2018 年 41 巻 3 号 p. 303-311
Fluidity and mosaicity are two critical features of biomembranes, by which membrane proteins function through chemical and physical interactions within a bilayer. To understand this complex and dynamic system, artificial lipid bilayer membranes have served as unprecedented tools for experimental examination, in which some aspects of biomembrane features have been extracted, and to which various methodologies have been applied. Among the lipid bilayers involving liposomes, planar lipid bilayers and nanodiscs, recent developments of lipid bilayer methods and the results of our channel studies are reviewed herein. Principles and techniques of bilayer formation are summarized, which have been extended to the current techniques, where a bilayer is formed from lipid-coated water-in-oil droplets (water-in-oil bilayer). In our newly developed method, termed the contact bubble bilayer (CBB) method, a water bubble is blown from a pipette into a bulk oil phase, and monolayer-lined bubbles are docked to form a bilayer through manipulation by pipette. An asymmetric bilayer can be readily formed, and changes in composition in one leaflet were possible. Taking advantage of the topological configuration of the CBB, such that the membrane’s hydrophobic interior is contiguous with the surrounding bulk organic phase, oil-dissolved substances such as cholesterol were delivered directly to the bilayer interior to perfuse around the membrane-embedded channels (membrane perfusion), and current recordings in the single-channel allowed detection of immediate changes in the channels’ response to cholesterol. Chemical and mechanical manipulation in each monolayer (monolayer technology) allows the examination of dynamic channel-membrane interplay.
The ion channel is an electric device that generates rapid and far-reaching electrical signals along the cell membrane.1,2) This unique function is based on the molecular properties of ion channels involving selective ion permeation and gating.3) For example, potassium channels allow selective K+ permeation over Na+,4–6) while the permeation rate was as high as 107 ions/s through a single-channel molecule.7–9) Also, some channels respond to the changes in the membrane potential with conformational changes.10–12) Such sophisticated proteins of ion channels, however, prohibit their ability to function in the absence of a membrane. For example, ion permeation through the channel is a passive process along the electrochemical potential gradient across the cell membrane,13) and membrane potential, posing the highest electric field in the biological system (107 V/m) to membrane proteins,14) can modify the gating, and affect the driving force of ion permeation. Accordingly, the membrane is not simply a supporting material for ion channels, but an integral counterpart in which chemical and physical features intervene in channel dynamics.15) In fact, the fluidity and mosaicity of biomembranes,16,17) involving varying local fluidity, asymmetry of leaflets,18,19) and the existence of local domains with different chemical compositions and physical phases,20) complicate channel-membrane interactions. Thus, even though studying the channel in the membrane is prerequisite, biomembranes are too complex a system to study easily and in isolation; an alternative simple system of study is necessary. The lipid bilayer is the system of choice for this purpose, in which lipid compositions are arbitrarily controlled, and purified channel proteins are reconstituted21–23) (Fig. 1). This provides a reconstitution method24) in which more sophisticated chemical and physical manipulations are possible.
Detergent-solubilized channels were purified and reconstituted into liposomes. These liposomes were added in one of the compartments separated by a preformed planar lipid bilayer. The liposomes underwent membrane fusion to the lipid bilayer, and the channels were reconstituted. Single-channel currents were recorded under the voltage-clamp condition. A. The conventional chamber method. B. The tip-dip method. C. The contact bubble bilayer method.
Various lipid bilayer methods21) are tentatively classified into two groups according to experimental techniques and analysis convention. The first category involves liposomes25,26) and nanodiscs,27,28) in which huge numbers of isolated membranes exist in a system, and the collective and ensemble behavior of the entire membrane system is examined. In the second category, which is the main topic of this review, the system studied involves only one membrane, in which only one or many channels are embedded. By taking advantage of the single membrane system, channels are imposed to the same membrane potential, which is critical for studying the molecular properties of channels, such as single-channel current recordings.29)
The channel-membrane system has changed over a long evolutionary period, eventually yielding varieties of cellular function.30) It originally began as a simple system of ions, water, channels and the membrane. Lipid bilayer experiments remind us of such primitive actions of the channel-membrane system. In turn, we have been able to examine secrets of the channel-membrane system by intentionally “evolving” this experimental system, by manipulating the membrane to become more complex in form and function. Here, we present such experimental evolution of the lipid bilayer system through sophisticated monolayer manipulation techniques.
We start with the planar lipid bilayer (PLB) method as a standard method,21,31) in which a free standing bilayer is formed in a small hole. Ion channels operate under the applied membrane potential, and the PLB method allows application of the membrane potential as high as ±400 mV.32) Various PLB methods have been developed, and further sub-categorized into those using chambers and pipettes. In the chamber methods (Fig. 2), painting33) and folding34) methods were developed for membrane formation on a hole in the chamber, whereas tip-dip32,35–38) and punch-out39,40) methods were developed for the formation of a membrane at the tip of a glass pipette. (Detailed methodology is described in detail elsewhere.21))
There is a small hole (ca. 100 um in diameter) in the septum (made of Teflon) separating two aqueous compartments, to which membrane lipids dissolved in decane were added. The lipids partitioned the oil–water interface to form a monolayer. The decane phase, sandwiched by two monolayers, escapes along the hydrophobic surface of the septum, and two monolayers are eventually attached to form a bilayer.
Even though the apparent procedures differ substantially, a common process underlies membrane formation.41) The monolayers are formed between an oil–water33,42,43) or air–water34) interface, and these are eventually attached through the spontaneous removal of the organic solvent between the monolayers (Fig. 2)33) or by the mechanical apposition of two monolayers (Fig. 5B).42) In the chamber methods, the monolayers are guided on the surface of the hydrophobic material to place them close together in order to form the bilayer. By contrast, in the pipette methods, a monolayer is attached at the tip and is bound to another monolayer through pipette manipulations. Both methods allow for the formation of membranes with arbitrary lipid compositions, and channels are reconstituted into the lipid bilayer via various methods.
2.2. From Chamber to Water-in-Oil DropletFunakoshi et al. proposed an alternative method for bilayer formation43) in which water-in-oil droplets, lined with a lipid monolayer, were guided by microfluidic circuitry to attach together to form the bilayer. Alternatively, two water-in-oil droplets were docked. In these methods, manipulation of the water droplets was performed in a bulk oil phase (Fig. 3), rather than that of oil droplets in a bulk aqueous phase. In fact, this method produces a topological inversion of the bulk phase from the electrolyte solution to the bulk oil phase. Consequently, both aqueous phases were enclosed in a closed space.
Once a water droplet is formed in the bulk oil phase containing phospholipids, a monolayer is formed at the oil–water interface. Two droplets coated with monolayers are attached to form the bilayer.
Bayley’s group also established a new DIB method.42,44) Shallow wells were filled with lipid-dispersed hexadecane. The tip of an Ag wire was made ball-shaped by melting; it was then coated with AgCl. The ball surface was further coated with agarose gel to prevent wetting to oil. One microliter electrolyte solution was hung on the ball immersed in the oil phase. The lipids in the oil immediately transferred to the oil-electrolyte interface, forming the monolayer. Two of these droplets were then attached to form the bilayer.
2.4. Droplet Transfer MethodPrior to the development of the water-in-oil droplet method, the droplet transfer method was developed45) (Fig. 4A), although this was not intended for use in electrophysiological measurements. An oil phase was layered on top of a water phase, and lipids dispersed in the oil phase were partitioned into the oil-water interface where a monolayer is formed. Water droplets were dispersed in the oil phase and the droplet surface was also covered with the monolayer. The monolayer coated droplets crossed the interface monolayer, upon which the relevant monolayer covered the surface monolayer of the droplets, forming a bilayer.46)
A. The droplet transfer method. A water droplet in oil is lined by a monolayer of membrane lipid, which is transferred across the monolayer formed at the water-oil interface. After the transfer, a liposome is formed. B. Topological transformation from the liposome to the contact bubble bilayer. The outer leaflet of the liposomes was peeled, and the inner surface of the leaflet came into contact with the oil. Extended peeling resulted in the enclosure of the water phase into a droplet. This led to the formation of monolayer-lined water droplets, which then came into contact and adhered into a bilayer formation.
Beginning with Funakoshi et al., including the DIB and droplet transfer methods, these methods are all based on the principle of the water-in-oil droplet. Thus, here we collectively term the bilayer formed via the water-in-oil droplets as a “water-in-oil bilayer.” Figure 4B summarizes the conceptual transformation, from conventional liposomes to the water-in-oil bilayer, in which the topological relationships of the aqueous and hydrophobic phases are inversed. Imaginary processes proceed from the liposome in an aqueous solution, and the outer leaflet is peeled, by which the inner hydrophobic surface of the leaflets is exposed to an oil phase. As the peeling advances, the outer leaflet is inverted and the water phase is enclosed. This corresponds to a reversal of the process involved in the droplet transfer method. The water-in-oil bilayer has a unique property. Whereas in liposomes the bilayer is an isolated system, each leaflet of the bilayer in the water-in-oil bilayer is continuous with the monolayer lining a droplet and is in equilibrium with each other. This monolayer phase provides unique features to the water-in-oil bilayer, as shown in the later sections.
2.6. The Contact Bubble Bilayer (CBB) MethodIn parallel to the patch-clamp methods, many methods using glass pipettes have been developed for planar lipid bilayer experiments, such as the tip-dip,35–38) punch-out,40) bubble patch,48) liposome patch49,50) and modified tip-dip32) methods (Fig. 5). In contrast to chamber methods, pipette methods generally form smaller membranes (<30 µm in diameter) with less electrical background noise. Inspired by the method by Funakoshi et al., we developed the CBB method,47) in which the manipulability of the bilayer was enhanced by the use of pipettes. A water “bubble” can be blown in air under zero gravity such as in the Spacelab of a space station, whereas in the presence of gravity, a water bubble can only be blown in an oil phase. A glass pipette with a 30 µm diameter tip was filled with an electrolyte solution in connection with the Ag/AgCl electrode (Fig. 6). The pipette was connected to a micro-injector, by which the bubble pressure could be controlled, and the tip of the pipette was placed in a hexadecane oil phase. Lipids were added either in the oil phase (lipid-out) or in the bubble, as liposomes (lipid-in),51) and the lipids were readily transferred to the oil–water interface to form the monolayer. Application of slight pressure (ca. 0.1 kPa) produced a bubble in an oil phase, and steady pressure was applied to maintain the bubble size throughout the experiment. This bubble is stably fixed at the tip of the 30 µm pipette, and is readily manipulated through the micromanipulator.
A. The punch-out method. The lipid bilayer is formed on the hole using the chamber method. Then, a glass pipette “patch-clamps” the bilayer. B. The tip-dip method. At the air–water interface, a monolayer is formed and the tip of the pipette is drawn across the interface. At the first withdrawal from the water phase, a monolayer is transferred to the tip of the pipette. Upon the next crossing, two monolayers are attached to form a bilayer. There are several modified methods involving hydrophobic treatment of glass surface. C. The liposome patch. The giant unilamellar liposome is patch-clamped. D. The bubble patch. A bubble is blown into an electrolyte solution from a plastic tube, and is patch-clumped. In the punch-out and in a modified version of the tip-dip methods, the surface of the glass pipette is coated with silicone.
Bubbles are blown into the hexadecane phase from two glass pipettes with 30 µm diameter tips, and pipette pressure is maintained thereafter (the holding pressure). Membrane lipids are either dispersed in the oil phase (lipid-out) or involved in the bubble as liposomes. The lipids transfer to the bubble surface to form a monolayer. Two bubbles are manipulated into contact in order to connect and form a bilayer.
An asymmetric membrane can be formed when different lipids are added in two droplets as liposomes (the lipid-in method52)).
2.7. CBB vs. DIBCompared to the DIB, there are fundamental differences in the CBB. First, rather than the isolated droplet in the DIB, in CBB, the aqueous phase in the bubble is open to the bulk aqueous phase in the pipette. Thus, the chemical composition, such as ion concentration, can be maintained throughout the experiment, even though the flux of ions between two bubbles was huge. Second, bubbles are maintained by applying steady pressure, and thus the bubble size can be changed by applying either positive or negative pressure. When the monolayer tension is gradually reduced as additional phospholipids partitioned to the interface, the bubble size increases slowly according to the law of Laplace,53,54)
 | (1) |
The CBB method is generally performed under inverted microscopy, whereas DIB is performed under a stereomicroscope. The smaller size and sharper focus of the image in the CBB method allows a clear view, from a tangential direction, of the merger of the bilayer and monolayer.
2.8. The Electrophysiological Features of the CBBCompared to the DIB method, the smaller bilayer in CBB reduces the electrical background noise of the system, and a higher membrane potential, of up to ±400 mV, can be applied.55) Using the kΩ level of the series resistance,56) high-resolution recordings can be performed.57)
In electrophysiological experiments on cells, the membrane potential is defined for the intracellular side, referenced by the extracellular side being zero. In lipid bilayer experiments, the channel-added compartment is considered the cis side (the other side is trans), but the polarity of the membrane potential is defined functionally.58) When channels were incorporated into the membrane with their cytoplasmic sides remaining on the cis side, then the membrane potential is defined for the cis side relative to the trans side.
The electrode potential (the potential arising at the interface between the metal surface and the electrolyte) must be adjusted by applying the offset potential. This is especially important when the electrolyte solution is asymmetrical (Cl− must be contained in both electrolyte solutions). This can be performed by immersing a pair of electrodes in a solution. This is attained by breaking the bilayer membrane through an application of high membrane potential (>1 V; electrical breakdown). Among the lipid bilayer methods presented here, the CBB method is favorable to breaking and reforming the bilayer repeatedly, thereby readily adjusting the offset potential.
Additionally, the liquid junction potential,59) the diffusion potential arising at the liquid interface of different electrolyte compositions, must be adjusted. The junction potential is generally small, and it is readily calculated from the given composition of the electrolyte solution.60) This adjustment is crucially important when the reversal potential is evaluated for ion selectivity.
Channel-membrane interaction has been studied extensively, and studies of the chemical and physical processes involved in this interaction are underway.15,61–64) Here, the chemical interaction of channels with membrane lipids is addressed using a canonical potassium channel, KcsA.65) In the crystal structure of the KcsA potassium channel,66) phosphatidylglycerol (PG) was co-crystallized,67) suggesting that PG may contribute to channel function. Thus, single-channel current recordings of the non-inactivating mutant (E71A68)) of the KcsA channel were performed in a bilayer membrane of pure lipid composition.47,69)
In these studies, channel proteins were reconstituted into liposomes of a distinct lipid coposition, which was included in one of the bubbles. The liposomes were then fused spontaneously to the bilayer, and the channel proteins were reconstituted into the bilayer. This procedure is a standard method for the channel incorporation into the bilayer, and the reconstituted liposomes reach the CBB readily because of the small size of the bubble.
The open probability of the gate was nearly 90% in the pure PG membrane, whereas it was only 10% in the pure PC membrane.47,69) Such a big variation in channel activity was not anticipated by a change in lipid composition.
Next, the sidedness of the lipid composition was examined. A planar lipid bilayer created with the folding method allows for the formation of an asymmetrical membrane,69) whereas the CBB method serves the asymmetric membrane more easily.47) For the formation of an asymmetric membrane using the CBB method,47) the lipid-in method52) is used. The channel incorporated liposomes with a defined lipid composition were included in one of the bubbles, while the other bubble contained liposomes of a different lipid composition (Fig. 7A).
The non-inactivating mutant E71A of the KcsA channel was used for examining the activation gate. Liposomes were formed from either PG (negatively charged) or PC (neutral), and they are suspended in KCl solution. Each bubble was formed in the presence of the liposomes, which extend at the water/oil interface, forming a monolayer of defined lipid composition. The single-channel current recordings of the KcsA channel at +100 mV revealed high open probability (90%) as far as PG exists in the inner leaflet of the membrane. The single-channel conductance was high in the PG-containing membranes because the surface potential generated by negatively charged PG accumulates local K+ concentration.32)
When the inner leaflet was formed with PG and the outer leaflet was formed with PC (PGin/PCout), the open probability is 90%, which is indistinguishable from that in the symmetrical PG membrane (Fig. 7B). In contrast, when the reversed asymmetric bilayer (PCin/PGout) was formed, the open probability is 10%, similar to that in the pure PC membrane. Accordingly, the PG in the membrane inner leaflet keeps the KcsA channel active.47,69) Further mutational study introduced into the N-terminal M0 helix elucidated that the positively charged residues on the M0 helix are responsible for the electrostatic interaction with the negatively charged lipid head groups in the inner leaflet. Fluorescence study revealed a novel mechanism of lipid sensing, such that the amphipathic M0 helix revolves around the helix axis upon opening and closing of the activation gate (the roll-and-stabilize mechanism).69) The role of amphipathic helices at the membrane interface has been previously proposed,70,71) but such dynamic behavior had not been elucidated until now.
3.1. Perfusion of the Aqueous SolutionPerfusion methods of the aqueous solution have been developed for the patch-clamp method,72,73) and a rapid temperature jump was also performed.74) On the other hand, in the chamber method, perfusion was not effective because of the large quantity of the aqueous solution (ca. 1 mL).22) In the case of CBB, the pL size bubble allows for rapid perfusion47) (Fig. 8). A glass micropipette was inserted into one of the bubbles, and the injection of solution was followed by a slight release of the holding pressure of the relevant pipette to allow the solution to drain.47) The solution jump was completed with a time constant of 20 ms. If the second micropipette inserted into the same bubble as the original bubble, then the injection from this pipette returned the solution to its original composition. Two successive injections from two pipettes generated the pulse application of a solute.
A. The perfusion method. A micropipette was inserted into a bubble. Injection of a hydrophilic substance followed by the drain to the bubble-holding pipette yielded rapid perfusion. An additional micropipette allows injection of the original bubble solution, by which the hydrophilic substance can be applied as a pulse-like fashion with an arbitrary pulse duration. B. pH pulse experiment for macroscopic KcsA current. Neutral pH jump immediately attenuated the channel activity with single channel decaying process in the inset. The current recovered immediately upon acidic pH jump.47)
In Fig. 8, the response of the KcsA channel to this pH pulse experiment is shown. The pH of the intracellular side was set as acidic, and a steady macroscopic current was observed. Upon conversion to neutral pH, the channel immediately closed, whereas the channel activity returned to its original level upon its return to an acidic state.
3.2. Perfusion of the “Membrane” InteriorThe water-in-oil droplets system is topologically reversed from the membrane in aqueous solution, and this system is generally governed by the following rules75) (Fig. 9A):
A. The water-in-oil droplet system is governed by three rules. First, one of the leaflets of the bilayer is contiguous to the monolayer lining the bubble, and lipids are freely diffusible between two phases. Second, the hydrophobic membrane interior is contiguous to the surrounding bulk oil phase, and an addition of hydrophobic substances in the oil phase leads to transfer towards the membrane interior. Third, integral membrane proteins remain in the bilayer and never move to the monolayer. B. Membrane perfusion of a hydrophobic substance. C. A representative current trace of the KcsA channel upon cholesterol perfusion.72)
For hydrophobic substances to reach the interior of conventional membrane, they must be dissolved in an aqueous solution, which are then partitioned into the membrane. In the CBB, Rule #2, the membrane interior is open to the bulk oil phase, is applied. Hydrophobic substances are dissolved in hexadecane, and this solution is sprayed over the CBB from the oil phase (Fig. 9B). Highly hydrophobic substances are immediately transferred to the membrane interior through partitioning into either the monolayer or bilayer (Rule #1).
Figure 9C shows an example of this membrane perfusion. During single-channel recording of the non-inactivating mutant (E71A) of the KcsA channel, cholesterol was perfused into the membrane, and the immediate closure of the activation gate was seen. As the perfusion stopped, cholesterol in the membrane redistributed to the bulk oil phase, and the channel activity recovered (Rule #3).75) Such rapid response of the channels to cholesterol in the membrane has never been previously observed.
Advantages of the CBB method include its manipulability of bubbles, by which monolayer-bilayer transitions are allowed. Here we briefly summarize applications of manipulability of the bilayer. Unlike other water-in-oil bilayer methods, the bubbles in CBB are maintained by pressure, and this pressure can be readily changed in either a symmetric or asymmetric manner. For example, bilayer tension has been applied by injecting the solution into a droplet.76,77)
4.1. Detach-Attach of the BilayerTo date, no one has imagined what happens to channel activity when a bilayer is detached and membrane embedded channels are withdrawn from the bilayer. Surely, one would imagine the channel activity would be lost in the absence of the bilayer membrane. However, in the case of the peptide channel polytheonamide B (pTB), channel activity immediately resumed when the monolayers were re-attached.58,78)
pTB is a 48-mer peptide toxin extracted from the marine sponge Theonella swinhoei. pTB is targeted to cell membranes, where it forms a cation-selective channel. Its β6.3 helical structure, with highly hydrophobic N-terminal and relatively hydrophilic C-terminal, allows for an oriented insertion into the membrane, led by the N-terminal.
pTB was added into a bubble, which was then spontaneously inserted into the membrane, where channel activity can be readily measured electrophysiologically (Fig. 9A). When the bilayer was detached, which is similar to the freeze-fracture method of separating the membrane,79) the channel current immediately disappeared, but then appeared again upon re-attaching the monolayers. Repeated attach-detach yielded a repeated on-and-off performance of the current. This strange behavior is generated by the following mechanism. As membrane insertion is led by the hydrophobic N-terminus, the N-terminus stays in the hydrophobic core of the bilayer, and eventually spans the bilayer. On the other hand, the hydrophilic C-terminus remains at the membrane interface of the peptide added side, and pTB forms a pore across the bilayer (Fig. 9B). When the bilayer is detached, the C-terminus is anchored to a water-monolayer interface, whereas the N-terminus is buried in the hydrophobic milieu. When the two monolayers are re-attached, the N-terminal is tucked in the bilayer and able to span the membrane. Even with repeated detach-attach activity, the orientation of the channel was maintained.47)
4.2. One-Bubble ApplicationIn the CBB method, each monolayer can be treated separately prior to bilayer formation. Here we show an example of using a one-bubble application. As shown in the detach-attach procedure, pTB is located at the monolayer interface, with its hydrophilic C-terminus anchored in the aqueous solution. Thus, pTB should be transferred to the monolayer once it has been added in the oil phase. When the two bubbles were separated from each other, and pTB was added exclusively to one of the bubbles, they were both distributed at the monolayer interface. Thus, once two bubbles are attached, the hydrophobic N-terminus spans the bilayer and forms a channel, and the orientation of the channel is determined by the bubble with the pTB addition.75)
A. Detach-attach and channel activities. pTB was added in the right bubble, which was inserted into the bilayer and formed the channel (b, c). Macroscopic outward current was recorded at +200 mV. When detached (a), no channel current flowed. During this process, pTB was uprooted from the bilayer and located at the interface of the right monolayer. Upon re-attachment (c), the current amplitude recovered to the original level. When the membrane potential was changed to −200 mV, the inward current was seen with less amplitude than the outward current (rectification). The detach-attach can be repeated. Application of high membrane potential broke the bilayer, and two bubbles were fused (d), by which pTB originally located at the surface of the right bubble diffused to the left side of the fused bubble. Afterward, upon creating a bilayer, pTB was located in both monolayers, and the orientation of the membrane-inserted pTB became random.47)
Biomembranes are highly complex systems involving bilayer asymmetry and heterogenous phases, in which dynamic interplay between the bilayer and channels is ongoing. Thus, in parallel to studying these complicated biomembranes, the lipid bilayer provides an alternative investigative approach from the perspective of “divide and conquer.” The membrane construction system–with defined lipid compositions reconstituted with purified channels–is a simple system to start with. The recent development of water-in-oil bilayers, including DIB and CBB, allows researchers to manipulate the bilayer more freely. Various maneuvers in the monolayer, such as forming an asymmetric membrane, as well as mechanical maneuvers, have opened to researchers an unprecedented variety of experimental venues for studying channel-membrane interactions. Membrane manipulation is not limited to physical and chemical procedures conducted on the bilayer itself, but now include alterations to membrane-incorporated channels, such as channel forming substances, which may deform the membrane upon channel formation.80–82) Our picture of channel-membrane interplay is more and more realistic through touching the membrane and designing more sophisticated membranes.
The author acknowledges JSPS KAKENHI (17H04017), Challenging Exploratory Research (16K15179) and Grant-in-Aid for Scientific Research on Innovative Areas (16H00759).
The authors declare no conflict of interest.
