Dendritic domains with hexagonal symmetry formed by X-shaped bolapolyphiles in lipid membranes

Article abstract of DOI:10.1002/chem.201405994

A novel class of bolapolyphile (BP) molecules are shown to integrate into phospholipid bilayers and self-assemble into unique sixfold symmetric domains of snowflake-like dendritic shapes. The BPs comprise three philicities: a lipophilic, rigid, π-π stacki

Full text of DOI:10.1002/chem.201405994

DOI: 10.1002/chem.201405994
Full Paper
&
Self-Assembly
Dendritic Domains with Hexagonal Symmetry Formed by
X-Shaped Bolapolyphiles in Lipid Membranes
Stefan Werner,[a, c] Helgard Ebert,[a] Bob-Dan Lechner,[a] Frank Lange,[b] Anja Achilles,[b]
Ruth Bärenwald,[b] Silvio Poppe,[a] Alfred Blume,[a] Kay Saalwächter,*[b] Carsten Tschierske,*[a]
and Kirsten Bacia*[a, c]
Abstract: A novel class of bolapolyphile (BP) molecules are
shown to integrate into phospholipid bilayers and self-as-
semble into unique sixfold symmetric domains of snowflake-
like dendritic shapes. The BPs comprise three philicities:
a lipophilic, rigid, p–p stacking core; two flexible lipophilic
side chains; and two hydrophilic, hydrogen-bonding head
groups. Confocal microscopy, differential scanning calorime-
try, XRD, and solid-state NMR spectroscopy confirm BP-rich
domains with transmembrane-oriented BPs and three to
four lipid molecules per BP. Both species remain well organ-
ized even above the main 1,2-dipalmitoyl-sn-glycero-3-phos-
phocholine transition. The BP molecules only dissolve in the
fluid membrane above 708C. Structural variations of the BP
demonstrate that head-group hydrogen bonding is a pre-
requisite for domain formation. Independent of the head
group, the BPs reduce membrane corrugation. In conclusion,
the BPs form nanofilaments by p stacking of aromatic cores,
which reduce membrane corrugation and possibly fuse into
a hexagonal network in the dendritic domains.
Introduction
polymers, such as amphiphilic di- or triblock copolymers, often
affect the membrane integrity and permeability and are, in
many cases, known to form separated microdomains with spe-
cific local structure and packing,[11–13] whereas low-molecular-
weight triblock molecules can stabilize bicellar structures.[14] Bi-
polar amphiphiles, so-called bolaamphiphiles, are incorporated
into membranes to provide stabilization[15] or facilitate transbi-
layer diffusion.[16] Moreover, these bolaamphiphiles can operate
as ligands for lectins[17] or form nanodomains.[18] Bolaamphi-
philes involving molecular rods, such as octaphenylenes,[19]
naphthalene or perylene diimide arrays,[20] and shorter linear
biphenyl,[21] tolane,[22,23] or polyene units,[24] provide useful scaf-
folds for ion or electron conductance through membranes[3] or
serve as fluorescent probes.[24] Apart from these p-conjugated
rods, nonconjugated linear oligospiroketals have also been in-
corporated into lipid membranes.[25] Although the ion trans-
port properties, in particular, have been intensively investigat-
ed for various kinds of molecules, there is only limited knowl-
edge about the modes of self-assembly of these synthetic mol-
ecules in membranes. This issue, which is of relevance to mem-
brane stability and for the performance of membrane
channels, could also contribute to the understanding of funda-
mental aspects of self-assembly in biomembranes, and is ad-
dressed herein.
Lipid membranes with their specific spatial variation of polarity
along the membrane normal provide a unique environment
for supramolecular structure-formation processes in two di-
mensions. Of particular biological relevance are, for instance,
the formation of well-defined pores and channels from pep-
tides, proteins,[1,2] or artificial molecules,[3] and the formation of
lipid domains, sometimes referred to as rafts.[4,5] Microdomain
formation and compartmentalization[6–9] are important in lend-
ing specific functions to biological membranes.[10] Synthetic
[a] S. Werner,+ Dr. H. Ebert,+ B.-D. Lechner,+ S. Poppe, Prof. Dr. A. Blume,
Prof. Dr. C. Tschierske, Prof. Dr. K. Bacia
Institut für Chemie, Martin-Luther-Universität Halle-Wittenberg
06120 Halle (Saale) (Germany)
E-mail: carsten.tschierske@chemie.uni-halle.de
kirsten.bacia@chemie.uni-halle.de
[b] F. Lange, A. Achilles, R. Bärenwald, Prof. Dr. K. Saalwächter
Institut für Physik - NMR
Martin-Luther-Universität Halle-Wittenberg
06120 Halle (Saale) (Germany)
E-mail: kay.saalwaechter@physik.uni-halle.de
[c] S. Werner,+ Prof. Dr. K. Bacia
ZIK HALOmem, Martin-Luther-Universität Halle-Wittenberg
06120 Halle (Saale) (Germany)
[+] These authors contributed equally to this work.
Specifically, we report on a novel self-organization phenom-
enon of X-shaped bolapolyphiles (BPs), namely, synthetic mole-
cules with rigid rod-like oligo(phenylene ethynylene)-based p-
conjugated hydrophobic cores,[26] identical hydrophilic groups
at each end, and two long aliphatic chains attached in the
middle,[27–29] within lipid membranes made of 1,2-dipalmitoyl-
sn-glycero-3-phosphocholine (DPPC). The unique feature is the
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201405994.
ꢀ 2015 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA.
This is an open access article under the terms of the Creative Commons At-
tribution-NonCommercial License, which permits use, distribution and re-
production in any medium, provided the original work is properly cited and
is not used for commercial purposes.
Chem. Eur. J. 2015, 21, 8840 – 8850
8840 ꢀ 2015 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
self-assembly of some of these X-shaped BPs into microdo-
mains of a clearly delineated dendritic shape with hexagonal
symmetry in DPPC membranes of giant unilamellar lipid vesi-
cles (GUVs). We used laser scanning confocal fluorescence mi-
croscopy (CFM) to image the regular dendritic domains and
probe the BP orientation in these domains. To characterize the
structural and dynamic features of the novel structures and
correlate these with the complex phase behavior of the BP/
DPPC/water system, as revealed by differential scanning calo-
rimetry (DSC), we applied different solid-state NMR spectrosco-
py techniques and XRD to multilamellar vesicle (MLV) prepara-
tions.
By using B12 with n=12 and glycerol head groups as a start-
ing point and keeping the core constant, further variations
were introduced: First, the length of the side chains was ex-
tended (B18, n=18). In addition to side-chain variation, both
glycerol head groups of B12 were replaced by simple hydroxyl
groups (C12), by head groups combining a hydrophilic and
flexible oligo(ethylene oxide) spacer unit (EO3) with glycerol
moieties (D12/3), or by oligo(ethylene oxide) (EO7) head
groups incapable of acting as a hydrogen-bond donor (E12/7;
see Scheme 1).
The polyphiles were synthesized by two strategies, as shown
in Scheme 2. Compounds Bn (n=12,18) with glycerol groups
were obtained in a similar way to that used previously for
structurally related compounds.[29] Herein, 1,2-O-isopropylidene
First, the molecular design concept and syntheses of the
new compounds are described, followed by CFM studies of
GUVs prepared from mixtures of DPPC with each of the BPs.
Next, the thermal behavior of one representative BP (B12) in
different concentrations of mixtures of DPPC/H2O studied by
DSC and XRD is presented. The remainder of the paper focuses
on a detailed spectroscopic characterization of B12 in DPPC by
using solid-state 31P and 1H NMR techniques to characterize
the mobility of the respective components to explain the com-
plex thermal behavior and investigate structural features. We
conclude with a discussion of tentative structural models.
Results
Design and synthesis of materials
The structures of the investigated BPs are shown in Scheme 1.
Because certain types of living organisms contain bolaamphi-
philic molecules in their phospholipid membranes (albeit with
different structures and without the central attachments[15,30]),
Scheme 2. Synthesis of compounds Bn–Em/n. Reagents and conditions:
i) [Pd(PPh3)4], CuI, Et3N, reflux, 7 h; ii) K2CO3, MeOH, CH2Cl2, RT, 10 h; iii) pyridi-
nium tosylate (Py·TsOH), THF, MeOH, 608C, 5 d; iv) tetra-n-butylammonium
fluoride (TBAF), THF, 258C, 1 h; v) K2CO3, DMF, 808C, 8 h (Y=Br and Y=OTs
in the syntheses of D12/3 and E12/7, respectively; R is shown in Scheme 1,
TIPS=triisopropylsilyl).
glycerol substituted phenylacetylene (2)[31] was coupled in a So-
nogashira reaction[32] with 4-bromotrimethylsilylethinylbenzene
(1a)[33] to obtain 4’-1,2-O-isopropylideneglycerol-functionalized
4-ethinyltolane 4a after basic cleavage (K2CO3/MeOH) of the C-
terminal trimethylsilyl group.[34] Sonogashira coupling of 4a
Scheme 1. Structures of the BPs B12–E12/7. In the nomenclature for Bn, Cn,
Dn/m, and En/m, n denotes the alkyl chain length and m is the number of
ethylene oxide groups.
(2 equiv) with 2,5-dibromohydroquinone ethers
5 (n=12,
18)[35] led to compounds 6a (n=12, 18), which were deprotect-
ed with Py·TsOH in methanol[36] to give the BPs B12 and B18,
respectively. This last deprotection step turned out to be the
bottleneck of this synthetic strategy, leading to low yields
(ꢀ20–50%) due to the sensitivity of the electron-rich oligo-
(phenylene ethynylene) core under acidic conditions, the long
reaction time required, and the incomplete deprotection of
both glycerol groups, giving rise to purification problems.
Therefore, an alternative strategy, which avoided this step, was
used for the synthesis of subsequent compounds.
we reasoned that X-type BPs of about 3 nm core length,[29]
roughly matching the hydrophobic DPPC bilayer core, might
integrate into phospholipid bilayers in a transmembrane orien-
tation. Additional structural diversity is conveyed by the central
side chains, leading to a combination of three philicities: the
core is lipophilic, rigid, and capable of p–p stacking interac-
tions; the side chains are lipophilic, but flexible; and the head
groups are hydrophilic and capable of hydrogen bonding.
Chem. Eur. J. 2015, 21, 8840 – 8850
www.chemeurj.org
8841 ꢀ 2015 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
For the synthesis of compounds D12/3 and E12/7 with ex-
tended polar groups the oligo(phenylene ethynylene)-based
biphenol C12 was synthesized first in two subsequent Sonoga-
shira coupling reactions, and the polar groups were attached
in the final step through etherification reactions. In the first
coupling reaction, the TIPS-protected 4-ethinylphenol 3[37] was
coupled with 4-iodophenylethinyltrimethylsilane (1b)[38] to give
the 4’-triisopropyloxy-substituted 4-ethinyltolane 4b after se-
lective base-catalyzed desilylation (K2CO3/MeOH) of the ethynyl
group.[34] Two equivalents of 4b were then coupled with 2,5-
dibromo-1,4-didodecyloxybenzene 5 (n=12) to yield the O-
TIPS-protected compound 6b. Removal of the TIPS groups
under fluoride ion catalysis[39] yielded the oligo(phenylene
ethynylene)-based biphenol C12. Etherification of C12 with 12-
bromo-3,6,9-trioxadodecane-1,2-diol
(RY,
Y=Br)
or
3,6,9,12,15,18,21-heptaoxadocosyl-p-toluenesulfonate (RY, Y=
OTs)[40] yielded compounds D12/3 and E12/7, respectively. Ana-
lytical data and details of the syntheses and purification are
described in the Supporting Information.
Figure 1. Confocal fluorescence images (shadow projections of axial image
stacks) of GUVs from DPPC (a) and of GUVs from DPPC and different BPs
(B12 (b), B18 (c), C12 (d), D12/3 (e), and E12/7 (f)) at a 10:1 molar ratio at
room temperature (228C). The BP autofluorescence is displayed in green.
The Rh-PE counterstain is shown in red. In b), separate images for the red
(Rh-PE) and green (B12 autofluorescence) channels are also shown. In d),
the BP fluorescence is not observable due to fast photobleaching of C12.
Notably, sixfold symmetric domains with a snowflake-like appearance are
observed with B12, B18, and C12. Scale bar=20 mm.
GUVs and CFM
GUVs represent free-standing bilayer model membranes,
which, by virtue of their large size, allow direct visualization of
the membrane morphology and phase separation by fluores-
cence microscopy. Domains with irregularly shaped boundaries
are observed when a solid and a fluid membrane phase coex-
ist, for instance, in GUVs from DPPC and 1,2-dilauroyl-sn-
glycero-3-phosphocholine (DLPC) at room temperature.[41] In
contrast, in fluid–fluid phase separation, the length of the in-
terfacial lines becomes minimized, leading to circular do-
mains.[42]
that GUV samples tended to be slightly heterogeneous in
quantitative compositions.[45] Most notably, the domains
formed by B12 had a striking, snowflake-like appearance,
which featured sixfold symmetry and dendritic branches (Fig-
ure 1b). The dendritic domains are probably formed through
a kinetically controlled nucleation and growth mechanism[46]
and bear a superficial resemblance to phase-separated struc-
tures observed in mixed-lipid GUVs[10] or supported bilayers.[6,7]
The clean sixfold symmetry observed for B12 is remarkable. It
suggests a regular packing structure of the BP and lipid com-
ponents, and most likely reflects its local symmetry.
Electroformation, that is, rehydration of a lipid film in the
presence of an alternating electric field, is a preferred method
for obtaining giant vesicle preparations with a large degree of
unilamellarity.[43] For the preparation of GUVs in this study, BPs
were mixed with DPPC at a 1:10 molar ratio and the solvent
was evaporated to yield a dry mixed film. GUVs containing
both BP and phospholipid were formed upon electroformation
(for details, see the Experimental Section).[44] GUVs were visual-
ized by means of CFM by exploiting the autofluorescence of
the BPs (absorption maxima at l=334 and 386 nm, emission
maximum at l=428 nm) and by counterstaining with a small
amount of a red fluorescent rhodamine-labeled lipid (1,2-dipal-
mitoyl-sn-glycero-phosphoethanolamine-N-(lissamine rhodami-
ne B sulfonyl), Rh-PE).
Chemical variations allow us to delineate necessary condi-
tions for the formation of the observed domains. A similar, six-
fold-symmetric snowflake-like appearance to that of B12 was
observed for B18 (Figure 1c), which was a polyphile with ex-
tended lateral alkyl chains, as well as with C12, which repre-
sented a polyphile with only a single hydroxyl head group at
each end (Figure 1d). A bulkier head group (D12/3), which
combines the glycerol moiety with an (EO)3 spacer unit, pre-
serves phase separation into domains, but removes the strict
symmetry, resulting in boomerang-shaped domains (Fig-
ure 1e). Only E12/7, which carries methoxy-terminated (EO)7
head groups that are incapable of acting as a hydrogen-bond
donor, does not form macroscopically segregated domains.
The nearly uniformly distributed green fluorescence indicates
that, in this case, individual BPs or small microdomains are ran-
domly distributed in the lipid bilayer. Hence, the ability of the
polyphiles to form hydrogen bonds appears to be a prerequi-
site for macroscopic segregation. Enlargement and increased
flexibility of the head groups appear to contribute to a reduced
size and regularity of the domains.
GUVs from pure DPPC can only be prepared by electrofor-
mation at a temperature above the main transition tempera-
ture. When the GUVs are cooled to room temperature for CFM,
the GUVs exhibit a faceted, corrugated shape and hole defects
due to the presence of the gel phase,[44] as shown in Figure 1a.
Next, GUVs were prepared from BPs and DPPC to test the mis-
cibility of the two components in the membrane plane (Fig-
ure 1b–f). Incorporation of any of the BPs was found to sup-
press the hole defects and reduce the corrugated appearance
of facetted GUVs. Some GUVs appeared smooth, whereas
others were still corrugated; this was accounted for by the fact
The characteristic angle dependence of the fluorescence in-
tensity along the perimeter of a macroscopically homogeneous
Chem. Eur. J. 2015, 21, 8840 – 8850
www.chemeurj.org
8842 ꢀ 2015 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 3. DSC thermograms of MLV preparations of pure DPPC and of B12
in mixtures with DPPC with different molar ratios. The indicated DH values
refer to the sum of the pre- and main transition of DPPC in units of kcal
molÀ1 lipid.
forms a gel phase at room temperature, and the DSC results
shown in Figure 3 demonstrate a reduced transition enthalpy
for the main lipid gel to fluid phase transition at 428C upon
mixing with increasing amounts of B12. Three new transitions
appear between 50 and 708C, which suggests the existence of
phase separation within the membrane. The whole set of three
new transitions shifts to higher temperature for a DPPC/B12
ratio of 4:1 and lower, which indicates a change in the compo-
sition of the domains with increasing B12 content. The re-
duced DH values for the main transition at 428C suggest that
a significant fraction of lipid molecules do not take part in the
main transition at this temperature, but become disordered
only at higher temperature. The sum of all peak areas in the
mixtures remains almost constant, which indicates that at
808C all DPPC molecules are in the disordered state. Because
the length of the domain boundaries in a phase-separated
system can also decrease the transition enthalpies, we refrain
from a quantitative interpretation of the reduced DH values,
and refer to our NMR spectroscopy based assessment de-
scribed below.
Figure 2. a) Structure of the BP core (substituents OR as in Scheme 1 and
RH =OC12H25). b) Confocal slice of GUVs prepared from a 1:10 ratio of E12/7
and DPPC, excited with polarized l=405 nm laser light. The direction of po-
larization is indicated by the blue double arrow. A multicolor palette (as de-
picted below) is used to highlight fluorescence intensity variations. Scale
bar=20 mm. Imaging was performed at room temperature (228C). c) The
fluorescence intensity along the perimeter of the largest GUV in b) as a func-
tion of angle g (black line: data; red line: cosine-squared fit function,
I=a+bcos2 (g)) shows maxima at the top and bottom of the image and
minima to the right and the left; this indicates an approximately transmem-
brane incorporation of the BP into the bilayer.
GUV, illuminated with linearly polarized light, is indicative of
the orientation of a fluorophore in a bilayer membrane.[47–49]
Therefore, BP fluorescence along the perimeter of the GUVs
obtained by confocal laser scanning microscopy with a polar-
ized laser beam was used to study how the BP was incorporat-
ed into the membrane. A single confocal slice of the GUVs
containing E12/7 from Figure 1 f is shown in Figure 2. The
transition dipole moment of BP is approximately parallel to the
long axis of the rod-like aromatic core. The systematic variation
in fluorescence intensity along the perimeter of a GUV with
fluorescence maxima appearing at the top and bottom of the
microscope image and minima appearing to the right and left
confirms that the BP is indeed integrated into the phospholip-
id bilayer in an approximately transmembrane orientation. The
pronounced noise-like fluctuations suggest that GUVs from the
1:10 mixture of E12/7 and DPPC are not perfectly homogene-
ous, but heterogeneous around the confocal resolution limit.
XRD investigation of a B12/DPPC mixture
The phase behavior and demixing phenomena found by DSC
were also investigated by using temperature-dependent
powder XRD to gain more insight into the structural organiza-
tion of the B12/DPPC system as a function of temperature. A
B12/DPPC sample with a molar mixing ratio of 1:10 was pre-
pared within a capillary and temperature-dependent powder
patterns were recorded (Figure 4). In the SAXS region, up to
four reflections with equidistant positions are visible, which in-
dicate the presence of a lamellar structure. The WAXS region
shows two strong reflections: the (020) and (110) reflections
from the chain lattice of the lipids. From these reflections, the
chain packing mode and tilt angle of the chains can be deter-
mined to be similar to pure DPPC.
The repeat distance (membrane plus interlamellar water
layer thickness) within the multilayer stacks was determined at
different temperatures. At room temperature, the lamellar
repeat distance of the mixture is slightly larger (d (1:10)=
7.25 nm) than the distance found for pure DPPC bilayers (d=
6.35 nm at 208C in the Lb’ phase and 6.70 nm at 508C in the La
phase).[50] The interlamellar water layers are thus increased in
DSC investigation of B12/DPPC mixtures
To obtain a better understanding of the origin of the unusual
mode of molecular self-assembly in the lipid layers, in-depth
investigations of B12/DPPC mixtures were performed. DPPC
Chem. Eur. J. 2015, 21, 8840 – 8850
www.chemeurj.org
8843 ꢀ 2015 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
perature. Upon cooling below room temperature, an increase
in splitting of the two WAXS reflections is observed, which in-
dicates a further change in chain packing. When the sample is
further cooled below 08C, the crystallization of excess water
leads to the appearance of sharp ice reflections. The freezing
of interlamellar water leads to reduced membrane repeat dis-
tances (d=6.06 nm). After heating the sample again to room
temperature, the initial repeat distance, and thus, the mem-
brane structure is restored, which indicates a reversible process
of excess water freezing (dehydration/hydration).
At higher temperatures, the two WAXS reflections change
into a broad scattering peak, which indicates the transition
from the gel to the fluid La phase. For the 1:10 mixture, this
occurs at 428C in coincidence with the main transition at Tm
observed by DSC. Above this transition, still intense WAXS
peaks are visible at slightly different reciprocal spacings and
additional WAXS peaks appear (sT>428C =2.25 and 2.39 nmÀ1).
These reflections indicate that the sample still contains some
ordered lipid chains, but the packing is in a different geometry
to that of the normal Lb’-phase with tilted chains.
Above 688C, fluid high-temperature phases are formed. No
sharp reflections, but only a broad halo characteristic for fluid
chains is visible in the WAXS region. The layer thickness is fur-
ther decreased, as expected for a fluid lipid bilayer (d (1:10)=
6.37 nm), and the high-temperature fluid phase appears to be
homogeneously mixed.
Figure 4. X-ray contour plot of aqueous suspensions of B12/DPPC (1:10)
with 50 wt% H2O. In the upper part, the scattering intensities are plotted in
grayscales (reciprocal spacing axis at the left-hand side) and the temperature
profile is plotted in the lower part (temperature axis at the lower right-hand
side). Arrows pointing upwards mark phase transitions, and arrows pointing
downwards show the freezing and melting of excess water. Scattering
curves at 08C are highlighted with white vertical lines. At temperatures be-
tween the freezing and melting of water, sharp ice reflections appear that
are indexed with horizontal arrows in the right-hand border. Small-angle X-
ray scattering (SAXS) reflections at temperatures I–IV are marked with white
horizontal dashes and respective reflections of the wide-angle X-ray scatter-
ing (WAXS) region with yellow or red dashes (two phases). The WAXS reflec-
tion indexing in the right-hand border (long horizontal arrows) refers to the
lipid-rich phase. The box at the bottom shows the repeat distances at the
marked temperatures (I–IV).
The XRD results support the assumptions based on the DSC
results that the mixtures seem to be phase separated. The XRD
results clearly show that a pure or very DPPC-rich lamellar
phase must exist, which transforms into a fluid La phase at
a temperature of 428C. Above this temperature, part of the
lipids are still ordered and probably form a type of complex
with B12, in which the chains are in a different chain lattice. A
homogeneous fluid phase is only formed at very high temper-
ature above 688C with a reduced lamellar repeat distance and
a typical broad halo in the WAXS region that is characteristic
for fluid chains. The X-ray results for the B12/DPPC (1:4) mix-
ture are similar (not shown), although the phase separation at
low temperature is not as pronounced. This is also evident in
the DSC curves, in which the low-temperature transition peak
characteristic for DPPC is much smaller.
thickness and/or the molecular tilt has decreased. An increase
in the water layer thickness can be caused by a change in hy-
dration of the head-group region due to the incorporation of
the B12 molecules, which, at the same time, can also induce
a reduction of the tilt angle of the lipid chains. For a phase-
separated system, we would expect two sets of lamellar repeat
distances in the SAXS region, provided that the phase-separat-
ed domains do not have identical thicknesses of lipid layers in-
cluding the water layer. Analysis of the profiles of the peaks in
the SAXS region indicate only that the peaks may have
a shoulder. Experiments that have been performed on oriented
bilayers by using X-ray reflectometry support this finding be-
cause the shoulders are more pronounced.[51] In addition, at
a higher temperature of approximately 508C, when the lipids
in the almost pure DPPC domains are fluid, the difference in
layer thickness between the two domain types becomes larger
and the two sets of SAXS reflections are more clearly separat-
ed.
NMR spectroscopy of B12/DPPC mixtures
Information on the physical properties of the membranes and
changes associated with the addition of B12 can be obtained
by static 31P NMR spectroscopy. The single 31P nucleus in the
lipid head group has a strongly orientation-dependent chemi-
cal shift (chemical-shift anisotropy (CSA)), which is sensitive to
rotational motions of the head group.[53,54] The width of the
CSA tensor, Dn, is preaveraged by fast axial rotations, leading
to a symmetric tensor with a principal axis parallel to the
membrane normal. The extent of rotational averaging depends
on the mobility of the lipid and primarily reflects the phase
state, that is, the more confined environment of the gel phase
and the more mobile environment of the fluid La phase. This is
demonstrated for experimental results on pure DPPC given in
In the WAXS region, the reciprocal spacings of 2.33 (110)
and 2.43 nmÀ1 (020) indicate the presence of an orthorhombic
(symmetry group Pbnm) herringbone lattice[52] with tilted
chains characteristic for bilayers in an Lb’ phase at room tem-
Chem. Eur. J. 2015, 21, 8840 – 8850
www.chemeurj.org
8844 ꢀ 2015 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
for pure DPPC. This Ds value is, of course, an average of two
components, but it indicates that the second fraction of immo-
bilized lipid displays larger-amplitude, fast head-group fluctua-
tions than a pure DPPC gel phase. This is in agreement with
the WAXS observations, which indicate a lattice that is different
from the normal Lb’-phase.
The intrinsic line broadening (refocused line width) is best
studied by Hahn-echo T2 relaxation time measurements, the re-
sults of which are shown in Figure 5c. The T2 decay was evalu-
ated at the isotropic-shift position of the CSA patterns (center
of gravity, close to d=0 ppm), corresponding to the magic-
angle orientation of the membrane normal in the vesicle prep-
arations with respect to the magnetic field B0, in which T2 is
sensitive to slow orientation fluctuations. The minimum in the
gel phase arises from the onset of free-lipid rotation on a time-
scale of (DsnL)À1 ꢀ0.1 ms, whereas the slow decay in the fluid
phase is associated with slower bending modes.[54] Again ex-
pectedly, the lower average T2 value in the mixture reflects in-
termediate behavior, for which the minimum of the unmodi-
fied DPPC fraction is barely visible. The most important effect
is the shifted minimum, which demonstrates that the larger
part of the lipid is still in a gel phase above 408C. Clearly, the
T2 value of the mixed system approaches the bulk value only
at temperatures above 758C, at which B12 is freely mobile in
the membrane. The latter can be demonstrated with the help
of quantitative 1H NMR spectra, which we now discuss.
Quantitative temperature-dependent magic-angle spinning
Figure 5. 31P NMR spectroscopy of a 1:4 mixture of B12 with DPPC and of
pure DPPC. a) Static spectra reflecting the CSA of the 31P of DPPC in the La
phase (47 ppm or 7.6 kHz between the dashed lines) and in the gel phase
(81 ppm or 13.1 kHz) of the pure lipid (left) and corresponding data for the
mixture (right). b) CSA values as a function of temperature. c) 13P T2 relaxa-
tion times evaluated at the magic-angle orientation (around d=0 ppm in
the spectra); the minimum arises from slow lipid rotations in the gel phase.
The slight temperature shifts of the NMR spectroscopy data relative to the
DSC traces (gray in the background) originate from weak radiofrequency-in-
duced heating due to 1H decoupling and possible errors in the temperature
calibration.
1
(MAS) H NMR spectra were taken to first quantify the amount
of lipid involved in the different phases, by taking advantage
of the fact that the aliphatic lipid resonances associated with
1
the ordered gel phase are broadened by strong H–1H dipole–
dipole couplings beyond detection at moderate MAS, namely,
5 kHz spinning frequency (Figure 6a). Integrations of the ali-
phatic and glycerol (C2 proton) resonances (Figure 6b) reveal
that, for the 4:1 and 10:1 mixtures, only around 10–15 and 50–
55%, respectively, of the total expected signal is visible above
the transition into the fluid La phase at 428C. This suggests
that significant amounts of DPPC do not participate in this
transition, but are associated with a rather immobile B12-rich
phase. Taking into account the contribution of B12 residues to
the two signal regions, we consistently estimate the mixed
phase to be composed of about 3.4 and 4.2 immobilized DPPC
molecules per B12 molecule for the 4:1 and 10:1 mixtures, re-
spectively.
Figure 5a and b, which show the well-documented spectral
changes.[54]
Qualitative changes are observed for the spectra of the 1:4
mixture (Figure 5a, right column, and b). In lieu of the steep
pretransitional drop to a lower CSA value above 408C, Ds in
the mixture exhibits a decay with indications of a two-step fea-
ture associated with the reduced main transition and the first
new transition in DSC. This is the first indication that a fraction
of DPPC is still within the gel phase above 408C, as further
substantiated below. Notably, the Ds values represent an aver-
age over the different DPPC fractions, that is, the one associat-
ed with B12 still in the gel phase and the mobilized fraction.
The intrinsic line width of the spectra is only sufficiently re-
duced at 658C, so that the two fractions can be discriminated
(see the connected points in Figure 5b). The fact that the Ds
value associated with the La phase falls below the bulk value
can be attributed to the inaccuracy related to the simple read-
ing off of the anisotropy values, which is ambiguous due to
the intrinsic line width. An important observation at low tem-
peratures is that the Ds value of the mixture is lower than that
Figure 6b also demonstrates that, within the error margin of
about 20% intrinsic to our comparison of different samples, all
resonances are visible in the MAS 1H NMR spectrum above
758C. Thus, the aliphatic chains of all membrane components,
as well as the B12 aromatic cores seen at dꢀ7 ppm, are com-
pletely mobilized and possibly dissolved in a homogeneous
membrane above the last transition. As determined from the
inflection points of the sigmoidal trends, the majority of the
still immobilized DPPC tails are consistently seen to become
mobile about 58C below the B12 cores. Thus, the two highest
DSC transitions are assigned to DPPC melting and B12 melt-
ing/dissolution, respectively, which confirms the interpretation
of the 31P NMR spectroscopy data of the head group.
Chem. Eur. J. 2015, 21, 8840 – 8850
www.chemeurj.org
8845 ꢀ 2015 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
with half-opening angles of 20 or 308, respectively. This reinfor-
ces the conclusion of a mainly transmembrane orientation of
B12 already inferred from the fluorescence measurements at
lower temperature.
For further refinement of the analysis, we use advanced 13C–
1H dipolar solid-state NMR spectroscopy techniques.[56] An ac-
count of these results is beyond the scope of this work and
will be reported separately. To summarize, similar 13C–1H order
parameter measurements unambiguously demonstrate that
B12 performs fast, well-defined 1808 flip motions in DPPC
MLVs at room temperature, as well as in its pure, crystalline
state. This is a common process for p-substituted, p–p-stacked
phenyl rings.[59] In contrast, above all thermal transitions in the
miscible phase (758C), the data confirm free rotation of B12.
The p–p interactions between aromatic rings are corroborat-
ed by our observation of a redshifted fluorescence of B12 in
the star-shaped domains of DPPC GUVs, when visualized with
a spectral detector, compared with B12 dissolved in organic
solvent (Figure 7). This redshift of the fluorescence due to p–p
interactions is also observed in fluorescence spectra of MLVs
taken at 208C, at which the emission observed for B12 in solu-
tion becomes shifted to longer wavelength when incorporated
into DPPC bilayers. Above the transition at 758C, the fluores-
cence spectrum becomes nearly identical to that recorded in
solution in chloroform; this indicates that the B12 molecules
are now molecularly dispersed in the bilayers.[51]
1
We finally note that H MAS spectra were also recorded for
an MLV mixture of DPPC and E12/7 (see Scheme 1). Here,
a very gradual mobilization of the E12/7 aromatic cores is also
only seen above the main DPPC phase-transition temperature,
which indicates that, in this case, when no macroscopic
domain formation was observed optically (see Figure 1 f), p-
stacked, rigid aggregates should also be present. Apparently,
these aggregates form microdomains that randomly pervade
the DPPC membranes, smooth out the GUV membranes, and
explain that the majority of giant vesicles lack visible corruga-
tion.
1
Figure 6. MAS H NMR data (5 kHz spinning) of hydrated pure DPPC and
1
DPPC/B12 MLV preparations. a) Exemplary quantitative H NMR spectra of
a 4:1 mixture obtained by using a recycle delay of 8 s at 20 and 708C. b) In-
tegrals of different spectral regions, normalized to 100% as the total expect-
ed signal, as a function of temperature overlaid with scaled DSC traces (right
vertical scale) and c) double-quantum (DQ) spinning sideband pattern
(10 kHz MAS) for the aromatic B12 resonance of a 4:1 mixture at 758C by
using the BaBa-xy16 pulse sequence[55] with a recoupling time of 4 rotor pe-
riods.
Details on the dynamic state (amplitude of motion) and ori-
entation of B12 can be inferred from the dynamic order pa-
rameter, S, associated with different internuclear vectors. It can
be probed by suitable NMR spectroscopy techniques that mea-
sure the respective residual (motion-averaged) dipole–dipole
coupling constants, Dres,[56] by using S=Dres/Dstat, in which Dstat
is the effective static-limit coupling constant in radsÀ1. Specifi-
cally, fast orientation fluctuations (which are much faster than
1/Dstat) of dissolved B12 can be characterized by the 1H–1H
dipole–dipole coupling of pairs of neighboring aromatic pro-
tons to exploit the fact that the internuclear vector points
along the long axis and is thus not affected by fast uniaxial ro-
tation. The value of Dres can be obtained from DQ[57] spinning
sideband analysis.[58] From the result at 758C (Figure 6c), in
combination with a reference coupling, Dstat/2p, of 8.2 kHz,
which corresponds to an H–H distance of 2.45 Š, we obtain
S=0.78Æ0.1, which characterizes orientation fluctuations of
the B12 long axis. This can be interpreted by models that
assume diffusive motion of the long axis on or within a cone
Discussion
Based on these insights, we discuss tentative models of self-as-
sembly of the X-shaped BPs in the lipid bilayers. XRD investiga-
tions and DSC results indicate that the B12/DPPC mixtures are
phase separated. The major phase is a pure or very DPPC-rich
lamellar Lß’ phase, which transforms into a fluid La phase at
a temperature of 428C. The B12 molecules self-assemble into
B12-rich domains that contain highly ordered, tightly p–p-
stacked B12 molecules oriented, on average, along the mem-
brane normal and are only able to perform p flips. In the B12-
rich domains, about 3–4 DPPC molecules per B12 molecule are
involved, which display enhanced head-group mobility, al-
though the chain order (probably on a different lattice) is re-
tained even above 428C, that is, the chains appear to be im-
mobilized.
At first sight, a simple explanation for the dendritic appear-
ance of the B12-enriched domains in the confocal images of
GUV membranes (Figure 1b) might seem to be defect lines
Chem. Eur. J. 2015, 21, 8840 – 8850
www.chemeurj.org
8846 ꢀ 2015 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
phase of DPPCs associated with the BPs. The simultaneously
enhanced head-group mobility of these immobilized lipids
suggests that the orthorhombic chain packing might be modi-
fied. The observed dendritic growth of the domains with hex-
agonal symmetry could result if the symmetry of chain packing
were changed to hexagonal in the BP-rich domains, at least in
time and space average.
Our observation of well-packed BPs with motions restricted
to p flips suggests that they form extended filaments with par-
allel p faces. An analogous filament formation has recently
been demonstrated for bolaamphiphiles without lateral chains
in the absence of lipids.[60] With the X-shaped bolaamphiles
used herein, the additional lateral alkyl chains interact and
preferentially mix with the alkyl chains of the DPPC molecules;
thus separating adjacent BP filaments. Moreover, the BP head
groups capable of hydrogen bonding are expected to addi-
tionally stabilize the string-like filaments. The strings may
become interconnected, resulting in a random network. The
lack of visible corrugation in the DPPC-rich parts of the GUVs
could be explained by a low density of randomly distributed
nanofilaments that smooth out and stabilize the membranes
by gluing together larger patches of DPPC gel phase. The six-
fold symmetric shape of the B12-rich domains could, in this
case, result from a fusion of these filaments into a regular hex-
agonal net. A tentative model that takes into consideration all
these restrictions, in which the walls are built of transmem-
brane-oriented, aligned, and p–p-packed B12 cores, is shown
in Figure 8 and in the Supporting Information. The resulting
hexagonal cells are filled by the lateral chains of the B12 and
DPPC molecules that become immobilized by this confine-
ment. As the overall effective head group to chain cross-sec-
tional area ratio of the superstructure is reduced (by the lateral
chains of BP and the smaller head groups of most BPs), lipid-
chain immobilization could indeed be associated with the ob-
served lipid head-group mobilization. Based on the maximum
observed ratio for B12/DPPC of around 1:3 (B12-rich domains)
and considering the dimensions of the involved molecules, it is
estimated that at least 7 molecules of B12 are arranged along
Figure 7. Spectrally resolved CFM of GUVs from B12/DPPC (1:10), counter-
stained with Rh-PE. a) Confocal slice image color-coded by wavelength, as
indicated by the color palette below (32 channels, l=414 to 690 nm,
8.9 nm resolution). Scale bar=5 mm. Imaging was performed at room tem-
perature. b) Corresponding spectral plots for the four regions indicated by
circles in the confocal image. Regions 1 and 2 are located in the B12-rich
star-shaped domains; regions 3 and 4 are in the surrounding phase. The
peak at lꢀ470 nm (which is higher within the domains) is due to B12 emis-
sion; the peak at lꢀ590 nm (which is higher outside the domains) is due to
Rh-PE emission. c) As a reference, the spectrally resolved fluorescence of
B12 in chloroform/methanol was recorded on the same setup, and showed
a peak at lꢀ430 nm.
around the gel-phase facets formed upon cooling and gel-
phase formation in DPPC GUVs. One might envision that these
may be occupied by the B12-enriched domains. However, if
this were the case, an irregular network of B12-filled cracks
would be expected to cover the surface of the GUVs, instead
of the impressively regular, hexagonal, star-shaped domains
observed in the B12, B18, and C12 samples.
It could also be envisaged that, according to the solid docks
hypothesis,[9] the BP microdomains favor chain crystallization
of the lipid molecules around them, as indicated by the ob-
served increase in the transition temperature to the fluid La
Figure 8. Model showing the organization of BPs (B12) in a hexagonal hon-
eycomb (cut perpendicular to the membrane normal).
Chem. Eur. J. 2015, 21, 8840 – 8850
www.chemeurj.org
8847 ꢀ 2015 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
each of the honeycomb walls; thus accommodating about 50
DPPC molecules in the hexagonal cells, as estimated by calcu-
lations given in the Supporting Information. This dense pack-
ing in hexagonal honeycombs could be favored by hydrogen
bonding between the relatively small head groups of the BPs
B12, B18, and C12, whereas larger head groups with reduced
capability for hydrogen bonding cannot fuse to form a dense
honeycomb, and thus, the BP filaments remain randomly dis-
tributed in the lipid matrix, as observed for E12/7.
The formation of polygonal honeycombs composed of paral-
lel aligned aromatic cores is a common feature of BPs, previ-
ously found as a dominating mode of self-assembly of such
molecules in liquid-crystalline (LC) bulk phases.[27–29,61] However,
in contrast to most reported LC honeycombs with aromatic
cores oriented tangentially around the lipophilic domains, in
the membranes, as a result of the confinement provided by
the membranes and the anchoring of polar groups at the sur-
faces, the aromatic cores align parallel to the cylinder long
axes. This indeed resembles the organization of some T-
shaped polyphiles in the hexagonal channeled layer LC
phases.[62] We therefore consider the honeycomb model to be
the most probable model, although further in-depth character-
ization of the mesoscale structure will be required for a final
conclusion to be drawn.
Experimental Section
Synthesis
Details of the syntheses and analytical data are described in the
Supporting Information.
GUV preparation
The lipid DPPC and red fluorescent lipid marker Rh-PE were pur-
chased from Avanti Polar Lipids (Alabaster, Alabama, USA) and
used without further purification. GUVs were prepared by a modi-
fied electroformation method.[64,44] With chloroform as a solvent,
the DPPC was mixed with one of the BPs at 10:1 molar ratio by
adding 0.5 mol% of Rh-PE as a red counterstain. The total concen-
tration was 15 mgmLÀ1. The mixture (15 mL) was spread on optical-
ly transparent indium tin oxide (ITO) coverslips (Gesim GmbH,
Grosserkmannsdorf, Germany) preheated to 608C. The coverslips
were assembled in
a capacitor-type configuration by using
a home-built perfusion chamber with 2 mm spacers. The chamber
was filled with deionized water and an alternating sinusoidal volt-
age (1.3 V effective voltage, 10 Hz) was applied for 4 h at 608C.
Prior to confocal microscopy, the perfusion chamber was allowed
to cool to room temperature (228C).
CFM measurements
CFM was carried out on an LSM 710 setup (Carl Zeiss Microimag-
ing, Jena, Germany) by using a C-Apochromat 40”/1.2 N.A. water
immersion objective at room temperature. The BPs were excited
with a diode laser at l=405 nm; the fluorescence emission was
collected from l=412 to 500 nm by using a photomultiplier detec-
tor. A l=561 nm DPSS laser was used to excite the lipid marker
Rh-PE; the fluorescence emission of the marker was collected from
l=566 to 681 nm. Spectrally resolved confocal imaging was per-
formed on an LSM 780 setup (Carl Zeiss Microimaging, Jena, Ger-
many) by using a C-Apochromat 63”/1.2 N.A. water immersion ob-
jective. Fluorescence was collected between l=414 and 690 nm at
8.9 nm resolution by using a 32-channel GaAsP detector. Spectral
plots were generated for selected regions of interest (ROIs). For
the reference measurement, an image of a solution of B12 in
CHCl3/MeOH (2:1) was acquired and a spectral plot for an arbitrary
ROI within the uniform image was generated.
Conclusion
We demonstrated that a new class of bolapolyphilic molecules
exhibited a unique propensity to form supramolecular struc-
tures in free-standing lipid bilayer membranes. We found that,
as a function of the head-group structure, BPs and lipids repro-
ducibly self-organized in either highly regular, sixfold symmet-
ric structures or less ordered 2D networks. The molecular
origin of the three first-order thermal transitions associated
with these structures, as seen in DSC, were elucidated by XRD
and NMR spectroscopy. By using distinct head groups, it was
demonstrated how supramolecular organization could be
tuned by molecular design. The mesoscale structure and mo-
lecular orientation of B12 in DPPC and in DMPC have been fur-
ther investigated by transmission electron microscopy, differ-
ent X-ray diffraction techniques, fluorescence and polarized in-
frared spectroscopy, as reported in a separate publication.[51]
Because the physical properties of self-assembled materials
depend crucially on their supramolecular organization, we en-
vision that a targeted design will enable novel applications in
nanotechnology, encapsulation, and drug delivery or template-
assisted engineering of scaffolds of p-conjugated rods, which
are of interest, for example, for molecular electronics and
sensor applications. The investigation of the self-assembly of
properly designed synthetic molecules in lipid membranes
could also contribute to the better understanding of microdo-
main formation in biomembranes[4–10] and to the design, stabi-
lization, and investigation of functional membranes and proto-
cells.[63]
DSC measurements
DSC thermograms were acquired by using a VP-DSC[65] instrument
(MicroCal/GE Inc. Northampton, USA) with heating/cooling rates of
60 KhÀ1 in a temperature range of 2–958C. The traces shown in
Figure 5 represent the seventh heating scan of mixed vesicles pre-
pared as follows: All components were mixed in CHCl3/MeOH (2:1)
at the required molar ratio; the organic solvent was then evaporat-
ed in a stream of N2 to obtain a lipid film. The film was further
dried under reduced pressure at 708C for about 3 h. After hydra-
tion with H2O, the sample had a total concentration of 2.0–2.5 mm.
The aqueous suspension was then vortexed and sonicated at
about 608C for 30 min to obtain vesicles. All DSC results were re-
producible for repeated sample preparations, and showed no sig-
nificant hystereses, except for the first few heating–cooling cycles,
after which the samples were in equilibrium.
XRD measurements
The XRD measurements of lipid mixtures were performed by using
monochromic CuKa1 radiation (l=0.154051 nm from a Ge(111)
Chem. Eur. J. 2015, 21, 8840 – 8850
www.chemeurj.org
8848 ꢀ 2015 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
monochromator (Seifert X-ray/GE Inc. Freiberg) and a curved linear
position-sensitive detector (range: 2q=0–408). Samples were pre-
pared according to the standard procedure followed by lyophiliza-
tion and rehydration to give a water content of 50 wt% for a good
signal to noise ratio of the scattering patterns. The samples were
transferred into glass capillaries, which were then flame-sealed.
The temperature was controlled by using a high-temperature
sample holder (STOE & CIE GmbH, Darmstadt). SAXS and WAXS
(s=1–4.7 nmÀ1) data were collected in the temperature range from
À35 to 808C. The temperature was varied stepwise (heating rates
of 1 KminÀ1) and the sample was equilibrated for 5 min at each
temperature before data acquisition with 10 min of exposure time
per diffractogram. Each scattering pattern was corrected by sub-
tracting the scattering of an empty capillary. All diffraction patterns
of one temperature series were then converted into a contour dia-
gram (intensities in grayscales).
sidebands, which were also integrated. At 758C and above, the
Curie-corrected integrals did not change further, which indicated
that all species were sufficiently mobile. The deviations of this
high-temperature data from 100% also demonstrated that the
measurements were subject to systematic errors on the order of
20%, which could be explained by baseline problems upon inte-
gration, and external calibration, for which a change of the sample
in a MAS rotor could lead to setup variations, and thus, deviations
in the absolute spectral intensities.
1H–1H dipole–dipole coupling constants for the adjacent aromatic
CH protons were determined by DQ[57] spinning sideband analy-
sis[58] by using the BaBa-xy16 pulse sequence[55] at 10 kHz MAS and
four rotor periods recoupling. The pattern was fitted to a numeri-
cally calculated and Fourier-transformed powder average of the
theoretical time domain signal by assuming a Gaussian distribution
of residual dipole–dipole couplings, Dres. The standard deviation
was always around 10% of the average, which indicated weak ap-
parent distribution effects.[67] Such effects did not necessarily arise
from an actual coupling distribution, but could also be assigned to
changes in the pattern shape that arose from couplings to remote
protons[58c] or from a bias in the assumed isotropic powder aver-
age[57] that arose from anisotropic T2.
NMR spectroscopy
1
Static 31P and MAS H NMR spectroscopy investigations were car-
ried out on MLV preparations made by co-dissolving lipid and B12
in methanol/chloroform, drying, and subsequent rehydration by
adding 50 wt% of H2O/D2O to the powder. All data were acquired
on a Bruker Avance III instrument with 400 MHz 1H Larmor frequen-
cy by using a Bruker static 5 mm double-resonance probe for 31P
spectra and a Bruker 4 mm MAS WVT double-resonance probe
head at 5 kHz spinning frequency (if not indicated otherwise) for
1H spectra, relying in both cases on a flow of heated air for tem-
perature regulation. Typical 908 pulse lengths were 3 ms for both
31P and 1H NMR spectra.
Acknowledgements
S.W. and B.D.L. performed and analyzed the confocal microsco-
py experiments; H.E. and S.P. designed and synthesized the
new compounds; B.D.L. performed and analyzed the DSC and
X-ray experiments; F.L., A.A. and R.B. performed and analyzed
the NMR experiments; A.B., K.S., C.T. and K.B. designed the re-
search, analyzed and interpreted data, and wrote the manu-
script. We thank Gunter Reuter for use of the LSM 780. We are
grateful to the Deutsche Forschungsgemeinschaft (DFG) for fi-
nancial support in the framework of the Forschergruppe 1145.
K.B. also acknowledges funding from ERDF (grant 1241090001)
and BMBF (FKZ 03Z2HN22).
31P direct-polarization spectra to characterize the motion of the
1
lipid head groups were taken with a 3 s repetition time and H de-
coupling achieved by using SPINAL64.[66] The 31P NMR chemical
shift was calibrated with 85% H3PO4 (d=0 ppm) as an external ref-
erence. In addition, T2 relaxation time measurements were per-
formed with a standard Hahn-echo sequence (908xÀt/2À1808yÀt/
1
2Àacq) with an eight-step phase cycle and H decoupling over the
whole sequence. The Fourier-transformed echo signals were inte-
grated within a narrow range around the magic-angle orientation
and fitted exponentially. The decoupling strength was adjusted to
around 10–30 kHz nutation frequency, which represented a good
compromise between sample heating at high power and addition-
al line broadening (T2 reduction) at low power.
Keywords: bolapolyphiles
interactions · self-assembly
·
lipids
·
membranes
·
pi
[1] D. A. Doyle, Science 1998, 280, 69–77.
[2] Y. Shai, Biochim. Biophys. Acta Biomembr. 1999, 1462, 55–70.
[3] N. Sakai, S. Matile, Langmuir 2013, 29, 9031–9040.
[4] D. A. Brown, E. London, J. Biol. Chem. 2000, 275, 17221–17224.
[5] K. Simons, W. L. C. Vaz, Annu. Rev. Biophys. Biomol. Struct. 2004, 33, 269–
295.
[6] J. M. Moran-Mirabal, D. M. Auberecht, H. G. Craighead, Langmuir 2007,
23, 10661–10671.
[7] U. Bernchou, J. H. Ipsen, A. C. Simonsen, J. Phys. Chem. B 2009, 113,
7170–7177.
[8] K. Simons, M. J. Gerl, Nat. Rev. Mol. Cell Biol. 2010, 11, 688–699.
[9] R. F. de Almeida, E. Joly, Front. Plant Sci. 2014, 5, 72
[10] L. A. Bagatolli, J. H. Ipsen, A. C. Simonsen, O. G. Mouritsen, Prog. Lipid
Res. 2010, 49, 378–389.
[11] W. H. Binder, V. Barragan, F. M. Menger, Angew. Chem. Int. Ed. 2003, 42,
5802–5827; Angew. Chem. 2003, 115, 5980–6007.
[12] A. Hädicke, A. Blume, J. Colloid Interface Sci. 2013, 407, 327–338.
[13] M. Schulz, Angew. Chem. Int. Ed. 2013, 52, 1829–1833; Angew. Chem.
2013, 125, 1877–1882.
[14] P. Scholtysek, A. Achilles, C.-V. Hoffmann, B.-D. Lechner, A. Meister, C.
Tschierske, K. Saalwächter, K. Edwards, A. Blume, J. Phys. Chem. B 2012,
116, 4871–4878.
Quantitative 1H spectra were taken at 5 kHz MAS. Weight-con-
trolled samples were prepared in 4 mm MAS rotors with Teflon
spacers for reproducible sample positioning. The recycle delay was
adjusted to a value much larger than T1 at each temperature, and
the spectral intensities were corrected for the Curie temperature
dependence, by multiplying them by T/Tref, and allowing for
a quantitative evaluation of the integrals. The aliphatic, aromatic,
and glycerol resonances of DPPC and B12 in the gel phase (208C)
at the given spinning frequency were subject to dipolar line broad-
ening down to the baseline level due to the well-defined packing
and low mobility; this precluded their observation in sufficiently
narrow integration ranges. Just above the main lipid phase transi-
tion, all resonances of pure hydrated DPPC became well resolved
and quantitatively detectable, although their intensity remained
lower for the mixtures. The resolved glycerol resonance of the
pure DPPC preparation at Tref =508C served as an external intensity
standard to define 100% of the signal for each group of resonan-
ces. Notably, a significant intensity was found in sharp spinning
Chem. Eur. J. 2015, 21, 8840 – 8850
www.chemeurj.org
8849 ꢀ 2015 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[15] J.-H. Fuhrhop, T. Wang, Chem. Rev. 2004, 104, 2901–2937.
[16] C. C. Forbes, K. M. DiVittorio, B. D. Smith, J. Am. Chem. Soc. 2006, 128,
9211–9218.
[17] S. Aleandri, A. Casnati, L. Fantuzzi, G. Mancini, G. Rispoli, F. Sansone,
Org. Biomol. Chem. 2013, 11, 4811.
[18] D. P. Brownholland, G. S. Longo, A. V. Struts, M. J. Justice, I. Szleifer, H. I.
Petrache, M. F. Brown, D. H. Thompson, Biophys. J. 2009, 97, 2700–
2709.
[19] N. Sakai, J. Mareda, S. Matile, Acc. Chem. Res. 2005, 38, 79–87.
[20] A. Vargas Jentzsch, A. Hennig, J. Mareda, S. Matile, Acc. Chem. Res. 2013,
46, 2791–2800.
[41] J. Korlach, P. Schwille, W. W. Webb, G. W. Feigenson, Proc. Natl. Acad.
Sci. USA 1999, 96, 8461–8466.
[42] K. Bacia, P. Schwille, T. Kurzchalia, Proc. Natl. Acad. Sci. USA 2005, 102,
3272–3277.
[43] N. Rodriguez, F. Pincet, S. Cribier, Colloids Surf. B 2005, 42, 125–130.
[44] M. Schulz, D. Glatte, A. Meister, P. Scholtysek, A. Kerth, A. Blume, K.
Bacia, W. H. Binder, Soft Matter 2011, 7, 8100–8110.
[45] S. L. Veatch, S. L. Keller, Biophys. J. 2003, 84, 725–726.
[46] X. Y. Liu, From solid-fluid interfacial structure to nucleation kinetics:
principles and strategies for micro/nanostructure engineering; in: Xiang
Yang Liu, James J. De Yoreo, Nanoscale structure and assembly at
solid-fluid interfaces, Vol. 1, Kluwer Academics Publishers 2004, ISBN
1 4020 7792 0.
[21] W. Wang, R. Li, G. W. Gokel, Chem. Eur. J. 2009, 15, 10543–10553.
[22] J. M. Moszynski, T. M. Fyles, Org. Biomol. Chem. 2010, 8, 5139–5149.
[23] T. Muraoka, T. Shima, T. Hamada, M. Morita, M. Takagi, K. Kinbara, Chem.
Commun. 2011, 47, 194–196.
[47] J. Yguerabide, L. Stryer, Proc. Natl. Acad. Sci. USA 1971, 68, 1217–1221.
[48] D. Axelrod, Biophys. J. 1979, 26, 557–574.
[24] E. Quesada, A. U. Acuna, F. Amat-Guerri, Angew. Chem. Int. Ed. 2001, 40,
2095–2097; Angew. Chem. 2001, 113, 2153–2155.
[49] R. R. Gullapalli, M. C. Demirel, P. J. Butler, Phys. Chem. Chem. Phys. 2008,
10, 3548–3560.
[25] J. Nikolaus, S. Czapla, K. Mçllnitz, C. T. Hçfer, A. Herrmann, P. Wessig, P.
Müller, Biochim. Biophys. Acta Biomembr. 2011, 1808, 2781–2788W.
[26] For previously reported oligo(phenylene ethynylenes) involving hydro-
philic groups, see: E. H. Hill, D. Sanchez, D. G. Evans, D. G. Whitten,
Langmuir 2013, 29, 15732–15737; S. R. Bull, L. C. Palmer, N. J. Fry, M. A.
Greenfield, B. W. Messmore, T. J. Meade, S. I. Stupp, J. Am. Chem. Soc.
2008, 130, 2742–2743; B. Erdogan, J. N. Wilson, U. H. F. Bunz, Macromo-
lecules 2002, 35, 7863–7864; R. L. Phillips, I.-B. Kim, B. E. Carson, B. Tid-
beck, Y. Bai, T. L. Lowary, L. M. Tolbert, U. H. F. Bunz, Macromolecules
2008, 41, 7316–7320; I.-B. Kim, R. Phillips, U. H. F. Bunz, Macromolecules
2007, 40, 5290–5293.
[27] a) P. Kieffer, M. Prehm, B. Glettner, K. Pelz, U. Baumeister, F. Liu, X. Zeng,
G. Ungar, C. Tschierske, Chem. Commun. 2008, 3861–3863; b) F. Liu, R.
Kieffer, X. Zeng, K. Pelz, M. Prehm, G. Ungar, C. Tschierske, Nat.
Commun. 2012, 3, 1104.
[28] a) C. Tschierske, Angew. Chem. Int. Ed. 2013, 52, 8828–8878; Angew.
Chem. 2013, 125, 8992–9047; b) C. Tschierske, C. Nürnberger, H. Ebert,
B. Glettner, M. Prehm, F. Liu, X. Zeng, G. Ungar, Interface Focus 2012, 2,
669–680.
[50] J. F. Nagle, S. Tristram-Nagle, Biochim. Biophys. Acta Rev. Biomembr.
2000, 1469, 159–195.
[51] B.-D. Lechner, H. Ebert, M. Prehm, S. Werner, A. Meister, G. Hause, A.
Beerlink, K. Saalwächter, K. Bacia, C. Tschierske, A. Blume, Langmuir
2015, 31, 2839–2850.
[52] G. Fçrster, A. Meister, A. Blume, Current Opinion in Colloid & Interface Sci-
ence 2001, 6, 294–302; Interface Science 2001, 6, 294–302.
[53] J. Seelig, Biochim. Biophys. Acta Rev. Biomembr. 1978, 515, 105–140.
[54] E. J. Dufourc, C. Mayer, J. Stohrer, G. Althoff, G. Kothe, Biophys. J. 1992,
61, 42–57.
[55] a) K. Saalwächter, F. Lange, K. Matyjaszewski, C.-F. Huang, R. Graf, J.
Magn. Reson. 2011, 212, 204–215; b) M. Feike, D. E. Demco, R. Graf, J.
Gottwald, S. Hafner, H. W. Spiess, J. Magn. Reson. Ser. A 1996, 122, 214–
221.
[56] D. Reichert, K. Saalwächter in NMR Crystallography (Encyc. Magn.
Reson.), (Eds.: R. K. Harris, R. E. Wasylichen, M. J. Duer), Wiley, Chichester
2009, pp. 177–193.
[57] K. Saalwächter, ChemPhysChem 2013, 14, 3000–3014.
[58] a) R. Graf, D. E. Demco, J. Gottwald, S. Hafner, H. W. Spiess, J. Chem.
Phys. 1997, 106, 885–895; b) I. Schnell, S. P. Brown, H. Y. Low, H. Ishida,
H. W. Spiess, J. Am. Chem. Soc. 1998, 120, 11784–11795; c) I. Schnell,
H. W. Spiess, J. Magn. Reson. 2001, 151, 153–227.
[59] I. Fischbach, T. Pakula, P. Minkin, A. Fechtenkçtter, K. Müllen, H. W.
Spiess, K. Saalwächter, J. Phys. Chem. B 2002, 106, 6408–6418.
[60] T. Rudolph, N. K. Allampally, G. Fernandez, F. H. Schacher, Chem. Eur. J.
2014, 20, 13871–13875.
[29] X. B. Zeng, R. Kieffer, B. Glettner, C. Nürnberger, F. Liu, K. Pelz, M. Prehm,
U. Baumeister, H. Hahn, H. Lang, G. A. Gehring, C. H. M. Weber, J. K.
Hobbs, C. Tschierske, G. Ungar, Science 2011, 331, 1302–1306.
[30] a) T. Benvegnu, M. Brard, D. Plusquellec, Current Opinion in Colloid and
Interface Science 2004, 8, 469–479; b) A. Meister, A. Blume, Current
Opinion in Colloid & Interface Science 2007, 12, 138–147; Interface Sci-
ence 2007, 12, 138–147.
[31] a) M. Kçlbel, T. Beyersdorff, C. Tschierske, S. Diele, J. Kain, Chem. Eur. J.
2000, 6, 3821–3837; b) B. Glettner, F. Liu, X. Zeng, M. Prehm, U. Bau-
meister, M. Walker, M. A. Bates, P. Boesecke, G. Ungar, C. Tschierske,
Angew. Chem. Int. Ed. 2008, 47, 9063–9066; Angew. Chem. Int. Ed. 2008,
120, 9203–9206.
[32] K. Sonogashira, Y. Tohda, N. Nagihara, Tetrahedron Lett. 1975, 16, 4467–
4470.
[33] C. Eaborn, A. R. Thompson, D. R. M. Walton, J. Chem. Soc. C 1967, 1364–
1366.
[61] a) B. Chen, X. Zeng, U. Baumeister, G. Ungar, C. Tschierske, Science 2005,
307, 96–99; b) C. Tschierske, Chem. Soc. Rev. 2007, 36, 1930–1970.
[62] a) B. Chen, X. Zeng, U. Baumeister, S. Diele, G. Ungar, C. Tschierske,
Angew. Chem. Int. Ed. 2004, 43, 4621–4625; Angew. Chem. 2004, 116,
4721–4725; b) B. Chen, U. Baumeister, G. Pelzl, M. K. Das, X, Zeng, G.
Ungar, C. Tschierske, J. Am. Chem. Soc. 2005, 127, 16578–16591 .
[63] S. Rasmussen, M. A. Bedau, L. Chen, D. Deamer, D. C. Krakauer, N. H.
Packard, P. F. Stadler, Protocells: Bridging Nonliving and Living Matter,
MIT Press, Cambridge, MA, 2009.
[34] a) W. B. Austin, N. Bilow, W. J. Kelleghan, K. S. Y. Lau, J. Org. Chem. 1981,
46, 2280–2286; b) D. L. Musso, M. J. Clarke, J. L. Kelley, G. E. Boswell, G.
Chen, Org. Biomol. Chem. 2003, 1, 498–506.
[35] a) T. Vahlenkamp, G. Wegner, Makromol. Chem. Phys. 1994, 195, 1933–
1952; b) A. Fkyerat, G.-M. Dibin, R. Tabacchi, Helv. Chim. Acta 1999, 82,
1418–1422.
[64] M. I. Angelova, D. S. Dimitrov, Faraday Discuss. Chem. Soc. 1986, 81,
303–311.
[65] V. V. Plotnikov, J. M. Brandts, L. N. Lin, J. F. Brandts, Anal. Biochem. 1997,
250, 237–244.
[66] B. M. Fung, A. K. Khitrin, K. Ermolaev, J. Magn. Reson. 2000, 142, 97–
101.
[36] R. Sterzycki, Synthesis 1979, 724–725.
[67] a) S. P. Brown, I. Schnell, J. D. Brand, K. Müllen, H. W. Spiess, J. Mol.
Struct. 2000, 521, 179–195; b) G. P. Holland, B. R. Cherry, T. M. Alam, J.
Magn. Reson. 2004, 167, 161–167.
[37] F. Mitzel, S. Fitzgerald, Chem. Eur. J. 2003, 9, 1233–1241.
[38] a) J. S. Moore, E. J. Weinstein, Z. Wu, Tetrahedron Lett. 1991, 32, 2465–
2466; b) O. Lavastre, O. Laurence, P. H. Dixneuf, Tetrahedron 1996, 52,
5495–5504.
[39] a) T. D. Nelson, R. D. Crouch, Synthesis 1996, 1031–1069; b) J. Cheng, U.
Hacksell, G. D. Daves, J. Org. Chem. 1986, 51, 3093–3098.
[40] A. Saha, S. Ramakrishnan, Macromolecules 2009, 42, 4956–4959.
Received: November 6, 2014
Published online on May 4, 2015
Chem. Eur. J. 2015, 21, 8840 – 8850
www.chemeurj.org
8850 ꢀ 2015 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim