Responsive vesicles from dynamic covalent surfactants

Article abstract of DOI:10.1002/anie.201007401

Breaking bilayers: Incorporation of dynamic covalent bonds in vesicle-forming surfactants leads to the formation of responsive vesicles, which can be switched back and forth between the bilayer state and the isotropic solution using either dilution or a change in the pH value as external stimuli. Copyright

Full text of DOI:10.1002/anie.201007401

DOI: 10.1002/anie.201007401
Self-Assembly
Responsive Vesicles from Dynamic Covalent Surfactants**
Christophe B. Minkenberg, Feng Li, Patrick van Rijn, Louw Florusse, Job Boekhoven,
Marc C. A. Stuart, Ger J. M. Koper, Rienk Eelkema, and Jan H. van Esch*
Herein we describe the use of dynamic covalent surfactants to
create dynamic vesicles that are highly responsive to changes
in their environment. Synthetic vesicles have drawn much
attention because of their close similarity to biological cells, of
which the membranes are generally composed of double-
tailed phospholipid surfactants. In living systems, bilayer
membranes are highly dynamic,[1] amongst others as a result
of embedded proteins and carotenoids, which is essential for
endo- and exocytosis, cell signaling, and cell division. In
contrast, synthetic vesicles tend to be very static, especially
when composed of double-chain surfactants.[2,3] A few specific
exceptions exist, including dynamic systems comprised single-
tailed surfactants, thus forming vesicles with critical aggrega-
tion concentrations (CAC) several orders of magnitude
higher than those of their double-chain surfactant analogues.
The building blocks for these systems are limited to surfac-
tants with relatively short tails,[4,5] and vesicles are only
formed within narrow pH windows.[6] Morphological transi-
tions have been effected using surfactants that are capable of
covalent structural modification in an aggregate environment.
However, these transitions tend to be unidirectional, and do
not allow formation from and reversal to isotropic solution.[7]
The low solubility of double-chain surfactants limits sponta-
neous vesicle formation in water, making sonication, film
hydration, and solvent injection techniques necessary to
induce vesicle formation. Double-chain surfactant systems
that are capable of switching reversibly between a non-
aggregated state and an aggregated, vesicular state remain
largely unexplored.
To change the dynamics of bilayers and vesicular assem-
blies, we opted to enhance the dynamics of formation and
dissociation of their surfactants through the introduction of a
reversible covalent bond. This should render these surfac-
tants, and thus their assemblies, responsive to changes in their
environment. The reversible and dynamic nature of the
dynamic covalent bond has proven to be a powerful tool for
the construction of adaptive and reversible systems[8] and
already led to a variety of supramolecular aggregates like
gels,[9] nanorods,[10] and micelles.[11,12]
Herein we report on the formation of dynamic vesicles, in
which the responsiveness of the bilayer membrane is regu-
lated by controlling the reversible formation of its amphi-
philic surfactant constituents. For this approach, we designed
a double-tailed surfactant that can form from two water-
soluble, not surface-active precursors by the formation of a
reversible covalent bond. As surfactant headgroup, we
synthesized cationic bisaldehyde A, which we mixed with an
apolar amine-functionalized tail B in water, creating cationic
surfactants in situ from water-soluble precursors through the
formation of reversible imine bonds. It was expected that both
single-tailed AB1 and double-tailed AB2 surfactants are
formed, which are in equilibrium with each other and with
both precursors (Scheme 1).
[*] C. B. Minkenberg, Dr. F. Li, Dr. P. van Rijn, L. Florusse, J. Boekhoven,
Dr. G. J. M. Koper, Dr. R. Eelkema, Prof. Dr. J. H. van Esch
Department of Chemical Engineering
Delft University of Technology
Julianalaan 136, 2628 BL, Delft (The Netherlands)
Fax: (+31)15-278-4289
E-mail: j.h.vanesch@tudelft.nl
Homepage: http://www.cheme.tudelft.nl/sas
Dr. M. C. A. Stuart
Biomolecular Sciences and Biotechnology Institute, University of
Groningen (The Netherlands)
[**] We are grateful to Lars van der Mee, Pieter Braams, Hao Dinh, and
Bart Homan for their help with the synthesis of the compounds, and
to Wim Kruizinga and Theodora Tiemersma-Wegman for the HRMS
analysis. This work was supported by the Netherlands Organization
for Scientific Research (NWO), STW/Nanoned (C.B.M.), the Euro-
pean Commission (Marie Curie European Reintegration Grant,
(R.E.), and COST Action D43 (G.J.M.K.).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201007401.
Scheme 1. Dynamic formation of imine vesicles. Imine surfactants AB1
and AB2 were prepared in situ by mixing A and B in water.
Angew. Chem. Int. Ed. 2011, 50, 3421 –3424
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3421

Communications
After mixing clear solutions of the precursors in water
([A]0 = 10 mm, [B]0 = 20 mm, solution pH = 9.5 Æ 0.2 because
of the added amine; [A]0 and [B]0 denote initial concentra-
tions), the solution turned opalescent within minutes, indicat-
ing aggregate formation. Dynamic light scattering (DLS)
experiments showed a bimodal distribution with hydrody-
namic radii of (95 Æ 11) nm and (5.2 Æ 1.6) mm in water
(Figure 2 in the Supporting Information). The aggregate
morphology was investigated by confocal laser scanning
microscopy (CLSM) and cryo-transmission electron micros-
copy (Cryo-TEM). CLSM studies showed the presence of 2–
5 mm vesicles (Figure 1a), and 50–100 nm unilamellar vesicles
were found by Cryo-TEM (Figure 1b), which is in good
agreement with the results of the DLS measurements.
Figure 1. a) CLSM image of A·B in water, showing 2–5 mm vesicles.
Hydrophobic Nile Red fluorescent probe (10 mm) was used for visual-
ization of the bilayer membrane. b) Cryo-TEM shows the 50–100 nm
vesicle population of A·B. Conditions: [A]0 =10 mm, [B]0 =20 mm,
pH 9.5, 258C.
*
Figure 2. a) Vesicle association in time followed by DLS ( ) and
fluorescence spectroscopy, in which 1 mm Nile Red was used as
*
fluorescent probe ( ); precursors A and B were mixed in water, thus
forming A·B vesicles ([A]0 =10 mm, [B]0 =20 mm). b) Vesicle dissocia-
tion followed in time with DLS; a vesicle solution ([A]0 =10 mm,
[B]0 =20 mm) in water was diluted 100 times.
Imine formation was confirmed for an A·B vesicle
solution ([A]0 = 10 mm, [B]0 = 20 mm; A.B represents an
unspecified mixture of A and B) by LC-MS analysis, thus
showing formation of both single- (AB1) and double-tailed
(AB2) surfactants. For quantification, it was found necessary
to quench the imine exchange of the equilibrated mixture by
using NaBH4, as such reducing A, AB1, and AB2 to their
corresponding nondynamic alcohols and amines (see Table 1
in the Supporting Information). 45% of A was converted into
imine surfactants, of which the double-tailed AB2 was the
major constituent (AB1: 14%; AB2: 86%; see Figure 3 in the
Supporting Information). The CAC (herein defined as
minimal precursor concentration needed for aggregate for-
mation when A and B are mixed in a 1:2 ratio) of [A]0 was
determined to be 0.4 mm using fluorescence spectroscopy
with the hydrophobic Nile Red probe (Figure 4 in the
Supporting Information). By measuring the surface tension
a CAC of 1.7 mm was determined (Figure 5 in the Supporting
Information). The fact that the CAC determined by fluores-
cence spectroscopy is significantly lower than the one
determined by surface tension can be explained by the
induction of hydrophobic interactions by the Nile Red
probe.[13]
by a rapid increase in scattering intensity, and a concomitant
blue shift of Nile Red fluorescence, indicating the formation
of hydrophobic domains (Figure 2a).
CLSM measurements showed that individual vesicles are
stable for at least 50 minutes (Figure 6 in the Supporting
Information). However, diluting an opalescent A·B vesicle
solution in water 100 times ([A]0 = 10 mm, [B]0 = 20 mm)
resulted in a transparent solution, thus suggesting that all
vesicles had dissociated into their surfactant precursors.
Dilution was followed by a rapid decrease in scattering
intensity with a timescale in the order of one minute
(Figure 2b). Fluorescence emission measurements showed
an emission shift from 630 nm to 660 nm; the latter corre-
sponds to Nile Red emission in pure water and thus to the
absence of hydrophobic domains.[13] A shift of the imine
equilibrium towards the surfactant building blocks was
confirmed by 1H NMR spectroscopy. This behavior is in
sharp contrast to that observed in both phospholipid and
conventional double-tailed quaternary ammonium surfactant
membranes, where the bilayers have a very high stability and
associated slow dissociation kinetics upon dilution, related to
their characteristic nano- to sub-nanomolar CACs.[2,14]
Vesicle dissociation was investigated in more detail by
CLSM. We found that upon dilution otherwise stable vesicles
Since the vesicles are formed by reversible association of
their non-amphiphilic precursors, we expected that a shift of
the imine equilibrium should lead to a fast response of the
vesicle bilayer. After mixing solutions of both precursors in
water, vesicles were formed within a few minutes, exemplified
3422
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 3421 –3424

fuse, forming larger vesicles (Figure 3a). Then, the interfacial
membrane thins to rupture, much in the same way as is
observed for fusing soap bubbles. In the meantime, after
fusion, the initially larger vesicle shrinks to a smaller vesicle
between bilayer and bulk for double-chain surfactants, it
seems likely that the double-chain surfactants in the bilayer
first dissociate into their precursors. Subsequently, the
surfactant precursors migrate from the bilayer to the bulk
solution. Because dilution initially decreases the precursor
concentrations in the external bulk solution, it is likely that
dilution first leads to the loss of surfactants from the outer
bilayer leaflet, leading to enhanced susceptibility towards
fusion between nearby vesicles. At later stages, the bilayer
stability might further decrease as a result of a larger
discrepancy between the outer and inner leaflet, leading to
enhanced permeability and/or surfactant flip-flop, and ulti-
mately shrinkage of the vesicles as a whole.
Comparison of the rate of vesicle dissociation with the
rate of dissociation of a water-soluble, non-aggregating imine
upon dilution shows similar time scales (Figure 8 in the
Supporting Information), indicating that aggregation does not
contribute significantly to the kinetic stability of the aggre-
Figure 3. Vesicle dissociation followed in time with CLSM: a) adjacent
vesicles fuse to form larger unilamellar vesicles which then start to
shrink; b) single vesicles shrink until no longer visible. Conditions:
10 mm [A]0 and 20 mm [B]0, pH 9.5, 258C. Fluorescent probe: 10 mm
Nile Red. Typically 50 mL samples were used and diluted with 90 mL
10 mm KBr solution containing 10 mm Nile Red. No significant pH
shift was measured (DpHꢀ0.2) upon dilution.
or dissolves completely, depending on the final concentration
(Figure 3b). The decrease in size of the single vesicles also
took place on a timescale in the order of one minute (Figure 7
in the Supporting Information), which is in good agreement
with the results of the DLS measurements.
For the mechanism of vesicle dissociation, it is assumed
that for vesicle-forming surfactants AB2, the concentration of
nonaggregated surfactant in bulk is generally negligibly low
and the exchange between bilayer and bulk phase is very slow
(Figure 4).[2] Lowering the concentration of the surfactant
precursors in the bulk surroundings by dilution drives the
dissociation of the dynamic surfactants, in accordance with Le
Chꢀtelierꢁs principle. Because of the low exchange rates
Figure 5. pH-triggered vesicle dissociation ([A]0 =20 mm,
[B]0 =40 mm): a) an A·B vesicle solution in water (pH 9.5Æ0.2) was
acidified by addition of an aqueous HBr solution. Vesicle dissociation
*
*
was studied with DLS ( ) and fluorescence spectroscopy ( ), in which
*
1 mm Nile Red was used as fluorescent probe; b) DLS ( ) showed that
vesicle dissociation of A·B is reversible going from alkaline (pH>8) to
&
Figure 4. Proposed mechanism of vesicle dissociation upon dilution:
a) surfactants in the bilayer dissociate into their precursors (1).
Subsequently, the precursors migrate from the bilayer to the bulk (2);
b) bilayer destabilization and shrinkage over time. Grey: cationic
headgroup; red: apolar tail.
acidic (pH<4) pH values. The vesicle size ( ), as determined by DLS
methods, remained constant after each cycle. It should be noted that
the scattering intensity at alkaline pH values decreased after each
cycle, which is caused by dilution effects (Figure 9 in the Supporting
Information).
Angew. Chem. Int. Ed. 2011, 50, 3421 –3424
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3423

Communications
gated imine surfactants. The similarity in time scales between
Keywords: aggregation · dynamic covalent chemistry · imines ·
self-assembly · vesicles
.
imine dissociation and vesicle shrinkage and dissolution
suggests that imine dissociation in the bilayer is the rate-
determining step. Overall, continuous surfactant dissociation
and subsequent diffusion of precursors to the bulk result in
destabilization of the bilayer and shrinkage of the vesicle
(Figure 4).
As an alternative means of control, vesicle response to a
change in pH was investigated. We titrated a concentrated
A·B vesicle solution in water with an aqueous HBr solution
and found that the vesicles remained stable down to pH 7.1 Æ
0.2 as observed with DLS and fluorescence spectroscopy
(Figure 5a).
At pH < 4 the vesicles had completely dissociated result-
ing from a gradual shift of the imine association equilibrium
towards its precursors as was evident from a significant
decrease in scattering and a red shift of the Nile Red probe
from 637 nm to around 655 nm (Figure 5a). Vesicle dissoci-
ation was entirely reversible, demonstrated by a significant
increase in scattering and regeneration of the original
vesicular size when the pH value of the solution was changed
back to the alkaline state and vice versa (Figure 5b and
Supporting Information).
In summary, we have shown how the introduction of
dynamic covalent bonds in double-tailed surfactants can lead
to the formation of vesicles that are highly responsive to
changes in their environment. The surfactants are formed
in situ from two water-soluble, not surface-active precursors
by the formation of a reversible imine bond. This dynamic
covalent imine bond results in vesicle assemblies with a highly
dynamic bilayer. Complete vesicle dissociation is realized by
dilution and change of the pH value. This study represents a
new approach for accessing dynamic vesicular architectures,
which are highly interesting candidates for dynamic encapsu-
lation and release vehicles because of their ease of formation,
fast dissociation, and switchable morphology.
[1] U. J. Meierhenrich, J. Filippi, C. Meinert, P. Vierling, J. P.
Dworkin, Angew. Chem. 2010, 122, 3826; Angew. Chem. Int.
Ed. 2010, 49, 3738.
[2] J. Israelachvilli, Intermolecular & Surface Forces, 2nd ed.,
Academic Press, London, 1991.
[3] J. Voskuhl, B. J. Ravoo, Chem. Soc. Rev. 2009, 38, 495.
[4] a) P. Walde, R. Wick, M. Fresta, A. Mangone, P. L. Luisi, J. Am.
Chem. Soc. 1994, 116, 11649; b) P. Stano, P. L. Luisi, Chem.
Commun. 2010, 46, 3639.
[5] F. Zhang, F. Kamp, J. A. Hamilton, Biochemistry 1996, 35, 16055.
[6] K. Morigaki, P. Walde, Curr. Opin. Colloid Interface Sci. 2007,
12, 75.
[7] a) K. Kurihara, K. Takakura, K. Suzuki, T. Toyota, T. Sugawara,
Soft Matter 2010, 6, 1888; b) T. Toyota, K. Takakura, Y.
Kageyama, K. Kurihara, N. Maru, K. Ohnuma, K. Kaneko, T.
Sugawara, Langmuir 2008, 24, 3037; c) T. Toyota, K. Takakura,
T. Sugawara, Chem. Lett. 2004, 33, 1442.
[8] a) S. Otto, R. L. E. Furlan, J. K. M. Sanders, Science 2002, 297,
590; b) I. Huc, J. M. Lehn, Proc. Natl. Acad. Sci. USA 1997, 94,
2106; c) J. M. Lehn, A. V. Eliseev, Science 2001, 291, 2331;
d) P. T. Corbett, J. Leclaire, L. Vial, K. R. West, J. Wietor,
J. K. M. Sanders, S. Otto, Chem. Rev. 2006, 106, 3652; e) J. M.
Lehn, Science 2002, 295, 2400; f) A. W. Jackson, D. A. Fulton,
Chem. Commun. 2010, 46, 6051; g) T. Ono, S. Fujii, T. Nobori,
J. M. Lehn, Chem. Commun. 2007, 4360.
[9] a) A. Ghoussoub, J. M. Lehn, Chem. Commun. 2005, 5763; b) N.
Sreenivasachary, J. M. Lehn, Proc. Natl. Acad. Sci. USA 2005,
102, 5938; c) R. J. Williams, A. M. Smith, R. Collins, N. Hodson,
A. K. Das, R. V. Ulijn, Nat. Nanotechnol. 2009, 4, 19.
[10] a) J. F. Folmer-Andersen, J. M. Lehn, Angew. Chem. 2009, 121,
7800; Angew. Chem. Int. Ed. 2009, 48, 7664; b) J. M. A. Carnall,
C. A. Waudby, A. M. Belenguer, M. C. A. Stuart, J. J. P. Peyr-
alans, S. Otto, Science 2010, 327, 5972.
[11] R. Nguyen, L. Allouche, E. Buhler, N. Giuseppone, Angew.
Chem. 2009, 121, 1113; Angew. Chem. Int. Ed. 2009, 48, 1093.
[12] C. B. Minkenberg, L. Florusse, R. Eelkema, G. J. M. Koper, J. H.
van Esch, J. Am. Chem. Soc. 2009, 131, 11274.
[13] a) M. C. A. Stuart, J. C. van de Pas, J. B. F. N. Engberts, J. Phys.
Org. Chem. 2005, 18, 929; b) I. Capek, Colloid Interface Sci.
2002, 97, 91.
[14] T. F. Svitova, Y. P. Smirnova, S. A. Pisarev, N. A. Berezina,
Colloids Surf. A 1995, 98, 107.
Received: November 24, 2010
Revised: January 26, 2011
Published online: March 16, 2011
3424
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 3421 –3424