Journal of Biomaterials and Nanobiotechnology, 2011, 2, 527-532
doi:10.4236/jbnb.2011.225063 Published Online December 2011 (http://www.scirp.org/journal/jbnb)
527
Targeted Macromolecules Delivery by Large
Lipidic Nanovesicles Electrofusion with
Mammalian Cells
Pascal Demange1,2, Valérie Réat1,2, Stefan Weinandy1,2, Remy Ospital1,2, Louise Chopinet-Mayeux1,2,
Pauline Henri1,2, Alain Milon1,2, Justin Teissié1,2*
1Centre National de la Recherche Scientifique (CNRS), Institut de Pharmacologie et de Biologie Structurale (IPBS), Toulouse,
France; 2Université de Toulouse, Université Paul Sabatier (UPS), IPBS, Toulouse, France.
Received October 3rd, 2011; revised November 4th, 2011; accepted November 14th, 2011.
ABSTRACT
Lipidic nanovesicles (so called liposomes) were one the earliest forms of nanovectors. One of their limits was our lack
of knowledge on the delivery pathway of their content to the target cell cytoplasm. The present communication de-
scribes an efficient way to enhance the delivery. Pulsed electric fields (PEF) are known since the early 80’s to mediate
a fusogenic state of plasma membranes when applied to a cell suspension or a tissue. Polykaryons are detected when
PEF are applied on cells in contact during or after the pulses. Heterofusion can be obtained when a cell mixture is
pulsed. When lipidic nanovesicles, either small unilamellar vesicles (SUVs) or large unilamellar vesicles (LUVs), are
electrostatically brought in contact with electropermeabilized cells by a salt bridge, their content is delivered into the
cytoplasm in electropermeabilized cells. The PEF parameters are selected to affect specifically the cells leaving the
vesicles unaffected. It is the electropermeabilized state of the cell membrane that is the trigger of the merging between
the plasma membrane and the lipid bilayer. The present investigation shows that the transfer of macromolecules can be
obtained; i.e. 20 kD dextrans can be easily transferred while a direct transfer does not take place under the same elec-
trical parameters. Cell viability was not affected by the treatment. As delivery is present only on electropermeabilized
cells, a targeting of the effect is obtained in the volume where the PEF parameters are over the critical value for elec-
tropermeabilization. A homogeneous cytoplasm labeling is observed under digitised videomicroscopy. The process is a
content and “membrane” mixing, following neither a kiss and run or an endocytotic pathway.
Keywords: Electrofusion, Delivery Systems in Cancer, Liposomes, Lipidic Nanovesicles
1. Introduction
New methods for large molecules delivery in cells re-
main a major problem in pharmacology. The main prob-
lem is to provide a pathway across the cell membrane
with no effect on the cell viability. Using liposomes (the
magic bullet) to target the delivery was a promising ap-
proach but it faces the limit of a poor transfer. Pulsed
electric fields (PEF) are known since the early 80’s to
mediate a fusogenic state of plasma membranes when
applied to a cell suspension or a tissue [1-3]. Polykaryons
are detected when PEF are applied on cells in contact
during or after the pulses [4]. Heterofusion can be ob-
tained when a cell mixture is pulsed. But the two partners
should have a similar size if the mixture is pulsed due to
the fact that the PEF effect is strongly size dependent.
Nevertheless we reported that when lipidic nanovesicles
(SUVs or LUVs) are electrostatically brought in contact
with electropermeabilized cells by a salt bridge, their
content was delivered in the cytoplasm [5]. This was
experimentally directly detected by the transfer of a
small soluble fluorescent dye, Pyranin, trapped in the
nanovesicles to the cell cytoplasm. A homogeneous cy-
toplasm labeling is observed under digitized videomi-
croscopy. The PEF parameters were adjusted to affect
specifically the cells leaving the vesicles unaffected by a
direct effect of the PEF. Under these specific experimen-
tal conditions, the nanovesicles were too small to be af-
fected by the field and thus only the electropermeabilized
state of the cell membrane was the trigger of the merging
between the plasma membrane and the lipid bilayer. No

Targeted Macromolecules Delivery by Large Lipidic Nanovesicles Electrofusion with Mammalian Cells 528
deleterious by-effect of PEF was detected on the cell
viability. The biophysical processes supporting the merg-
ing between intact nanovesicles and electropermeabilized
(fusogenic) cells is still under investigation. Membrane
fusion is a key process in life. It occurs when two mem-
branes, brought in close contact, merge in a single one. A
close approach of two membranes (as needed for fusion
to occur) is prevented by repulsive forces [6-8]. In the
case of synaptic vesicles, SNAREs bring the contact but
the following steps where a transient Ca2+ interfacial in-
crease occurs, are unclear [9,10]. The means, by which
the repulsive forces between the two partners are over-
come, remain poorly understood. But clearly a membrane
destabilization should occur [11]. It may result from
structural alterations of the target (plasma) membrane
under a transient electric stress [12].
Taking into account, on one hand that membranes are
electrically charged interfaces and on the other hand that
electrical modulations bring cell membranes to a destabi-
lized but spontaneously competent for fusion state, the
hypothesis of Rosenheck [13] that a rate limiting step in
membrane coalescence is due to the contact of electri-
cally destabilized membranes with small vesicles, can be
relevant of our observation. The model of Rosenheck
was recently extended to vesicles [14]. If electrofusion
between cells and lipid vesicles obeys the Rosenheck’s
model, then macromolecules transfer should be obtained
as in the case of exocytosis. This was checked in the
present study. Due to their large inner volume, large
unilamellar vesicles (LUV) were selected as nanovesicles
to give a significant delivery to the cytoplasm. Fluores-
cent dextrans were selected to monitor the resulting
transfer by digitized fluorescent microscopy and to valid
the proof of concept.
2. Materials and Methods
2.1. Cells
The wild type Toronto cells (WTT), is derived from
Chinese hamster ovary cells (CHOs) and was first intro-
duced in the 1960s. It can grow in suspension and in cul-
ture flasks (generation time = 18 - 20 hours).
The property to grow in suspension prevents the ne-
cessity of trypsinisation.
MEM 0111 buffer (Eurobio France (ref: CM1MEM-
40K-BP)) with Foetal Calf Serum 8%, (SVF EUROBIO,
ref: CVFSVF00-01, lot no: S155839), D (+) – Glucose
45%, (3.5 g/l) (Sigma, USA), Tryptose phosphate (2.95
g/l), Vitamins (GIBCO, ref: 043-01040) and Antibiotics
(penicillin 100 units/ml, streptomycin 100 mg/ml, L-
glutamin 0.58 mg/ml) is used to cultivate the CHO cells
under slow agitation (70 to 100 rpm, 37˚C). Cells stay in
the exponential growth phase by a control of the cell
number (dilution from 0.55 up to 0.7 × 106 cells every
day).
2.2. Vesicles
The lipids were Cholesterol (362794, SIGMA), L-α-
Phosphatidylserine (840032, AVANTI), 1,2-Dioleoyl-sn-
Glycero-3-Phosphocholine (DOPC, 850375, AVANTI).
8.5 mg of PC/PS/Cholesterol lipid mixture (6/1/3 mo-
lar ratio) was dissolved in chloroform. Lipids were dried
under nitrogen flow and vacuum (30 min). They were
then resuspended in 5 mM Hepes (pH 7.2) containing
Fluorescein labeled 20 kD dextrans (labeled-FD20) and
vortexed to form large multilamellar vesicles (LMV).
The temperature of HEPES buffer was kept above gel-
liquid crystal transition temperature of the lipid mixture
(i.e. −20˚C for DOPC). The suspension was sonified (3
times 5 minutes, with 3 min break between each treat-
ment) to obtain smaller unilamellar vesicles (SUV), then
centrifuged at 13,000 g to remove of the titanium debris
which came from the sonication tip. A 5 times freez-
ing/thawing process followed to form bigger vesicles.
Large unilamellar vesicles were obtained by extrusion
through 0.2 μm pore polycarbonate membrane (The
Mini-Extruder (AVANTI)). The size homogeneity was
checked by dynamic light scattering (Proteinsolution
Dynapro)
2.3. Electropulsation
Cells were suspended in a 10 mM HEPES buffer (pH
7.2), 250 mM sucrose and 5 mM CaCl2 at a high cell
density (4 × 107 c/ml). The “pulsing” buffer contained
CaCl2 to ensure the connection between cells and lipo-
somes via electrostatic interaction. Low conductivity was
important to decrease the Joule effect. A unipolar GHT
1287 generator (Jouan, France) was used to provide rec-
tangular pulses as high as 1 kV with a constant intensity
over the 100 μs pulse duration. An oscilloscope (Metrix
OX 520 B, France) ensured the monitoring of the pulse.
The generator was connected to two flat parallel stainless
steel electrodes with an inter-electrode distance of 4 mm.
The edges of the electrodes were in contact with the bot-
tom of a plastic Petri dish (Nunc 35 × 10) to build an
open pulsing chamber. Cells (0.1 ml at 4 × 107 c/ml)
were put between the two electrodes, together with a
labelled-Dextran (FD20) loaded LUV suspension and/or
HEPES buffer. 10 or more pulses (480 V = 1.2 kV/cm,
100 μs, frequency 1 Hz) were applied. After pulse deliv-
ery cells were kept 10 minutes at 25˚C.
A “washing” PBS buffer (Dulbecco’s phosphate buff-
ered saline, Eurobio, France, pH 7.2, -Ca2+, -Mg2+) was
added. It contained phosphate ions to chelate calcium and
remove the electrostatic bridges. The unfusioned nano-
vesicles were not bound anymore to the surface of the

Targeted Macromolecules Delivery by Large Lipidic Nanovesicles Electrofusion with Mammalian Cells529
cells and could be washed out by centrifugation (5 min,
100× g).
The cell pellet was resuspended once again in PBS and
analysed.
2.4. Analysis
Cells were analysed with the fluorescence digitized mi-
croscope Leica DM IRB (Wetzlar, Germany) using the
filter set for fluorescein. Video monitoring was possible
with a cooled CCD camera (Princeton Instruments, NJ,
USA). The pictures were taken with the Metavue soft-
ware (Molecular Devices, USA). Images were analysed
with ImageJ (for more information http://rsb.info.nih.gov
/ij/index.html). No filtering was operated on the raw data.
A ROI was selected and was analysed by the “surface
plot” operation.
3. Results
3.1. LUV Characterization
Fluorescein labeled 20 kD dextrans were used to make
sure that the electric treatment was not just inducing an
exchange (by a “kiss and run” process) of small mole-
cules as previously shown with Pyranin [5].
To eliminate the labeled 20 kD dextrans that was not
trapped in the vesicles, the extrusion method was fol-
lowed by a G200 chromatography. The elution profiles
are shown in Figure 1. LUVs were in the excluded frac-
tion volume (fraction 10 to 15). Strong fluorescein ab-
sorption was detected on a broad range of fractions,
starting from fraction 16. It was due to the elution of free
labeled 20 kD dextrans (untrapped in vesicles).
Furthermore, one single maximum in fluorescein ab-
sorption was detected suggesting homogeneity of the
fluorescent specie. This was indicative that no free fluo-
rescein (without dextran, i.e. with a low molecular
weight) was trapped in the LUVs.
LUVs were used after the gel chromatography.
Figure 1. Elution profiles at 400 and 480 nm Extruded LUV
solutions were analysed by gel chromatography. “Absorp-
tion” at 400 nm is due to the scattering of light by LUVs.
Absorption at 480 nm is due to fluorescein bound to dex-
trans.
3.2. Fusion
Labeled 20 kD dextrans could not be directly electro-
loaded in CHO cells under the electro pulsing conditions
of 10 × 0.1 ms [15].
The cells and LUVs (3/1, V/V) were premixed in
HEPES 10 mM, pH 7.2, sucrose 0.25 M, CaCl2 5 mM
(“pulsing” buffer). An electrostatic bridge bound the
negatively charged liposomes to the cells. No pulse was
applied. After 10 min incubation, the mixture was
washed in PBS buffer to chelate Ca2+ ions and remove
the free liposomes. Cells were then incubated at 37˚C up
to 4 hours. As shown in Figure 2, no evidence of fusion
was obtained as no fluorescence could be detected in the
cytoplasm. No endocytotic process was indeed detected
as no fluorescent spot could be detected even with the
very sensitive detection obtained with the cooled CCD
camera and the image processing. In order to get elec-
trofusion, a similar protocol was followed but 10 or 20
square pulses (1.2 kV/cm, 0.1 ms, 1 Hz) were applied on
the cell/LUVs mixture in the “pulsing” buffer. Fusion
was then detected by the content mixing making the cell
cytoplasm fluorescent as observed under a digitized
fluorescent microscope (Figure 3).
For both experiments, fluorescence was present in a
rather homogeneous pattern (an egg shell surface profile
was observed). The roughness of the distribution (Fig-
ures 3(a) and 3(b) surface plot) can be just relevant of
the noise of the signal as the magnitude of the “hot” spots
Contrast Fluorescence
Figure 2. No spontaneous fusion is present between cells
and LUVs Cells and LUVs were just mixed in the pulsing
buffer. No fusion with LUVs was observed without elec-
tropermeabilization as no fluorescence was detected in the
cytoplasm as shown on the right by the surface plot dia-
gram. The bar is 20 μm long.

Targeted Macromolecules Delivery by Large Lipidic Nanovesicles Electrofusion with Mammalian Cells 530
(a)
(b)
Figure 3. Cells after LUVs electrofusion observed under
fluorescence microscopy. (a) 10 pulses were applied. A gen-
eral view at moderate magnification, a zoomed view of a
single cell and its “surface plot” analysis are given; (b) 20
pulses were applied. A general view at moderate magnifica-
tion, a zoomed view of two “single” cells and their “surface
plot” analysis are given. Cells were observed 4 hours after
PEF delivery. The bar is 20 m long.
Is similar to what is observed in the background signal
detected outside of the cell. This may nevertheless sug-
gest that some endocytotic vesicles are present. The gen-
eral view at moderate magnification obtained on cells
treated by 20 pulses (Figure 3(b)) shows “hot” spots
very clearly (cells in the lower part of the figure). How-
ever, this observation is done with a wide field micro-
scope and the intensity of these “hot” spots is very high
(saturation of the camera) and so they may correspond to
stuck unfused-LUVs on the cell surface, rather than en-
docytotic vesicles. It is important to note that the ho-
mogenous cytoplasm labeling was observed only after 4
h incubation at 37˚C in complete culture medium (nano-
vesicles being washed out). No content mixing was ob-
served if cells were kept at 4˚C after PEF delivery. Fur-
thermore, more fluorescent cells in the pulsed population
are detected when increasing the number of pulses
(compared Figures 3(a) and 3(b)). There was however a
limit due to the loss of viability even if a protective effect
of the nanovesicles was present. Homogeneous FD entry
meant that the macromolecules were able to penetrate
within the nucleus. This organelle is rather large (see
Figure 2 the contrast picture to see its size) and should
give a poorly fluorescent area in the fluorescence image
if the macromolecules were not able to penetrate. This is
indeed not the case as displayed in Figure 3.
4. Conclusions
Fusion was obtained by using the electrical parameters
previously described to induce the cytoplasmic transfer
of the small sized Pyranin [5]. The homogeneity of the
cytoplasm fluorescence is due to its content mixing with
the internal volume of liposomes. Content mixing with
large molecules is the evidence of fusion of the nanove-
sicles, The process is a content and “membrane” mixing,
following neither a “kiss and run” nor an endocytotic
pathway [16-18]. The increase in fusion with the number
of pulses is linked to an increased fusogenicity with an
increase in electrically induced membrane defects. This
has been already observed in the fusion of cells with
Pyranin loaded vesicles [5] or in cell homofusion [19].
Content mixing for large molecules labeled 20 kD dex-
trans is obtained only after a long incubation suggesting
that a membrane-cytoskeleton reorganization must take
place.
Delivery by lipidic nanovesicles is associated to a tar-
geting and avoids the free circulation of the drug. The
lack of a direct electrotransfer of labeled 20 kD dextrans
was already reported [15]. LUV electrofusion can be
targeted either by a biological method (immunolipo-
somes) or simply by a targeted field delivery through a
proper localization of the electrodes.
Electrically enhanced binding of liposomes to cells

Targeted Macromolecules Delivery by Large Lipidic Nanovesicles Electrofusion with Mammalian Cells531
was reported 20 years ago [20]. The spontaneous desta-
bilization of the lipid bilayer of LUVs remains to be ex-
plained [21]. The field pulse induces a membrane desta-
bilization which is prone to a spontaneous fusion as pro-
posed by Rosenheck. Our results may be considered
within the framework of elucidating the physical basis of
exocytosis. The conclusion is that electropermeabiliza-
tion (and associated fusion) may be endogenously pre-
sent as already reported for peptide membrane permeabi-
lization [22]. A key feature of our delivery method is that
the transfer is obtained after a very short contact period
(10 min in this report), a major advantage for drug deliv-
ery. A final advantage is that the electrical conditions can
be used in clinical applications [23].
5. Acknowledgements
Supports from the region Midi Pyrénées, ANR Cemirbio,
PCV CNRS, ARC and Erasmus programs should be ac-
knowledged.
REFERENCES
[1] E. Neumann, G. Gerisch and K. Optaz, “Cell Fusion In-
duced by High Electric Impulse Applied to Dictyos-
telium,” Naturwissenschaften, Vol. 67, 1980, pp. 414-415.
doi:10.1007/BF00405493
[2] J. Teissie, V. P. Knutson, T. Y. Tsong and M. D. Lane,
“Electric Pulse-Induced Fusion of 3T3 Cells in Mono-
layer Culture,” Science, Vol. 216, 1982, pp. 537-538.
doi:10.1126/science.7071601
[3] H. Mekid and L. M. Mir, “In Vivo Cell Electrofusion,”
Biochimica et Biophysica Acta (BBA)—General Subjects,
Vol. 1524, No. 2-3, 2000, pp. 118-130.
doi:10.1016/S0304-4165(00)00145-8
[4] J. Teissie and M. P. Rols, “Fusion of Mammalian Cells in
Culture Is Obtained by Creating the Contact between
Cells after Their Electropermeabilization,” Biochemical
and Biophysical Research Communications, Vol. 140, No.
1, 1986, pp. 258-266.
doi:10.1016/0006-291X(86)91084-3
[5] C. Ramos, D. Bonato, M. Winterhalter, T. Stegmann and
J. Teissié, “Spontaneous Lipid Vesicle Fusion with Elec-
tropermeabilized Cells,” FEBS Letters, Vol. 518, 2002,
pp. 135-138. doi:10.1016/S0014-5793(02)02676-5
[6] J. Heuving, F. Pincet and S. Cribier, “Hemifusion and
Fusion of Giant Vesicles Induced by Reduction of In-
ter-Membrane Distance,” The European Physical Journal
E: Soft Matter and Biological Physics, Vol. 14, No. 3,
2004, pp. 269-276. doi:10.1140/epje/i2003-10151-2
[7] D. Tareste, F. Pincet, E. Perez, S. Rickling, C.
Mioskowski and L. Lebeau, “Energy of Hydrogen Bonds
Probed by the Adhesion of Functionalized Lipid Layers,”
Biophysical Journal, Vol. 83, No. 6, 2002, pp. 3675-3681.
doi:10.1016/S0006-3495(02)75367-8
[8] F. Pincet, L. Lebeau and S. Cribier, “Short-Range Spe-
cific Forces Are Able to Induce Hemifusion,” European
Biophysics Journal, Vol. 30, No. 2, 2001, pp. 91-97.
doi:10.1007/s002490100131
[9] C. G. Schuette, K. Hatsuzawa, M. Margittai, A. Stein, D.
Riedel, P. Küster, M. König, C. Seidel and R. Jahn, “De-
terminants of Liposome Fusion Mediated by Synaptic
SNARE Proteins,” Proceedings of the National Academy
of Sciences USA, Vol. 101, No. 9, 2004, pp. 2858-2863.
doi:10.1073/pnas.0400044101
[10] A. Cypionka, A. Stein, J. M. Hernandez, H. Hippchen, R.
Jahn and P. J. Walla, “Discrimination between Docking
and Fusion of Liposomes Reconstituted with Neuronal
Snare-Proteins Using FCS,” Proceedings of the National
Academy of Sciences USA, Vol. 106, No. 44, 2009, pp.
18575-18580. doi:10.1073/pnas.0906677106
[11] W. Xu and F. Pincet, “Quantification of Phase Transitions
of Lipid Mixtures from Bilayer to Non-Bilayer Structures:
Model, Experimental Validation and Implication on
Membrane Fusion,” Chemistry and Physics of Lipids, Vol.
163, No. 3, 2010, pp. 280-285.
doi:10.1016/j.chemphyslip.2009.12.002
[12] S. Martens, M. M. Kozlov and H. T. McMahon, “How
Synaptotagmin Promotes Membrane Fusion,” Science,
Vol. 316, No. 5828, 2007, pp. 1205-1208.
doi:10.1126/science.1142614
[13] K. Rosenheck, “Evaluation of the Electrostatic Field
Strength at the Site of Exocytosis in Adrenal Chromaffin
Cells,” Biophysical Journal, Vol. 75, 1998, pp. 1237-
4123. doi:10.1016/S0006-3495(98)74043-3
[14] P. Luitel, D. F. Schroeter and J. W. Powell, “Self-Elec-
troporation as a Model for Fusion Pore Formation,”
Journal of Biomolecular Structure & Dynamics, Vol. 24,
No. 5, 2007, pp. 495-503.
[15] M. P. Rols and J. Teissie, “Electropermeabilization of
Mammalian Cells to Macromolecules: Control by Pulse
Duration,” Biophysical Journal, Vol. 75, No. 3, 1998, pp.
1415-1423. doi:10.1016/S0006-3495(98)74060-3
[16] E. R. Travis and R. M. Wightman, “Spatio-Temporal
Resolution of Exocytosis from Individual Cells,” Annual
Review of Biophysics and Biomolecular Structure, Vol.
27, 1998, pp. 77-103.
doi:10.1146/annurev.biophys.27.1.77
[17] E. Alés, L. Tabares, J. M. Poyato, V. Valero, M. Lindau
and G. Alvarez de Toledo, “High Calcium Concentrations
Shift the Mode of Exocytosis to the Kiss-and-Run Me-
chanism,” Nature Cell Biology, Vol. 1, No. 1, 1999, pp.
40-44. doi:10.1038/9012
[18] S. O. Rizzoli and R. Jahn, “Kiss-and-Run, Collapse and
‘Readily Retrievable’ Vesicles,” Traffic, Vol. 8, No. 9,
2007, pp. 1137-1144.
doi:10.1111/j.1600-0854.2007.00614.x
[19] J. Teissie and C. Ramos, “Correlation between Electric
Field Pulse Induced Long-Lived Permeabilization and
Fusogenicity in Cell Membranes,” Biophysical Journal,
Vol. 74, 1998, pp. 1889-1898.
doi:10.1016/S0006-3495(98)77898-1
[20] L. V. Chernomordik, D. Papahadjopoulos and T. Y.
Tsong, “Increased Binding of Liposomes to Cells by

Targeted Macromolecules Delivery by Large Lipidic Nanovesicles Electrofusion with Mammalian Cells 532
Electric Treatment,” Biochimica et Biophysica Acta
(BBA)—General Subjects, Vol. 1070, No. 1, 1991, pp.
193-197. doi:10.1016/0005-2736(91)90163-3
[21] J. Teissie, M. Golzio and M. P. Rols, “Mechanisms of
Cell Membrane Electropermeabilization: A Minireview
of Our Present (Lack of ?) Knowledge,” Biochimica et
Biophysica Acta (BBA)—General Subjects, Vol. 1724, No.
3, 2005, pp. 270-280. doi:10.1016/j.bbagen.2005.05.006
[22] M. Miteva, M. Andersson, A. Karshikoff and G. Otting,
“Molecular Electroporation: A Unifying Concept for the
Description of Membrane Pore Formation by Antibacte-
rial Peptides, Exemplified with NK-lysin,” FEBS Letters,
Vol. 462, No. 1-2, 1999, pp. 155-158.
doi:10.1016/S0014-5793(99)01520-3
[23] T. Hampton, “Electric Pulses Help with Chemotherapy
may Open New Paths for Other Agents,” Journal of the
American Medical Association (JAMA), Vol. 305, No. 6,
2011, pp. 549-551. doi:10.1001/jama.2011.92