Invited Feature Article pubs.acs.org/Langmuir
Perspectives: Synthetic Bilayer Membrane and Giant Nanomembrane Toyoki Kunitake* Kyushu University, 744 Moto-oka, Nishi-ku, Fukuoka 819-0395, Japan ABSTRACT: This is a short story of the beginning of our research on bilayer assembly and its subsequent developments. Inspired by the fluid mosaic model of the cell membrane, we started in the mid 1970’s to look for the synthetic counterpart of membrane-forming biolipids. It was soon found by electron microscopy that the stable molecular bilayer was formed from double−chain ammonium amphiphiles. This finding was extended to a number of related types of compounds, thus establishing that the bilayer formation itself was general physicochemical phenomena. The same concept was applicable to non-aqueous systems. The molecular scale thicknessone of the unique features of the biological membranewas partially reproduced by development of giant nanomembranes. Finally, prospects and some unsolved problems are discussed.
1. BEGINNING In the early 1960s, I started my research on enzyme-like polymers. The design of enzyme models was a hot research topic at that time because the molecular mechanism of miraculous enzyme catalysis was just being unveiled. Our first successful design of the model system of α-chymotrypsin was copolymers of vinylimidazole and vinylpyrrolidone, which catalyzed the hydrolysis of p-nitrophenyl acetate, a popular ester substrate, according to the enzyme-like Michaelis− Menten kinetics.1 We could subsequently develop highly effective, multifunctional polymer catalysts as models of the multifunctional active site of α-chymotrypsin.2 Such enzyme models are chemically interesting and challenging, but their activities and specificities were inferior to those of natural enzymes. One of the intrinsic differences was apparently the lack of precise molecular organization in the artificial counterpart. Natural protein molecules are made of intricate folding of polypeptide chains, but the artificial polymer catalysts exist as random coil without a precise steric arrangement of functional units. Therefore, we started to search for artificial systems of organized molecules. Aqueous micelles assume globular shapes, but component surfactant molecules exist only as mobile aggregates. Liquid crystals were interesting candidates, but their organization was stable only on the macroscopic scale. A turning point came when I happened to read a publication by Singer and Nicolson on the fluid mosaic model of a biological membrane.3 This model properly explained the thenknown basic features of the biological membrane. In this model, fluid bilayers are formed from component lipid molecules, and mobile protein molecules are buried in the lipid bilayer as a mosaic pattern. I was curious as to why the biological lipid (particularly lecithin) should be unique in providing the stable bilayer structure. My hypothesis was that the unique molecular feature of the biological lipid came from the biological requirements: available precursor molecules, biosynthetic © XXXX American Chemical Society
pathway, and biological decomposition. If that is the case, then we should be able to find artificial molecules that give bilayer structures similar to the biological one. In 1976, we started to search for amphiphilic compounds that would produce spontaneously organized molecular aggregates.
2. DISCOVERY The presence of the bilayer lipid leaflet had been reported on the basis of electron microscopic observation, and various features of the bilayer or black lipid membrane (BLM) were studied by the unique pinhole setup as a model of a biological membrane.4 The fluid mosaic model proposes that membrane stability is maintained by the lipid bilayer alone without the support of a protein layer. In fact, the formation of liquidcrystalline aggregates had been observed by electron microscopy. It was also necessary, in our case, to find direct evidence for the existence of a single leaflet of a molecular bilayer, and electron microscopy was the only means for this purpose. Bilayer-forming lipid molecules are characterized by a combination of a polar headgroup and nonpolar double-alkyl chains. We thus selected double-chain ammonium compounds as the simplest artificial model in our first trial because such compounds were well studied for their aggregation properties. These compounds gave transparent aqueous dispersions by sonication. After intensive efforts for several months, Dr. Okahata, our co-worker at that time, succeeded in finding vesicular objects with a bilayer sheath in an electron micrograph. Other less-direct measurements were consistent Special Issue: Tribute to Toyoki Kunitake, Pioneer in Molecular Assembly Received: May 29, 2016 Revised: August 26, 2016
A
DOI: 10.1021/acs.langmuir.6b02030 Langmuir XXXX, XXX, XXX−XXX
Langmuir
Invited Feature Article
with the formation of aqueous vesicles.5 This was the first example of totally synthetic bilayer membranes (Figure 1).
lost interests in such studies and instead were attracted more to the basic principle of membrane formation. At first, we simply assumed that our stable bilayer dispersion was a simplified model membrane of biolipids. However, a question arose. The unique molecular structure of biolipids must be restricted mostly by the biological requirement and not necessarily by the abiotic physicochemical requirement. The latter requirement is composed of two items: (1) superior alignment of highly aggregating molecules and (2) sufficient stability of the interface between molecular-scale assembly and the surrounding medium. Ammonium and other polar headgroups such as sulfonate and oxyethylene chain could provide sufficient interfacial stability for single-leaflet bilayer assembly. In contrast, it is not clear why the double-chain portion is the indispensable unit for 2D molecular ordering. Two-dimensional molecular alignment may be achieved by other types of hydrophobic units. Thus, we examined other types of amphiphilic compounds as shown in Figure 2. Spontaneous bilayer aggregation is observed in water as a result of improved molecular alignment, when hard molecular segmentsbiphenyl and an aromatic Shiff base as in thermotropic liquid-crystalline materialsare introduced into the middle of the alkyl chain potion. Triple-chain ammonium amphiphiles may appear to produce reverse micelles. This is not the case, however, when 2D ordering is preferred by dint of connecter units of proper chemical structure. We used the same logic to design a four-chain ammonium amphiphile that is schematically illustrated in Figure 2. The trifunctional glutamic acid unit was used doubly as one connector. These results unambiguously confirmed that bilayer organization is based on the above-mentioned physicochemical requirements, and the unique chemical structure of biolipids is a result of the additional biological requirements.8 Further generalization of bilayer formation was possible. An aqueous environment is most suitable for creating effective molecular assemblies because hydrophobic effects are quite efficient for inducing the aggregation of organic molecules. Such aggregation is feasible in other fluid media, as shown by old examples of micellar formation in organic solvents. It has been known that fluorocarbons are not readily miscible with hydrocarbons. Such immiscibility led to the use of solvophobicity in spontaneous bilayer formation. As shown in Figure 3, new types of amphiphiles were synthesized, where one type of amphiphile possesses a combination of two fluorocarbon chains as a solvophobic unit and one flexible alkyl chain as a solvophilic unit. This amphiphile forms a bilayer structure spontaneously in organic media.9 In the third type of bilayer, a hydrophilic head and hydrophobic alkyl chains are coupled by
Figure 1. Electron micrograph of bilayer vesicles. Aqueous dispersion of didodecyldimethylammonium bromide.
It must be pointed out that Tanford previously discussed that the steric balance of hydrophilic and hydrophobic segments was critical in determining the aggregate morphology of alkylammonium amphiphiles: globular and layered aggregation arises from amphiphiles with single and double alkyl chains, respectively.6 Shinoda and Kunieda in Japan also reported the liquid-crystalline nature of aqueous double-chain ammonium amphiphiles. These studies were concerned with a thermodynamically stable lamellae dispersion but did not treat thermodynamically metastable bilayer vesicles. The cell membrane and its artificial counterpart (synthetic bilayer membrane) are never thermodynamically stable entities, although they have lifetimes sufficient for acting as functional structures. Fortunately, our finding attracted the interest of other chemists, and many publications followed. A monograph that covered this new research area was published in 1982 with the main title of Membrane Mimetic Chemistry by Janos Fendler.7
3. GENERALIZATION Synthetic bilayer vesicles were convenient molecular systems for modeling biological events occurring on cell membranes. We therefore tested various cell-model chemical processes to explore this new research area. The study of catalytic hydrolysis in and on the synthetic bilayer membrane was the first case in relation to the membrane characteristics. Photochemistry and energy transfer were other examples. However, we gradually
Figure 2. Molecular modules of one-, two-, three-, and four-chain amphiphiles. B
DOI: 10.1021/acs.langmuir.6b02030 Langmuir XXXX, XXX, XXX−XXX
Langmuir
Invited Feature Article
Figure 3. Three motifs of bilayer membranes. Figure 4. Epoxy nanomembrane (3 cm × 3 cm, thickness 50 nm) being sucked into a pipet hole.
complementary hydrogen bonding. That is, noncovalent hydrogen bonds are good enough to maintaining the stable bilayer organization.
bilayer is versatile, but their modes of assembly are much more restricted. From a materials standpoint, the bilayer membrane is unique because of its two major characteristics: regular molecular ordering and ultimate thinness. These features provide interesting hints for the design of innovative molecular systems. As for molecular ordering, we previously found that spectral properties and energy transfer efficiency depended specifically on the mode of chromophore orientation in the azobenzenecontaining bilayers. Unfortunately, such observation has never been applied to problems of practical interest until recently. This situation is now changing. Kimizuka and co-workers found that the triplet state was very effectively transmitted among well-designed, organized molecules, leading to highly efficient photon up-conversion via triplet−triplet annihilation.11 This breakthrough would much improve the use of solar energy by covering the wide spectral region. In the case of ultimate thinness, the sub-100-nm thickness of a giant nanomembrane could be a superior scaffold for constructing membrane functions that are based on individual molecular events, as found with the biological membrane. Totally artificial molecular devices may be advantageously constructed on/in the ultrathin support. Molecular events are thus directly connected to macroscopic functions, thanks to size matching. Another advantage is its use as a permselective membrane. In the case of nonporous nanomembranes, the microphase separation of mixed homopolymers and block copolymers could lead to interesting membranes that are robust as well as functional. The sub-100-nm thickness of the nanomembrane is advantageous for channel formation from functional nanodomains and molecules. We have already found unique proton-conducting behavior for a sub-100-nm aluminosilicate membrane and explained it by a percolation model.12 Electron microscopy proved the presence of two different nanodomains in this case, and the intrinsic proton conductivity was enhanced with decreasing membrane thickness as a result of better connection of the active domain. The engineering of nanopores may also be possible by the molecular imprinting technique.13 The designed pore can be size-specific and/or affinity-based. In either case, percolation theory predicts the preferred channel formation from nanometer-thick membranes. A straight molecular channel across the membrane is an interesting possibility. The mean-free-path of gas molecules at atmospheric pressure is estimated to be 30−100 nm; therefore, sub-100-nm membranes could provide a site for the ballistic
4. GIANT NANOMEMBRANE As described above, our studies had been directed mostly to the principles of bilayer formation. Versatile functions of the biological system are often related to the unique assembly of the biological membrane, and there is a fantastic possibility that the artificial bilayer membrane is useful for mimicking or replacing the role of the biological membrane. There are a few examples by other research groups in that direction, but our research took a different course. One of the outstanding aspects of biomembranes from a practical viewpoint (apart from medical uses) is its robustness in spite of its extreme thinness of 5−10 nm. If such robustness is realized in macroscopic size, then the application potential would be enormous. In the case of biomembranes, the supporting protein network is often essential for robustness. Similarly, molecular cross-linking is conceivably effective for preparing robust nanometer-thick membranes of macroscopic size. Our first successful case was the formation of nanometer-thick films of (highly cross-linked) metal oxides and their isolation from underlying substrates. These inorganic films of less than 100 nm thick were freestanding and flexible. We extended this experience to organic/ inorganic hybrid precursors and thermosetting resins, and in all cases, stable nanometer-thick membranes were obtainable.10 The aspect ratio (size/thickness) of these membranes can be greater than one million, and that is why we call them giant nanomembranes. The mechanical property of a giant nanomembrane of an epoxy resin, for example, was similar to that of the corresponding macroscopic resin after thickness correction. Apparently, the basic mechanical property remains unchanged by thickness reduction to 10 nm. 5. PROSPECTS AND UNSOLVED PROBLEMS It is interesting to compare the general feature of a 2Dstructured bilayer membrane and a 1D linear chain of polymers in terms of artificial or natural origins. Natural polymers such as proteins, nucleic acids, and polysaccharides are composed of limited types of unit molecular structures, but the mode of combination of those structural units is highly specific and of extremely rich variety. In contrast, the material base of synthetic polymers is quite versatile whereas its mode of combination is rather limited (e.g., homopolymers and copolymers). A similar contrast is found between natural and synthetic bilayer membranes. The kind of component molecules of the artificial C
DOI: 10.1021/acs.langmuir.6b02030 Langmuir XXXX, XXX, XXX−XXX
Langmuir
Invited Feature Article
transport of gas molecules. It will be controlled by its interaction with chemically modified pore walls.
■
AUTHOR INFORMATION
Corresponding Author
*Tel +81-93-400-6882. E-mail
[email protected]. Notes
The author declares no competing financial interest.
■
ACKNOWLEDGMENTS This article is a summary of the extensive research activity of our group at Kyushu University and at Institute of Physical and Chemical Research. The author extends sincere thanks to the named and unnamed co-workers for their dedication.
■
REFERENCES
(1) Kunitake, T.; Shimada, F.; Aso, C. Imidazole Catalyses in Aqueous Systems. I. The Enzyme-Like Catalysis in the Hydrolysis of a Phenyl Ester by Imidazole-Containing Copolymers. J. Am. Chem. Soc. 1969, 91, 2716−2723. (2) Kunitake, T.; Okahata, Y. Multifunctional Hydrolytic Catalyses. IV. The Catalytic Hydrolysis of p-Nitrophenyl Acetate by Copolymers Containing Complementary Functional Groups (Hydroxamate and Imidazole). Macromolecules 1976, 9, 15−22. (3) Singer, S. J.; Nicolson, G. L. The fluid mosaic model of the structure of cell membranes. Science 1972, 175, 720−731. (4) Ti Tien, H. Bilayer Lipid Membranes (BLM) Theory and Practice; Marcel Dekker: New York, 1974; Chapter 1. (5) Kunitake, T.; Okahata, Y. A Totally Synthetic Bilayer membrane. J. Am. Chem. Soc. 1977, 99, 3860. (6) Tanford, C. The Hydrophobic Effect: Formation of Micelles and Biological Membranes, 2nd ed.; Wiley-Interscience, 1980; Chapter 3. (7) Fendler, J. H. Membrane Mimetic Chemistry: Characterizations and Applications of Micelles, Microemulsions, Monolayers, Bilayers, Vesicles, Host-Guest Systems, and Polyions; Wiley-Interscience, 1982. (8) Kunitake, T. Synthetic Bilayer Membranes: Molecular Design and Self Organization. Angew. Chem., Int. Ed. Engl. 1992, 31, 709−726. (9) Ishikawa, Y.; Kuwahara, H.; Kunitake, T. Self-Assembly of Bilayers from Double-Chain Fluorocarbon Amphiphiles in Aprotic Organic Solvents: Thermodynamic Origin and Generalization of the Bilayer Assembly. J. Am. Chem. Soc. 1994, 116, 5579−5591. (10) (a) Vendamme, R.; Onoue, S.; Nakao, A.; Kunitake, T. Robust Free-Standing Nanomenbranes of Organic/Inorganic Interpenetration Networks. Nat. Mater. 2006, 5, 494−501. (b) Watanabe, H.; Vendamme, R.; Kunitake, T. Development of Fabrication of Giant Nanomembranes. Bull. Chem. Soc. Jpn. 2007, 80, 433−440. (11) Ogawa, T.; Yanai, N.; Monguzzi, A.; Kimizuka, N. Highly Efficient Photon Upconversion in Self-Assembled Light-Harvesting Molecular Systems. Sci. Rep. 2015, 5, 10882. (12) Aoki, Y.; Muto, E.; Nakao, A.; Kunitake, T. Efficient Proton Conductivity of Gas-Tight Nanomembranes of Silica-Based Double Oxides. Adv. Mater. 2008, 20, 4387−4393. (13) Lee, S.-W.; Ichinose, I.; Kunitake, T. Molecular Imprinting of Azobenzene Carboxylic Acid on a TiO2 Ultrathin Film by the Surface Sol-Gel Process. Langmuir 1998, 14, 2857−2863.
D
DOI: 10.1021/acs.langmuir.6b02030 Langmuir XXXX, XXX, XXX−XXX
