Tailoring Bicelle Morphology and Thermal Stability with Lanthanide

Aug 16, 2016 - Stéphane Isabettini , Marianne Liebi , Joachim Kohlbrecher , Takashi Ishikawa , Peter Fischer , Erich J. Windhab , Peter Walde , Simon...
0 downloads 0 Views 5MB Size
Download PDF
Article pubs.acs.org/Langmuir

Tailoring Bicelle Morphology and Thermal Stability with LanthanideChelating Cholesterol Conjugates Stéphane Isabettini,† Marianne Liebi,‡ Joachim Kohlbrecher,§ Takashi Ishikawa,∥ Erich J. Windhab,† Peter Fischer,† Peter Walde,⊥ and Simon Kuster*,† †

Laboratory of Food Process Engineering, ETH Zurich, Schmelzbergstrasse 9, 8092 Zurich, Switzerland Laboratory for Macromolecules and Bioimaging, §Laboratory for Neutron Scattering and Imaging, and ∥Biomolecular Research Laboratory, Paul Scherrer Institute, 5232 Villigen PSI, Switzerland ⊥ Department of Materials, ETH Zurich, Vladimir-Prelog-Weg 5, 8093 Zurich, Switzerland ‡

S Supporting Information *

ABSTRACT: Bicelles composed of DMPC and phospholipids capable of chelating lanthanide ions, such as 1,2-dimyristoyl-snglycero-3-phospho-ethanolamine-diethylene triaminepentaacetate (DMPE-DTPA), are highly tunable magnetically responsive soft materials. Further doping of these systems with cholesterol-DTPA conjugates complexed to a lanthanide ion considerably enhances the bicelle’s size and magnetic alignability. The high value of these cholesterol conjugates for bicelle design remains largely unexplored. Herein, we examine how molecular structural alterations within the cholesterol-DTPA conjugates lead to contrasting selfassembled polymolecular aggregate structures when incorporated into DMPC/DMPE-DTPA/Tm3+ bilayers. The nature of the linker connecting the DTPA-chelating moiety to the sterol backbone is examined by synthesizing conjugates of various linker lengths and polarities. The incorporation of these compounds within the bilayer results in polymolecular aggregate geometries of higher curvature. The increasing degrees of freedom for conformational changes conveyed to the chelator headgroup with increasing linker atomic length reduce the cholesterol-DTPA conjugate’s critical packing parameter. Consequently, an inverse correlation between the number of carbon atoms in the linker and the bicelle radius is established. The introduction of polarity into the carbon chain of the linker did not cause major changes in the polymolecular aggregate architecture. Under specific conditions, the additives permit the formation of remarkably temperature-resistant bicelles. The versatility of design offered by these amphiphiles gives rise to new and viable tools for the growing field of magnetically responsive soft materials.

■

INTRODUCTION

components and their interaction with each other and with the solvent. Conelike molecules will preferentially form aggregates with high curvature such as micelles. On the other hand, molecules with cylindrical geometry, such as phospholipids, will result in bilayer substructures, as is the case, for example, in vesicles. The packing parameter is commonly reported to characterize amphiphile geometry.7,8 Bicelles are generally formed from a mixture of two types of lipids.1,9 Studies have also reported monolipidic10 or lipid−detergent11 systems. In the most characterized case, 1,2-dimyristoyl-snglycero-3-phosphocholine (DMPC), with a packing parameter close to unity, constitutes the planar part of the bicelle. The regions of high curvature on the edge are covered by 1,2dihexanoyl-sn-glycero-3-phosphocholine (DHPC).

Self-assembled aggregate structures formed from amphiphilic molecules in solution are the result of physicochemical interactions in complex colloidal systems. Disklike assemblies composed of a planar bilayer and a micellar edge are referred to as bicelles.1,2 They are commonly used in research as membrane models and for the study of membrane-associated biological macromolecules.3 The possibility of aligning the polymolecular assemblies in magnetic fields and of forming oriented liquid crystals enables the measurement of anisotropic spin interactions by both solution-state and solid-state NMR.4−6 The structural characteristics of bicelles are dependent on numerous parameters including temperature, pressure, pH, and lipid concentration. The ongoing search for novel bicelle systems aims at enhancing their combined versatility and accessibility and ascertaining their value as quintessential soft materials for various applications. The final architecture of polymolecular aggregates is governed by the molecular geometry of the amphiphile © 2016 American Chemical Society

Received: May 24, 2016 Revised: July 27, 2016 Published: August 16, 2016 9005

DOI: 10.1021/acs.langmuir.6b01968 Langmuir 2016, 32, 9005−9014

Article

Langmuir

lanthanide ions, as an alternative to DMPE-DTPA. The reported structure delivers a substantial gain in bicelle size and magnetic alignability. The additional degrees of freedom induced by these specially designed multilipidic systems call for further investigation. Engineering of polymolecular aggregates is feasible only by understanding the chemical and physical forces that govern both the hydrophilic and hydrophobic interactions between the individual molecules that compose them. Systematic studies are required because minor changes on the molecular level have the potential to induce very different packing behaviors and result in unique polymolecular responses.24 Herein, we explore the morphology of bicelles resulting from the incorporation of a series of synthesized cholesterol-DTPA conjugates. The range of attainable selfassembled polymolecular architectures is revealed by cryo-TEM micrographs. On the basis of the observed geometries, we propose an appropriate form factor for fitting the considered sample’s radially averaged SANS curve. This provides for a more detailed and statistically relevant description of the achievable polymolecular assemblies. Bicelle geometries are favored through the combined addition of cholesterol and cholesterol-DTPA conjugates to a DMPC/DMPE-DTPA/ Tm3+ bilayer. The number of carbon atoms in the linker offers control over the bicelle’s hydrodynamic radius. Furthermore, polarity is introduced into the carbon chain of the linker to identify the forces responsible for the formation of specifically sized bicelles. The resulting magnetic alignability of the aggregates is characterized by birefringence experiments. In an ultimate step, we demonstrate the potential of these novel polymolecular assemblies by focusing on thermally stable systems. The present study aims at expanding the toolbox for the intelligent design of bicelle structures on a molecular level.

The incorporation of 1,2-dimyristoyl-sn-glycero-3-phosphoethanolamine-diethylene triaminepentaacetate (DMPE-DTPA) complexed to a lanthanide ion within the bicelle’s bilayer allows considerable enhancement and fine-tuning possibilities of the polymolecular aggregate’s alignment in magnetic fields. The high molecular diamagnetic susceptibility anisotropy Δχ of the phospholipid-lanthanide complex contributes considerably to the cumulative magnetic energy of the bicelle (Emag). This 1 energy is described by Emag = − 2 μ0 nΔχB2 , where B is the magnetic field strength, μ0 is the magnetic constant, and n is the aggregation number.12 The latter accounts for the cumulative property of the individual Δχ of the phospholipids composing the planar part of the bicelle. The addition of paramagnetic lanthanide ions to DHPC/DMPC bicelles engenders positive ordering of the colloidal suspension when it is subjected to an external magnetic field.13 The lanthanide ions are known to interact with the lipid phosphate groups. However, these loose interactions result in paramagnetic shifts that considerably decrease the NMR spectra resolution. Complexing the lanthanides with chelating agents such as DMPE-DTPA reduces these paramagnetic shifts, allowing us to obtain highly resolved NMR spectra of larger membrane proteins in the bilayer.14 Moving toward a solely bilipidic polymolecular aggregate architecture in which DMPE-DTPA is no longer used as a minor additive but as an alternative to DHPC gives rise to an entire new family of tunable soft materials.15 Analogous to DHPC, the chemically engineered DMPEDTPA/Tm3+ phospholipid will preferentially cover the edge part of the bicelle as a result of its conelike geometry, resulting from a comparatively larger headgroup. Creating bicelles from the longer 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) phospholipid and its lanthanide ion chelating counterpart DPPE-DTPA confirms the versatility of these systems and opens doors for the development of smart materials such as hydrogels.16 Concerning alternative ways of tuning the properties of bicelles, numerous molecules are known to interact with phospholipid bilayers and may be valuable tools. Cholesterol is an important component of eukaryotic membranes that has a major impact on the mechanics of the bilayer17 and offers advantages for bicelle design.18 The polar hydroxyl headgroup of cholesterol is partially responsible for the molecule’s affinity for lipid bilayers.19 However, the interaction and positioning of cholesterol within the hydrophobic interior of the bilayer is presumably governed by van der Waals’ forces acting between its sterol backbone and the neighboring lipid’s aliphatic tails.20,21 In the case of our bicelle systems, it has been shown that cholesterol acts as a spacer in the lipid layer between lanthanide ion phospholipid chelators and it considerably enhances the bicelle’s dimensions.18 Both an increase in aggregate number and a larger proportion of species with a high Δχ in the planar part of the bicelle result in an increase in Emag. The compatibility of the sterol moiety of cholesterol with phospholipid bilayers offers a fertile playground for further chemical modifications of the molecule through its hydroxyl group.22 We recently reported on the introduction of a cholesterol-diethylenetriaminepentaacetate compound within the bicelle bilayer.23 The sterol backbone of cholesterol was chemically bound to a DTPA moiety using the amino acid lysine as a linker. The cholesterol-DTPA conjugate acts as an anchor for further doping of the lipid plane with paramagnetic

■

EXPERIMENTAL SECTION

Materials. Phospholipids 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) and the hexa-ammonium salt of 1,2-dimyristoyl-snglycero-3-phosphoethanolamine-N-diethylenetriaminepentaacetic (DMPE-DTPA) were purchased from Avanti Polar Lipids (Alabaster, AL) in powdered form (99%) and were used without further purification. Chloroform (stabilized by ethanol), anhydrous thulium(III) chloride (99.9%), and methanol (99.8%) were purchased from Sigma-Aldrich (Buchs, Switzerland). D2O (99.9 atom % D) was purchased from ARMAR Chemicals (Döttingen, Switzerland). All materials used for the synthesis of lanthanide-chelating cholesterol conjugates are detailed in the Supporting Information. Synthesis of Cholesterol-DTPA Conjugates. The Chol-CnDTPA and Chol-C2-O-C2-DTPA compounds presented in Figure 1 were synthesized by multistep reactions involving cholesteryl chloroformate (precursor of the sterol moiety labeled chol), a diamine-bearing linker (precursor of Cn and C2-O-C2), and DTPA bisanhydride (precursor of the DTPA-chelator moiety).25 The C2-OC2-BOC linker was prepared from 2,2′-oxybis(ethylamine) as described by Yasushi et al.26 BOC-Cn-Chol was synthesized by a condensation reaction of cholesteryl chloroformate with a monoprotected amine-bearing linker structure. Deprotection followed using acetyl chloride in ethanol to yield Chol-Cn or Chol-C2-O-C2 with a free amino group after 72 h of reaction.27 Finally, the DTPA-chelator moiety was attached by reacting with DTPA bisanhydride followed by appropriate purification.28 Detailed synthesis procedures, reaction schemes, and analytical data are provided in the Supporting Information. Molecular Modeling. Chol-C2-DTPA and Chol-C6-DTPA were independently sketched and investigated with ChemBioDraw Ultra (Cambridge-Soft, 2009). The geometrical optimization of the structures was done with molecular modeling routine MM2 in vacuum without any interference of solvent or other phospholipids. 9006

DOI: 10.1021/acs.langmuir.6b01968 Langmuir 2016, 32, 9005−9014

Article

Langmuir

(Quantifoil, Jena, Germany) at either 5, 25, or 40 °C before being flash frozen in liquid ethane at 77 K using a cryoplunge 3 system (Gatan, Pleasanton, CA, USA). Samples were analyzed with a JEM2200FS (JEOL, Japan) equipped with a field emission gun and an in-column energy filter (JEOL) operated at 200 kV. A 4096 × 4096 CMOS camera (F416, TVIPS, Germany) with a 5−10 μm underfocus was used to increase the contrast. Samples were kept under liquid nitrogen during analysis in a cryo-holder (model 626, Gatan). Contrast-rich particles on the presented micrographs are ice crystals resulting from the freezing procedure. Birefringence. Birefringence measurements of the bicelle samples exposed to a 5.5 T field were undertaken as an alternative means of quantifying the orientation of the colloidal suspensions as described previously.18,29 The sample was in a temperature-controlled quartz cuvette (Hellma, Germany) and placed in the magnet. Light from a diode laser (Newport, Irvine, CA, USA) with a wavelength of 635 nm was directed through the first polarizer and a photoelastic modulator (PEM-90, Hinds Instruments, Hillsboro, OR, USA) before passing through the sample horizontally. The laser beam passed through a second polarizer oriented 90° relative to the first one before hitting a photodetector (Hinds Instruments). Two lock-in amplifiers (SR830, SRS, Sunnyvale, CA, USA) were employed for the recording of the first and second harmonics of the ac signal used to calculate the birefringence. The magnetic field strength at which the order parameter S reaches one-half of its maximum value B(S1/2) was determined in a similar setup at the HFML (Nijmegen, Netherlands) under a 33 T field as described previously.33,35

Figure 1. Molecular structures of the synthesized and employed CholCn-DTPA (n = 2, 5, 6) and Chol-C2-O-C2-DTPA compounds. Bicelle Preparation. Bicelles were prepared following our previously reported protocols.15,16,18,23,29 The phospholipids, cholesterol, and cholesterol-DTPA conjugates were suspended in chloroform stock solutions of 10 mg/mL. A 10 mM stock solution of thulium(III) chloride in methanol was prepared to supply the lanthanide ions. After the addition of the correct amount of each compound to a roundbottomed flask, a dry lipid film was obtained by evaporation under reduced pressure at 40 °C. The dry lipid film was then hydrated by the addition of 50 mM sodium phosphate buffer at a pH value of 7.5. For SANS experiments, a D2O phosphate buffer of corresponding concentration and a pH value of 7.0 were used. A total lipid concentration of 15 mM is required with DMPC/Chol/Chol-CnDTPA/DMPE-DTPA/Tm3+ (molar ratio 16:2:2:5:7) or DMPC/ Chol-Cn-DTPA/DMPE-DTPA/Tm3+ (molar ratio 16:4:5:9). The freshly hydrated samples were subject to five consecutive freeze− thaw cycles consisting of plunging in liquid nitrogen before heating to 50 °C. In a final step, the samples were extruded (Lipex Biomembranes, Vancouver, Canada) 10 times through a polycarbonate membrane (Sterico, Dietikon, Switzerland) with a pore size of 200 nm and another 10 times through a polycarbonate membrane with a pore size of 100 nm. Small-Angle Neutron Scattering. SANS measurements were conducted on the SANS-I beamline at PSI, Villigen, Switzerland. The neutron wavelength was fixed at 0.8 nm, and the detector distance was set at 2, 6, or 18 m. A 2D 3He detector was used to cover a q range of between 0.03 and 1.5 nm−1. Data were corrected for the blank cell, transmission, and detector efficiency. Radially averaged scattering curves were constructed from scattering patterns of samples measured in the absence of any magnetic fields. The SASfit software package was used for the fitting of the scattering curves.30 Bicelle geometries were fitted with Porod’s approximation for cylinders as a form factor in combination with an appropriate size distribution.18,31 The hydrodynamic radius was computed for comparison with DLS measurements.32 Bicelles with a hole were fitted with a form factor of a flat cylindrical shell.33 A log-normal distribution was used when needed to account for the size distribution of concentric holes. The SANS measurements were mainly conducted at 5 °C for comparison with other techniques. However, measurements were also carried out at 10, 13, 25, and 40 °C as specified throughout the article and in the figure captions. Dynamic Light Scattering. DLS measurements were conducted on a Malvern Zetasizer Nano ZS equipped with a He−Ne laser fixed at 90°. Samples were measured undiluted by using noninvasive backscattering technology (NIBS). All DLS measurements were conducted at 5 °C. Cryo-Transmission Electron Microscopy. Samples measured by cryo-TEM were prepared as described previously.34 The bicelle mixture was suspended as a thin aqueous films on holey carbon grids

■

RESULTS Structure of the Polymolecular Assemblies Composed of DMPC/Chol-C n -DTPA/DMPE-DTPA/Tm 3+ (16:4:5:9). Chol-C2-DTPA and Chol-C6-DTPA were integrated within the phospholipid bilayers composed of DMPC and DMPE-DTPA/Tm3+ at a total lipid concentration of 15 mM. The resulting polymolecular aggregate structures were investigated by cryo-TEM and DLS. In this subsection, all samples mentioned were composed of DMPC/Chol-CnDTPA/DMPE-DTPA/Tm3+ (16:4:5:9). The linker length n was varied, and the samples will be referred to by the Chol-CnDTPA molecule name for simplicity. For the Chol-C2-DTPA system, cryo-TEM micrographs revealed that long cylindrical aggregates were preferentially formed over bicelles at both investigated temperatures in Figure 2A,B. Tilting the sample holder by 30° showed aggregates with

Figure 2. Cryo-TEM micrographs of DMPC/Chol-C2-DTPA/DMPEDTPA/Tm3+ (16:4:5:9) flash frozen at (A) 5 °C and at (B) 40 °C. (C) Cryo-TEM micrographs of DMPC/Chol-C6-DTPA/DMPEDTPA/Tm3+ (16:4:5:9) flash frozen at 5 °C. If not specified, the sample holder tilt angle was 0°. Each scale bar represents 200 nm.

the same thickness and contrast as in the nontilted view in Figure 2B, supporting the existence of a round cross section in these structures. Cryo-TEM micrographs of samples containing Chol-C6-DTPA kept at 5 °C prior to flash freezing revealed the presence of small bicelles and several long and thin ribbonlike aggregates in Figure 2C. DLS measurements supported the presence of bicelles with a hydrodynamic radius of 18.3 nm at 5 °C. Both cholesterol-DTPA molecules induced polymolecular 9007

DOI: 10.1021/acs.langmuir.6b01968 Langmuir 2016, 32, 9005−9014

Article

Langmuir

white spots on the micrographs. Bright patches may be distinguished on some of the polymolecular assemblies presented in Figure 3A. Small holes were found to a larger extent at 40 °C as emphasized with arrows in Figure 3B. By introducing cholesterol, bicelles formed preferentially over the cylindrical aggregates. Cholesterol, with its inverted conelike molecular geometry, probably acted against the curvature induced by the cholesterol-DTPA conjugates.18 In a similar way to the Chol-C2-DTPA molecule, the presence of Chol-C5-DTPA induced curvature that resulted in smaller assembled structures. DLS measurements revealed an average hydrodynamic radius of 47 nm at 5 °C, which is 15% smaller than for bicelles containing only cholesterol.18 The existence of several non-disk-shaped aggregates that contain multiple small holes within the bilayer was evidenced by the lighter spots in the cryo-TEM micrograph of Figure 3C. The appearance of perforations within the bilayer may have resulted from higher local concentrations of the conelike lipids.36 Such a possibility is conceivable when considering the diversity of forces induced by the coexistence of four structurally distinct amphiphiles in the bilayer. Systems containing Chol-C6-DTPA revealed long ribbonlike aggregates together with bicelles with concentric holes as shown in Figure 3D. The part of the ribbon indicated with a tilted arrow in Figure 3D appeared in face-on view and in edgeon view if the sample holder was tilted from 0 to 30°, respectively. Ribbonlike structures increase the edge area of the aggregates when comparing to bicelles of equivalent surface area. The replacement of Chol-C6-DTPA with 50% cholesterol was not sufficient to counterbalance the curvature-inducing nature of the Chol-C6-DTPA molecule. Nevertheless, an increased aggregate size was evident from cryo-TEM micrographs and was confirmed by DLS measurements as the hydrodynamic radius was, at 40.9 nm, about twice as large as in the mixture without cholesterol. The existence of perforations within the bilayer of bicelles containing Chol-C2-DTPA was supported by SANS measure-

aggregate structures with a higher degree of curvature than bicelles. Structure of the Polymolecular Assemblies Composed of DMPC/Chol/Chol-Cn-DTPA/DMPE-DTPA/Tm3+ (16:2:2:5:7). To reduce the curvature resulting from the incorporation of the cholesterol-DTPA conjugates and benefit from the molecule’s potential for generating magnetically alignable bicelle structures, we replaced 50% of the cholesterolDTPA with cholesterol molecules. The resulting polymolecular aggregate structures were investigated by cryo-TEM, DLS, and SANS. In this subsection, all samples mentioned were composed of DMPC/Chol/Chol-Cn-DTPA/DMPE-DTPA/ Tm3+ (16:2:2:5:7). The linker length n was varied, and the samples will be referred to by the Chol-Cn-DTPA molecule name for simplicity. For the mixture containing Chol-C2-DTPA, bicelles were observed in both edge-on and face-on views at 5 °C in Figure 3A. The presence of holes within the bicelle disk appears as

Figure 3. Cryo-TEM micrographs of DMPC/Chol/Chol-C2-DTPA/ DMPE-DTPA/Tm3+ (16:2:2:5:7) flash frozen at (A) 5 °C and at (B) 40 °C. (C) Cryo-TEM micrographs of DMPC/Chol/Chol-C5-DTPA/ DMPE-DTPA/Tm3+ (16:2:2:5:7) flash frozen at 5 °C and (D) DMPC/Chol/Chol-C6-DTPA/DMPE-DTPA/Tm3+ (16:2:2:5:7) at 5 °C. The appearance of perforations within the bilayer is shown with arrows in micrographs B and C. The tilted arrows in micrographs D highlight the planar nature of the ribbonlike polymolecular aggregate as the sample holder was tilted from 0 to 30°. If not specified, the sample holder tilt angle was 0°. Each scale bar represents 200 nm.

Figure 4. Radially averaged SANS curve (×) and fitting () of (A) DMPC/Chol/Chol-C2-DTPA/DMPE-DTPA/Tm3+ and (B) DMPC/Chol/ Chol-C5-DTPA/DMPE-DTPA/Tm3+ (16:2:2:5:7) measured at 5 °C. The respective form factors employed for the fittings are schematically represented. (C) The geometric difference between the two samples is emphasized in the SANS curves, where the intensity I was multiplied by the square of the scattering vector q2. 9008

DOI: 10.1021/acs.langmuir.6b01968 Langmuir 2016, 32, 9005−9014

Article

Langmuir

Figure 5. (A) Schematic comparison of the molecular shape of Chol-C2-DTPA-Tm3+ and Chol-C6-DTPA-Tm3+. The possible positions for the DTPA-Tm3+ headgroup are highlighted. The longer linker offers a larger range of movement, which results in a stronger conelike geometry (smaller packing parameter). The molecular sketches illustrate some of the possible conformational changes that may occur in the linker. These bond rotations are at the source of the DTPA-Tm3+ headgroup’s ability to occupy more or fewer positions in space. The resulting conformers are labeled and shown in another color. The 3D models of the cholesterol-DTPA conjugates gained by an MM2 geometry optimization are shown with carbon atoms in gray, hydrogen atoms in white, oxygen atoms in red, nitrogen atoms in blue, and Tm3+ in yellow. The possibility of forming a hydrogen bond in the Chol-C2-DTPA-Tm3+ molecule is highlighted with a dashed line. (B) Magnetic field required to reach half the order parameter B(S1/2) at 5 and 50 °C for DMPC/Chol/Chol-C2-DTPA/DMPE-DTPA/Tm3+ (16:2:2:5:7), DMPC/Chol/Chol-C6-DTPA/DMPE-DTPA/Tm3+ (16:2:2:5:7), and DMPC/Chol-C6-DTPA/DMPE-DTPA/Tm3+ (16:4:5:9).

ments at 5 °C as presented in Figure 4A. The best fit was achieved with a form factor describing a disk with a concentric hole of radius 13.7 nm surrounded by a rim of 56.7 nm. The total radius of 70.4 nm was in good agreement with the hydrodynamic radius of 71.5 nm obtained from DLS measurements at 5 °C. The geometric diversity of bicelles containing Chol-C5-DTPA was confirmed by SANS measurements in Figure 4B. The radially averaged scattering curve of the sample at 5 °C was fitted with a form factor for flat cylindrical disks. A log-normal size distribution proved ineffective for fitting the data. Instead, we propose a fractal distribution to account for the large size disparity that was particularly pronounced in favor of smaller species. A bilayer thickness of 4.8 nm resulted from the fitting procedure and was in good agreement with the literature.23,37 Furthermore, the computed hydrodynamic radius of 47 nm was in agreement with the DLS results. The scattering intensity of both samples decays with q−2, which is characteristic of planar disklike structures.18 The intensity was multiplied by q2 and plotted as a function of the scattering vector q in Figure 4C. This emphasizes scattering resulting from other geometric considerations. The proposed contrasting nature in the bicelles formed from the Chol-C2-DTPA and Chol-C5-DTPA systems is confirmed by the observable difference in these SANS curves. Molecular Modeling of Chol-C2-DTPA/Tm3+ and CholC6-DTPA/Tm3+ Molecules. Chol-C2-DTPA and Chol-C6DTPA complexed with thulium were geometrically optimized using molecular modeling routine MM2. This model was employed to identify the energetic ground configuration of the molecule in order to further speculate on reasonable conformational changes that could occur in the bilayer’s environment.

This simplified picture offered a basis for understanding the observed polymolecular architectures without proceeding to costly and time-consuming molecular dynamics simulations. The number of thermally induced conformational changes and electrostatic forces that the molecules would be exposed to are not considered in this first step. However, by observing interactions already occurring within the molecule in a simplified environment, we were able to identify sites prone to interact with the surroundings. We then constructed a realistic average molecular geometry for the cholesterol-DTPA conjugates based on the range of conformations they may adopt. The results are presented and schematically analyzed in Figure 5A. When comparing Chol-C2-DTPA and Chol-C6-DTPA, it was evident that the length of the carbon chain of the linker has a significant impact on the geometry of the molecules. The alltrans conformation of the hydrocarbon chain suggested by MM2 calculations is highly unlikely in the bicelle environment. Different conformations induced by the surrounding environment’s thermal energy or electrostatic forces are highly conceivable. The low energy barrier for rotation around carbon−carbon single bonds essentially allows free rotation about their axis in the temperature range of this study. van der Waals interactions between adjacent lipids in the bilayer would remove this conformational freedom from the aliphatic chains of the phospholipids that are in the solid-ordered phase. The carbon chains forming the linkers would not benefit from such stabilizing interactions because they are likely exposed to the unfavorable hydrophilic environment of the bilayer. Consequently, they have freedom to undergo conformation changes. The resulting molecular reorientations deviate from 9009

DOI: 10.1021/acs.langmuir.6b01968 Langmuir 2016, 32, 9005−9014

Article

Langmuir

Figure 6. Hydrodynamic radius RH of the bicelles as determined by DLS (black circles) and their birefringence signal at 5 °C and 5.5 T (red squares) as a function of the number of carbon atoms, n, in the Chol-Cn-DTPA linker. Filled markers correspond to the mixtures with cholesterol: DMPC/ Chol/Chol-Cn-DTPA/DMPE-DTPA/Tm3+ (16:2:2:5:7). Empty markers are from a DMPC/Chol-C2-DTPA/DMPE-DTPA/Tm3+ (16:4:5:9) sample. A schematic representation of the birefringence experiment is presented on the right-hand side. The polarized laser shines through the sample cuvette exposed to an external magnetic field B. A sample composed of magnetically aligned bicelles will result in a strong birefringence signal (top closeup), while weakly aligned bicelles do not (bottom closeup).

be surpassed by the magnetic energy of the bicelles for alignment to occur.35 An increase in magnetic alignability upon replacement of 50% of the Chol-C6-DTPA with cholesterol was observed as the B(S1/2) values dropped when comparing the filled and empty circles in Figure 5B. The enhanced alignment is in line with the aforementioned increase in bicelle size. Moreover, samples containing Chol-C2-DTPA exhibited considerably lower B(S1/2) values, confirming its superior capacity of creating larger and more alignable aggregate structures when compared to the six-carbon-long linker CholC6-DTPA compound. These results were further supported by comparing the hydrodynamic radius and magnetic alignability of bicelles at 5.5 T containing Chol-Cn-DTPA with n = 2, 5, and 6 at 5 °C. The results of birefringence measurements were plotted as a function of the number of carbon atoms in the linker in Figure 6. There exists an inverse correlation between the number of carbon atoms in the linker separating the sterol backbone from the chelator moiety and the subsequent mean bicelle radius as revealed by DLS measurements. The enhanced birefringence signal with increasing radius of the polymolecular aggregates supported the existence of this trend. Larger magnetic alignability was correlated to the increased cumulative contribution of the magnetic orientation energy Emag of individual molecules in the bicelle when it increased in size. The results suggested that curvature increased within the polymolecular aggregates with increasing number of carbons in the linker chain. Cholesterol’s interactions with neighboring phospholipids in the bilayer are dominated by van der Waals forces between the hydrophobic sterol backbone and the aliphatic chains of the phospholipids. The position of cholesterol within the bilayer is fixed and primarily dictated by these forces.20,38 Therefore, increasing the linker chain length would tend to push the DTPA-chelator headgroup out into the hydrophilic environment surrounding the bicelle. The degrees of freedom that the amphiphile’s headgroup has to adopt for different conformations increase with the number of

the fully extended zigzag arrangement and lead to a larger effective headgroup size, which attributes a conelike geometry to the amphiphilic cholesterol-DTPA conjugate. Moreover, the number of different conformations that the six-carbon-long linker may adopt is considerably larger than for the two-carbon counterpart. The added degrees of freedom of the Chol-C6DTPA headgroup may further reduce the packing parameter of the amphiphile. Consequently, the tendency to form polymolecular architectures with an enhanced degree of curvature is highly probable. Moreover, MM2 calculations revealed the possibility of forming a hydrogen-bond-stabilized ring structure between the carbamate and the amide bond closest to the DTPA moiety in Chol-C2-DTPA. This ring formation could further reduce the length of the spacer and limit the molecule’s conformational freedom, forcing it to assume a conelike geometry (packing parameter < 1). Correlation between the Linker Length of Chol-CnDTPA and Bicelle Geometry and the Effect of Chol-C2-OC2-DTPA. To test the hypothesis that arose from molecular modeling, we measured the order parameter S of the polymolecular assemblies at a higher magnetic field strength of 33 T. Such a high magnetic field strength allows full saturation of the alignment of the bicelle samples to be reached. These conditions permit the determination of the order parameter S and quantitative data analysis as discussed by Liebi et al.33 For bicelle systems containing Tm3+, the maximum value of S is 1. The magnetic field strength B at which the order parameter reaches half of its maximum value (S1/2 = 0.5) is readily computed and directly related to the product nΔχ of the number of molecules n in one bicelle and the molar diamagnetic susceptibility anisotropy Δχ. This results in B(S1/2), which is a direct measurement of the magnetic alignability of the bicelles. The magnetic field required to reach half of the order parameter B(S1/2) is reported in Figure 5B. The lower B(S1/2) was, the better aligned the bicelles were in the magnetic field. The increase in B(S1/2) when moving from 5 to 50 °C may be explained by higher thermal energy that must 9010

DOI: 10.1021/acs.langmuir.6b01968 Langmuir 2016, 32, 9005−9014

Article

Langmuir

radius of 48 nm. This was in strong agreement with a mean hydrodynamic radius of 49 nm as measured by DLS. These values were analogous to those obtained for bicelle systems containing Chol-C5-DTPA. Both cholesterol-DTPA conjugates interacted in similar ways with the neighboring components in the bilayer, leading to complementary polymolecular structures. The introduction of polarity within the linker’s chain did not permit control of the bicelle’s size; the linker’s chain length and its resulting freedom for conformational change are the determining parameters. Thermal Resistance of Bicelles Containing Chol-CnDTPA. The preparation of bicelles based on DMPC and DMPE-DTPA/Tm3+ phospholipids that are capable of withstanding temperatures as high as 40 °C was possible by doping the bilayer with cholesterol. Cholesterol’s capacity to induce a liquid-ordered phase could be the source of the bicelle’s thermoresistivity.17,39 Without cholesterol, the thermoreversible collapse into vesicles would occur at the phase-transition temperature of DMPC at 24 °C when the lipids move from the solid-ordered to the liquid-disordered state.18 With cholesterol and cholesterol-based DTPA-Tm3+ anchors, we exploited the molecule’s ability to enhance the range of temperature in which disklike aggregates exist. The thermal resistance of the multilipidic systems was explored with DMPC/Chol/Chol-C5-DTPA/DMPE-DTPA/ Tm3+ (16:2:2:5:7) by heating the samples to temperatures above 5 °C. The appearance of a large concentric hole within the bicelles was observed at 25 °C in the cryo-TEM micrographs presented in Figure 8A. The similarity observed at 5 °C between this Chol-C5-DTPA system and the Chol-C2O-C2-DTPA system in DMPC/Chol/Chol-C2-O-C2-DTPA/ DMPE-DTPA/Tm3+ (16:2:2:5:7) was also evident at 25 °C in

atoms in the linker. This induces an increase in the effective headgroup size, resulting in smaller packing parameters as the molecules take on more conelike geometries and generates curvature in the final polymolecular aggregates. These results are in line with the hypothesized schematic based on the MM2 calculation discussed in Figure 5A. To further support these affirmations, Chol-C2-O-C2-DTPA was synthesized and incorporated into the bilayer. This molecule differs from Chol-C5-DTPA in that the carbon atom in the middle of the linker chain was replaced with an oxygen atom as shown in Figure 1. By introducing polarity into the carbon chain of the linker, we aimed to investigate the ability of the hydrophilic environment to stabilize the linker structure and hence limit its conformational degrees of freedom. Cryo-TEM micrographs of the sample at 5 °C presented in Figure 7A supported the existence of bicelles upon the

Figure 7. DMPC/Chol/Chol-C2-O-C2-DTPA/DMPE-DTPA/Tm3+ (16:2:2:5:7) at 5 °C. (A) Cryo-TEM micrograph of the flash-frozen sample with a sample holder tilt angle of 0°; the scale bar represents 200 nm. (B) Radially averaged SANS curve fitted with a form factor for flat cylindrical disks.

incorporation of Chol-C2-O-C2-DTPA. These bicelle structures were similar to the ones formed by Chol-C5-DTPA as revealed in the cryo-TEM micrographs in Figure 3C. The presence of Chol-C2-O-C2-DTPA in the bilayer yielded systems with alignability comparable to that of the ones with Chol-C5DTPA as evidenced by their similar birefringence signal: 3.31 × 10−6 for the Chol-C5-DTPA-containing bicelles compared to 3.42 × 10−6 for the Chol-C2-O-C2-DTPA system at 5 °C. The similarity between the bicelle architectures resulting from these two compounds was further confirmed by SANS measurements. The radially averaged SANS curve of the DMPC/Chol/ Chol-C2-O-C2-DTPA/DMPE-DTPA/Tm3+ (16:2:2:5:7) sample at 5 °C in Figure 7B yielded a computed hydrodynamic

Figure 8. Cryo-TEM micrograph of (A) DMPC/Chol/Chol-C5DTPA/DMPE-DTPA/Tm3+ (16:2:2:5:7) and (B) DMPC/Chol/ Chol-C2-O-C2-DTPA/DMPE-DTPA/Tm3+ (16:2:2:5:7) flash frozen at 25 °C with a sample holder tilt angle of 0°; the scale bars represent 200 nm. Radially averaged SANS curve of (C) DMPC/Chol/Chol-C5DTPA/DMPE-DTPA/Tm3+ (16:2:2:5:7) at 10 and 25 °C and of (D) DMPC/Chol/Chol-C2-O-C2-DTPA/DMPE-DTPA/Tm3+ (16:2:2:5:7) at 13 and 25 °C. A form factor for a disk with a concentric hole was employed in all fittings. 9011

DOI: 10.1021/acs.langmuir.6b01968 Langmuir 2016, 32, 9005−9014

Article

Langmuir the cryo-TEM micrograph in Figure 8B. Analogously, large concentric holes appeared within the bicelles. The transformation was noticeable by SANS for the system containing Chol-C5-DTPA in Figure 8C where a change in gradient appeared at a q value of 2.5 at 10 °C and, more pronounced, at 25 °C. These curves confirmed the appearance of a disk with a concentric hole as the data at 25 °C was fitted with a form factor of a flat cylindrical shell of thickness 15 nm and a core radius with a log-normal distribution of μ = 15 nm and σ = 0.5. The presence of this architecture was also evident at 10 °C, where SANS fittings yielded the equivalent parameters except for a larger shell of 20 nm. Furthermore, the radially averaged scattering curves for bicelles containing Chol-C2-O-C2-DTPA in Figure 8D showed the same trends as for Chol-C5-DTPA along the studied temperature range. The SANS curve at 25 °C was fitted with a form factor of a flat cylindrical shell of thickness 13 nm and a core radius with a log-normal distribution of μ = 19 nm and σ = 0.5. The appearance of a concentric hole in the bilayer was previously reported for DMPC/Chol/DMPE-DTPA/Tm3+ bicelles at an equivalent temperature range upon heating.18 This suggests that the mechanics of lipid rearrangement within the bilayer remain highly influenced by cholesterol-phospholipid interactions. The formation of a single hole at the core of the bicelle could be a result of the coalescence of the smaller perforations observed at lower temperatures. Under certain conditions of bilayer line tension, such a phenomenon is common and has been investigated by Monte Carlo simulation.40 Chol-C2-DTPA and Chol-C5-DTPA in a 1:1 mixture with cholesterol in the phospholipid bilayer resulted in bicelle structures at 40 °C as evidenced in the cryo-TEM micrographs in Figure 2D and in Figure 9A, respectively. The planar nature of the polymolecular aggregate was observed from a side-on view where they appear as a black line on the micrographs. The system containing Chol-C2-O-C2-DTPA continued to yield polymolecular aggregates with architecture similar to that of those containing Chol-C5-DTPA at 40 °C as evidenced in the cryo-TEM micrograph in Figure 9B. The ability to form stable bicelles at temperatures as high as 40 °C was highlighted by the SANS data in Figure 9C. Unlike previously reported DMPC/ DMPE-DTPA/Tm3+ bicelle systems, the collapse into vesicles did not occur at this temperature because there was no evidence of a characteristic kink at q = 0.2 nm−1 for both the Chol-C5-DTPA and Chol-C2-O-C2-DTPA systems.15,18 The introduction of cholesterol and cholesterol-DTPA conjugates within the bilayer enhanced the thermal stability of the polymolecular aggregates by exploiting the stabilization effect induced by the sterol backbone while simultaneously benefitting from the added functionality of an engineered polar headgroup. The combined possibility of fine tuning the cholesterol-DTPA conjugate’s molecular structure to generate desired polymolecular aggregate architectures in a specific temperature range results in a unique engineering toolbox for lanthanide ion-doped, magnetically responsive soft materials.

Figure 9. Cryo-TEM micrographs of (A) DMPC/Chol/Chol-C5DTPA/DMPE-DTPA/Tm3+ (16:2:2:5:7) and (B) DMPC/Chol/ Chol-C2-O-C2-DTPA/DMPE-DTPA/Tm3+ (16:2:2:5:7) flash frozen at 40 °C with a sample holder tilt angle of 0°; the scale bar represents 200 nm (C) The samples’ radially averaged SANS curves at 40 °C.

chain resulted in a larger effective headgroup size by an increased degree of freedom for adopting different conformations. The introduction of a polar component with an ether bond within the linker structure of Chol-C2-O-C2-DTPA was not sufficient to stabilize the headgroup through a reduction of repulsive hydrophobic interactions with the surrounding environment. The inverse correlation established between the number of carbon atoms in the linker and the bicelle radius provides guidelines for tailoring the polymolecular aggregate’s geometry with cholesterol-DTPA conjugates following a simple, packing-parameter-based model. Short linker structures, such as in Chol-C2-DTPA, were required to guarantee the largest and most magnetically alignable bicelle structures. The ideas conveyed in this work could provide essential building blocks toward a more systematic approach for the design of optimal bicelle architectures. The resulting degree of control could, for example, be most welcome for the study of bilayerassociated molecules by NMR spectroscopy, where the magnetic alignment of the employed bicelles must be carefully tailored on the basis of the nature of the study. Although this is commonly achieved by controlling the bicelle’s size through the ratio of the composing phospholipids and the overall lipid concentration, working with cholesterol-DTPA conjugates offers alternative possibilities.6 Furthermore, the possibility of forming stable bicelle structures at temperatures equivalent to physiological conditions at up to 40 °C is a valuable feature offered by these cholesterol-based bilayer-doping compounds.

■

CONCLUSIONS The design of DMPC/DMPE-DTPA/Tm3+-based bicelle systems with cholesterol-DTPA conjugates demands a careful consideration of the molecular structure of the linker unit employed. The number of atoms in the linker chain was a key parameter in controlling the adopted geometry of the molecule and tailoring the degree of curvature introduced into the polymolecular aggregate’s geometry. Adding atoms to the linker 9012

DOI: 10.1021/acs.langmuir.6b01968 Langmuir 2016, 32, 9005−9014

Article

Langmuir

(8) Israelachvili, J. N.; Mitchell, D. J.; Ninham, B. W. Theory of selfassembly of hydrocarbon amphiphiles into micelles and bilayers. J. Chem. Soc., Faraday Trans. 2 1976, 72, 1525−1568. (9) Yamaguchi, T.; Suzuki, T.; Yasuda, T.; Oishi, T.; Matsumori, N.; Murata, M. NMR-based conformational analysis of sphingomyelin in bicelles. Bioorg. Med. Chem. 2012, 20, 270−278. (10) Meyer, H. W.; Richter, W.; Rettig, W.; Stumpf, M. Bilayer fragments and bilayered micelles (bicelles) of dimyristoylphosphatidylglycerol (DMPG) are induced by storage in distilled water at 4°C. Colloids Surf., A 2001, 183−185, 495−504. (11) Ujwal, R.; Bowie, J. U. Crystallizing membrane proteins using lipidic bicelles. Methods 2011, 55, 337−341. (12) Yamaguchi, M.; Tanimoto, Y. Magneto-Science: Magnetic Field Effects on Materials: Fundamentals and Applications; Springer Series in Materials Science; Springer: Berlin, 2006. (13) Prosser, R. S.; Hwang, J. S.; Vold, R. R. Magnetically aligned phospholipid bilayers with positive ordering: a new model membrane system. Biophys. J. 1998, 74, 2405−2418. (14) Prosser, R. S.; Bryant, H.; Bryant, R. G.; Vold, R. R. Lanthanide chelates as bilayer alignment tools in NMR studies of membraneassociated peptides. J. Magn. Reson. 1999, 141, 256−260. (15) Beck, P.; Liebi, M.; Kohlbrecher, J.; Ishikawa, T.; Rüegger, H.; Fischer, P.; Walde, P.; Windhab, E. Novel type of bicellar disks from a mixture of DMPC and DMPE-DTPA with complexed lanthanides. Langmuir 2010, 26, 5382−5387. (16) Liebi, M.; Kuster, S.; Kohlbrecher, J.; Ishikawa, T.; Fischer, P.; Walde, P.; Windhab, E. J. Magnetically enhanced bicelles delivering switchable anisotropy in optical gels. ACS Appl. Mater. Interfaces 2014, 6, 1100−1105. (17) Róg, T.; Pasenkiewicz-Gierula, M.; Vattulainen, I.; Karttunen, M. Ordering effect of cholesterol and its analogues. Biochim. Biophys. Acta, Biomembr. 2009, 1788, 97−121. (18) Liebi, M.; Kohlbrecher, J.; Ishikawa, T.; Fischer, P.; Walde, P.; Windhab, E. J. Cholesterol increases the magnetic aligning of bicellar disks from an aqueous mixture of DMPC and DMPE-DTPA with complexed thulium ions. Langmuir 2012, 28, 10905−10915. (19) Björkbom, A.; Róg, T.; Kaszuba, K.; Kurita, M.; Yamaguchi, S.; Lönnfors, M.; Nyholm, K. M. J.; Vattulainen, I.; Katsumura, S.; Slotte, J. P. Effect of sphingomyelin head group size on molecular properties and interactions with cholesterol. Biophys. J. 2010, 99, 3300−3308. (20) Ohvo-Rekilä, H.; Ramstedt, B.; Leppima, P.; Slotte, J. P. Cholesterol interactions with phospholipids in membranes. Prog. Lipid Res. 2002, 41, 66−97. (21) Mannock, D. A.; Lewis, R. N. A H.; McMullen, T. P. W.; McElhaney, R. N. The effect of variations in phospholipid and sterol structure on the nature of lipid-sterol interactions in lipid bilayer model membranes. Chem. Phys. Lipids 2010, 163, 403−448. (22) Hosta-Rigau, L.; Zhang, Y.; Teo, B. M.; Postma, A.; Städler, B. Cholesterol - a biological compound as a building block in bionanotechnology. Nanoscale 2013, 5, 89−109. (23) Liebi, M.; Kuster, S.; Kohlbrecher, J.; Ishikawa, T.; Fischer, P.; Walde, P.; Windhab, E. J. Cholesterol-diethylenetriaminepentaacetate complexed with thulium ions integrated into bicelles to increase their magnetic alignability. J. Phys. Chem. B 2013, 117, 14743−14748. (24) Mezzenga, R.; Schurtenberger, P.; Burbidge, A.; Michel, M. Understanding foods as soft materials. Nat. Mater. 2005, 4, 729−740. (25) Rui, M.; Guo, W.; Ding, Q.; Wei, X.; Xu, J.; Xu, Y. Recombinant high-density lipoprotein nanoparticles containing gadolinium-labeled cholesterol for morphologic and functional magnetic resonance imaging of the liver. Int. J. Nanomed. 2012, 7, 3751−3768. (26) Suzuki, H.; Kanai, A.; Uehara, T.; Guerra Gomez, F. L.; Hanaoka, H.; Arano, Y. Facile synthesis and evaluation of Cfunctionalized benzyl-1-oxa-4,7,10-triazacyclododecane-N,N′,N″-triacetic acid as chelating agent for 111In-labeled polypeptides. Bioorg. Med. Chem. 2012, 20, 978−984. (27) Nudelman, A.; Bechor, Y.; Falb, E.; Fischer, B.; Wexler, B. A.; Nudelman, A. Acetyl chloride-methanol as a convenient reagent for: A) quantitative formation of amine hydro- chlorides B) carboxylate

Tuning the temperature-resistance of self-assembled polymolecular aggregates is of great importance in numerous applications including drug delivery for dermal applications,41 protein characterization by NMR,42 and the development of smart optical gels.16 Furthermore, the synthesized cholesterolDTPA conjugates alone may find use as contrast-enhancing agents in magnetic resonance imaging. Their ability to integrate into phospholipid bilayers is of particular interest for such studies.23,25

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.6b01968. Detailed synthesis procedures, reaction schemes, analytical data, and a further description of the fractal size distribution employed for SANS fittings (PDF)

■

AUTHOR INFORMATION

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS

■

REFERENCES

We acknowledge the Swiss National Science Foundation for funding (SMhardBi project number 200021_150088/1). Part of this work was supported by EuroMagNET under EU contract no. 228043. The SANS experiments were performed at the Swiss spallation neutron source SINQ, Paul Scherrer Instute, Villigen, Switzerland. The birefringence measurements under high magnetic fields were performed at the high-field magnetic laboratory HFML, Nijmegen, The Netherlands. We acknowledge C. Holenstein for her contribution to birefringence experiments and the synthesis of the cholesterol-DTPA compounds. We thank Dr. S. Bolisetty for his help with DLS measurements on the Zetasizer Nano ZS.

(1) Sanders, C. R.; Schwonek, J. P. Characterization of magnetically orientable bilayers in mixtures of dihexanoylphosphatidylcholine and dimyristoylphosphatidylcholine by solid-state NMR. Biochemistry 1992, 31, 8898−8905. (2) Glover, K. J.; Whiles, J. A.; Wu, G.; Yu, N.; Deems, R.; Struppe, J. O.; Stark, R. E.; Komives, E. A.; Vold, R. R. Structural Evaluation of Phospholipid Bicelles for Solution-State Studies of MembraneAssociated Biomolecules. Biophys. J. 2001, 81, 2163−2171. (3) Prosser, R. S.; Evanics, F.; Kitevski, J. L.; Al-Abdul-Wahid, M. S. Current applications of bicelles in NMR studies of membraneassociated amphiphiles and proteins. Biochemistry 2006, 45, 8453− 8465. (4) Losonczi, J. A.; Prestegard, J. H. Improved dilute bicelle solutions for high-resolution NMR of biological macromolecules. J. Biomol. NMR 1998, 12, 447−451. (5) Tjandra, N.; Bax, A. Direct measurement of distances and angles in biomolecules by NMR in a dilute liquid crystalline medium. Science 1997, 278, 1111−1114. (6) Dürr, U. H. N.; Gildenberg, M.; Ramamoorthy, A. The magic of bicelles lights up membrane protein structure. Chem. Rev. 2012, 112, 6054−6074. (7) Kumar, V. V. Complementary molecular shapes and additivity of the packing parameter of lipids. Proc. Natl. Acad. Sci. U. S. A. 1991, 88, 444−448. 9013

DOI: 10.1021/acs.langmuir.6b01968 Langmuir 2016, 32, 9005−9014

Article

Langmuir ester formation C) mild removal of N-t-BOC-protective group. Synth. Commun. 1998, 28, 471−474. (28) Aikawa, K.; Nishigaki, J.; Takahashi, K. Glycerophosphoric Acid Ester Derivative Having Polyfunctional Metal Chelate Structure. EP1795208A1 2007. (29) Gielen, J. C.; Shklyarevskiy, O. I.; Schenning, A. P. H. J.; Christianen, P. C. M.; Maan, J. C. Using magnetic birefringence to determine the molecular arrangement of supramolecular nanostructures. Sci. Technol. Adv. Mater. 2009, 10, 014601. (30) Bressler, I.; Kohlbrecher, J.; Thünemann, F. A. SASfit: a tool for small-angle scattering data analysis using a library of analytical expressions. J. Appl. Crystallogr. 2015, 48, 1587−1598. (31) Porod, G. Die Abhängigkeit der Röntgenkleinwinkelstreuung von Form und Grösse der kolloiden Teilchen in verdünnten Systemen IV. Acta Phys. Austriaca 1947, 255−292. (32) Rübe, A.; Hause, G.; Mäder, K.; Kohlbrecher, J. Core-shell structure of Miglyol/poly(D,L-lactide)/Poloxamer nanocapsules studied by small-angle neutron scattering. J. Controlled Release 2005, 107, 244−252. (33) Liebi, M.; van Rhee, P. G.; Christianen, P. C. M.; Kohlbrecher, J.; Fischer, P.; Walde, P.; Windhab, E. J. Alignment of bicelles studied with high-field magnetic birefringence and small-angle neutron scattering measurements. Langmuir 2013, 29, 3467−3473. (34) Ishikawa, T.; Sakakibara, H.; Oiwa, K. The architecture of outer dynein arms in situ. J. Mol. Biol. 2007, 368, 1249−1258. (35) Christianen, P. C. M.; Shklyarevskiy, I. O.; Boamfa, M. I.; Maan, J. C. Alignment of molecular materials in high magnetic fields. Phys. B 2004, 346−347, 255−261. (36) Almgren, M. Stomatosomes: perforated bilayer structures. Soft Matter 2010, 6, 1383−1390. (37) Pencer, J.; Nieh, M. P.; Harroun, T. A.; Krueger, S.; Adams, C.; Katsaras, J. Bilayer thickness and thermal response of dimyristoylphosphatidylcholine unilamellar vesicles containing cholesterol, ergosterol and lanosterol: A small-angle neutron scattering study. Biochim. Biophys. Acta, Biomembr. 2005, 1720, 84−91. (38) McMullen, T. P.; McElhaney, R. N. Physical studies of cholesterol-phospholipid interactions. Curr. Opin. Colloid Interface Sci. 1996, 1, 83−90. (39) De Meyer, F.; Smit, B. Effect of cholesterol on the structure of a phospholipid bilayer. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 3654− 3658. (40) Shillcock, J. C.; Seifert, U. Thermally induced proliferation of pores in a model fluid membrane. Biophys. J. 1998, 74, 1754−1766. (41) Barbosa-Barros, L.; Rodríguez, G.; Barba, C.; Cócera, M.; Rubio, L.; Estelrich, J.; López-Iglesias, C.; De la Maza, A.; López, O. Bicelles: Lipid nanostructured platforms with potential dermal applications. Small 2012, 8, 807−818. (42) Shapiro, R. A.; Brindley, A. J.; Martin, R. W. Thermal stabilization of DMPC/DHPC bicelles by addition of cholesterol sulfate. J. Am. Chem. Soc. 2010, 132, 11406−11407.

9014

DOI: 10.1021/acs.langmuir.6b01968 Langmuir 2016, 32, 9005−9014