Lipophilic nucleic acids--a flexible construction kit for organization and functionalization of surfaces.

Advances in Colloid and Interface Science 208 (2014) 235–251

Contents lists available at ScienceDirect

Advances in Colloid and Interface Science journal homepage: www.elsevier.com/locate/cis

Lipophilic nucleic acids — A flexible construction kit for organization and functionalization of surfaces Matthias Schade a, Debora Berti b, Daniel Huster c, Andreas Herrmann a, Anna Arbuzova a,⁎ a b c

Humboldt-Universität zu Berlin, Institut für Biologie, Invalidenstr. 42, 10115 Berlin, Germany Dipartimento di Chimica, Universita' di Firenze & CSGI, Via della Lastruccia 3, 50019 Sesto Fiorentino, Firenze, Italy Universität Leipzig, Institut für Medizinische Physik und Biophysik, Härtelstr. 16-18, 04107 Leipzig, Germany

a r t i c l e

i n f o

Available online 5 March 2014 Keywords: Lipophilic nucleotides Self-assembly DNA Liposomes Membranes Cells

a b s t r a c t Lipophilic nucleic acids have become a versatile tool for structuring and functionalization of lipid bilayers and biological membranes as well as cargo vehicles to transport and deliver bioactive compounds, like interference RNA, into cells by taking advantage of reversible hybridization with complementary strands. This contribution reviews the different types of conjugates of lipophilic nucleic acids, and their physicochemical and self-assembly properties. Strategies for choosing a nucleic acid, lipophilic modification, and linker are discussed. Interaction with lipid membranes and its stability, dynamic structure and assembly of lipophilic nucleic acids upon embedding into biological membranes are specific points of the review. A large diversity of conjugates including lipophilic peptide nucleic acid and siRNA provides tailored solutions for specific applications in bio- and nanotechnology as well as in cell biology and medicine, as illustrated through some selected examples. © 2014 Elsevier B.V. All rights reserved.

Contents 1.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1. Dynamical assembly due to weak interactions . . . . . . . . . 1.2. Lipophilic nucleic acids . . . . . . . . . . . . . . . . . . . 1.3. Self-assembly of LiNAs . . . . . . . . . . . . . . . . . . . . 2. LiNAs and membranes . . . . . . . . . . . . . . . . . . . . . . . 2.1. Binding energies of various lipophilic anchors . . . . . . . . . 2.2. Binding of LiNA to model membranes . . . . . . . . . . . . . 2.3. Formation of Watson–Crick base pairs . . . . . . . . . . . . . 2.4. Thermodynamic stability of hybrids from modified molecules . . 3. Organization on the membrane surface . . . . . . . . . . . . . . . 3.1. Membrane interaction of lipophilic nucleosides and nucleotides . 3.2. Orientation of nucleic acids on the surface of model membranes . 3.3. Functionalization of membranes with LiNA . . . . . . . . . . 3.4. DNA structures and origami on membranes . . . . . . . . . . 3.5. Lateral distribution of LiNA in model membranes . . . . . . . . 3.6. Lipophilic nucleic acids as a model of SNARE protein fusion system 3.7. Interaction of LiNA with cells . . . . . . . . . . . . . . . . . 4. Conclusion and outlook . . . . . . . . . . . . . . . . . . . . . . . 5. Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

⁎ Corresponding author. E-mail address: [email protected] (A. Arbuzova).

http://dx.doi.org/10.1016/j.cis.2014.02.019 0001-8686/© 2014 Elsevier B.V. All rights reserved.

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

236 236 236 237 239 239 240 241 241 242 242 243 243 243 245 247 247 248 248 249 249

236

M. Schade et al. / Advances in Colloid and Interface Science 208 (2014) 235–251

1. Introduction 1.1. Dynamical assembly due to weak interactions Life is based on dynamical supramolecular assembly ruled by the interplay of many weak intermolecular interactions. Constant reformation of the weak interactions allows control and adjustment for the appropriate structure, whereas strong interactions would produce amorphous static structures [1,2]. While the covalent bonds between atoms have interaction energies of 100–400 kJ/mol, those of weak forces are in the order of or slightly higher than a few kBT (2.5 kJ/mol at room temperature, with kB the Boltzmann constant and T the absolute temperature). Weak forces include hydrogen bonding (4–120 kJ/mol), van der Waals interactions (b 5 kJ/mol), π stacking (b1–50 kJ/mol), forces that arise due to the hydrophobic effect, and to some degree electrostatic forces. These interactions are responsible for numerous structures found in nature, but also for the recognition of chemical signals and their translation into a biological response. One important example demonstrating the structural power of weak forces is the hydrophobically driven selfassembly of lipids and detergents into membranes, micelles or other supramolecular assemblies [3]. Weak forces are not only sufficient for formation of these assemblies but are also responsible to provide them with appropriate lateral fluidity, they allow for shape changes in the process of spontaneous curvature formation, endo- or exocytosis, caveolae formation, and virus entry into cells [4–7]. They also allow for binding and dissociation of molecules, e.g. to lipid membranes, without very large energy barriers separating the two states. The balance of a wealth of single weak interactions leads in some other cases to bistability phenomena, with coexistence of different assemblies (i.e. phase-separated lipid domains in fluid bilayers, raft-like domains [8]). In general, most biological functions are accomplished by a molecular switch mechanism, based on the interplay between multiple weak forces, e.g. protein complex formation, oligomerization, change of phosphorylation state, or reversible lipid modification. Also, nucleic acid strands are held together by multiple hydrogen bonds and π-stacking giving rise to a large variety of DNA and RNA structures, while simultaneously allowing for the easy separation of DNA in the process of transcription or cell division. A chemical modification of nucleic acids permits further degree of organization, e.g. an attachment of a hydrophobic moiety allows assembly in micelles or presentation on lipid membrane surfaces. 1.2. Lipophilic nucleic acids A lipophilic conjugate is a chemical structure of one or many hydrophobic (lipid-like) moieties with a polar molecule, e.g. nucleoside, nucleotide, peptide, protein, nucleic acid, or synthetic polymer. Some lipophilic conjugates exist in nature, e.g. cytidine diphosphate diacylglycerol is a precursor of phosphatidylglycerol and phosphatidylinositol synthesis [9] and lipid modified proteins constitute an essential part of cell-signaling system [10–12]. Other conjugates are synthesized de novo on demand as described in, e.g. Rosemeyer, 2005; Kaczmarek, 2008; Gissot, Barthélémy, 2008; Berti, 2011, Soft Matter; and Allain, Couvreur, 2012 [9,13–16]. Another example of a functional head group conjugated to lipids is nitrilotriacetic acid (NTA), a well characterized chelator that allows isolation of His-tagged proteins. Starting from the first lipophilic modification of NTA introduced by Tampe and coworkers [17], multivalent conjugates with highaffinity for His-tagged protein were developed [18] and further tested for analytical and drug delivery applications [19–21]. Recently protein targeting into cell rafts by a multivalent NTA lipophilic conjugate was reported [22]. The discovery of natural RNAzymes and synthesis of DNAzymes, DNA origami, and aptamers showed clearly the versatility of nucleic

acid structures [23–26]. The double helix size, its bending rigidity, the high fidelity and cooperativity of base pairing, provide a material with high (subnanometer) spatial resolution and addressability. Moreover the paired structures can be (dis)-assembled via reversible thermal cycling in accessible temperature ranges (20–80 °C). Nucleic acid nanostructures can be functionalized with single or multiple chemical units, whose positioning and distance within the nano-constructs are programmed by design. Conjugation of lipophilic moieties with a nucleic acid results in lipophilic nucleic acid (LiNA) combining self-assembly properties of the anchors and specific recognition of the nucleic acid strand. Some LiNA selfassemble into defined structures capable of enhanced affinity binding due to multivalency, e.g. aptamer-presenting micelles [27], and others allow easy and controllable functionalization of surfaces, including cell membranes [15,28–34]. LiNA were used to enhance gene delivery and gene silencing by siRNA and antisense PNA [35–39]. LiNA were found to inhibit viruses, e.g. Hepatitis C virus translation [37] and HIV fusion with cells [40]. They were used for development of drug delivery systems [16,27,41], as a detection tool for microRNA in living cells [42], and for DNA detection [43,44]. For recent reviews describing applications of LiNA see [45,46]. In this review, we discuss biophysical aspects of lipophilic nucleic acid self-organization, interaction with lipid membranes, and formation of structures on the surfaces. A LiNA conjugate consists of an anchor, linker, and a molecular recognition unit (nucleic acid). Fig. 1 presents several examples of lipophilic conjugates used in recent studies, which will be discussed below. Different types of backbone structures were used as shown in Fig. 1 (B1, B2). The respective aim of a study (and availability) determines the choice of the nucleic acids. DNA and RNA can be easily synthesized, and their charged phosphate groups make them well water soluble. Solubility in water is important if working in aqueous solutions, especially for living systems. DNA is chemically more stable than RNA and lipophilic DNA is often used in biophysical studies, assembly studies, and for building of nucleic acid detection devices [47]. RNA is more vulnerable to enzymatic degradation by RNases that are both ubiquitous and highly resistant against denaturation, but has a higher thermal stability than DNA [48]. RNA is a regulating molecule in cells and, therefore, lipid modifications of antisense, micro, and small interfering RNA were studied [45]. Stability against nucleases can be significantly enhanced while increasing target affinity and specificity by attaching or substituting nucleobases with artificial nucleobases, such as locked nucleic acid (LNA), which is stable even in cellular environment [49,50]. One of the most stable forms of nucleic acid is peptide nucleic acid (PNA). It consists of a peptide backbone conjugated to the canonical nucleobases [51] (Fig. 1). PNA is not only chemically highly stable [52], there are also no enzymes known capable of degrading PNA [38]. Also, DNA/PNA hybrids are highly insensitive to changes in ionic strength [53]. Lacking phosphate groups, PNAs are uncharged and natively hardly water soluble. Solubility of PNA can be significantly enhanced by extending the PNA strand with charged amino acids [54]. To prevent self-aggregation of neutral PNA, a negatively charged linker is reasonable to use. Addition of a total of four glutamic acids to the sequence was used recently for improving solubility of dipalmitoylated PNA [30,55]. Further increase in solubility can be achieved by prehybridizing a lipophilic PNA with the charged complementary DNA. Choosing an anchor, a linker, and a nucleic acid adequate for the aim of a study is always a compromise between the probability of a conjugate to aggregate, kinetics and stability of incorporation into membranes, and accessibility of the nucleic acid for hybridization with complementary strands. Linkage of a lipophilic anchor is now possible to any position of a DNA or RNA sequence, e.g. using phosphoramidites of the lipophilic nucleotides or “click chemistry”. Often just 5′end or 3′ end are modified [28,56]. Commercially available 5′end or 3′end modifications with tri-ethylene-glycol cholesteryl (TEG-cholesteryl) are used often for membrane targeting and cell uptake studies as synthesis of


237

Fig. 1. Composition and components of lipophilic nucleic acids. (A1) Schematic presentation of LiNA interaction with model membrane: two different lipophilic nucleic acids with lateral segregation into either the liquid disordered (ld, left) or liquid ordered (lo, right) domain of a lipid membrane are depicted as three component sketch: “nucleic acid”–“linker”–“anchor”; cholesterol is enriched in the lo domain and shown in dark gray (for details see below the Lateral distribution of LiNA in model membranes section 3.5). Nucleic acids: deoxyribonucleic acid (DNA, B1), peptide nucleic acid (PNA, B2), and locked nucleic acid (LNA, B3); only the respective backbones are shown; ‘b’ denotes any canonical nucleic base. Two linker types: polyethylene glycol (C1) and polyphenylethynylene (C2). End-groups are denoted as NA and A for nucleic acid and anchor, respectively. Typical lipophilic moieties are shown as α-tocopherol (D1), cholesteryl (D2), palmitoyl (D3), and porphyrin (D4).

LiNA [15,57,58] requires chemical expertise. Both ends of the sequence can be also modified, e.g. for studies on controlled aggregation of liposomes [59–62]. When an anchor is introduced into the central part of the nucleic acid sequence, part of the DNA can be oriented more parallel to the membrane [63,64]. Furthermore, Okahata and coworkers [65] reported on a self-aggregating lipophilic construct, in which every phosphate group was modified with azetidinium lipids. A hydrophobic anchor can allow incorporation into lipid membranes, which can be reversible or stable for the time of the experiments; however the anchor can also cause aggregation of conjugates, which if stable prevents incorporation. Additionally, a hydrophobic anchor could flip back to the nucleic acid and bind to the hydrophobic cleft [24,63]. The latter can be especially problematic for lipophilic PNA, which itself is rather hydrophobic [30]. There are several strategies either to prevent undesirable aggregation and back flipping or to use the effects for regulation of processes. The first strategy is using a rigid anchor which cannot flip back and which also decreases a probability of aggregation. A drawback is that a rigid anchor might lead to decreased accessibility for hybridization. The second strategy is using a shorter anchor, however, this leads to a weaker, reversible attachment of the conjugates to a lipid membrane; therefore, two shorter anchors are often required to allow sufficiently high and stable incorporation into lipid membranes for the time of experiment [28,30,66]. The third strategy is to pre-hybridize the LiNA with the corresponding complementary strand before addition to the membranes: hybridization changes the conformation of NA to a stiff rod and the anchor cannot bind to the hybrid [63]. Taking into account electrostatic repulsion of DNA/RNA from zwitterionic or negatively charged lipid membranes,

one should consider a longer neutral linker such as TEG or HEG, bringing the charged nucleic acid further away from the membrane, hence, decreasing the electrostatic repulsion. Using a longer linker is also often required for an assembly of nanostructures in order to provide a sufficiently flexible kink between the double-stranded segments allowing for the specific special organization. 1.3. Self-assembly of LiNAs The intrinsically amphiphilic nature of LiNA calls for a deeper understanding of their self-assembly in solution; first of all, the aggregational properties of the bio-conjugates present a fundamental interest for soft matter science, because they shed light on how the presence of a nucleic polar head affects the overall phase behavior and the morphology of the assemblies. The second reason is on more practical grounds that the self-aggregation of LiNAs will compete with insertion into preformed bilayers, resulting in a partition between the lipid phase and LiNAs assemblies. Amphiphilic molecules self-assemble in water to form a rich variety of microstructures, like globular or cylindrical micelles, vesicles, and liquid crystals. According to the classical approach [3,67], it is the geometry of the individual amphiphile that governs the aggregate morphology and the final shape of the assemblies. In addition, self-assemblies are intrinsically dynamic in nature due to the continuous exchange between the non-aggregated monomers and those forming the aggregate. The residence times in the aggregate vary dramatically, according to the morphology of the assembly, being in the microsecond regime in classical surfactant micelles [68] up to

238


several hours for the desorption of lipid molecules from a bilayer of vesicles [69]. Therefore, the insertion of a LiNA into the host phospholipid membrane will be ruled by the thermodynamics of LiNA self-aggregation and will be affected by the typical residence times of LiNAs in their self-assemblies. Additionally, both monomer and micelles can interact with the bilayer [70]. Therefore, the changes in composition of the bilayer and the final presentation of LiNAs on the surface will depend directly on the properties of LiNAs' assemblies in solution. Concerning the fundamental aspects, the investigations reported on LiNA self-assemblies in the last decade have mainly focused on derivatives with a single nucleoside or nucleotide on the polar head, often termed nucleolipids. In this case, the phase behavior is mainly dictated by the choice of the hydrophobic block and of the linker, if present. These LiNAs are classical amphiphilic molecules, with an additional degree of freedom for the phase behavior, provided by the stacking and H-bonding properties of the polar head. Therefore, while the balance between the cross sections of the polar and apolar portions dictates the nature of the aggregates, the nucleic hydrophilic corona provides further properties in terms of lateral interactions within the aggregate and of possible interactions with single and double strands of nucleic acids in solution. The aggregation properties of nucleolipids have been reviewed recently [15,45,71] and the interested reader is referred to these contributions.

Although the microstructure of the assemblies is mainly dictated by the hydrophobic building block, several interesting properties arise from the presence of the nucleotide polar head. For instance, 1,2lauroyl-phosphatidyl nucleosides in aqueous solutions aggregate in locally cylindrical aggregates, as predictable from geometric considerations on the amphiphile molecular structure and MD simulations [72] (Fig. 2, top panel); however, changing the polar head from uridine to adenosine switches the self-assembly from flexible wormlike micelles to twisted micellar ribbons with evident hierarchical interactions between micellar units (Fig. 2, bottom panel). Therefore, the rather minimal chemical variation, due to higher self-stacking constants of adenosine with respect to uridine has a cascade effect on the mesoscale and is a clear example of the synergy of intermolecular interactions in selfassemblies. Recently, the self-organization of two anionic nucleolipids, thymidine3′-(1,2-dipalmitoyl-sn-glycero-3-phosphate) and adenosine-3′(1,2-dipalmitoyl-sn-glycero-3-phosphate), with triethylammonium [(C2H5)3NH+] as the counterion, was studied at the air/water interface through Langmuir isotherms, imaging ellipsometry and polarization modulation infrared reflection absorption spectroscopy. The authors observed the formation of a quasi-hexagonal packing of bilayer domains at relatively low surface pressures in water and Tris–Ca 2 + buffer, while they assemble as a simple monolayer in Tris–Na + buffer. The most probable organization of the bilayers

Fig. 2. Supramolecular assemblies of mononucleoside LiNA. Top panels: 1,2-dilauroyl-phosphatidyl-adenosine molecules arranged in cylindrical micelles, as obtained from MD simulations. Starting from a random spatial distribution (A), 60 molecules self-assemble in the first 4 ns of simulation as a cylindrical micelle (B and C, cross section and side views, respectively). The MD result, i.e. the preference towards locally cylindrical packing, is in full agreement with experimental results from scattering and Cryo-TEM techniques. Bottom panels: Cryo-TEM images (bar = 100 nm) of the aggregates formed by 1,2-dilauroyl-phosphatidyl-adenosine and 1,2-dilauroyl-phosphatidyl-uridine in PBS; though both locally cylindrical, the assemblies show dramatically different morphology on the mesoscale. Adapted from “Collective headgroup conformational transition in twisted micellar superstructures”, F. B. Bombelli, D. Berti, S. Milani, M. Lagi, P. Barbaro,G. Karlsson, A. Brandt, P. Baglioni, Soft Matter, 2008, 4, 1102–1113 with permission from The Royal Society of Chemistry and D. Berti, F. Baldelli Bombelli, M. Fortini, P. Baglioni; Amphiphilic Self-Assemblies Decorated by Nucleobases, J. Phys. Chem. B 2007, 111, 11734–11744. Copyright © 2007 American Chemical Society.


with respect to the interface seems to be an inverted bilayer, where the two leaflets are joined through the polar head portions [73]. A switch of counterion, from triethylammonium to Na+ is sufficient to disrupt the bilayers, whose formation is clearly a very delicate balance of hydrophobic, H-bonding and stacking interactions. For a lipophilic oligonucleotide (LiON) the phase behavior is also modulated by the hybridization state. For instance, a vesicle-tomicelle transition has been observed for the assemblies formed by a 9mer single-stranded (ss) DNA1 conjugated to a 3,4-di(octadecyloxy) benzoic acid, when a copolymer PEG-DNA2 is added to the dispersion, Fig. 3 [74]. The ss DNA1 is relatively short with respect to the apolar portion and drives the aggregation towards bilayers, with zero interfacial curvature; the hybridization increases the charge density and the steric hindrance of the hydrophilic portion and favors higher interfacial curvature, i.e. micellar assemblies. Likewise, Pokholenko et al. [41] have shown that a hydrophobic drug (paclitaxel), trapped in the apolar compartment of lipophilic LiONs micelles, was released from the micelles upon hybridization with the complement. This study provides a nice proof-of-principle for a drug delivery system, where the payload release is triggered by a mild yet highly selective stimulus, i.e. hybridization. Hence, the molecular design of the amphiphilic architecture is central both in terms of phase behavior and of responsitivity to external stimuli, i.e. temperature, ionic strength and, in this particular case, the presence of the complementary strand in solution. The same is true also when LiNA has to be inserted into a fluid surface, like a lipid membrane, for two reasons. First, the hydrophobic anchor should partition into the lipid bilayer, as discussed in Section 2. Second, the extent and nature of the aggregation state can determine kinetics and thermodynamics of the insertion into the bilayer. As more and more LiNAs insert into the fluid interface, further inclusion of highly charged amphiphilic molecules can be slowed down or even prevented. In other words, there will be a saturation threshold that can be expressed as number density of LiNAs per area of interface (or per lipid). These aspects

239

were studied using oligonucleotide (ON)-cholesteryl derivatives [75] as discussed below in Section 2.2. 2. LiNAs and membranes 2.1. Binding energies of various lipophilic anchors Membrane association by lipophilic groups that insert into the membrane is a common motif in nature in particular for proteins involved in signal transduction [76,77]. Several lipophilic groups are known that provide a soluble (protein) molecule sufficient membrane binding energy (Table 1). It is commonly accepted that a single lipid modification is not sufficient for permanent membrane binding [78,79] and nature has evolved two mechanisms to permanently associate proteins with the membrane: (i) two or more lipophilic groups are attached to the molecule of interest or (ii) the combination of one lipophilic membrane anchor and a second membrane binding motif (such as a patch of positively charged amino acids that can bind electrostatically to the negatively charged plasma membrane) [80]. The thermodynamics of membrane insertion of lipophilic molecules have been widely discussed [78,79,81,82]. Partitioning of a lipophilic group into a membrane decreases the free energy due to the hydrophobic effect [83]. Depending on the chemical nature of the membrane anchor, specific free energies for the transfer of the anchor from aqueous environment into the membrane can be calculated and are listed in Table 1. Although quite sizable energies are reported, it should be noted that at a lipid concentration of 200 μM, which is typical for the experimental situation, a free energy gain of ΔG0 = 31.1 kJ/mol is required for the association of 50% of the lipid modified molecule. Therefore, the need for a second binding motif for stable membrane association of LiNA results and in most applications of lipophilic oligonucleotides and membrane surfaces, molecules that carry at least two lipophilic groups or hybrids/oligomers of single lipid modified LiNAs have been used.

Fig. 3. DNA-directed vesicle-to-micelle phase transition, as monitored by the decrease in hydrodynamic diameter Dh, (A) or with TEM (B, C). Reprinted with permission from Matthew P. Thompson, Miao-Ping Chien, Ti-Hsuan Ku, Anthony M. Rush, and Nathan C. Gianneschi, Smart Lipids for Programmable Nanomaterials, Nano Letters, American Chemical Society, 2010, Nano Lett. 2010 14; 10, 2690–2693. Copyright (2010) American Chemical Society.

240


Table 1 Overview of the gain in free energy for transferring a lipophilic anchor molecule from the aqueous phase to the lipid membrane. Data adapted from literature [79,81]. Anchor molecule

ΔG0//kJ/mol

C12:0 chain C14:0 chain C16:0 chain Farnesyl chain Geranylgeranyl chain Cholesterol moiety

−14.7 −21.8 −28.1 −31.8 −40.4 −49.7a

a For cholesterol, the hydrophobic free energy of partitioning one CH2 group into the membrane (−3.45 kJ/mol) was divided by the area of a CH2 group (31 Å2) and multiplied with the hydrophobic area of cholesterol (447 Å2) to derive this value [151].

2.2. Binding of LiNA to model membranes Learning from nature and the available data, one can assume that cholesterol, tocopherol, porphyrin, and dipalmitoyl (C16)2 are the anchors with the best membrane incorporation properties (Table 1, Fig. 1). In addition to membrane incorporation properties, porphyrin is a multifunctional anchor, which can be used for electron/energy transfer and coordination of ligands. Albinsson and coworkers [66] surmise that porphyrins have, similar to cholesterol, physical interaction with lipid membranes as they have similar structure and similar binding constants (cholesterol modified ON — 6 × 107 M−1 [28]; porphyrin modified ON — 3 × 106 M− 1 [66]). One cannot, however, make a general rule, that molecules of a similar structure with cholesterol will bind similarly to lipid membranes: detailed studies of cholesterol analogs interaction with lipid membranes showed that strong incorporation of native cholesterol into lipid membranes is a unique property and cannot be a priori assumed for any homologues molecules [84]. Moreover, one should be careful in comparing binding constants of hydrophobic anchors to lipid membranes from different studies, as different assumptions are made. Usually either a partitioning into a hydrophobic interior [78] or interaction with “binding sites” on the membrane [28,66] is assumed. As discussed above, one can expect that conjugates with a single acyl chain will bind only transiently to a lipid membrane and, hence, cannot be used for a stable attachment of NA to a lipid membrane. Two weakly binding anchors separated by a spacer can be used for a stable attachment. To a first approximation, binding energy of a single anchor and distance between the two anchors provide a guess on the binding energy of the conjugate according to the ball-string model [85]. However, determination of binding energies by measurements is usually needed as the interaction depends not only on the anchors type and the distance between them, but also on the linker and nucleic acid structure, charge, flexibility, and possible interaction with membrane. Molecular modeling, including experimental information on the interaction between molecules, water, and the membrane, as well as non-linear approximation of electrostatic interactions, reveals binding energies that agree better with the experimentally measured [86,87]. Anchors of tocopherol or cholesterol modified nucleic acids insert quickly (within seconds) into model lipid membranes, the attachment is reversible, the conjugates relocate to other membranes accessible, or the anchor can be pulled out of the membrane upon hybridization [28,30,88]. Anchors of dipalmitoyl or distearoyl modified nucleic acids insert slowly (within hours) into model lipid membranes, the attachment is practically irreversible: the conjugates do not relocate to other membranes accessible within the time of experiments as expected from data on spontaneous intermembrane transfer of lipids [69,89]. For cholesterol and tocopherol modified DNA similar binding to membranes of the conjugates and of the LiNA-hybrids with unmodified complementary strands was observed [28,30,88].

The self-assembly properties in aqueous solution of a cholesteryltetraethylenglycol ss 18-mer oligonucleotide (ON1 TEG-Chol) and its spontaneous insertion in fluid phospholipid membranes, either monodisperse phospholipid vesicles or supported lipid bilayers were studied [75]. Baglioni and coworkers [90] investigated the self-assembly of this derivative and the insertion into bilayers performed at concentration comparable to the critical aggregation concentration. They found that the saturation onto the bilayer was reached for grafting densities comparable to the Flory radius of the hydrophilic portion, and that the conformation of the oligonucleotide portion was strictly controlled by the average distance between anchoring sites, passing from a quasirandom coil to a more rigid configuration, as concentration increases [90]. These conformational details affect in a straightforward fashion the hybridization kinetics. The anchoring of this cholesterol tagged oligonucleotide to phospholipid bilayers has been then compared to a multiple cholesteryl modified oligonucleotide, differing in the architecture and hydrophobicity of the lipophilic moiety, where on average 3.5 cholesteryl units per amphiphile were present. The self-assembly properties in solution of the two derivatives were also considered and evaluated as competitive with respect to the adsorption at fluid or solid interfaces and with respect to hybridization with the complementary sequence in solution [90,91]. The onset of aggregation is 10 μM for the single cholesteryl LiNA and 0.2 μM for the multiple cholesteryl derivatives. The self-assembly is not a single-step cooperative process, but consists of a gradual multistep association of monomers and/or oligomers into larger aggregates. For the multiple cholesteryl tag, the aggregation does not presumably result in complete segregation of the hydrophobic groups from the aqueous medium. The single cholesteryl LiNA inserts as a unimer into the membrane and eventually saturates it, both for vesicles, where the hydrodynamic thickness is monitored as a function of the ratio [LiNA]/[host lipid], and for supported lipid bilayers, where a pseudo-Langmuir adsorption isotherm is found for the same stoichiometric ratios (Fig. 4, bottom panel, left). Conversely, the multiple cholesteryl LiNA shows three clearly distinct regimes in terms of bilayer insertion, which strictly depend on the selfassembly state in solution, i.e. on LiNA concentration (Fig. 4, top panel); the first regime, pertaining to low concentrations, results in monolayer saturation, and does not follow a Langmuir-type isotherm, but rather a cooperative trend, indicating that insertion is activated by the presence of guest molecules in the bilayer. Beyond this concentration regime, the unimers approach the surface in random orientation and can be adsorbed either onto free spots or onto preexisting host patches; this results in a marked increase of the adlayer thickness around the bilayer. In the third regime, observed only for supported lipid bilayers, a membrane-assisted association mechanism is likely to result in aggregate formation at the interface. This example highlights the importance of taking into account the self-assembly properties of LiNA in designing their insertion into lipid membranes; the presentation and orientation of the oligonucleotide polar head strictly depend on the aggregation properties of the derivatives and on their hybridization state. Dipalmitoylated PNA was shown to incorporate into liposomal membrane within 1 h at room temperature when pre-hybridized with a complementary DNA strand, and it took at least twice as long without pre-hybridization (M. Schade, A. Arbuzova, unpublished results). Although MD simulations predict that unmodified PNA will adsorb to phospholipid membranes at low ionic strengths, an interaction of PNA with membranes at physiological ionic conditions was not observed [92]. Leakage studies suggest that PNA can cross membranes at a slower rate than DNA of comparable length and that PNA adsorbs to the waterlipid interface [93]. Incorporation of porphyrin modified DNA (with a linker of two or three phenylethynylene moieties, Fig. 1, C2) into membranes was more than an order of magnitude slower than that of the hybrids with


241

Fig. 4. Association and organization of cholesteryl LiON in membranes. Insertion patterns for a multiply cholesteryl LiON as a function of its concentration and self-assembly state in solution into a preformed bilayer (top panel). The oligonucleotide portion lies rather flat on the surface (bottom panel, right) with respect to the single cholesteryl derivative (bottom panel, left). Reprinted with permission from {F. Gambinossi, M. Banchelli, A. Durand, D. Berti, T. Brown, G. Caminati and P. Baglioni, Modulation of Density and Orientation of Amphiphilic DNA on Phospholipid Membranes Part I: Supported Lipid Bilayers; J. Phys. Chem. B, 2010, 114, 21, 7338–7347; M. Banchelli, F. Gambinossi, A. Durand, G. Caminati, T. Brown, D. Berti and Piero Baglioni, Modulation of Density and Orientation of Amphiphilic DNA on Phospholipid Membranes. Part II: Vesicles; J. Phys. Chem. B, 2010, 114, 21, 7348–7358.} Copyright (2010) American Chemical Society.

a complementary strand. Interestingly, binding of double-stranded (ds) three phenylethynylene modified porphyrin DNA to supported bilayers was found to be almost irreversible (Kd = 10 nM on supported bilayers versus Kd = 120 nM on liposomes) [63]. Analogous assembly of complementary LiNAs and nucleic acids giving complexes with multiple lipid modifications, can enhance the membrane partitioning, e.g. when lipid modifications are on one surface of the complex, for details see Sections 3.2 and 3.3. It is known that protein complex formation/oligomerization is a biological switch mechanism regulating membrane binding of proteins cycling on and off the membrane and their lateral distribution upon activation. It can be a driving force for specific lateral membrane partitioning, e.g. in signal transduction cascades dimerization of palmitoylated receptors leads to relocation of the complexes to rafts [12,76,94]. Similarly, binding of Histagged proteins to tris-NTA modified with saturated alkyl chains results in an efficient raft partitioning, when His-tags of more than 10 histidine residues were used allowing crosslinking/dimerization of the lipid analogs [22]. As discussed below in Section 3.5 crosslinking of LiNAs with different lipid anchors can change the partitioning drastically. 2.3. Formation of Watson–Crick base pairs The structure of the membrane-associated complex of membrane anchored DNA with a complementary DNA strand was investigated by 1 H NMR spectroscopy [29,88]. First, complementary DNA segments without lipophilic anchor were studied in solution. Upon duplex formation, characteristic changes in the 1H chemical shift of the sugar and nucleobases were observed due to the structural changes and πstacking in the complex. Next, LiONs with the same sequence bound to liposomes were studied by 1H MAS NMR. The detected chemical shifts were identical to those of the free ss DNA, suggesting that

membrane binding had no effect on the structure of the nucleotide. Upon duplex formation with the complementary strand, chemical shift changes identical to those observed for the DNA complex in solution were observed. These results confirmed that also the membranebound DNA was fully able to form a duplex with complementary DNA by means of Watson–Crick base pairing. 2.4. Thermodynamic stability of hybrids from modified molecules An important issue with regard to the application of LiNAs in biotechnology concerns the question if all nucleobases of the LiON are involved in duplex formation, when the molecule is bound to the membrane surface. Due to the lipophilic properties of the ss DNA, contacts with the membrane are quite frequent and consequently, the nucleobases in the direct vicinity to the membrane anchor may not be available for duplex formation with complementary DNA. This problem was approached by differential scanning calorimetry (DSC). This technique is an excellent tool to characterize thermally induced dissociation of oligonucleotide duplexes [95,96]. The melting of the DNA duplexes produces an exothermic peak in the DSC thermogram. DSC thermograms were analyzed for the cholesterylTEG anchored DNA-duplexes, described by the Höök group [28]. These duplexes consist of two complementary ss cholesteryl-TEG anchored DNA strands of varying lengths with a sticky end open for further functionalization [28]. Both duplexes gave individual DSC melting peaks that could be assigned to the dissociation of the long and of the short segment of the duplex [97]. In further work, we investigated the length of the duplex formation in LiONs containing two lipophilic nucleosides as membrane anchors (A. Bunge and D. Huster, unpublished results). A master curve with the melting temperature of double-stranded DNA as a function of duplex

242


LiNA, but also on the charge of the lipid membrane: due to the long range electrostatic repulsion assemblies containing negatively charged lipids melt at lower temperatures [101,102]. These exemplary studies demonstrate that thermodynamically stable assemblies of lipid vesicles with complementary DNA strands are formed, which can be used to build larger nanotechnological assemblies or for nucleic acid detection. 3. Organization on the membrane surface 3.1. Membrane interaction of lipophilic nucleosides and nucleotides

Fig. 5. Dependence of the melting temperature of DNA oligonucleotides with varying number of A–T base pairs. From this master curve, the number of base pairs of oligonucleotides and LiONs can be determined. The complementary oligonucleotide to the two given structures was an A25mer. The melting curves revealed that in the absence of lipophilic anchors all 23 T bases formed Watson Crick pairs with the A bases, but in the presence of the lipid anchor, only the free end (i.e. the 17 T bases) could couple to the complementary strand. (A. Bunge and D. Huster, unpublished results).

length was generated and showed an increase in melting temperature for longer duplexes (Fig. 5). Then, we studied lipophilic duplex DNA strands (25mer) with lipophilic anchors in positions 1 and 8. Complementary unmodified DNA was added and the melting temperature determined. For the unmodified DNA in solution, all complementary bases formed base pairs. However, for the LiON of the same sequence but bound to lipid membranes via two α-tocopherol anchors, the measured melting temperatures for the duplex revealed that only about 17 base pairs were involved in the duplex formation, suggesting that only the part of the bases after the second anchor (compare blue vs. green high line in Fig. 5) was involved in duplex formation, but not the bases between the two lipid anchors. Thus, the design where the lipophilic moieties are attached to the end of the oligonucleotide strand [28,98] appears more favorable as all nucleobases contribute to the formation of the duplex as shown by DSC measurements [98]. At the current stage of the study, we cannot rule out that the ΔH of base pairing in LiONs is somewhat reduced, which could also give rise to the reduced melting temperatures. Assembly of large unilamellar vesicles or lipid nanodiscs carrying LiNAs can be induced by DNA hybridization [61,62,99–101]. The thermodynamic stability of these aggregates tethered by DNA hybrids was studied by UV absorption measurements [62,101,102]. Quite sharp thermal transitions were observed for the (dis)-assembly of the aggregates [62,101]. Assemblies of other nucleic acid functionalized structures, e.g. gold nanoparticles [60], were also found to exhibit steeper melting curves. This is due to the cooperative melting of multiple strands necessary for disassembly of the suprastructures. For the lipid vesicle aggregates, higher melting temperatures were measured compared to the free DNA sequences [102]. This can be explained in the framework of the DLVO theory, which describes interactions between charged surfaces in liquid medium. Obviously, the interparticle forces (van der Waals attraction, electrostatics, entropic forces) modulate the thermodynamics of DNA-mediated self-assembly of the vesicle assembly by contribution to the binding free energy of the system. Correspondingly, the melting temperature was dependent not only on the

Molecular details of membrane binding of LiNA have been investigated by solid-state NMR spectroscopy. In particular, the influence of the lipophilic molecules on the membrane structure and dynamics was probed. Static 31P NMR spectroscopy provides information about the phase state and the headgroup orientation and dynamics of the phospholipids in the membrane [103–105]. The presence of LiNAs did not affect the phase state of the membrane; however, some alterations in the headgroup orientation have been observed [13,97,106–108]. The 31 P NMR spectra of LiONs showed the broad anisotropic line shape indicative of an axially symmetric shielding tensor, characteristic for the lamellar liquid crystalline line shape of the membrane superimposed with an isotropic NMR signal [88,97]. The latter could be assigned to the phosphate groups in the DNA sequence of the nucleotide suggesting that this part is highly mobile undergoing large amplitude reorientations. Also, the influence of the LiNA on membrane packing properties was studied. To this end, static 2H NMR on the phospholipids of the membrane as well as on the lipophilic nucleoside was employed, providing insights into the lipid chain order and into the lateral organization of the lipids in the membrane [104,109]. In contrast to the relatively moderate influence of lipophilic nucleosides on phospholipid headgroup orientation [108], their impact on lipid chain order varied between the investigated molecules and sometimes led to a very significant influence on membrane packing properties [13,106,108]. There was a general tendency that the LiNAs induced an increase in lipid chain order leading to an increase in membrane packing density for molecules where the lipid chain was attached to the 2′ and to the 5′ position of the sugar [13,106]. In contrast, lipid moieties attached to the nucleobase induced a decrease in lipid order parameter and packing density, due to the lipophilic nature of the bases [108]. LiONs did only very moderately alter the lateral packing density of membranes [88,97] suggesting that anchoring of larger structures on the membrane surface does not compromise the barrier function of the cell membrane. To further explore the localization of the sugar/base moiety in the lipid membrane, 1H magic-angle spinning (MAS) NOESY experiments were carried out [110]. Under MAS, molecules with high mobility can be detected just as in solution NMR, leading to highly resolved spectra that allow proximity measurements between neighboring protons exploiting the nuclear Overhauser effect (NOE) [111]. In 1H MAS NMR spectra of lipid membranes with LiNAs or LiONs, signals from the sugar as well as the nucleobase could be detected. This allows intermolecular NOE measurements between these molecular segments and the phospholipids, from which the localization of the respective segment relative to the z-direction of the membrane can be determined [111]. A localization of the sugar/nucleobase moiety in the lipid–water interface of the membrane, that is the headgroup/glycerol/upper chain region was found [13,97,106,108,112] emphasizing the lipophilic character of the nucleobase. The exact localization of individual segments of the nucleosides was subject to a broad distribution over several Ångström reiterating the large degree of molecular disorder and conformational flexibility in the membrane. However, subtle differences in the exact localization of the nucleobase/sugar moieties were found. There was a clear tendency that the functional groups of the


nucleosides were more deeply buried in the membrane when the nucleoside contained more than one acyl chain [13]. As such, a localization of the functional group would likely compromise molecular recognition by a complementary DNA strand; for nucleosides that featured a molecular spacer consisting of two polyethylene glycol (PEG) units in the headgroup, the localization of the sugar/nucleobase group was more shifted towards the aqueous phase [13]. 3.2. Orientation of nucleic acids on the surface of model membranes Single DNA or RNA strands attached to a zwitterionic or negatively charged lipid membrane are flexible and repelled from the lipid membrane surface due to electrostatic repulsion; therefore, the nucleic acid is usually available for hybridization with a complementary strand. Nevertheless, intermolecular NOEs between segments and lipophilic oligonucleotides were also observed indicating that the ss DNA chain of the oligonucleotide was localized in the direct proximity of the membrane surface [88]. In particular, cross peaks of the nucleobase/sugar moiety not only with the lipid headgroup, but also with prominent signals from the upper chain were detected. Therefore, even the negatively charged DNA oligonucleotides interact with lipid surfaces, as was previously observed for cationic membranes as well [113]. A hybrid is significantly more rigid (persistence length for double-stranded DNA is of approx. 40 nm) and presumably perpendicular to the membrane surface. DNA-membrane contacts were not observed for double-stranded DNA attached to the membrane via two cholesteryl anchors on the end of the duplex indicating that the duplex was in a more upright conformation rendering DNA-phospholipid contacts rather unlikely [97]. Nordén and coworkers [64] reported that DNA constructs with one anchor are oriented perpendicular to the membrane surface, whereas the constructs with two anchors at the ends are laying parallel to the membrane surface. Recently, Vanderlick and coworkers [101] observed directly – using electron microscopy – that LiNAs were oriented perpendicular between the lipid nanodiscs. 3.3. Functionalization of membranes with LiNA LiNAs were used for reversible or stable attachment of functionalities on membranes via hybridization. Small and giant vesicles were attached to supported bilayers [114,115]. Binding of the bilayer attached vesicles displaying complementary DNA to each other upon lateral diffusion on the supported lipid bilayer was observed [116]. Using a set of lipophilic oligonucleotides, nanoparticles could be covered by several layers of small liposomes whereby each layer could be composed of a defined population of liposomes [117]. GUVs are less stable and, depending on conditions, both attachments of single GUVs to supported lipid bilayers and, by higher coverage of the surface with GUVs, rupture of GUVs and formation of tethered lipid layers on a support were described [118,119]. Recently quasi-one-dimensional stacks of membrane nanodiscs were formed using diacyl modified DNAs [101]. Kwiat et al. [120] introduced functionalization of a hydrophobic brush plate with lipophilic oligonucleotides. HEK293 cells were successfully patterned on ss DNA-SH modified glass via plasma membrane-bound ss DNAPEG-lipids [121]. This topic was recently reviewed by Beales and Vanderlick [46]. 3.4. DNA structures and origami on membranes In this section, we review the recent reports on the construction of complex DNA architectures on membranes. This exciting field of research has extended the use of LiONs above the application as powerful and selective glue between bilayers: the build-up of complex 2D and 3D DNA structures on membranes represents a key development of soft nanotechnology and an exciting perspective in membrane biophysics. The binding of complexes to solid interfaces (i.e. mica or gold) that is necessary for many applications, such as nanopatterning or

243

nanoelectronics, as recently reviewed by Howorka [122], is achieved through electrostatic forces or covalent grafting, which affects the native conformation of the nanostructures, and often results in loss of in-plane diffusion freedom and freezing of structural defects. Assembly of DNA at fluid interfaces, such as bilayer membranes, overcomes these disadvantages. Whereas the first assembly of several short strands into larger ordered structures at the membrane has been reported [15]; now the origami design principle is mainly used [123]. DNA origami (the name for derived from Japanese art of paper folding) are nanoscale architectures, with typical size up to 100 nm, where one long scaffold strand is folded into tailored unique 2D or 3D structures by the addition of shorter oligonucleotide DNA staple strands [124,125]. Both approaches have been reported for DNA and bilayer membranes. In the first case a few examples are available in the recent literature, where DNA nanostructures can be assembled step-by-step with membrane assistance or preformed in solution [126] and then inserted in the membrane. In the second option, reported for the first time recently [127,128], the DNA origami was assembled in solution, purified with gel electrophoresis from byproducts and then tethered to the lipid bilayer. The first report [129] showing a complex DNA pattern on bilayer surfaces made use of single cholesteryl LiNA as discussed above in Sections 1.3 and 2.3. The derivative was inserted as a hydrophobic mobile anchor into a preformed bilayer and the sequential addition of complementary DNA sequences, as shown in Fig. 6, resulted in the formation of a closed pseudo-hexagonal DNA nanostructures grafted onto lipid membranes (supported planar bilayers or liposomes). Interestingly, the structural features of the final lipid/DNA nanohybrids are independent of the sequence of preparation, i.e. stepwise on the membrane or addition of preformed hexagons, up to a threshold of density on the surface or vesicle number density. After an initial region of almost linear dependence of the hydrodynamic thickness on the grafting density, there was a saturation threshold around the predicted maximum occupancy number (N = 100); the occupancy number being the number of LiNAs per liposome. The liposomes decorated with DNA pseudo-hexagons were structurally stable for weeks and could thus be further exploited or specifically addressed. The orientation of the hexagons with respect to the membrane was tuned by the occupancy number. While the formation of DNA hexagons in solution required a lengthy and delicate annealing; the step-by-step assembly on liposomes could be performed at room temperature. This is a key feature that nicely illustrates the templating/assisting role of the membrane in the correct presentation of the DNA structures for access by complementary strands in solution. In this first example, at high liposomal loading, the hexagon plane lay orthogonal to the membrane plane. A later report made use of a porphyrin nucleoside to align the DNA nanoconstruct onto the membrane surface [130]. The hexagonal core, with sides of 10 bases, was formed by six 22-mer oligonucleotides, consisting of two stretches of 10 bases separated by two unpaired thymines, which act as hinges to provide the system with bending flexibility. Each hexagon has three protruding arms, i.e. oligonucleotides of 39 bases, complementary to the porphyrin modified oligonucleotide. Using a double-chained glyceryl-bis-C16-hexaethyleneglycol hydrophobic anchor, the Nordén group has recently achieved a controllable tethering of the hexagonal DNA nanostructures on the lipid bilayers in two distinct orientations: the DNA nanostructures lay preferentially in a parallel alignment with the lipid surface when two anchors in para conformation were attached, while a perpendicular orientation was observed when only one anchor was used [64]. The insertion of a DNA origami into fluid membranes has been reported recently: a large rod-like DNA nanostructure with eight cholesteryl-tetraethylene glycol anchors precisely located on one side of this nanoarchitecture was included in model lipid membranes

244


Fig. 6. Lipid bilayers decorated with DNA hexagons ds-DNA nanostructures on liposomes, using a cholesteryl modified oligonucleotide as the mobile anchor in POPC double layers; Left: monitoring the hydrodynamic radii for the step-by-step construction of DNA pseudo-hexagons on liposomes for symmetric (circles) and asymmetric (squares) protocols. Right: Hydrodynamic radii of liposomes (radius = 50 nm)/DNA hybrids vs occupancy number. The dotted line is the saturation limit, based on geometrical considerations. The solid symbols refer to hybrids, whose thickness has been investigated with QCM on supported lipid bilayers. Adapted from Baldelli Bombelli, F., F. Betti, Gambinossi, G. Caminati, T. Brown, P. Baglioni and D. Berti, “Closed Nanoconstructs Assembled by step-by-step ss-DNA Coupling Assisted by Phospholipid Membranes”, Soft Matter, 2009, 5, 1639–1645 with permission from The Royal Society of Chemistry.

[128]. As previously mentioned, the DNA bundle formed by six helices, as shown in Fig. 7 A and B, was built separately in solution through hybridization of predefined staple segments to the long scaffold and isolated with agarose gel electrophoresis. Each staple strand has a unique “address” inside the DNA origami and can be independently functionalized either with a hydrophobic anchor or with a fluorescent reporter, as illustrated in Fig. 7 D. The origami was tethered to the membrane plane of giant unilamellar vesicles (GUV) of various phospholipid compositions, to determine possible preferences towards different liquid crystalline phases in membranes featuring lateral phase separation, which is discussed in a further paragraph. The system has also been characterized with Fluorescence Correlation Spectroscopy (FCS); depending on the fluorescent labeling scheme,

either at the center of mass of the rod or on one of its ends (Fig. 7 D, stars), the technique samples different dynamics, i.e. the translational diffusion or the rotational motions on the membranes. This report nicely demonstrates how DNA origami structures can be employed as a tool to probe complex dynamic features of molecules or supramolecular structures bound to membranes. A remarkable example of this concept is the realization of a synthetic ion channel, consisting in a DNA origami spanning a lipid bilayer membrane [127]. The design of this nanopore is inspired by α-hemolysin, a bacterial protein that perforates red blood cells, causing ion leaching. The origami spontaneously inserts in the correct orientation into small unilamellar vesicles. The channel contains a stem, consisting of six helices that form a hollow tube with size 42 × 2 nm2, which penetrates and

Fig. 7. DNA origami structures. (A) 3D side and cross-section sketches of the bundle formed by six DNA helices, where the ellipses represent the cholesteryl-TEG anchors and the asterisks the fluorophore. (B) A TEM image of the DNA origami (bar = 200 nm) (C) The chemical structure of the hydrophobic anchor. (D) Two fluorescent labeling figures with a single fluorophore placed at the center and at one side of the origami, which allow distinguishing diffusional and rotational dynamics of the rod on the free standing model membrane. Graphic from: Czogalla, A., Petrov, E. P., Kauert, D. J., Uzunova, V., Zhang, Y., Seidel, R., & Schwille, P. (2013). Switchable domain partitioning and diffusion of DNA origami rods on membranes. Faraday Discussions, 161(89), 31. Reproduced by permission of The Royal Society of Chemistry.


spans the lipid membrane, and a barrel-shaped unit that adheres to the lipid bilayer through the insertion of 26 cholesterol units. The level of structural control and complexity of this structure is indeed unprecedented, as supported by the TEM images. This synthetic ion channel possesses gating behavior when a potential is applied, with a stochastic noise which recalls natural channels, but with a significantly higher stability with respect to the natural counterparts. A considerably simpler DNA nanopore has been reported more recently, built with a DNA origami formed by a bundle of six helices interconnected by crossovers to improve structural stability (Fig. 8) [131]. Only two tetraphenylporphyrin hydrophobic anchors, which are also fluorescent, are necessary to stably tether the DNA nanostructure into the hydrophobic core of the membrane. This elegant strategy can be applied to other DNA designs, and will surely push experimental research in the field of DNA origami pores. The resulting nanopore has a width of 5.5 nm, a height of 14 nm, while the channel has an inner diameter of 2 nm. The positioning of the porphyrin tags at the end of the barrel and their hydrophobicity guarantee bilayer insertion and the correct orientation with respect to the membrane. Assembly of LiNAs into suprastructures influences the interaction with lipid membranes, e.g. kinetics and energy of incorporation, diffusion in the membrane, and lateral partitioning. These aspects require further investigation.

3.5. Lateral distribution of LiNA in model membranes Artificial membranes are used as lower-complexity models to study the lateral organization of lipids and partitioning of molecules in cell membranes: ternary lipid mixtures can form coexisting liquid ordered (lo) and liquid disordered (ld) domains at physiological conditions. Liquid ordered domains are preferentially enriched in cholesterol, saturated lipids, and sphingolipids, whereas liquid disordered domains are less densely packed, largely devoid of cholesterol and enriched in unsaturated phospholipids [132]. Depending on their modification, lipophilic molecules incorporated into phase-separated systems often show a preferential lateral segregation into either domain. In the following, reports on LiNA with an asymmetrical lateral distribution between the ld and the lo domain are reviewed. LiNA structure, used membrane compositions, found partitioning ratios (lo/ld), and references are listed in Table 2 and, unless stated otherwise, all results refer to experiments on artificial membranes. For note, accepted concentrations of LiNAs without disturbing the native membrane order lie at ≤1 mol% of LiNA/lipids.

245

Kurz et al. [29] investigated tocopherol modified DNA, a molecule containing two tocopherol units at the 5′-end of a (dT) 23mer. In lipid domain forming GUVs, this double tocopherol modified LiNA stably inserted into the membrane and was almost exclusively enriched in the ld domain (Table 2). This lateral distribution was the same for ss and ds tocopherol modified DNA with only insignificant alterations in the membrane order itself [88], without influencing GUV stability. Beales and Vanderlick [133] studied the effect of single and double cholesteryl modifications of a DNA strand (Table 2) using TEG as a linker. While single cholesteryl-TEG conjugated DNA was found only slightly enriched in the lo domain, double cholesteryl-TEG conjugated DNA as well as two pre-hybridized single cholesteryl-TEGconjugated DNA strands were twice as much enriched in the lo than in the ld phase. This distribution is consistent with the enrichment of native cholesterol in the lo phase [134], and this observation is notable as most lipophilic fluorophores strongly partition into the ld domain [135]. As discussed above, Schwille and coworkers [128] have functionalized large DNA-origami based stiff helices (6HB) with multiple cholesteryl-TEG anchors on the one side and with fluorophore labeled DNA (CH-TEG-6HB) on the solvent-facing side (Fig. 7). For GUVs with ternary lipid mixture a strong partitioning into the ld domain was found in the absence of bivalent ions (lo/ld b 1/100). This changed dramatically to a strong enrichment in the lo domain (lo/ld ~ 49/1) upon the addition of 10 mM Mg2+ and was reversible by chelation of Mg2+ via EDTA over about 60 min. Interestingly, Mg2 + ions had no effect once positively charged DOTAP with an ld preference was part of the lipid mixture locking the DNA complex in the ld phase. Surprisingly, 6HB alone, lacking any cholesteryl-TEG, still weakly adhered to the lo domain but not to the ld domain of the membrane. Consistent with other studies reported above, single stranded oligonucleotides did not attach to the membrane and cholesteryl-TEG-conjugated DNA was distributed homogeneously over the membrane independent of the presence of Mg2+ ions. For unsaturated C18 dialkyl-conjugated DNA, an enrichment in lo domains (lo/ld = 2–3) was found [136]. Here, increasing the fraction of various unsaturated lipids also increased the enrichment in the lo domain, reaching a maximum enrichment of lo/ld N 10 for cardiolipincontaining GUVs for both dialkyl-conjugated DNA as well as for two hybridized single-cholesterol-TEG-conjugated DNAs. By choosing two complementary strands of DNA and incubating two population of domain-forming GUVs with only one of these, the authors were also able to create salt dependent size limited assemblies of Janus-vesicles,

Fig. 8. DNA-nanopores. (A) The hydrophobic tag, based on tetraphenylporphyrin (TPP). (B) A cartoon of the DNA nanopore composed of six interconnected duplexes, with the staple DNA oligonucleotides shown in green. Reproduced from Angew. Chem. Int. Ed. 2013, 52, 12069–12072 with permission from Wiley.

246


Table 2 Building blocks of lipophilic nucleic acids in the literature involved in lateral segregation: lipophilic moieties as membrane anchors, linker, membrane composition and partitioning ratio of the respective LiNA. Anchor

#

Linker

Nucleic acid

Membrane composition

Partitioning lo/ld

Ref.

Cholesteryl Cholesteryl Cholesteryl Cholesteryl Cholesteryl α-Tocopherol α-Tocopherol α-Tocopherol α-Tocopherol Palmitoyl Palmitoyl Palmitoyl Stearoyl

1 1a 1a 1a 6 2 2 2 2 2 2 2 2

TriEG TriEG TriEG TriEG TriEG – – – – – – – TriEG

DNA DNA DNA DNA DNA DNA DNA DNA DNA PNA PNA PNA DNA

DOPC/SSM/CH 1:1:1 DOPC/DPPC/CH 1:1:1 DOPC/DPPC/CH 3:3:2 DOPC/DPPC/CH 3:3:2b DOPC/BSM/CH 2:2:1 POPC/SSM/CH 1:1:1 DOPC/SSM/CH 1:1:1 POPC/SSM/CH 1:1:1 [GPMV] DOPC/SSM/CH 1:1:1 POPC/SSM/CH 1:1:1 [GPMV] DOPC/DPPC/CH 3:3:2e

~1(ssDNA) ~2(dsDNA) ~2(dsDNA) N3(dsDNA) 0.01–5c ≪1 ≪1 ≪1 ≪1 5–10d ~19 ≫1 N10

[97] [133] [133] [136] [128] [29,88] [30,55] [30] [30] [30,55] [30] [30] [136]

EG = ethylene glycol; (a) double-stranded DNA with 1 Chol each; (b) replacing up to 10% of DOPC with CL; (c) dependent upon [Mg2+]; (d) re-localization into ld possible via hybridization or loss of it due to temperature change or DNase activity; (e) only for CL at 10 mol%, otherwise at 2–3 mol%

whereas for non-domain forming vesicles uncontrollably large aggregates were reported [136]. With palmitoylation being a key component in the lateral membrane organization of various proteins [94], Loew et al. [30] studied dipalmitoyl-conjugated LiNA based on PNA instead of DNA (PNA_C16, Fig. 9). When using a complementary DNA reporter strand, the PNA_C16/ DNA-hybrid was found highly enriched in the lo domain of GUVs, roughly at a ratio of lo/ld as 19/1 at an estimated overall lipid ratio of LiNA/lipid ratio of 1/3000 as shown in Fig. 9 [30]. This partition ratio is remarkably higher than the one reported for distearyl-TEG-DNA [136] in the absence of cardiolipin. It also raises the question of the individual attributions of anchor, linker, and backbone to the partitioning of the LiNA. Notably, the lateral segregation reported by Loew et al. [30] was repeatedly reversible by heating above and cooling below the lipid transition temperature, upon which the domains dissolved and reformed, respectively. Another switchable LiNA-based system was reported in Schade et al. 2012 [55] employing two LiNAs: a double-tocopherol-conjugated DNA [29] and the double palmitoylated PNA [30]. In agreement with previous studies, these LiNA specifically located to the ld and lo domains respectively in the same domain-forming GUVs without major membrane

perturbation when used at ratios of each 1:300 of LiNA/lipid. Addition of DNA strands designed to hybridize both LiNAs into a larger fourcomponent complex, resulted in the PNA_C16 relocating from the lo domain into the ld domain (reaching lo/ld ~ 0.3) without any apparent concentration in the tension-line between the domains (Fig. 9, upper panel). Using a nuclease, the release and re-partitioning of PNA_C16 from this complex again into the lo domain (lo/ld = 6) occurred over the course of 2 min at room temperature (final state: Fig. 10, bottom panel). This enzyme based approach demonstrates a possible control-mechanism for the enrichment or release of LiNAs in specific domains. Going from artificial membrane compositions to cell-derived giant plasma membrane vesicles (GPMV), this double palmitoylated LiNA was found colocalized with GPI-anchors [30] indicating its recruitment to the liquid ordered domain of GPMVs [137,138], and it distributed complementary to the previously established ld domain staining double-tocopherol-conjugated DNA [29]. Thus, LiNAs have been shown to be versatile tools for targeting both the ld and the lo domains of artificial membranes or of GPMVs, as well as being reliable platforms for carrying nucleic acid based moieties in membranes. As suggested for multivalent NTA lipids crosslinking of His-tagged proteins at the cell surface [22], controllable crosslinking of

Fig. 9. Lateral distribution of a double palmitoyl-anchored PNA/DNA hybrid in a domain-forming GUV. (A) A Cartoon summarizes the observation that C6-NBD-PC is localized to the ld domain (light green), while PNA_C16/DNAc1 (light red) is recruited to the cholesterol enriched lo domain (rose marked, gray rectangles: Chol). (B) Confocal fluorescence microscopy overlay image of C6-NBD-PC fluorescence (green) visualizing the ld phase and the membrane-inserted PNA_C16/DNAc1 molecules (red), which are localized in the lo phase of the GUV. Bar corresponds to 5 μm. Adapted with permission from Loew, M., Springer, R., Scolari, S., Altenbrunn, F., Seitz, O., Liebscher, J., Huster, D., Herrmann, A., Arbuzova, A. (2010). Lipid domain specific recruitment of lipophilic nucleic acids: a key for switchable functionalization of membranes. Journal of the American Chemical Society, 132(45), 16066-72. Copyright 2010 American Chemical Society.


247

Fig. 10. Redistribution of LiNAs upon cleavage of four-component complexes by EcoR1-HF. Cartoons on the left illustrate the four-component complex and the DNA cleavage products with tags shown in italics. Images A and B show the same vesicle before and after addition of EcoR1-HF. (A) PNA_C16/DNAc1_Rh (red) and DNA_tocopherol/DNAc2_FITC (green) are colocalized as the four-component complexes in one domain as clearly seen in the merged image on the right. (B) Between 5.5 and 45 min after addition of EcoR1-HF, fluorescence signals were not any longer co-localized but gradually separated from each other, which was also indicative of a successful cleavage. Ratios between the fluorescence intensities lo/ld are given. Scale bars correspond to 10 μm. Reprinted with permission from Schade, M., Knoll, A., Vogel, A., Seitz, O., Liebscher, J., Huster, D., Arbuzova, A. (2012). Remote control of lipophilic nucleic acids domain partitioning by DNA hybridization and enzymatic cleavage. Journal of the American Chemical Society, 134(50), 20490-7. Copyright 2012 American Chemical Society.

LiNAs through hybridization [55] can be used to mimic protein complex formation at the cell membrane and its influence on the lateral partitioning. 3.6. Lipophilic nucleic acids as a model of SNARE protein fusion system SNARE protein complex formation triggers close apposition and subsequent fusion of synaptic vesicles with plasma membranes leading to efficient fusion of the membranes and release of the vesicles content. LiNAs have been used to mimic the function of SNARE proteins: Stengel et al. [139] pioneered use of the antiparallel hybrids of lipophilic DNA as a model for SNARE mediated fusion. Depending on the length of nucleic acids and orientation of the conjugates, hybridization of the LiNAs attached to different membranes led to membrane fusion. For example, fusion of attached vesicles was observed when short LiNAs of 8 nucleotides, but not when longer LiNAs of 24 nucleotides were used [99]. Cholesteryl and diacyl modified DNA hybridization led mostly only to the merger of the outer but not of the inner monolayers of the superimposed membranes, i.e. to hemifusion; only less than 10% of the vesicles fused completely as revealed from content mixing [119,140,141]. The low yield of full fusion was ascribed to the fact that the lipophilic anchors span only one monolayer and that destabilization of the hemifusion diaphragm requires membrane anchors spanning the entire membrane. However, Diederichsen and coworkers [142] using conjugates of PNA with transmembrane peptides spanning the bilayer also observed mostly hemifusion events. It is not known why fusion was essentially arrested at the hemifusion state. Perhaps, the conjugates did not provide the stiffness even upon hybridization to break the hemifusion diaphragm. 3.7. Interaction of LiNA with cells When LiNAs are added to cells, it is usually expected that the anchors partition into the membrane. However, interaction with cells

is more complex than with model membranes: the negatively charged glycocalyx can repel DNA/RNA, preventing attachment or hindering hybridization [143], or LiNA can bind to proteins instead of inserting into lipid bilayer [36]. Whether or not the nucleic acid of the LiNA will be available for hybridization with the complementary strand depends not only on the conjugate properties but also on the type of the cells studied. Borisenko et al. [144] reported quick (within 3 min) incorporation of 6-carboxyfluorescein-labeled LiNA of 18 dT modified with stearoyl (C18) into Jurkat and HL-60 floating and macrophage-like J774 adhesive cells. No difference in attachment of a 18mer and 25mer modified with either C16 or C18 was observed. Although these short LiNAs incorporated into plasma membranes and hybridized with the complementary strands, for cell–cell attachment much longer oligonucleotides (about 70mer) were necessary, which is in good agreement with the necessity of at least a 60mer linker reported by Gartner and coworkers [31]. Borisenko et al. [144] also suggested that, even though no exchange was detected, one acyl chain was probably not sufficient for a stable attachment; therefore, a diacyl anchor (a lipid analog) is required for cell membrane attachment. In agreement, Palte and Raines [143] observed reversible attachment of LiNAs modified with one acyl chain (C12, C18, C26) to HeLa cells. The amount of LiNA incorporated into the cells was increasing with increasing length of the acyl chain, which indicates hydrophobic nature of interaction and insertion of anchors into the plasma membrane. The amount of LiNA incorporated into the cells was decreasing with increasing length of the oligonucleotide as suggested due to electrostatic repulsion of negatively charged DNA by the negatively charged glycocalyx. Interestingly, they observed no stable binding of LiNA modified with distearoyl ((C18)2) to HeLa cells. This is in agreement with Selden et al. [31] reporting almost no binding of distearoyl modified LiNA to cells: low extent of attachment to Jurkat Tlymphocytes using flow cytometry and no cell–cell adhesion was observed. Presumably distearoyl modified LiNAs were organized in

248


stable micelles. As expected from studies on model membranes, no binding of single acyl chain modified NA was observed. Dipalmitoyl modified LiNA rapidly (half-binding at 5–10 min of incubation) associated with the cells and could be detected on the surface of cells incubated at 37 °C even after 3 h. A long linker of at least 60-mer poly(dT) was necessary to achieve cell–cell or cell–surface attachment [31]. Lipophilic modification of NA is usually of advantage for cellular uptake in comparison to the corresponding unmodified NA [37,145,146]. Fluorescently-labeled lipophilic RNA was detected in cytoplasm after 30 min, whereas corresponding RNA without the lipophilic modification did not penetrate cells even after a 24 hour incubation at 37 °C [37]. Uptake into T24 human carcinoma cells of porphyrin and cholesteryl modified LiNAs was increased 5–6 and 30–100 fold compared to the uptake of the unmodified oligonucleotides, respectively [145]. Uptake via non-receptor endocytosis was suggested as LiNA did not penetrate through the membrane and 80% remained attached at the outer layer of the plasma membrane even after 6 h of incubation [146]. Other reports are in line with this suggestion. Borisenko et al. [144] also concluded that uptake of the LiNA occurs via endocytosis and is followed by nuclease degradation. Godeau et al. [37] surmised energy dependent uptake of lipophilic RNA as cholesterol modified RNA was found on the cell surface after incubation of cells with LiNA at 4 °C. Sciuto and colleagues [98] compared degradation of LiNA modified with phospholipids and corresponding unmodified oligonucleotides by a phosphodiesterase. The LiNAs, presumably assembled in aggregates, were slowly (within 24 h) degraded by snake venom phosphodiesterase, whereas corresponding oligonucleotides were completely degraded after 20 min. They also tested whether phospholipase C or D could cleave the LiNAs. No degradation by either of the phospholipases was observed: LiNA remained intact. Notably, intracellular uptake of LiNA was not accompanied by cytotoxicity [34,37,144,145]. Likewise, viability and proliferation of Jukart cells were not impaired upon incubation with dipalmitoyl modified 100mer DNA LiON [31]. Barthélémy and coworkers observed neither cytotoxicity of both stearoyl- and cholesterol modified RNA nor a change in proliferation of Huh7 cells even after a 4-day incubation [37]. From the available studies, one can conclude that LiNA, which incorporates spontaneously into model membranes, generally incorporates also into cell membranes, but often less efficiently. Hybridization of lipophilic oligonucleotides on the cell surface can be reduced or even inhibited by the negatively charged glycocalyx and cell proteins. Therefore, significantly longer linkers are required to achieve cell–cell or cell–surface attachment via hybridization [31,34]. How fast lipophilic nucleic acids are removed from the surface of cells via endocytic uptake and what happens upon uptake depends on the cell type. Lateral partitioning of proteins is a crucial part of the membrane trafficking machinery: changing the partitioning attributes changes the destiny of a protein [147,148]. Preliminary results indicate an inhomogeneous partitioning of dipalmitoylated PNA in cells [30]. Specific lateral partitioning of LiNAs into lipid domains in plasma membrane will influence the cellular uptake. Research on LiNAs lateral partitioning and sorting in cells is yet to be done. 4. Conclusion and outlook Recent research has demonstrated the potential of lipophilic nucleic acids (LiNAs) for application in bio(nano)technology and medicine. Modifying molecular properties of these building blocks offers unprecedented control over shape, size, fluidity, mechanical flexibility, or surface modification, to name just a few. The oligonucleotide portion can be tailored to perform pre-programmed tasks, thanks to the highfidelity base pairing with complementary strands or to the binding properties of nucleic acids to biological components, e.g. proteins. On the other side, the lipid anchor endows the conjugates with self-

assembly properties and affinity for lipid membranes. As of today LiNA provides a smart molecular glue that can be used to build various assemblies of vesicles and other lipid phases, keep vesicles at welldefined distances or induce fusion between them. One of the key characteristics of nucleic acid structures is its sequence specific binding to proteins [25,149], which allows the buildup of protein arrays, biosensors, and diagnostic tools. In addition, single strands can be designed to contain or bind molecular components (optically or electronically active) at precise sites, and to assemble into complex arrays with positioning of the active molecules and addressability with subnanometer precision. Taking advantage of the increasing number of the impressive results on complex arrangements of nucleic acids and of the flexibility of the synthesis of LiNAs, various strategies and approaches for structuring and functionalization of membranes, for defined and quantitative arrangement of different vesicular populations, and for targeted delivery of bioactive compounds have been designed and described. Specific selection of the nucleic acid, linker and lipophilic modifications allows for the possibility of this broad variety of applications. In particular, selection of the lipophilic anchor enables regulation of the membrane interaction as well as the arrangement of LiNAs on membranes in a lateral heterogeneous manner using the pattern of lateral lipid distribution. Being able to control the mixing and demixing of LiNAs in heterogeneous lipid membranes by using a parameter as simple as temperature change by a few Kelvin will allow designing applications such as microreactors, controlled enzyme function, and microcatalysis. Further studies on lipophilic aptamers will not only enable functionalization of specific lateral domains in synthetic membranes, but also address distinct compartments of cells membranes: ligand binding might induce changes in the partitioning due to a redistribution allowed by the lipid conjugate lateral diffusion in the membrane. The applications of LiNAs in bio(nano)technology and biophysics are an emerging interdisciplinary field, whose potential is starting to be explored lately. Therefore, most of the studies reviewed here provide just the proof of principle. A recent example is the complex arrangement of lipophilic nucleic acids to form a membrane anchored structure mimicking the properties of membrane channels [131]. The translation into efficient and affordable applications, in particular on the level of living cells and organisms, is still a challenging task. Chou et al. [150] demonstrated recently that DNA controlled assembly of colloidal particles into superstructures reduced particle retention by macrophages improving tumor accumulation. Instead of DNA-conjugated colloidal particles, future studies along this line might use lipophilic DNA-connected superstructures of lipid vesicles for cargo delivery and better biocompatibility. 5. Abbreviations

Chol ds CL DSC FCS GPMV GUV HEG LiNA LiON LNA NOE NTA ON PNA

cholesterol double-stranded cardiolipin differential scanning calorimetry Fluorescence Correlation Spectroscopy giant plasma membrane vesicle giant unilamellar vesicle hexaethylene glycol lipophilic nucleic acid lipophilic oligonucleotide locked nucleic acid Nuclear Overhauser effect nitrilotriacetic acid oligonucleotide; PEG — polyethylene glycol peptide nucleic acid


QCM ss TEG

quartz crystal microbalance single-stranded triethylene glycol

Acknowledgment This work was supported by the grants of the Deutsche Forschungsgemeinschaft (AR 783/1-1 (AA); SFB 765 (AH), and the Graduate School BuildMoNa (DH)), MIUR (Ministero Italiano dell'Università e della Ricerca, PRIN 2010–2011 grant 2010BJ23MN (DB)), the Federal Ministry of Education and Research (BMBF INUNA 0312027 (DH, AH)) and by the Leibniz Graduate School of Molecular Biophysics (fellowship to MS; # 02208701). References [1] Whitesides GM, Grzybowski B. Self-assembly at all scales. Science 2002;295:2418–21. [2] Lehn J-M. Toward complex matter: supramolecular chemistry and selforganization. Proc Natl Acad Sci U S A 2002;99:4763–8. [3] Israelachvili JN, Mitchell DJ, Ninham BW. Theory of self-assembly of hydrocarbon amphiphiles into micelles and bilayers. J Chem Soc Faraday Trans 2 1976;72: 1525–68. [4] Wiese W, Harbich W, Helfrich W. Budding of lipid bilayer vesicles and flat membranes. J Phys Condens Matter 1992;4:1647–57. [5] Helfrich W. Elastic properties of lipid bilayers: theory and possible experiments. Z Naturforsch C n.d.;28:693–703 [6] Seifert U, Berndl K, Lipowsky R. Shape transformations of vesicles: phase diagram for spontaneous-curvature and bilayer-coupling models. Phys Rev A 1991;44:1182–202. [7] Kozlov MM. Biophysics: joint effort bends membrane. Nature 2010;463:439–40. [8] Simons K, Gerl MJ. Revitalizing membrane rafts: new tools and insights. Nat Rev Mol Cell Biol 2010;11:688–99. [9] Rosemeyer H. Nucleolipids: natural occurrence, synthesis, molecular recognition, and supramolecular assemblies as potential precursors of life and bioorganic materials. Chem Biodivers 2005;2:977–1063. [10] Resh MD. Trafficking and signaling by fatty-acylated and prenylated proteins. Nat Chem Biol 2006;2:584–90. [11] Resh MD. Targeting protein lipidation in disease. Trends Mol Med 2012;18:206–14. [12] Levental I, Lingwood D, Grzybek M, Coskun U, Simons K. Palmitoylation regulates raft affinity for the majority of integral raft proteins. Proc Natl Acad Sci U S A 2010;107:22050–4. [13] Kaczmarek O, Brodersen N, Bunge A, Löser L, Huster D, Herrmann A, et al. Synthesis of nucleosides with 2′-fixed lipid anchors and their behavior in phospholipid membranes. Eur J Org Chem 2008;2008:1917–28. [14] Gissot A, Camplo M, Grinstaff MW, Barthélémy P. Nucleoside, nucleotide and oligonucleotide based amphiphiles: a successful marriage of nucleic acids with lipids. Org Biomol Chem 2008;6:1324–33. [15] Berti D, Montis C, Baglioni P. Self-assembly of designer biosurfactants. Soft Matter 2011;7:7150–8. [16] Allain V, Bourgaux C, Couvreur P. Self-assembled nucleolipids: from supramolecular structure to soft nucleic acid and drug delivery devices. Nucleic Acids Res 2012;40:1891–903. [17] Schmitt L, Dietrich C, Tampe R. Synthesis and characterization of chelator-lipids for reversible immobilization of engineered proteins at self-assembled lipid interfaces. J Am Chem Soc 1994;116:8485–91. [18] Lata S, Reichel A, Brock R, Tampe R, Piehler J. High-affinity adaptors for switchable recognition of histidine-tagged proteins. J Am Chem Soc 2005;127:10205–15. [19] Huang Z, Park JI, Watson DS, Hwang P, Szoka Jr FC. Facile synthesis of multivalent nitrilotriacetic acid (NTA) and NTA conjugates for analytical and drug delivery applications. Bioconjug Chem 2006;17:1592–600. [20] Platt V, Huang Z, Cao L, Tiffany M, Riviere K, Szoka Jr FC. Influence of multivalent nitrilotriacetic acid lipid–ligand affinity on the circulation half-life in mice of a liposome-attached His6-protein. Bioconjug Chem 2010;21:892–902. [21] Watson DS, Platt VM, Cao L, Venditto VJ, Szoka Jr FC. Antibody response to polyhistidine-tagged peptide and protein antigens attached to liposomes via lipid-linked nitrilotriacetic acid in mice. Clin Vaccine Immunol 2011;18:289–97. [22] Beutel O, Nikolaus J, Birkholz O, You C, Schmidt T, Herrmann A, et al. High-fidelity protein targeting into membrane lipid microdomains in living cells. Angew Chem Int Ed Engl 2014;53:1311–5. [23] Krishnan Y, Simmel FC. Nucleic acid based molecular devices. Angew Chem Int Ed Engl 2011;50:3124–56. [24] Stulz E. DNA architectonics: towards the next generation of bio-inspired materials. Chemistry 2012;18:4456–69. [25] Mayer G. The chemical biology of aptamers. Angew Chem Int Ed Engl 2009;48:2672–89. [26] Smith D, Schüller V, Engst C, Rädler J, Liedl T. Nucleic acid nanostructures for biomedical applications. Nanomedicine (Lond) 2013;8:105–21. [27] Wu Y, Sefah K, Liu H, Wang R, Tan W. DNA aptamer-micelle as an efficient detection/delivery vehicle toward cancer cells. Proc Natl Acad Sci U S A 2010;107:5–10.

249

[28] Pfeiffer I, Höök F. Bivalent cholesterol-based coupling of oligonucletides to lipid membrane assemblies. J Am Chem Soc 2004;126:10224–5. [29] Kurz A, Bunge A, Windeck A-K, Rost M, Flasche W, Arbuzova A, et al. Lipidanchored oligonucleotides for stable double-helix formation in distinct membrane domains. Angew Chem Int Ed Engl 2006;45:4440–4. [30] Loew M, Springer R, Scolari S, Altenbrunn F, Seitz O, Liebscher J, et al. Lipid domain specific recruitment of lipophilic nucleic acids: a key for switchable functionalization of membranes. J Am Chem Soc 2010;132:16066–72. [31] Selden NS, Todhunter ME, Jee NY, Liu JS, Broaders KE, Gartner ZJ. Chemically programmed cell adhesion with membrane-anchored oligonucleotides. J Am Chem Soc 2012;134:765–8. [32] Yoshina-Ishii C, Miller GP, Kraft ML, Kool ET, Boxer SG. General method for modification of liposomes for encoded assembly on supported bilayers. J Am Chem Soc 2005;127:1356–7. [33] Zhang G, Farooqui F, Kinstler O, Letsinger RL. Informational liposomes: complexes derived from cholesteryl-conjugated oligonucleotides and liposomes. Tetrahedron Lett 1996;37:6243–6. [34] Tokunaga T, Namiki S, Yamada K, Imaishi T, Nonaka H, Hirose K, et al. Cell surfaceanchored fluorescent aptamer sensor enables imaging of chemical transmitter dynamics. J Am Chem Soc 2012;134:9561–4. [35] Soutschek J, Akinc A, Bramlage B, et al. Therapeutic silencing of an endogenous gene by systemic administration of modified siRNAs. Nature 2004;432:173–8. [36] Wolfrum C, Shi S, Jayaprakash KN, Jayaraman M, Wang G, Pandey RK, et al. Mechanisms and optimization of in vivo delivery of lipophilic siRNAs. Nat Biotechnol 2007;25:1149–57. [37] Godeau G, Staedel C, Barthélémy P. Lipid-conjugated oligonucleotides via “click chemistry” efficiently inhibit hepatitis C virus translation. J Med Chem 2008;51:4374–6. [38] Koppelhus U, Nielsen PE. Cellular delivery of peptide nucleic acid (PNA). Adv Drug Deliv Rev 2003;55:267–80. [39] Koppelhus U, Shiraishi T, Zachar V, Pankratova S, Nielsen PE. Improved cellular activity of antisense peptide nucleic acids by conjugation to a cationic peptide–lipid (CatLip) domain. Bioconjug Chem 2008;19:1526–34. [40] Xu L, Cai L, Chen X, Jiang X, Chong H, Zheng B, et al. DNA duplexes with hydrophobic modifications inhibit fusion between HIV-1 and cell membranes. Antimicrob Agents Chemother 2013;57:4963–70. [41] Pokholenko O, Gissot A, Vialet B, Bathany K, Thiéry A, Barthélémy P. Lipid oligonucleotide conjugates as responsive nanomaterials for drug delivery. J Mater Chem B 2013;1:5329–34. [42] Aimé A, Beztsinna N, Patwa A, Pokolenko A, Bestel I, Barthélémy P. Quantum dot lipid oligonucleotide bioconjugates: toward a new anti-microRNA nanoplatform. Bioconjug Chem 2013;24:1345–55. [43] Gunnarsson A, Sjövall P, Höök F. Liposome-based chemical barcodes for single molecule DNA detection using imaging mass spectrometry. Nano Lett 2010;10:732–7. [44] Werz E, Korneev S, Montilla-Martinez M, Wagner R, Hemmler R, Walter C, et al. Specific DNA duplex formation at an artificial lipid bilayer: towards a new DNA biosensor technology. Chem Biodivers 2012;9:272–81. [45] Patwa A, Gissot A, Bestel I, Barthélémy P. Hybrid lipid oligonucleotide conjugates: synthesis, self-assemblies and biomedical applications. Chem Soc Rev 2011;40:5844–54. [46] Beales PA, Vanderlick TK. Application of nucleic acid–lipid conjugates for the programmable organisation of liposomal modules. Adv Colloid Interface Sci 2014. http://dx.doi.org/10.1016/j.cis.2013.12.009. [47] Kool ET, Morales JC, Guckian KM. Mimicking the structure and function of DNA: Insights into DNA stability and replication. Angew Chem Int Ed Engl 2000;39:990–1009. [48] Guo PX. The emerging field of RNA nanotechnology. Nat Nanotechnol 2010;5:833–42. [49] Vester B, Wengel J. LNA (locked nucleic acid): high-affinity targeting of complementary RNA and DNA. Biochemistry 2004;43:13233–41. [50] Patwa A, Salgado G, Dole F, Navailles L, Barthelemy P. Tuning molecular interactions in lipid–oligonucleotides assemblies via locked nucleic acid (LNA)-based lipids. Org Biomol Chem 2013;11:7108–12. [51] Egholm M, Buchardt O, Nielsen PE, Berg RH. Peptide nucleic acids (PNA). Oligonucleotide analogs with an achiral peptide backbone. J Am Chem Soc 1992; 114:1895–7. [52] Eriksson M, Christensen L, Schmidt J, Haaima G, Orgel L, Nielsen PE. Sequence dependent N-terminal rearrangement and degradation of peptide nucleic acid (PNA) in aqueous solution. New J Chem 1998;22:1055–9. [53] Nielsen PE, Egholm M. An introduction to peptide nucleic acid. Curr Issues Mol Biol 1999;1:89–104. [54] Marques BF, Schneider JW. Sequence-specific binding of DNA to liposomes containing di-alkyl peptide nucleic acid (PNA) amphiphiles. Langmuir 2005; 21:2488–94. [55] Schade M, Knoll A, Vogel A, Seitz O, Liebscher J, Huster D, et al. Remote control of lipophilic nucleic acids domain partitioning by DNA hybridization and enzymatic cleavage. J Am Chem Soc 2012;134:20490–7. [56] Chan YM, van Lengerich B, Boxer SG, Van Lengerich B. Effects of linker sequences on vesicle fusion mediated by lipid-anchored DNA oligonucleotides. Proc Natl Acad Sci U S A 2009;106:979–84. [57] Gissot A, Di Primo C, Bestel I, Giannone G, Chapuis H, Barthélémy P. Sensitive liposomes encoded with oligonucleotide amphiphiles: a biocompatible switch. Chem Commun (Camb) 2008:5550–2. [58] Köhler O, Jarikote DV, Seitz O. Forced intercalation probes (FIT Probes): thiazole orange as a fluorescent base in peptide nucleic acids for homogeneous single-nucleotide-polymorphism detection. Chembiochem 2005;6:69–77.

250


[59] Tabaei SR, Jönsson P, Brändén M, Höök F. Self-assembly formation of multiple DNAtethered lipid bilayers. J Struct Biol 2009;168:200–6. [60] Mirkin CA, Letsinger RL, Mucic RC, Storhoff JJ. A DNA-based method for rationally assembling nanoparticles into macroscopic materials. Nature 1996;382:607–9. [61] Jakobsen U, Vogel S. Assembly of liposomes controlled by triple helix formation. Bioconjug Chem 2013;24:1485–95. [62] Jakobsen U, Simonsen AC, Vogel S. DNA-controlled assembly of soft nanoparticles. J Am Chem Soc 2008;130:10462–3. [63] Woller JG, Börjesson K, Svedhem S, Albinsson B. Reversible hybridization of DNA anchored to a lipid membrane via porphyrin. Langmuir 2012;28:1944–53. [64] Lundberg EP, Feng B, Saeid Mohammadi A, Wilhelmsson LM, Nordén B. Controlling and monitoring orientation of DNA nanoconstructs on lipid surfaces. Langmuir 2013;29:285–93. [65] Yonamine Y, Kawasaki T, Okahata Y. Covalent modification of DNA with azetidinium lipids. Chem Lett 2010;39:84–5. [66] Börjesson K, Wiberg J, El-Sagheer AH, Ljungdahl T, Mårtensson J, Brown T, et al. Functionalized nanostructures: redox-active porphyrin anchors for supramolecular DNA assemblies. ACS Nano 2010;4:5037–46. [67] Israelachvili JN. Intermolecular and surface forces. London: Academic Press; 1992. [68] Hunter RJ. Foundations of colloid science. 2nd ed. New York: Oxford University Press; 2000. [69] Jones JD, Thompson TE. Spontaneous phosphatidylcholine transfer by collision between vesicles at high lipid concentration. Biochemistry 1989;28:129–34. [70] Evans E, Rawicz W, Hoffman AF. Lipid bilayer expansion and mechanical disruption in solutions of water-soluble bile acid. In: Hoffman AF, Paumgartner G, Stiehl A, editors. Bile acids gastroenterol.: Basic Clin. Adv. Dordrecht, Boston, London: Kluwer Academic Publishers; 1995. p. 59–68. [71] Berti D. Self assembly of biologically inspired amphiphiles. Curr Opin Colloid Interface Sci 2006;11:74–8. [72] Baldelli Bombelli F, Berti D, Milani S, Lagi M, Barbaro P, Karlsson G, et al. Collective headgroup conformational transition in twisted micellar superstructures. Soft Matter 2008;4:1102–13. [73] Desbat B, Arazam N, Khiati S, Tonelli G, Neri W, Barthélémy P, et al. Unexpected bilayer formation in Langmuir films of nucleolipids. Langmuir 2012;28:6816–25. [74] Thompson MP, Chien M-P, Ku T-H, Rush AM, Gianneschi NC. Smart lipids for programmable nanomaterials. Nano Lett 2010;10:2690–3. [75] Banchelli M, Betti F, Berti D, Caminati G, Bombelli FB, Brown T, et al. Phospholipid membranes decorated by cholesterol-based oligonucleotides as soft hybrid nanostructures. J Phys Chem B 2008;112:10942–52. [76] Brunsveld L, Waldmann H, Huster D. Membrane binding of lipidated Ras peptides and proteins—the structural point of view. Biochim Biophys Acta 2009;1788:273–88. [77] Casey PJ. Protein lipidation in cell signaling. Science 1995;268:221–5. [78] Peitzsch RM, McLaughlin S. Binding of acylated peptides and fatty acids to phospholipid vesicles: pertinence to myristoylated proteins. Biochemistry 1993;32:10436–43. [79] Pool CT, Thompson TE. Chain length and temperature dependence of the reversible association of model acylated proteins with lipid bilayers. Biochemistry 1998; 37:10246–55. [80] Murray D, Ben-Tal N, Honig B, McLaughlin S. Electrostatic interaction of myristoylated proteins with membranes: simple physics, complicated biology. Structure 1997;5:985–9. [81] Silvius JR. Lipidated peptides as tools for understanding the membrane interactions of lipid-modified proteins. In: Simon SA, McIntosh TJ, editors. Peptide–lipid interactions. , Elsevier; 2002. p. 371–95. [82] Weise K, Huster D, Kapoor S, Triola G, Waldmann H, Winter R. Gibbs energy determinants of lipoprotein insertion into lipid membranes: the case study of Ras proteins. Faraday Discuss 2013;161:549–61. [83] Tanford C. The hydrophobic effect: formation of micelles and biological membranes. New York: John Wiley & Sons; 1980. [84] Scheidt HA, Müller P, Herrmann A, Huster D. The potential of fluorescent and spin-labeled steroid analogs to mimic natural cholesterol. J Biol Chem 2003; 278:45563–9. [85] Buser CA, Sigal CT, Resh MD, McLaughlin S. Membrane binding of myristylated peptides corresponding to the NH2 terminus of Src. Biochemistry 1994;33:13093–101. [86] Murray D, Arbuzova A, Hangyás-Mihályné G, Gambhir A, Ben-Tal N, Honig B, et al. Electrostatic properties of membranes containing acidic lipids and adsorbed basic peptides: theory and experiment. Biophys J 1999;77:3176–88. [87] Kosloff M, Alexov E, Arshavsky VY, Honig B. Electrostatic and lipid anchor contributions to the interaction of transducin with membranes: mechanistic implications for activation and translocation. J Biol Chem 2008;283:31197–207. [88] Bunge A, Kurz A, Windeck A-K, Korte T, Flasche W, Liebscher J, et al. Lipophilic oligonucleotides spontaneously insert into lipid membranes, bind complementary DNA strands, and sequester into lipid-disordered domains. Langmuir 2007; 23:4455–64. [89] Silvius JR, Leventis R. Spontaneous interbilayer transfer of phospholipids: dependence on acyl chain composition. Biochemistry 1993;32:13318–26. [90] Gambinossi F, Banchelli M, Durand A, Berti D, Brown T, Caminati G, et al. Modulation of density and orientation of amphiphilic DNA anchored to phospholipid membranes. I. Supported lipid bilayers. J Phys Chem B 2010;114:7338–47. [91] Banchelli M, Gambinossi F, Durand A, Caminati G, Brown T, Berti D, et al. Modulation of density and orientation of amphiphilic DNA on phospholipid membranes. II. Vesicles. J Phys Chem B 2010;114:7348–58. [92] Weronski P, Jiang Y, Rasmussen S. Molecular dynamics study of small PNA molecules in lipid–water system. Biophys J 2007;92:3081–91. [93] Wittung P, Kajanus J, Edwards K, Haaima G, Nielsen PE, Nordén B, et al. Phospholipid membrane permeability of peptide nucleic acid. FEBS Lett 1995;375:27–9.

[94] Levental I, Grzybek M, Simons K. Greasing their way: lipid modifications determine protein association with membrane rafts. Biochemistry 2010;49:6305–16. [95] Breslauer KJ, Frank R, Blöcker H, Marky LA. Predicting DNA duplex stability from the base sequence. Proc Natl Acad Sci U S A 1986;83:3746–50. [96] Duguid JG, Bloomfield VA, Benevides JM, Thomas GJ. DNA melting investigated by differential scanning calorimetry and Raman spectroscopy. Biophys J 1996;71:3350–60. [97] Bunge A, Loew M, Pescador P, Arbuzova A, Brodersen N, Kang J, et al. Lipid membranes carrying lipophilic cholesterol-based oligonucleotides—characterization and application on layer-by-layer coated particles. J Phys Chem B 2009;113:16425–34. [98] Chillemi R, Greco V, Nicoletti VG, Sciuto S. Oligonucleotides conjugated to natural lipids: synthesis of phosphatidyl-anchored antisense oligonucleotides. Bioconjug Chem 2013;24:648–57. [99] Maruyama T, Yamamura H, Hiraki M, Kemori Y, Takata H, Goto M. Directed aggregation and fusion of lipid vesicles induced by DNA-surfactants. Colloids Surf B Biointerfaces 2008;66:119–24. [100] Beales PA, Vanderlick TK. Specific binding of different vesicle populations by the hybridization of membrane-anchored DNA. J Phys Chem A 2007;111: 12372–80. [101] Beales PA, Geerts N, Inampudi KK, Shigematsu H, Wilson CJ, Vanderlick TK. Reversible assembly of stacked membrane nanodiscs with reduced dimensionality and variable periodicity. J Am Chem Soc 2013;135:3335–8. [102] Beales PA, Vanderlick TK. DNA as membrane-bound ligand–receptor pairs: duplex stability is tuned by intermembrane forces. Biophys J 2009;96:1554–65. [103] Cullis PR, de Kruijff B. The polymorphic phase behaviour of phosphatidylethanolamines of natural and synthetic origin. A 31P NMR study. Biochim Biophys Acta 1978;513:31–42. [104] Huster D. Solid-state NMR spectroscopy to study protein–lipid interactions. Biochim Biophys Acta 2014. http://dx.doi.org/10.1016/j.bbalip.2013.12.002. [105] Schiller J, Muller M, Fuchs B, Arnold K, Huster D. 31P NMR spectroscopy of phospholipids: from micelles to membranes. Curr Anal Chem 2007;3:283–301. [106] Brodersen N, Li J, Kaczmarek O, Bunge A, Löser L, Huster D, et al. Nucleosides with 5′-fixed lipid groups — synthesis and anchoring in lipid membranes. Eur J Org Chem 2007;2007:6060–9. [107] Montanha E, Pavinatto FJ, Caseli L, Kaczmarek O, Liebscher J, Huster D, Oliveira ON. Properties of lipophilic nucleoside monolayers at the air–water interface. Colloids Surf B Biointerfaces 2010;77:161–5. [108] Scheidt HA, Flasche W, Cismas C, Rost M, Herrmann A, Liebscher J, et al. Design and application of lipophilic nucleosides as building blocks to obtain highly functional biological surfaces. J Phys Chem B 2004;108:16279–87. [109] Davis JH. Deuterium magnetic resonance study of the gel and liquid crystalline phases of dipalmitoyl phosphatidylcholine. Biophys J 1979;27:339–58. [110] Forbes J, Husted C, Oldfield E. High-field, high-resolution proton “magicangle” sample-spinning nuclear magnetic resonance spectroscopic studies of gel and liquid crystalline lipid bilayers and the effects of cholesterolt; 1988 2903–9. [111] Scheidt HA, Huster D. The interaction of small molecules with phospholipid membranes studied by 1H NOESY NMR under magic-angle spinning. Acta Pharmacol Sin 2008;29:35–49. [112] Kaczmarek O, Scheidt HA, Bunge A, Föse D, Karsten S, Arbuzova A, et al. 2′-Linking of lipids and other functions to uridine through 1,2,3-triazoles and membrane anchoring of the amphiphilic products. Eur J Org Chem 2010;2010:1579–86. [113] Rädler JO, Koltenover I, Salditt T, Safinya CR. Structure of DNA-cationic liposome complexes: DNA intercalation in multilamellar membranes in distinct interhelical packing regimes. Science 1997;275:810–4. [114] Benkoski JJ, Höök F. Lateral mobility of tethered vesicle-DNA assemblies. J Phys Chem B 2005;109:9773–9. [115] Van Lengerich B, Rawle RJR, Boxer SG. Covalent attachment of lipid vesicles to a fluid-supported bilayer allows observation of DNA-mediated vesicle interactions. Langmuir 2010;26:8666–72. [116] Chan Y-HM, Lenz P, Boxer SG. Kinetics of DNA-mediated docking reactions between vesicles tethered to supported lipid bilayers. Proc Natl Acad Sci U S A 2007;104:18913–8. [117] Loew M, Kang J, Dähne L, Hendus-Altenburger R, Kaczmarek O, Liebscher J, et al. Controlled assembly of vesicle-based nanocontainers on layer-by-layer particles via DNA hybridization. Small 2009;5:320–3. [118] Chung M, Boxer SG. Stability of DNA-tethered lipid membranes with mobile tethers. Langmuir 2011;27:5492–7. [119] Rawle RJ, van Lengerich B, Chung M, Bendix PM, Boxer SG. Vesicle fusion observed by content transfer across a tethered lipid bilayer. Biophys J 2011;101:L37–9. [120] Kwiat M, Elnathan R, Kwak M, de Vries JW, Pevzner A, Engel Y, et al. Non-covalent monolayer-piercing anchoring of lipophilic nucleic acids: preparation, characterization, and sensing applications. J Am Chem Soc 2012;134:280–92. [121] Sakurai K, Hoffecker IT, Iwata H. Long term culture of cells patterned on glass via membrane-tethered oligonucleotides. Biomaterials 2013;34:361–70. [122] Howorka S. DNA nanoarchitectonics: assembled DNA at interfaces. Langmuir 2013;29:7344–53. [123] Seeman NC. DNA in a material world. Nature 2003;421:427–31. [124] Rothemund PWK. Folding DNA, to create nanoscale shapes and patterns. Nature 2006;440:297–302. [125] Castro CE, Kilchherr F, Kim D, Shiao EL, Wauer T, Wortmann P, et al. A primer to scaffolded DNA origami. Nat Methods 2011;8:221–9. [126] Bombelli FB, Gambinossi F, Lagi M, Berti D, Caminati G, Brown T, et al. DNA closed nanostructures: a structural and Monte Carlo simulation study. J Phys Chem B 2008;112:15283–94.

M. Schade et al. / Advances in Colloid and Interface Science 208 (2014) 235–251 [127] Langecker M, Arnaut V, Martin TG, List J, Renner S, Mayer M, et al. Synthetic lipid membrane channels formed by designed DNA nanostructures. Science 2012;338:932–6. [128] Czogalla A, Petrov EP, Kauert DJ, Uzunova V, Zhang Y, Seidel R, et al. Switchable domain partitioning and diffusion of DNA origami rods on membranes. Faraday Discuss 2013;161:31–43. [129] Bombelli FB, Betti F, Gambinossi F, Caminati G, Brown T, Baglioni P, et al. Closed nanoconstructs assembled by step-by-step ss-DNA coupling assisted by phospholipid membranes. Soft Matter 2009;5:1639–45. [130] Börjesson K, Lundberg EP, Woller JG, Nordén B, Albinsson B. Soft-surface DNA nanotechnology: DNA constructs anchored and aligned to lipid membrane. Angew Chem Int Ed Engl 2011;50:8312–5. [131] Burns JR, Göpfrich K, Wood JW, Thacker VV, Stulz E, Keyser UF, et al. Lipid-bilayerspanning DNA nanopores with a bifunctional porphyrin anchor. Angew Chem Int Ed Engl 2013:1–5. [132] Veatch SL, Keller SL. Separation of liquid phases in giant vesicles of ternary mixtures of phospholipids and cholesterol. Biophys J 2003;85:3074–83. [133] Beales PA, Vanderlick TK. Partitioning of membrane-anchored DNA between coexisting lipid phases. J Phys Chem B 2009;113:13678–86. [134] Veatch SL, Polozov IV, Gawrisch K, Keller SL. Liquid domains in vesicles investigated by NMR and fluorescence microscopy. Biophys J 2004;86:2910–22. [135] Baumgart T, Hunt G, Farkas ER, Webb WW, Feigenson GW. Fluorescence probe partitioning between Lo/Ld phases in lipid membranes. Biochim Biophys Acta 2007;1768:2182–94. [136] Beales PA, Nam J, Vanderlick TK. Specific adhesion between DNA-functionalized “Janus” vesicles: size-limited clusters. Soft Matter 2011;7:1747–55. [137] Johnson SA, Stinson BM, Go MS, Carmona LM, Reminick JI, Fang X, et al. Temperature-dependent phase behavior and protein partitioning in giant plasma membrane vesicles. Biochim Biophys Acta 2010;1798:1427–35. [138] Scolari S, Engel S, Krebs N, Plazzo AP, De Almeida RFM, Prieto M, et al. Lateral distribution of the transmembrane domain of influenza virus hemagglutinin revealed by time-resolved fluorescence imaging. J Biol Chem 2009;284:15708–16. [139] Stengel G, Zahn R, Höök F. DNA-induced programmable fusion of phospholipid vesicles. J Am Chem Soc 2007;129:9584–5.

251

[140] Stengel G, Simonsson L, Campbell RA, Höök F. Determinants for membrane fusion induced by cholesterol-modified DNA zippers. J Phys Chem B 2008; 112:8264–74. [141] Chung M, Lowe RD, Chan YM, Ganesan PV, Boxer SG. DNA-tethered membranes formed by giant vesicle rupture. J Struct Biol 2009;168:190–9. [142] Lygina AS, Meyenberg K, Jahn R, Diederichsen U. Transmembrane domain peptide/ peptide nucleic acid hybrid as a model of a SNARE protein in vesicle fusion. Angew Chem Int Ed Engl 2011;50:8597–601. [143] Palte MJ, Raines RT. Interaction of nucleic acids with the glycocalyx. J Am Chem Soc 2012;134:6218–23. [144] Borisenko GG, Zaitseva MA, Chuvilin AN, Pozmogova GE. DNA modification of live cell surface. Nucleic Acids Res 2009;37:e28. [145] Boutorine AS, Boiziau C, Le Doan T, Toulmé JJ, Hélène C. Effect of the terminal phosphate derivatization of beta- and alpha-oligodeoxynucleotides on their antisense activity in protein biosynthesis, stability and uptake by eucaryotic cells. Biochimie 1992;74:485–9. [146] Boutorine AS, Kostina EV. Reversible covalent attachment of cholesterol to oligodeoxyribonucleotides for studies of the mechanisms of their penetration into eucaryotic cells. Biochimie 1993;75:35–41. [147] Prior IA, Hancock JF. Ras trafficking, localization and compartmentalized signalling. Semin Cell Dev Biol 2012;23:145–53. [148] Surma MA, Klose C, Simons K. Lipid-dependent protein sorting at the trans-Golgi network. Biochim Biophys Acta 1821;2012:1059–67. [149] Garvie C, Wolberger C. Recognition of specific DNA sequences. Mol Cell 2001;8:937–46. [150] Chou LY, Zagorovsky K, Chan WC. DNA assembly of nanoparticle superstructures for controlled biological delivery and elimination. Nat Nanotechnol 2014;9:148–55. [151] Heerklotz H, Epand RM. The enthalpy of acyl chain packing and the apparent water-accessible apolar surface area of phospholipids. Biophys J 2001; 80:271–9.

Solvent-accessible surfaces of nucleic acids.

Preparation of luminescent chemosensors by post-functionalization of vesicle surfaces.

Orthogonal chemical functionalization of patterned gold on silica surfaces.

A microgel construction kit for bioorthogonal encapsulation and pH-controlled release of living cells.

Cellular self-organization on micro-structured surfaces.

Construction of a kit of reference strains for rapid genetic mapping in Bacillus subtilis 168.

High-risk HPV nucleic acid detection kit-the careHPV test -a new detection method for screening.

Biomolecular functionalization for enhanced cell-material interactions of poly(methyl methacrylate) surfaces.

Spine interbody implants: material selection and modification, functionalization and bioactivation of surfaces to improve osseointegration.

Flexible CRISPR library construction using parallel oligonucleotide retrieval.

Flexible Teflon nanocone array surfaces with tunable superhydrophobicity for self-cleaning and aqueous droplet patterning.

Scaling Up Nature: Large Area Flexible Biomimetic Surfaces.

Comparison of Bayesian methods for flexible modeling of spatial risk surfaces in disease mapping.

Self-organization of gold nanoparticles on silanated surfaces.

organs.

Formation of highly stable self-assembled alkyl phosphonic acid monolayers for the functionalization of titanium surfaces and protein patterning.

polysaccharide coatings on titanium surfaces: a study of functionalization and stability.

Enantioselective Alcohol C-H Functionalization for Polyketide Construction: Unlocking Redox-Economy and Site-Selectivity for Ideal Chemical Synthesis.

A Flexible Arrayed Eddy Current Sensor for Inspection of Hollow Axle Inner Surfaces.

Construction of an interpretive pattern directory for the API 10 S kit and analysis of its diagnostic accuracy.

Nanoengineered surfaces for focal adhesion guidance trigger mesenchymal stem cell self-organization and tenogenesis.

Combining plasmachemical emulsion-templating with ATRP to create macroporous lipophilic surfaces.

Structure, domain organization, and different conformational states of stem cell factor-induced intact KIT dimers.

Pharmacokinetic models for lipophilic compounds.