View Article Online / Journal Homepage / Table of Contents for this issue

REVIEW www.rsc.org/softmatter | Soft Matter

Vesicles in electric fields: Some novel aspects of membrane behavior†

Rumiana Dimova,*a Natalya Bezlyepkina,a Marie Domange Jord€o,b Roland L. Knorr,a Karin A. Riske,c
Margarita Staykova,a Petia M. Vlahovska,d Tetsuya Yamamoto,a Peng Yang‡a and Reinhard Lipowskya
Received 30th January 2009, Accepted 18th May 2009
First published as an Advance Article on the web 30th June 2009
Published on 30 June 2009. Downloaded by Universitat Autonoma de Barcelona on 28/10/2014 16:30:22.

DOI: 10.1039/b901963d

This review focuses on the effects of electric fields on giant unilamellar vesicles, a cell-size membrane
system. We describe various types of behavior of vesicles subjected to either alternating fields or strong
direct current pulses, such as electrodeformation, -poration and -fusion. The vesicle response to
alternating fields in various medium conditions is introduced and the underlying physical mechanisms
are highlighted, supported by theoretical modeling. New aspects of the response of vesicles with
charged or neutral membranes, in fluid or gel-phase, and embedded in different solutions, to strong
direct current pulses are described including novel applications of vesicle electrofusion for nanoparticle
synthesis.

1. Introduction response of the membrane to electric fields; for a partial overview

see Dimova et al.32 Membrane behavior in electric fields is a topic
The response of biological membranes and cells to electric fields of active research. Here, we will focus mainly on new develop-
has received a lot of attention, both because of fundamental ments by our group. Even though we will attempt to cite all
interest and because of potential practical applications. External important contributions in the field, the selection is subjective
electric fields, whether weak alternating (AC) fields or strong and far from being exhaustive.
direct current (DC) pulses, have emerged as a powerful method The paper is organized as follows; first we introduce some
for cell manipulation in biomedical and biotechnological appli- basic timescales that govern the interaction of electric fields with
cations. For example, electric fields are employed in novel in-vivo membranes. Then we consider the response of vesicles to AC
and in-situ applications for tissue ablation, wound healing and fields and DC pulses, discussing some new observations (both
cancer treatment.1–6 Strong electric fields can cause a significant reported and not yet published). We conclude with a short
increase in the electric conductivity and permeability of the cell outlook.
plasma membrane. This phenomenon, also referred to as elec-
troporation or electropermeabilization, is used for introducing
various molecules into the cell, to which the membrane is
otherwise impermeable.7,8 Because of its efficiency, this method is 2. Membranes in electric fields: some relations
rapidly becoming an established approach for treatment of
carcinoma, melanoma and connective tissue cancer,9–12 and it The response of membranes to electric fields involves dynamic
also holds great promise for gene therapy.13,14 Membrane elec- physical processes occurring on different time scales. Free
troporation and electrofusion are of particular interest because charges accumulate on boundaries separating media with
these methods are widely used in cell biology and biotechnology different electric properties. A spherical vesicle polarizes on the
as means for cell hybridization.15 Maxwell–Wagner time scale33
Synthetic lipid vesicles provide biomembrane models suitable 3in þ 23ex
tMW ¼ (1)
for systematic investigations of the impact of electric fields on lin þ 2lex
lipid bilayers. Studies on small vesicles with a size about
where 3in and 3ex are the dielectric constants, and lin and lex are
100 nm16–18 and on giant unilamellar vesicles with a diameter of
the conductivities of the solutions inside and outside the vesicle,
several tens of microns,19–31 have been performed to elucidate the
respectively.
The lipid bilayer is impermeable to ions and free charges pile
a
Max Planck Institute of Colloids and Interfaces, Science Park Golm, up on both membrane surfaces. Hence, the vesicle acts as
14424 Potsdam, Germany. E-mail: Rumiana.Dimova@mpikg.mpg.de; a capacitor, which charges on a time scale34,35
Fax: +49 331 567 9615; Tel: +49 331 567 9612  
b
Niels Bohr Institute, University of Copenhagen, Blegdamsvej 17, 2100 1 1
tc ¼ RCm þ (2)
Copenhagen, Denmark lin 2lex
c
Depto. de Biofı́sica, Universidade Federal de São Paulo, R. Botucatu, 862,
CEP 04023-062 São Paulo, Brazil where R is the vesicle radius and Cm is the membrane capaci-
d
Thayer School of Engineering, Dartmouth College, Hanover, NH, 03755, tance.
USA The capacitor charging time tc is typically much longer than
† This paper is part of a Soft Matter themed issue on Membrane
Biophysics. Guest editor: Thomas Heimburg. the Maxwell–Wagner time tMW. For example, we can estimate
‡ Present address: Department of Biomedical Engineering, Duke tc  10 ms and tMW  0.01 ms for conditions corresponding to
University, Durham, NC, 27705, USA. experiments on vesicles in 1 mM NaCl, namely 3in  3ex ¼ 8030,

This journal is ª The Royal Society of Chemistry 2009 Soft Matter, 2009, 5, 3201–3212 | 3201
 View Article Online

where 30 is the vacuum permittivity, lin  lex  10 mS/m,

Cm  0.01 F/m2, and R  10 mm.
These time scales are a key to understanding the dynamic
response of vesicles subjected to short electric pulses discussed in
section 4, as well as frequency dependence of vesicle deformation
discussed in section 3. Note that characteristic angular frequen-
cies are defined as the inverse of the time scales in eqn (1) and (2),
e.g. uMW ¼ 1/tMW. The experimental frequency, n, is related to
Published on 30 June 2009. Downloaded by Universitat Autonoma de Barcelona on 28/10/2014 16:30:22.

the angular one via n ¼ u/2p.

3. Vesicle response to AC fields

When subject to AC fields, cells and vesicles can deform. Studies
of cell deformation in AC fields have been carried out by many
groups and effects on the cell shape, motility and orientation
have been reported.26,36–38 Vesicle deformations have been treated
both experimentally and theoretically, but a comprehensive
description reconciling observations and calculations is still to
emerge.
A detailed understanding of the membrane behavior in AC
fields is important for various electromanipulation techniques, as
well as for vesicle electroformation protocols39,40 (including some
of the recent developments41–45). Even though vesicle electro-
formation is widely used, the underlying mechanism is not well
understood.46,47 This motivates further studies on effects of AC
fields on membranes. Most of the current research in this direc-
tion as well as the following sections 3.1 and 3.2 has focused on
lipid bilayers with only a few components, whereas biological
membranes contain a large number of different components.
Bridging the gap would require the exploration of more complex
systems.

Fig. 1 Morphological diagram of the shapes of vesicles at different

3.1 Vesicle deformation in AC fields conductivity conditions and various field frequencies (a) as determined
experimentally, and (b) theoretically predicted for lin ¼ 6.5 mS/m. The
The deformation of vesicles subjected to AC fields depends on
symbols in (a) correspond to different internal conductivity, lin, in units
the field frequency u (or n) and the conductivity conditions. The mS/m: 1.5 (solid squares), 6.5 (open circles), 13 (solid triangles), 1000
latter can be described by the ratio between the internal and the (open squares). The dashed lines are guides to the eye and the shaded
external conductivities lin and lex: areas indicate zones of specific morphology. The four types of morpho-
logical transitions are discussed in the text. The dotted vertical line in (a)
x ¼ lin/lex (3) shows the experimentally accessible frequency limit (n ¼ 2  107 Hz).
Schematic views of the vesicle shapes are included as insets and the
Systematically varying the field frequency and solution electric field is indicated by an arrow.
conductivities allowed us to construct a morphological diagram
of the shape transitions observed in phosphatidylcholine vesi-
cles;48 see Fig. 1a. At high frequencies, the vesicles are spherical towards realistic theoretical modeling is discussed in the next two
independently of x. As the frequency decreases, vesicles with x > sections.
1, i.e., with the internal salinity higher than the external one
become prolate ellipsoids corresponding to transition 1 in 3.1.1 Morphological diagram: energy minimization approach.
Fig. 1a, while vesicles with x < 1 adopt oblate shapes after Vesicle shapes in AC fields can be investigated within the
undergoing transition 2. Further decrease in frequency changes framework of the energy minimization approach introduced by
the vesicle shape at transition 4 from oblate to prolate for x < 1. Winterhalter and Helfrich.24 The original work, however, is
For intermediate frequencies an oblate vesicle can become limited to symmetric conductivity conditions with x ¼ 1 and thus
prolate at transition 3. the model predicts only prolate shapes independent of x. We
Theoretical studies of vesicle deformation in AC fields have extended the Winterhalter–Helfrich model to asymmetric
been limited to rather simple systems. For example, these studies conductivity conditions with x s 1.
omit the asymmetry in the media conductivities,24,25,27 and their The electric field deforms a vesicle from a sphere with radius R
theoretical predictions are at odds with experiments; see e.g. the into an ellipsoid. The vesicle deformation s2 is assumed to be
supplementary material of Aranda et al.48 Our recent progress small with s2  R; see also Fig. 2a for definition of s2. Prolate and

3202 | Soft Matter, 2009, 5, 3201–3212 This journal is ª The Royal Society of Chemistry 2009
 View Article Online

respectively. Depending on the polarity of Q, fq is directed either

towards the poles or the equator, and fr is directed towards or
away from the electrodes, leading to prolate or oblate vesicle
shapes as sketched in Fig. 2b, c.49
In the high frequency regime, u > uMW, the electric charges
cannot follow the oscillations of the electric fields. As a result, the
net charge density, Q, as defined in eqn (4), decreases with the
field frequency. This relaxes the shape of the vesicle from prolate
Published on 30 June 2009. Downloaded by Universitat Autonoma de Barcelona on 28/10/2014 16:30:22.

(x > 1) or oblate (x < 1) to spherical (transitions 1 and 2 in Fig. 1).

Fig. 2 (a) The vesicle geometry and (b, c) the net charge distribution Q at In summary, the charging dynamics of the membrane surfaces
the vesicle interfaces at intermediate frequencies. Due to the difference in and the radial and shear Maxwell stresses play a key role in
the conductivity conditions, the net charges across the membrane, illus- determining the vesicle morphology in AC fields. The frequency
trated with pluses and minuses, differ depending on the value of the of transition 4 in Fig. 1a corresponds to the inverse charging time
conductivity ratio x. The forces (fr and fq) applied to the charges by the of the membrane capacitor, 1/tc, and frequency transitions 1 and
normal and the tangential electric fields deform the vesicles into prolates
2—to the Maxwell–Wagner frequency uMW.
for x > 1 (b) and oblates for x < 1 (c).
Quantitatively, the present theory provides reasonable values
of the relative vesicle deformation s2/R ( 0.1) for small vesicles
oblate shapes correspond to s2 > 0 and s2 < 0, respectively. The with size of the linear order of 1 mm. For giant vesicles (R  10–
free energy of a vesicle in AC electric fields can be presented as F 100 mm), the theory gives unreasonably large values for s2/R (
¼ Fbend  W, where Fbend is the bending energy of the vesicle in 103–106!). However, the shapes of the boundaries in the
the elliptic deformation, and W is the work done by the Maxwell morphological diagram and the order of the transition frequen-
stresses arising from the electric fields. The deformation s2 can be cies agree with the experiment very well; see Fig. 1. The work W
determined by minimizing the free energy F or by balancing done by the Maxwell stresses is small at the vicinity of the
stresses exerted on the membrane as in section 3.1.2. Fig. 1b transition frequencies and for small vesicles. Therefore, the
shows the morphological diagram predicted by the model just present theory shows quantitative agreements with the experi-
described. The shapes of the boundaries and the transition ments when W is small. It is necessary to take into account
frequencies agree well with the experimentally determined tension and hydrodynamic forces in order to achieve quantitative
morphological diagram as shown in Fig. 1a. agreement with the experiments as discussed in the following
The physical mechanism responsible for the vesicle electro- section.
deformation is the interplay between the electric field partition-
ing in normal and tangential components, and the charging of 3.1.2 Vesicle deformation: force balance approach. Another
the membrane interfaces. The lipid bilayer is an insulator, and method to determine the vesicle deformation in electric fields is
acts as a capacitor. At low frequencies, u  1/tc, the large based on the balance of all forces exerted on the membrane49
membrane impedance blocks current from flowing into the
 
vesicle interior and the electric field lines are tangent to the   dFbend
n  Tex  Tin ¼ 2sH  n þ Vs s (5)
membrane. The vesicle is squeezed at the equator and pulled at dr
the poles by the radial Maxwell stress or pressure arising from the
tangential electric field. As a result, the vesicle adopts a prolate
shape. Here n a is normal vector to the vesicle membrane, Tex and Tin
At intermediate frequencies, 1/tc < u < uMW, the membrane is denote the exterior and interior Maxwell stress, H is the mean
capacitively short-circuited and displacement currents flow curvature, r is the radial coordinate, and s is the membrane
through it. The electric field lines penetrate the vesicle interior tension.
and the electric field acquires a component normal to the An essential feature of this approach is the consideration of
membrane. Because of x s 1, i.e., of the asymmetry of a variable membrane tension. First, flattening of the shape fluc-
the internal and external conductivities, the charge densities on tuations due to vesicle elongation increases the homogeneous
the inner and outer membrane interfaces become imbalanced. part of the tension. Second, because the membrane is nearly
Within the continuum theory, these charges arise from the incompressible, the tension can become nonuniform along the
discontinuity of the permittivities across the interfaces and surface. The resulting gradients in the tension, Vss, are particu-
represent local accumulation of cations and anions at these larly important in the intermediate frequency regime, 1/tc < u <
interfaces. The resulting net free charge density Q is given by uMW, in which the shear Maxwell stresses are significant and
  oblate shapes are observed.
lex lin 3ex 3in cosðut þ fÞ For small deviations from sphericity, eqn (5) yields
QðtÞ ¼ 3E0 cosq  pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi (4)
2lex þ lin lex lin 1 þ u2 =uMW 2
3pel  2sel
where E0 is the magnitude of the external electric field, and q and s2 ¼ R (6)
12ð6 þ sh Þ
f are the polar and azimuthal angles, respectively; see Fig. 2a. A
schematic snapshot of Q is sketched in Fig. 2b and 2c. If 3ex  3in, where sh is the homogeneous tension (which is independent of
the sign of Q is determined solely by the conductivity ratio. The position along the vesicle surface) and pel and sel are the maximal
interaction of the tangential and normal electric fields with the values of the difference of the radial and shear Maxwell stresses
free charges produces lateral and normal forces, fq and fr, across the membrane (the expressions for pel and sel can be found

This journal is ª The Royal Society of Chemistry 2009 Soft Matter, 2009, 5, 3201–3212 | 3203
 View Article Online

in Vlahovska et al.49). For a quasi-spherical vesicle, the homo-

geneous tension sh increases with the apparent area as50,51
 
16p k 2
sh ¼ s0 exp s2 (7)
5 kB T
where s0 is the initial membrane tension, k is the membrane
bending stiffness and kBT is thermal energy. Eqn (6) is a gener-
alization of the Kummrow–Helfrich result28 (see eqn (10)
Published on 30 June 2009. Downloaded by Universitat Autonoma de Barcelona on 28/10/2014 16:30:22.

in section 3.1.3), which is strictly valid only for low frequencies

(u < 1/tc), where the shear Maxwell stress is zero. If x < 1, the
numerator in eqn (6) changes sign at u ¼ 1/tc, which marks the
prolate–oblate transition (transition 4 in Fig. 1a).
Vesicle shapes computed from eqn (6) are consistent with the
experiment, see Fig. 3 (note that the vesicle semiaxis is simply a ¼
R + s2). The discrepancy in the high-frequency oblate–sphere
transition frequency for x < 1 is presumably due to electric
double layer effects.49 The mechanical approach also explains
why the energy approach (section 3.1.1) overestimates the vesicle
deformation in strong fields. The reason is that the free energy
balance does not take into account the tension. In strong electric
fields, the membrane tension controls the extent of vesicle
deformation.

3.1.3 Electrodeformation of vesicles as a method to determine

the membrane bending stiffness. Vesicle deformation induced by
AC fields can be used to measure the bending stiffness of
membranes following the approach developed by Helfrich and
coworkers.28,30 The protocol of such measurements consists of
subjecting a vesicle to an AC electric field of increasing strength
and recording the induced deformation. One example of vesicle
elongation due to stepwise increase in the field strength is shown Fig. 4 Vesicle electrodeformation as a method for measuring the
in Fig. 4a. The degree of deformation is expressed as the aspect membrane bending stiffness. (a) Degree of deformation, a/b (as indicated
in the inset) induced on a vesicle made of dipalmitoylphosphatidylcholine
ratio a/b, where a and b are the vesicle semiaxes along and
: cholesterol 9 : 1 (molar ratio) subjected to AC field with frequency
perpendicular to the field direction, respectively; see inset in
200 kHz. The applied electrical potential is increased every 10 s with
Fig. 4a. Observations of the response time of different vesicles a step of 0.5 V (10 V/cm), as indicated. The first two seconds after
suggest that typically 2 s are sufficient to reach the equilibrium changing the field are excluded from averaging over the shape in time. (b)
deformation after changing the field strength. Images recorded in Relative area change of a vesicle subjected to AC field (300 kHz) as
a function of the membrane tension. Each data point is a result of
averaging the relative area change over 90 images. The solid line is
a linear least squares fit, which slope yields k ¼ (9.5  0.6)  1020 J for
the bending stiffness and the intercept gives s00 ¼ (1.7  1.4)  106 N/m.

the following 3–8 s can be time-averaged to achieve better

precision in a/b. For the conductivity conditions and frequency
range (between 1 kHz and 300 kHz) in such experiments, the
vesicles adopt prolate deformation as discussed in the previous
sections.
The vesicle deformation is associated with a change in the
apparent area due to flattening the membrane fluctuations. Area
stored in small membrane undulations is pulled out and made
optically visible. The changes in apparent area is modulated by
the membrane tension50,51 (note that this equation is equivalent
to eqn (7)):
 
A  A0 kB T sh
ah ¼ log (8)
A0 8pk s0 0
Fig. 3 Comparison between experimental data, symbols (exp), from
Aranda et al.48 and theory, solid curves (th), as introduced by eqn (6), for where A is the area of the ellipsoid, A0 is the area of the sphere
vesicle shapes in AC fields at conditions given in the legend. The initial with the same volume, and s00 is the positive parameter obtained
tension s0 is an adjustable parameter for the theoretical curves. by extrapolation to a ¼ 0. Note that s00 can be larger than the

3204 | Soft Matter, 2009, 5, 3201–3212 This journal is ª The Royal Society of Chemistry 2009
 View Article Online

actual tension at zero field strength. The tension of the deformed To study the lipid flow dynamics in AC fields, we used giant
vesicle, sh, can be obtained from the electric stresses. The normal vesicles with mixed lipid bilayers, which, at room temperature,
electric stress at the equator of the vesicle as given in the work of phase separate in liquid ordered (lo) and liquid disordered (ld)
Helfrich and coworkers28,30 is: phases,55 leading to the formation of lo and ld domains on the
9 vesicles. A small fraction of fluorescent dye was added, which
ðTrr Þeq ¼  3w E02 (9) preferentially partitions in the ld phase. The lipid ratio was such
8
that the lo phase appeared as dark circular patches in the
where 3w is the dielectric constant of water, and E0 is the field surrounding fluorescently labeled ld phase.
Published on 30 June 2009. Downloaded by Universitat Autonoma de Barcelona on 28/10/2014 16:30:22.

strength far away from the vesicle. Since the pressure inside the The membrane flow pattern was resolved by following the
vesicle must be constant, combining the Laplace equation at the motion of the lo patches with confocal microscopy.56 The top or
poles and in the equator gives the bottom part of the vesicle were recorded as shown on the
micrographs in Fig. 5a–c. The inner and outer vesicle solutions
(c1 + c2)eqsh  (Trr)eq ¼ (c1 + c2)polesh (10) were 0.1 M sucrose and glucose, respectively. This ensures
osmotic balance, i.e. constant vesicle volume, and causes the
where c1, c2 are the principal curvatures taken either at the
vesicles to sediment at the bottom of the chamber. The electric
equator (eq) or the pole (pole), and therefore measurable from
field was applied between two parallel cylindrical electrodes with
the geometry of the vesicle.
a diameter of 200 mm and an inter-electrode gap of 500 mm. In
Logarithmic plot of the membrane lateral tension obtained
AC fields, smaller vesicles experience lifting due to negative
from eqn (10) against the change in apparent area gives a straight
dielectrophoretic forces, but the larger ones (R $ 50 mm), also
line with slope related to the bending rigidity as described in eqn
being heavier, remain at the chamber bottom. The proximity of
(8). One example of this protocol applied to a vesicle composed
the bottom glass to the vesicle, as shown in Fig. 5d, leads to an
of dipalmitoylphosphatidylcholine and cholesterol is given in
asymmetric field distribution at the membrane surface. The field
Fig. 4b. A linear least squares fit of the dependence of the relative
strength is much higher at the lower vesicle part, facing the glass,
area change as a function of the applied tension following eqn (8)
than at the top part.56
yields for the slope k ¼ 9.5  1020 J. Repeating the measurement
Such asymmetric field distribution leads to special membrane
on the same vesicle shows reproducibility within about 22%
flow patterns, consisting of concentric closed trajectories orga-
deviation from the value of k. Scatter within about 25% is
nized in four symmetric quadrants, each extending from the
observed when the measurements are performed on different
bottom to the top of the vesicle; see Fig. 5d, e. The flow is fastest
vesicles with the same composition. The obtained value for the
bending stiffness is consistent with published data.52
Note that this method does not apply to vesicles containing
charged lipids and for vesicles embedded in salt solutions. In
these cases, the Maxwell stress tensor used to evaluate the
membrane tension has to account for the media conductivity (as
discussed in the previous two sections) and the charges at the
membrane surface.

3.2 Electrohydrodynamic flows in vesicles induced by non-

homogeneous AC fields
As discussed above, electric fields induce forces at the vesicle
interface, due to the difference in the media polarizabilities. At
intermediate frequencies, 1/tc < u < uMW, as shown in section
3.1, the lateral force is responsible for the vesicle deformation. In
addition, this force may also lead to fluid flows, analogous to the
flows induced in liquid droplets.53 However, there is a funda- Fig. 5 Micrographs obtained on a confocal microscope (fully opened
mental difference between droplets and vesicles, which arises pinhole) illustrating the membrane flow on the bottom part (a–c) of
from the properties of the lipid bilayer.54 The membrane behaves a giant vesicle with a diameter of about 150 mm induced by an AC field
as a two dimensional nearly incompressible fluid. Under stress, it (360 V/cm, 80 KHz), at external and internal conductivities of 25 mS/m
develops tension to keep its surface area constant. In uniform AC and 0.3 mS/m, respectively. The vesicle was prepared from a mixture of
fields, membrane flow in the vesicle is not expected because the dioleoylphosphatidylcholine : dipalmitoylphosphatidylcholine : choles-
lateral electric stress is counterbalanced by the resulting axially terol, 4.8 : 3.2 : 2 in mole fractions. The time between the consecutive
snapshots is approximately 1.3 s. The yellow dashed arrows indicate the
symmetric gradient in the membrane tension. In inhomogeneous
trajectories of selected domains in the consecutive snapshots. The scale
fields however, this force balance is broken and a flow of lipids
bar corresponds to 50 mm. The vesicle is located close to the bottom of the
occurs in order to restore it. Note that in most experimental observation chamber as illustrated in (d), where the vesicle top and
conditions used for electromanipulation, vesicles, cells or other bottom parts, the poles and the field direction are indicated. The side and
particles are exposed to inhomogeneous fields, arising from the bottom views of the flow lines are sketched in (d) and (e), respectively.
screening by neighboring particles, sedimentation or chamber The length of the arrows in (d) roughly corresponds to the amplitude of
geometry. the flow velocity.

This journal is ª The Royal Society of Chemistry 2009 Soft Matter, 2009, 5, 3201–3212 | 3205
Published on 30 June 2009. Downloaded by Universitat Autonoma de Barcelona on 28/10/2014 16:30:22. View Article Online

Fig. 6 (a) A phase-contrast image and (b–g) confocal cross sections of a giant lipid vesicle enclosing dextran-rich droplets (green fluorescence) in a PEG-
rich phase. The cross section in (b), corresponding to the image in (a), is taken close to the equatorial plane of the vesicle and shows only the droplets in
focus. Application of an inhomogeneous AC field (460 V/cm, 80 KHz) at an external conductivity of 40 mS/m leads to a vesicle shape deformation and
an internal flow in the direction perpendicular to the plane of the image (c–g). The flow is visualized by following the motion of droplets 1, 2 and 3, which
come in focus and go out of focus. The time period is 2.5 s between images (c–d) and (d–e) and 5 s between (e–f) and (f–g). The field direction is indicated
by the arrow in (c).

at the periphery of the quadrant and at the bottom of the vesicle. lived shape deformations. We have previously explored the
The top and the bottom of the vesicle are stagnation points. The characteristic times associated with vesicle relaxation and pora-
velocity of the domains reaches about 30 mm/s corresponding to tion,60,61 as well as electrofusion of vesicles induced by DC pul-
laminar flows. The velocity can be further increased by the field ses.62,63 This section will be dedicated to some novel observations
strength and the conductivity of the external solution. Interesting in this direction. In particular, we will discuss the influence of
effects are observed when the field frequency is varied. At several other factors on the vesicle response to DC pulses: (i)
frequencies less than about 3 MHz, the motion in the circular presence of charged lipids in the membrane, see section 4.1; (ii)
trajectories is directed downwards past the poles and upwards particles in the vesicle solution, see section 4.2; and (iii) phase
along the equator as sketched in Fig. 5d but reverses its direction state of the membrane, see section 4.3. Finally, we will introduce
at higher frequencies.56 an interesting new application of electrofusion, namely for the
Calculations of the lateral electric stress or surface force synthesis of nanoparticles in vesicles; see section 4.4.
density on the membrane suggest that the vesicle experiences
significant shear stress in the vicinity of the solid substrate.56 As
a result, a non-uniform and non-symmetric membrane tension 4.1 Unusual behavior of charged membranes exposed to DC
builds up. It triggers lipid flow towards the regions of highest pulses: vesicle bursting
tension, in analogy to Marangoni flows in monolayers.
Strong electric pulses applied to single component giant vesicles
The flow in the membrane is coupled to fluid flows in the
made of phosphatidylcholine induce the formation of pores,
internal and external media. To visualize the effect of the
which reseal within milliseconds.60 The mechanism of this pore
membrane flow on the internal medium we used vesicles con-
formation, i.e., electroporation, can be understood in terms of
taining aqueous solution of the water-soluble polymers poly-
the stress in the bilayer created by the electric field.32 In the
(ethylene glycol) (PEG) and dextran. At specific polymer
presence of this field, the accumulated charges across the
concentration, this solution undergoes phase separation57,58
membrane create a transmembrane potential, which induces an
producing droplets of dextran-rich phase, which can be visual-
effective electrical tension19,60,64 as defined by the Maxwell stress
ized e.g. by fluorescently labeled dextran. The droplets gradually
tensor. Fluid membranes rupture if the tensions exceed about
coarsen. Before the coarsening is completed we subject such
10 mN/m19,65 also known as lysis tension.
vesicles to non-uniform AC fields. As expected, the droplets
Studying phosphatidylcholine membranes is motivated by the
move since they are coupled to the membrane flow. Therefore,
fact that phosphatidylcholines are the most abundant lipids
when a cross section of the vesicle is observed with confocal
found in mammalian cells. In order to better mimic biological
microscopy as in Fig. 6, the droplets are observed to come into
membranes, we investigated the behavior of multi-component
focus and to go out of focus again.
vesicles containing a fraction of negatively charged lipids in
Membrane labelling via domains allows visualization of lipid
different medium conditions.66
motion and this approach should be helpful in order to elucidate
Two different types of charged vesicles were used: vesicles
other membrane phenomena such as membrane dynamics during
composed of mixtures of synthetic or natural lipids. In the first
electroformation of vesicles, or in the membrane behavior in
case, palmitoyloleoylphosphatidylcholine (POPC) and palmi-
vesicles subjected to shear flows59 or mechanical stresses.
toyloleoylphosphatidylglycerol (POPG), which is negatively
Furthermore, the AC field-induced flows in the membrane and
charged, were used. In the second case, the vesicles were made of
the interior of the vesicles may find application in microfluidic
lipid extract (LE) from the plasma membrane of red blood cells,
technologies. We have already demonstrated the effectiveness of
which contains approximately 10 mol% anionic lipids, mainly
the membrane flow for lipid mixing.56
phosphatidylserines. When working with charged membranes,
the medium pH and ionic strength are very important, as they
can tune the bilayer electrostatic properties. Thus, three types of
4. Vesicle response to DC pulses
solutions for the vesicle preparation were considered: water,
As discussed in section 3, vesicles exposed to AC fields can adopt 1 mM Hepes buffer (pH 7.4) with 0.1 mM EDTA, and 0.5 mM
stationary shapes. The application of DC pulses induces short- NaCl, which provides the same ionic strength as the buffered

3206 | Soft Matter, 2009, 5, 3201–3212 This journal is ª The Royal Society of Chemistry 2009
 View Article Online

solution. Below, we refer to these solutions as non-buffered, As already mentioned, the ionic strengths of the buffered and
buffered and salt solutions, respectively. To ensure good optical the salt solutions were identical. Then, strictly speaking, the only
contrast, the vesicles in all preparations also contained 0.2 M composition difference between the two solutions is the presence
sucrose inside and isotonic glucose solution outside. of Hepes (1 mM) and EDTA (0.1 mM) in the buffer. To test
Under certain conditions, POPC : POPG mixtures behave in which of the two components was responsible for preventing the
the same way as pure PC vesicles,60 i.e., the pulses induce opening bursting, we prepared giant vesicles (GUVs) composed of 1 : 1,
of macropores with a diameter up to about 10 mm, which reseal POPC : POPG in 1 mM Hepes only as well as in 0.1 mM EDTA
within 50 ms. This behavior was observed for mixed vesicles in only. The experiments show that vesicles burst in the presence of
Published on 30 June 2009. Downloaded by Universitat Autonoma de Barcelona on 28/10/2014 16:30:22.

buffered solutions at all molar ratios used: 9 : 1, 4 : 1 and 1 : 1, 1 mM Hepes (without EDTA). On the other hand, in solutions
and for non-buffered or salt solutions at low POPG content (9 : 1 containing 0.1 mM EDTA the conventional poration–resealing
and 4 : 1). A very surprising response was observed for 1 : 1, behavior was observed. Thus, EDTA is the essential component
POPC : POPG vesicles in non-buffered and salt solutions: they preventing vesicle bursting induced by the electric pulse. EDTA
disintegrated after electroporation;66 see Fig. 7a. Typically, one is a chelating agent, which is generally added in solutions to bind
macropore formed and expanded in the first 50–100 ms at a very possible multivalent ions present as impurities in the solution,
high speed of approximately 1 mm/s. The entire vesicle content is like calcium.70 However, supplementing the 0.1 mM EDTA
released and is seen as darker liquid in Fig. 7a. In order to better vesicle solution with excess of CaCl2 (0.5 mM) to block the
resolve the membrane reorganization after rupture, we used EDTA did not recover the bursting phenomenon.
fluorescent labeling and confocal microscopy, as shown in Plasma membranes should exhibit similar bursting behavior as
Fig. 7b. The bursting was followed by restructuring of the that of the LE vesicles, because their lipid composition is similar.
membrane into what seemed to be interconnected bilayer frag- However, cell membranes are subjected to internal mechanical
ments in the first seconds, and a tether-like structure in the first constraints imposed by the cytoskeleton, which prevents their
minute. Then the membrane stabilized into interconnected disintegration even if their membranes are prone to disruption
micron-sized tubules and small vesicles. These observations when subjected to pulses. Instead, the pores in the cell membrane
suggest that the vesicle bursting and membrane instability is are stable for a long time71 and can either lead to cell death by
related to the large amount of POPG in the bilayer and to the lysis or reseal depending on the media.8,72 The latter is the key to
medium. No vesicle disintegration was observed in buffered efficient electroporation-based protocols for drug or gene
solution and for lower content of POPG. Thus, we considered the transfer in cells. The results reported here suggest that membrane
hypothesis that vesicle bursting and membrane instability is charge as well as minute amounts of molecules such as EDTA
related to the charged state of the bilayer.67–69 might be important but not yet well understood regulating agents
Interestingly, LE vesicles behave in the same way as synthetic in these protocols.
1 : 1, POPC : POPG vesicles. Conventional poration–resealing
was observed in buffered solution, whereas the unusual bursting
4.2 Vesicle behavior in the presence of nanoparticles
occurred in non-buffered and salt solutions. These results suggest
that the bursting is not specific to PG but to the charged state of Gold and silver nanoparticles, as well as quantum dots are
the membrane. The LE membranes contained approximately attractive tools for visualizing processes in cells. One possible
10 mol% anionic lipids, which was enough to induce membrane application involves their employment in optical trapping as
destabilization. In the synthetic membranes 50 mol% of PG was handles for force measurements inside living cells.73,74 Another
needed to lead to the same effect. appealing feature is that magnetic and charged particles can be
The amount of PG in the bilayer is not the only factor trig- manipulated by electromagnetic fields. Thus, we were interested
gering bursting of the synthetic membranes. In particular, vesi- in the response of lipid membranes to electric fields in the pres-
cles with the same high content of PG (50 mol%) do not burst in ence of nanoparticles. For this purpose, we used GUVs made
buffered solution. Even though the main difference between from the conventional lipid egg lecithin (L-a-phosphatidylcho-
buffered solution and the non-buffered and salt solutions seems line), and gold nanoparticles, 80 nm in diameter. The vesicles
to be the pH, significant protonation of PG should occur only for were electroformed in a sucrose solution and subsequently
pH lower than 5.5, which is below the working pH values in this diluted in an isotonic glucose solution containing the particles at
study. Thus, with respect to pH, the solutions are not very a concentration up to 2.2  1010 particles/ml. We applied DC
different. pulses with a duration of 200 ms and a field strength of 3.4 kV/cm.

Fig. 7 Bursting of charged (POPC : POPG, 1 : 1) vesicles subjected to electric pulses. The time after the beginning of the pulse is marked on each image.
(a) Phase contrast microscopy snapshots from fast camera observation of a vesicle in salt solution subjected to a pulse with field strength 1.2 kV/cm and
duration 200 ms. The field direction is indicated in the first snapshot. The vesicle bursts and disintegrates. (b) Confocal cross-sections of a vesicle, which
has been subjected to an electric pulse and has burst and rearranged into a network of tubes and smaller vesicles.

This journal is ª The Royal Society of Chemistry 2009 Soft Matter, 2009, 5, 3201–3212 | 3207
 View Article Online

Fig. 8 Vesicle response to DC pulses in the presence and absence of salt and gold particles. The direction of the field is indicated by the arrow on the left.
Published on 30 June 2009. Downloaded by Universitat Autonoma de Barcelona on 28/10/2014 16:30:22.

The first snapshot (a) shows a vesicle deformation in the absence of salt and particles. In this case, the DC pulse duration is 200 ms and the field strength is
1.4 kV/cm. The applied pulse is sufficient to porate the vesicle as visualized by dark eruptions of sucrose solution leaking out of the vesicle. In the
presence of 0.03 mM salt in the vesicle exterior and no particles present, the vesicles adopt cylindrical shapes as shown in (b). In the latter, the DC pulse
duration is 200 ms and the field strength is of 2 kV/cm. Snapshots (c) to (h) show vesicles deforming in the presence of 80 nm gold particles for a DC pulse
with a duration of 200 ms and a field strength of 3.4 kV/cm. The gold concentration was successively increased from 1.1  108 particles/ml in (c), to 5.5 
108 in (d), 1.1  109 in (e), 2.2  109 in (f), 1.1  1010 in (g), and 2.2  1010 particles/ml in (h). There is a clear concentration dependence of the shape
adopted by the deformed vesicles. Snapshots (a), (c), (e) and (f) were taken 150 ms after the beginning of the pulse and (b), (d), (g) and (h) after 200 ms. All
scale bars correspond to 15 micrometres.

Previous work, where GUVs were exposed to DC pulses, has concentration conditions. This suggests that the deformation
shown cylindrical deformations when salt was present in the mechanism in the presence of gold nanoparticles and salt is the
vesicle exterior;61 note that in the absence of salt in the external same. Indeed, both ions and particles are charged. By measuring
solution, the vesicles deform only into prolates, see Fig. 8a. By the electrophoretic mobility of the gold colloids, we could esti-
applying a DC pulse and systematically varying the concentra- mate their zeta potential to be slightly below 50 mV. This
tion of gold nanoparticles outside the vesicles, we observed very would indicate that the nanoparticles migrate towards the anode
similar morphologies. Since the lifetime of these cylindrical during the DC pulse. In the case of salt, Na+ and Cl move in
deformations is very short, between a few hundred microseconds opposite directions, while the gold colloids move only in one
and a few milliseconds, we used a fast digital camera recording at direction. This might explain the observed asymmetry in the
20 000 frames per second, i.e., an acquisition speed that corre- deformed vesicles, especially at the higher particle concentrations
sponds to one image every 50 ms. where the vesicles adopt a disc-like shape with a trapezoidal cross
By varying the concentration of gold nanoparticles in the section; see Fig. 8g and h. The area of the side of the disc facing
surrounding media, we could influence the shape adopted by the anode seems to be larger than the one facing the cathode.
a vesicle exposed to a DC pulse, as shown in Fig. 8c–h. We As discussed in a previous report,61 one possible explanation
observed an overall elongation or contraction of the GUV in could be that ions or particles flatten the equatorial zone of the
the direction of the electric field. The images in Fig. 8c–h show deformed vesicle. At least during the first part of the pulse there is
how the vesicles respond to an increase in the concentration of an inhomogeneity in the membrane tension due to the fact that
gold nanoparticles. It should be noted that the particles are the electric field is the strongest at the poles of the vesicle, and
only present in the external medium. For the lowest explored almost zero close to the equator. The kinetic energy of the
concentration cmin ¼ 1.1  108 particles/ml, shown in Fig. 8c, accelerated ions hitting the equatorial region of the vesicle is
the vesicles exhibit a similar behavior as in the absence of ions higher than the energy needed to bend the membrane, thus
or particles, compare with the image in Fig. 8a. The vesicles leading to the observed deformation. In addition, particle-driven
elongate only in the direction of the field into a prolate shape. flows may be inducing membrane instability giving rise to higher
By increasing the gold concentration we could observe a flat- order modes of the vesicle shape.75 Yet another possible expla-
tening of the vesicle equatorial region; the vesicles adopt nation may be related to a change in the spontaneous curvature
the shape of a cylinder with rounded caps. This is similar to the of the bilayer due to the particle (or ion) asymmetry across the
vesicle response in the presence of ions; compare with the membrane.76 During the pulse, local and transient accumulation
image in Fig. 8b. The particle concentration influences the type of particles in the membrane vicinity can occur. The mechanism
of the cylindrical deformations observed. At concentrations driving the cylindrical deformations might be a combination of
slightly above cmin, the vesicles adopt tube-like shapes parallel nanoparticle electrophoresis and changes in the membrane
to the direction of the electric field. When the gold concen- spontaneous curvature.
tration was increased tenfold, 10  cmin, coexistence of ‘‘discs’’ The idea that the balance between the particle concentration in
and ‘‘tubes’’ occurred during the DC pulse, some of them the inner and outer media influences the type of deformation is
almost looking ‘‘square’’ (Fig. 8e, f). At even higher particle supported by the observation that repeated exposure of the same
concentrations, the vesicles adopted only a disc-like shape vesicles to many consecutive DC pulses leads to coexistence
(Fig. 8g, h). between ‘‘tubes’’ and ‘‘discs’’. Poration of the lipid membrane is
Parallels can be drawn between the above observations and the frequent at these pulse strengths and durations60 and depends,
prolate and oblate shapes of vesicles subjected to AC fields among other factors, on the vesicle radius and proximity to the
described in section 3.1, but even more so to the shapes adopted electrode, e.g. larger vesicles porate at weaker pulses than smaller
by vesicles subjected to DC pulses in the presence of NaCl.61 The ones. The pulses might induce permeation of gold particles into
overall behavior is the same, disc-like, square-like or tube-like the interior of some of the vesicles, which would explain the
deformation depending on the outer (and inner) salt or particle variation in the cylindrical deformations.

3208 | Soft Matter, 2009, 5, 3201–3212 This journal is ª The Royal Society of Chemistry 2009
 View Article Online

In summary, when subjected to DC pulses, vesicles in the degree of deformation depends on the initial vesicle tension and
presence of nanoparticles respond similarly to vesicles in salt excess area,60 which are both unknown a priori.
solutions.61 The mechanisms behind these responses are still to be The responses of the two vesicles differ significantly. The fluid
clarified, and it remains to be seen whether the processes gov- vesicle gradually deforms and reaches maximum deformation at
erning them are the same. the end of the pulse. The gel-phase vesicle responds significantly
faster, and exhibits a relaxation with a decay time of about 30 ms
during the pulse. To our knowledge, such intra-pulse relaxation
has not been previously reported. The vesicles had similar size
Published on 30 June 2009. Downloaded by Universitat Autonoma de Barcelona on 28/10/2014 16:30:22.

4.3 Electrodeformation and poration of vesicles in the gel-phase

and were both in salt-free solutions with conductivity of about
In the previous two sections, the response of membranes in the 1 mS/m. For a fluid vesicle with radius R ¼ 20 mm, the charging
fluid state has been discussed. The mechanical and rheological time is tc y 550 ms, see eqn (2), i.e., longer than the pulse
properties of such membranes differ significantly from those of duration. Gel-phase membranes are thicker, and thus, have
membranes in the gel-phase; for a concise comparison see lower membrane capacitance,83 leading to charging times shorter
Dimova et al.77 For example, the bending stiffness and the shear or comparable to the duration of the pulse. The faster response of
surface viscosity of gel-phase membranes are orders of magni- the gel-phase vesicle as shown in Fig. 9 correlates with the shorter
tude higher than those of membranes in the fluid phase,78–81 and charging time as compared to the fluid vesicle.
membranes in the gel-phase are thicker.82 These differences After the end of the pulse, the relaxation of the gel DPPC
introduce new features in the response of gel-phase membranes vesicle is also much faster than that of the fluid membrane. The
to electric fields, which we discuss next. relaxation behavior depends on whether the membrane was
We studied POPC and dipalmitoylphosphatidylcholine porated or not.60 In the example given in Fig. 9, no microscopic
(DPPC) membranes, which undergo their main transition at 2 pores were detected, but it is plausible that in the gel-phase

C and 41.6  C, respectively. We compared the response to vesicle pores with sizes in the sub-optical range were formed
square wave DC pulses of a POPC vesicle, which at room during the pulse. The formation of such pores may explain the
temperature is in the fluid phase, with the response of a vesicle intra-pulse relaxation in the vesicle deformation.
made of DPPC, which is in the gel-phase. The applied DC pulses If DC pulses of field strength larger than the discussed above
were weak enough not to induce formation of microscopic pores are applied, the gel-phase vesicles rupture: the pores resemble
in the membranes and no leakage of the internal sucrose solution micrometre-sized cracks on a solid shell.32 Contrary to pores in
outside the vesicle was observed. Fig. 9 shows the deformation of fluid membranes, which reseal within tens of milliseconds,60 the
one POPC and one DPPC vesicle in response to DC pulses 300 ms cracks in gel-phase vesicles are stable and seem not to reseal.
long. The deformation is characterized by the ratio of the two Understanding the response of the gel-phase membranes will
semiaxes, a/b, of the vesicles. To achieve similar maximal degree require thorough consideration of the membrane mechanical and
of deformation, stronger pulses had to be applied to the gel- rheological properties as well as the interaction of electric fields
phase vesicle as compared to the fluid one. Pulses with field with such membranes. Both the intra- and after-pulse relaxations
strength about 1 kV/cm produce deformations in gel-phase of the vesicles in gel-phase are poorly understood and will be the
vesicles, which are not detectable optically, while strong pulses object of further investigation.
about 5 kV/cm applied to the fluid-phase vesicles cause poration.
The latter influences the relaxation dynamics.60 Note that the 4.4 Vesicle electrofusion as a method for nanoparticle synthesis
in vesicles
Strong electric pulses induce electrical breakdown of fluid lipid
bilayers leading to formation of transient pores. The vesicles
become permeable for a certain time. When two such porated
vesicles are in close contact, fusion can occur. The concept to fuse
two GUVs in order to initiate content mixing reactions has been
proposed earlier.84,85 In this section, we introduce the application
of fusion of giant vesicles for the synthesis of nanoparticles in
closed compartments.
The principle of fusion-mediated synthesis is simple: the
starting reagents are separately loaded into different vesicles, and
then the reaction is triggered by the fusion of these vesicles, which
allows the mixing of their contents. The success of this approach
is guaranteed by two important factors. First, the lipid
membrane is impermeable to the reactants such as ions or
macromolecules. Second, fusion can be initiated by a variety of
Fig. 9 Deformation response of a gel-phase DPPC vesicle with a radius fusogens such as membrane stress,86,87 ions or synthetic fusogenic
of 22 mm, and a fluid phase POPC vesicle with a radius of 20 mm to DC molecules,62,88–90 fusion proteins,91 laser beam radiation,85 or
pulses with a duration of 300 ms. The pulse duration is indicated by the electric fields.63,92 Among the fusion methods listed above, elec-
shaded zone. The field strength of the pulses was 5 kV/cm and 0.8 kV/cm trofusion becomes increasingly important because of its reliable,
for the DPPC and the POPC vesicle, respectively. fast and easy handling.63 An immediate benefit of this strategy is

This journal is ª The Royal Society of Chemistry 2009 Soft Matter, 2009, 5, 3201–3212 | 3209
Published on 30 June 2009. Downloaded by Universitat Autonoma de Barcelona on 28/10/2014 16:30:22. View Article Online

Fig. 10 Electrofusion of vesicles as a method for nanoparticle synthesis: vesicles containing Na2S and fluorescently labelled in red, and vesicles con-
taining CdCl2, labelled in green, are mixed in Na2S- and CdCl2-free environment and subjected to an AC field to align them in the direction of the field
and bring them close together. A DC pulse initiates the electrofusion of the two vesicles and the reaction between Na2S and CdCl2 proceeds to the
formation of CdS nanoparticles encapsulated in the fused vesicle. The last snapshot is a confocal scan of a vesicle obtained by fusion of a vesicle loaded
with 0.3 mM Na2S (red part of the fused vesicle) and a vesicle loaded with 0.3 mM CdCl2 (labelled in green). The fluorescence signal from the synthesized
CdS nanoparticles in the vesicle interior is visible as indicated by the arrow.94

that the precise temporal and spatial control on the synthesis In general, vesicle fusion provides many unexplored oppor-
process can be easily achieved. tunities for protein biosynthesis, enzyme-catalyzed reactions,
According to our electrofusion protocol, two vesicle pop- and biomineralization processes.101
ulations are mixed, one loaded with Na2S and labeled with one
fluorescent dye (red), the other loaded with CdCl2 and labeled
differently (green). The vesicle external media is Na2S- or CdCl2- 5. Conclusions
free, which can be achieved either by significant dilution of the
The results reported in this review demonstrate that cell-sized
starting vesicle solutions or by exposure to ion-exchange resins.
giant vesicles provide a very useful model for resolving the effect
Application of AC field aligns the vesicles in the direction of the
of electric fields on lipid membranes because vesicle dynamics
field due to dielectric screening, similarly to pearl-chain forma-
can be directly observed with optical microscopy. We have
tion in suspensions of cells.15 In order to monitor the nano-
examined the behavior of giant vesicles exposed to AC fields of
particle formation process, we locate a red-and-green vesicle
various frequencies and elucidated the underlying physical
couple (approximately half of the couples fall in this category)
mechanism for the vesicle deformations as well as stress-induced
and apply a DC pulse strong and long enough to porate each of
lipid flows in inhomogeneous AC fields. We have shown that the
the vesicles. For egg lecithin vesicles, pulses of 0.5–2 kV/cm field
vesicle response to electric fields can be exploited to evaluate the
strength and 150–300 ms duration are sufficient. The steps of this
mechanical properties of the membrane.
protocol are schematically illustrated in Fig. 10.
Until recently, the dynamics of vesicle relaxation and poration,
Fluorescence in the interior of the fused vesicle was observed,
which occur on microsecond time-scales, has eluded direct
see Fig. 10, which indicates formation of CdS nanoparticles.
observation because the temporal resolution of optical micros-
Fluorescence in the wavelength range between 400 and 800 nm has
copy observations with analog video technology is in the range of
been previously reported for CdS particles with diameters in the
milliseconds. We used fast digital imaging to discover new
range 1–25 nm.93 Because the confocal sections show only fluo-
features in the membrane response arising from the presence of
rescence from a thin slice of the vesicle, out of focus fluorescence,
charged lipids in the membrane, nanoparticles in the surrounding
which might be emitted from the upper and lower part of the
media, and compared the response of gel-phase membranes to
vesicle, is not detected. The obtained product was also investigated
fluid ones. Finally, we introduced a novel application of
using transmission electron microscopy and selected area electron
membrane electrofusion, which allowed us to perform nano-
diffraction, which showed the presence of dispersed nanoparticles
particle synthesis in vesicles.
of diameters ranging between 4 and 8 nm.94 A noticeable advance
In conclusion, the reported observations demonstrate that
of the above approach is that the whole reaction could be viewed
giant vesicles can help advance fundamental knowledge about
and monitored in real time under the optical microscope.
the complex behavior of cells and membranes in electric fields
Cells and microorganisms are able to synthesize inorganic
and can inspire novel practical applications.
nanoparticles.95–97 The tentative interpretation of this observa-
tion is related to the involvement of specific molecules such as
inorganic-binding peptides.98–100 Our experiments suggest that
Acknowledgements
nanoparticles could be synthesized in biological compartments
even without the mediation of biomacromolecules. For example, We thank Yanhong Li for the help with the experiments on
the fusion of small vesicles with the cell membranes could be vesicles loaded with two-phase systems, Andrew Richardson for
a possible mechanism for the cell-based synthesis of nano- the experiments with gold nanoparticles, Said Aranda and
particles. The necessary condition according to such a scenario is Ruben S. Gracia for the data acquisition on vesicles in AC fields,
that the vesicles are loaded with one reagent, while the local and Carmen Remde for the technical support. We acknowledge
concentration of the other chemical at the cell is suitably the financial support of the German Research Foundation
matched. Low concentrations in the submillimolar range are (Deutsche Forschungsgemeinschaft), the Max Planck Society,
sufficient to produce CdS nanoparticles.94 and FAPESP.

3210 | Soft Matter, 2009, 5, 3201–3212 This journal is ª The Royal Society of Chemistry 2009
 View Article Online

References 37 A. W. Friend Jr., E. D. Finch and H. P. Schwan, Science, 1975, 187,

357–359.
1 Electromanipulation of cells, ed. U. Zimmermann and G. A. Neil, 38 F. J. Iglesias, M. C. Lopes, C. Santamaria and A. Dominguez,
CRC Press, Boca Raton, 1996. Biophys. J., 1985, 48, 712–726.
2 Electroporation and electrofusion in cell biology, ed. E. Neumann, A. 39 M. I. Angelova and D. S. Dimitrov, Faraday Discuss. Chem. Soc.,
E. Sowers and C. A. Jordan, Plenum Press, New York, 1989. 1986, 81, 303–311.
3 Guide to electroporation and electrofusion, ed. D. C. Chang, B. M. 40 M. I. Angelova, S. Soleau, P. Meleard, J.-F. Faucon and P. Bothorel,
Chassey, J. A. Saunders and A. E. Sowers, Academic Press, New Prog. Colloid Polym. Sci., 1992, 89, 127–131.
York 1992. 41 D. J. Estes and M. Mayer, Biochim. Biophys. Acta, 2005, 1712,
4 R. H. W. Funk and T. K. Monsees, Cells Tissues Organs, 2006, 182, 152–160.
Published on 30 June 2009. Downloaded by Universitat Autonoma de Barcelona on 28/10/2014 16:30:22.

59–78. 42 T. Pott, H. Bouvrais and P. Meleard, Chem. Phys. Lipids, 2008, 154,
5 J. Voldman, Annu. Rev. Biomed. Eng., 2006, 8, 425–454. 115–119.
6 M. Zhao, B. Song, J. Pu, T. Wada, B. Reid, G. P. Tai, F. Wang, 43 L. R. Montes, A. Alonso, F. M. Goni and L. A. Bagatolli, Biophys.
A. H. Guo, P. Walczysko, Y. Gu, T. Sasaki, A. Suzuki, J., 2007, 93, 3548–3554.
J. V. Forrester, H. R. Bourne, P. N. Devreotes, C. D. McCaig and 44 P. Taylor, C. Xu, P. D. I. Fletcher and V. N. Paunov, Phys. Chem.
J. M. Penninger, Nature, 2006, 442, 457–460. Chem. Phys., 2003, 1732–1733.
7 J. Teissie, M. Golzio and M. P. Rols, Biochim. Biophys. Acta, 2005, 45 P. Taylor, C. Xu, P. D. I. Fletcher and V. N. Paunov, Chem.
1724, 270–280. Commun., 2003, 5, 4918–4922.
8 E. Tekle, R. D. Astumian and P. B. Chock, Proc. Natl. Acad. Sci. 46 M. I. Angelova, in Giant Vesicles, ed. P. L. Luisi and P. Walde, John
U.S. A., 1994, 91, 11512–11516. Wiley & Sons Ltd., Chichester, 2000, ch. 3, pp. 27–43.
9 R. Heller, R. Gilbert and M. J. Jaroszeski, Adv. Drug Deliv. Rev., 47 P. Sens and H. Isambert, Phys. Rev. Lett., 2002, 88, 128102.
1999, 35, 119–129. 48 S. Aranda, K. A. Riske, R. Lipowsky and R. Dimova, Biophys. J.,
10 A. Gothelf, L. M. Mir and J. Gehl, Cancer Treat. Rev., 2003, 29, 2008, 95, L19–L21.
371–387. 49 P. M. Vlahovska, R. S. Gracia, S. Aranda and R. Dimova, Biophys.
11 R. Nuccitelli, U. Pliquett, X. Chen, W. Ford, R. J. Swanson, J., 2009, 96, 4789–4803.
S. J. Beebe, J. F. Kolb and K. H. Schoenbach, Biochem. Biophys. 50 W. Helfrich and R. M. Servuss, Nuovo Cimento D, 1984, 3, 137–151.
Res. Commun., 2006, 343, 351–360. 51 E. Evans and W. Rawicz, Phys. Rev. Lett., 1990, 64, 2094–2097.
12 L. M. Mir, P. Devauchelle, F. Quintin-Colonna, F. Delisle, 52 J. J. Pan, T. T. Mills, S. Tristram-Nagle and J. F. Nagle, Phys. Rev.
S. Doliger, D. Fradelizi, J. Belehradek Jr and S. Orlowski, Brit. Lett., 2008, 100, 198103.
J. Cancer, 1997, 76, 1617–1622. 53 G. I. Taylor, Proc. R. Soc. London, Ser. A, 1966, 291, 159–166.
13 M. Golzio, J. Teissie and M. P. Rols, Proc. Natl. Acad. Sci. U. S. A., 54 D. Lacoste, G. I. Menon, M. Z. Bazant and J. F. Joanny, Eur. Phys.
2002, 99, 1292–1297. J. E, 2009, 28, 243–264.
14 M. Golzio, M. P. Rols and J. Teissie, Methods, 2004, 33, 126–135. 55 S. Veatch and S. Keller, Phys. Rev. Lett., 2002, 89, 268101.
15 U. Zimmermann, Rev. Physiol. Biochem. Pharmacol., 1986, 105, 56 M. Staykova, R. Lipowsky and R. Dimova, Soft Matter, 2008, 4,
176–256. 2168–2171.
16 J. Teissie and T. Y. Tsong, Biochemistry, 1981, 20, 1548–1554. 57 M. Helfrich, L. K. Mangeney-Slavin, M. S. Long, K. Djoko and
17 R. W. Glaser, S. L. Leikin, L. V. Chernomordik, V. F. Pastushenko C. D. Keating, J. Am. Chem. Soc., 2002, 124, 13374–13375.
and A. I. Sokirko, Biochim. Biophys. Acta, 1988, 940, 275–287. 58 Y. Li, R. Lipowsky and R. Dimova, J. Am. Chem. Soc., 2008, 130,
18 S. Kakorin, T. Liese and E. Neumann, J. Phys. Chem. B., 2003, 107, 12252–12253.
10243–10251. 59 M. Abkarian and A. Viallat, Soft Matter, 2008, 4, 653–657.
19 D. Needham and R. M. Hochmuth, Biophys. J., 1989, 55, 60 K. A. Riske and R. Dimova, Biophys J., 2005, 88, 1143–1155.
1001–1009. 61 K. A. Riske and R. Dimova, Biophys. J., 2006, 91, 1778–1786.
20 D. V. Zhelev and D. Needham, Biochim. Biophys. Acta, 1993, 1147, 62 C. K. Haluska, K. A. Riske, V. M. Artzner, J. M. Lehn, R. Lipowsky
89–104. and R. Dimova, Proc. Natl. Acad. Sci. U. S. A., 2006, 103,
21 O. Sandre, L. Moreaux and F. Brochard-Wyart, Proc. Natl. Acad. 15841–15846.
Sci. U. S. A., 1999, 96, 10591–10596. 63 K. A. Riske, N. Bezlyepkina, R. Lipowsky and R. Dimova, Biophys.
22 E. Tekle, R. D. Astumian, W. A. Friauf and P. B. Chock, Biophys. Rev. Lett., 2006, 4, 387–400.
J., 2001, 81, 960–968. 64 I. G. Abidor, V. B. Arakelyan, L. V. Chernomordik,
23 C.-H. Lee, Y.-F. Chang, C.-H. Tsai and P.-H. Wang, Langmuir, Y. A. Chizmadzhev, V. F. Pastushenko and M. R. Tarasevich,
2005, 21, 7186–7190. J. Electroanal. Chem., 1979, 104, 37–52.
24 M. Winterhalter and W. Helfrich, J. Colloid Interface Sci., 1988, 122, 65 K. Olbrich, W. Rawicz, D. Needham and E. Evans, Biophys. J.,
583–586. 2000, 79, 321–327.
25 P. Peterlin, S. Svetina and B. Žek&sacute, J. Phys.: Condens. 66 K. A. Riske, R. L. Knorr and R. Dimova, Soft Matter, 2009, 5,
Matter., 2007, 19, 136220. 1983–1986.
26 H. Engelhardt, H. Gaub and E. Sackman, Nature, 1984, 307, 67 M. D. Betterton and M. P. Brenner, Phys. Rev. Lett., 1999, 82,
378–380. 1598–1601.
27 M. Mitov, P. Meleard, M. Winterhalter, M. I. Angelova and 68 V. Kumaran, Phys. Rev. Lett., 2000, 85, 4996–4999.
P. Bothorel, Phys. Rev. E., 1993, 48, 628–631. 69 H. Isambert, Phys. Rev. Lett, 1998, 80, 3404–3407.
28 M. Kummrow and W. Helfrich, Phys. Rev. A, 1991, 44, 70 K. A. Riske, H. G. Dobereiner and M. T. Lamy-Freund, J. Phys.
8356–8360. Chem. B, 2003, 107, 5391–5392.
29 W. Harbich and W. Helfrich, Z. Naturforsch. A, 1979, 34, 71 K. Schwister and B. Deuticke, Biochim. Biophys. Acta, 1985, 816,
1063–1065. 332–348.
30 G. Niggemann, M. Kummrow and W. Helfrich, J. Phys. II, 1995, 5, 72 K. Kinosita and T. Y. Tsong, Nature, 1977, 268, 438–441.
413–425. 73 P. M. Hansen, V. K. Bhatia, N. Harrit and L. Oddershede, Nano
31 S. Lecuyer, W. Ristenpart, O. Vincent and H. Stone, Appl. Phys. Lett., 2005, 5, 1937–1942.
Lett., 2008, 92, 104105. 74 L. Jauffred, A. C. Richardson and L. B. Oddershede, Nano Lett.,
32 R. Dimova, K. A. Riske, S. Aranda, N. Bezlyepkina, R. L. Knorr 2008, 8, 3376–3380.
and R. Lipowsky, Soft Matter, 2007, 3, 817–827. 75 V. Kantsler, E. Segre and V. Steinberg, Phys. Rev. Lett., 2007, 99,
33 T. B. Jones, Electromechanics of particles, Cambridge University 178102.
Press, New York, 1995. 76 R. Lipowsky and H.-G. Dobereiner, Europhys. Lett., 1998, 43,
34 H. P. Schwan, in Interaction between EM fields and cells, ed. A. 219–225.
Chiabrera, C. Nivolin and H. P. Schwan, Plenum Press, 1985, pp. 77 R. Dimova, S. Aranda, N. Bezlyepkina, V. Nikolov, K. A. Riske
75–97. and R. Lipowsky, J. Phys.: Condens. Matter, 2006, 18,
35 C. Grosse and H. P. Schwan, Biophys. J., 1992, 63, 1632–1642. S1151–S1176.
36 H. Engelhardt and E. Sackman, Biophys. J., 1988, 54, 495–508. 78 T. Heimburg, Biochim. Biophys. Acta, 1998, 1415, 147–162.

This journal is ª The Royal Society of Chemistry 2009 Soft Matter, 2009, 5, 3201–3212 | 3211
 View Article Online

79 R. Dimova, B. Pouligny and C. Dietrich, Biophys. J., 2000, 79, 340– 91 R. Jahn, T. Lang and T. C. S€ udhof, Cell, 2003, 112, 519–533.
356. 92 A. Str€
omberg, F. Ryttsen, D. T. Chiu, M. Davidson, P. S. Eriksson,
80 C.-H. Lee, W.-C. Lin and J. Wang, Phys. Rev. E, 2001, 64, 020901. C. F. Wilson, O. Orwar and R. N. Zare, Proc. Natl. Acad. Sci.
81 K. R. Mecke, T. Charitat and F. Graner, Langmuir, 2003, 19, 2080– U. S. A., 2000, 97, 7–11.
2087. 93 H. Weller, Angew. Chem., 1993, 105, 43–55.
82 J. F. Nagle and S. Tristram-Nagle, Biophys. Biochem. Acta, 2000, 94 P. Yang, R. Lipowsky and R. Dimova, Small, DOI: 10.1002/
1469, 159–195. smll.200900560.
83 M. A. Stuchly, S. S. Stuchly, R. P. Liburdy and D. A. Rousseau, 95 D. Bhattacharya and R. K. Gupta, Crit. Rev. Biotechnol., 2005, 25,
Phys. Med. Biol., 1988, 33, 1309–1324. 199–204.
84 D. T. Chiu, C. F. Wilson, F. Ryttsen, A. Stromberg, C. Farre, 96 C. Sanchez, H. Arribart, M. Madeleine and G. Guille, Nat. Mater.,
Published on 30 June 2009. Downloaded by Universitat Autonoma de Barcelona on 28/10/2014 16:30:22.

A. Karlsson, S. Nordholm, A. Gaggar, B. P. Modi, A. Moscho, 2005, 4, 277–288.

R. A. Garza-Lopez, O. Orwar and R. N. Zare, Science, 1999, 283, 97 D. Mandal, M. E. Bolander, D. Mukhopadhyay, G. Sarkar
1892–1895. and P. Mukherjee, Appl. Microbiol. Biotechnol., 2006, 69,
85 S. Kulin, R. Kishore, K. Helmerson and L. Locascio, Langmuir, 485–492.
2003, 19, 8206–8210. 98 N. Kr€oger, M. B. Dickerson, G. Ahmad, Y. Cai, M. S. Haluska,
86 F. S. Cohen, M. H. Akabas and A. Finkelstein, Science, 1982, 217, K. H. Sandhage, N. Poulsen and V. C. Sheppard, Angew. Chem.,
458–460. 2006, 118, 7397–7401.
87 J. C. Shillcock and R. Lipowsky, Nat. Mater., 2005, 4, 225–228. 99 N. Kr€oger, M. B. Dickerson, G. Ahmad, Y. Cai, M. S. Haluska,
88 D. J. Estes, S. R. Lopez, A. O. Fuller and M. Mayer, Biophys. J., K. H. Sandhage, N. Poulsen and V. C. Sheppard, Angew. Chem.,
2006, 91, 233–243. Int. Ed., 2006, 45, 7239–7243.
89 B. R. Lentz, Eur. Biophys. J., 2007, 36, 315–326. 100 R. R. Naik, S. J. Stringer, G. Agarwal, S. E. Jones and M. O. Stone,
90 M. Kunishima, M. Tokaji, K. Matsuoka, J. Nishida, M. Kanamori, Nat. Mater., 2002, 1, 169–172.
K. Hioki and S. Tani, J. Am. Chem. Soc., 2006, 128, 14452–14453. 101 P. Walde and S. Ichikawa, Biomol. Eng., 2001, 18, 143–177.

3212 | Soft Matter, 2009, 5, 3201–3212 This journal is ª The Royal Society of Chemistry 2009