 mM EDTA, pH 8.0), Tris-SDS (10 mM SDS, pH 8.0). All buffers were filtered through 0.2 μm filters and degassed. Protein solutions, BSA (50 μg/mL), and streptavidin (50 μg/ mL), were prepared in Tris buffer. 2390
dx.doi.org/10.1021/jp406502b | J. Phys. Chem. B 2014, 118, 2389−2395
The Journal of Physical Chemistry B
Article
electrostatic repulsion between the negatively charged liposomes and the surface is overcome by the presence of the Ca2+ ions, and the fluidity of the lipid membrane allows adsorbed liposomes to rupture and form a supported lipid membrane as a critical surface coverage is reached, following a well-established pathway.14,15 After formation of a supported membrane from the high Tm liposomes is completed, as indicated by the characteristic values of a lipid membrane14 (Δf = −25 Hz and ΔD = 0.5 ×10−6 after 4 min in Figure 1B), attachment of intact liposomes to the membrane was observed indicated by a decrease in frequency and an increase in dissipation (Figure 1B, 5−8 min). This binding is due to the bridging effect of the Ca2+ ions between the DPPS head groups exposed both at the supported membrane surface and at the liposomes. Excess liposomes could be removed by rinsing with Tris buffer containing 2 mM EDTA to disrupt the intermembrane bridging by Ca2+ ions. Eventually, stable supported membranes were formed after 12 min (Δf = −25 Hz, ΔD = 0.2 × 10−6, structure I in Figure 1B). The formation of supported membranes using the described protocol was verified using the DPI technique (detailed in the Supporting Information). The adsorbed lipid mass was quantified from both QCM-D and DPI data and the obtained values were found to be in good agreement with theoretical and previously obtained values (Table 1).
Because the QCM-D response is sensitive to temperatureinduced changes in the sensor properties, it is important to subtract the inherent contribution of the QCM-D crystal to the recorded signals. In all cases, a reference measurement was performed using the same blank sensor exposed to the same conditions. When calculating the mass deposited on QCM-D crystals, supported lipid membranes are conventionally treated as rigid films,19 and the mass of the membranes can be estimated using the Sauerbrey equation.22 For the layers of adsorbed proteins, the obtained reference-subtracted (to compensate for inherent temperature effects, see above) QCM-D signals (harmonics n = 3, 5, 7, 9, 11, 13) were fitted to the viscoelastic model implemented in QTools (Q-sense AB, Sweden), yielding effective values of the film mass, shear modulus, and viscosity with the density as the only input parameter.23 2.4. Dual Polarization Interferometry (DPI) Measurements. DPI experiments were performed in an Analight 4D instrument (Farfield Group Ltd., U.K.). The sensor surface Anachip (Silicon oxynitride) (Farfield Group Ltd., U.K.) was sputter-coated with TiO2 (3.5 nm) by the ion sputtering method (40 sccm Argon, 4 sccm Oxgen, 5 × 10−3 mbar, 1 kW DC power) using FHR MS 150 (FHR Anlagenbau GmbH, Germany). Before deposition of the TiO2 layer, sensor surfaces were treated for 30 min by UV-ozone and 30 s by O2 plasma (TePla 300 PC, PVA TEPLA, Germany) (1 mbar O2, 250 W). One day prior to the experiment, the sensors were treated in UV-ozone for 30 min, and then stored in water overnight. The next day, sensors were dried by a gentle stream of nitrogen and mounted into the instrument. Chip calibrations were performed prior and after the TiO2 coating. The running buffer (20 μL/min) in the experiments was Tris, and all injections were performed using a syringe with volume of 200 μL (setting a limit to the maximum volume used) and with a flow rate of
