Bioelectrochemistry 54 Ž2001.

91–95
www.elsevier.comrlocaterbioelechem

Cell membrane electropermeabilization by symmetrical bipolar

rectangular pulses q
Part II. Reduced electrolytic contamination
Tadej Kotnik a,b,) , Damijan Miklavcic
ˇ ˇ a, Lluis M. Mir b
a
Faculty of Electrical Engineering, UniÕersity of Ljubljana, Trzaska 25, SI-1000 Ljubljana, SloÕenia
b
UMR 8532 CNRS, Institute GustaÕe-Roussy, F-94805 Villejuif, France
Received 16 February 2001; accepted 18 April 2001

Abstract

The paper presents a comparative study of the contamination of a cell suspension by ions released from aluminum cuvettes ŽAl 3q . and
stainless steel electrodes ŽFe 2qrFe 3q . during cell membrane electropermeabilization by unipolar and by symmetrical bipolar rectangular
electric pulses. A single pulse and a train of eight pulses were delivered to electrodes at a 2-mm distance, with 100-ms and 1-ms pulse
durations, and amplitudes ranging from 0 to 400 V for unipolar, and from 0 to 280 V for bipolar pulses. We found that the released
concentrations of Al 3q and Fe 2qrFe 3q were always more than one order of magnitude lower with bipolar pulses than with unipolar
pulses of the same amplitude and duration. We then investigated the viability of DC-3F cells after 1 h of incubation in the medium
containing different concentrations of Al 3q or Fe 2qrFe 3q within the range of measured released concentrations Žup to 2.5 mM for both
ions., thus separating the effects of electrolytic contamination from the effects of electropermeabilization itself. For Fe 2qrFe 3q, loss of
cell viability became significant at concentrations above 1.5 mM, while for Al 3q, no effect on cell survival was detected within the
investigated range. Still, reports on the biochemical effects of released Al 3q also suggest that with aluminum cuvettes, electrolytic
contamination can be detrimental. Our study shows that electrolytic contamination and its detrimental effects can be largely reduced with
no loss in efficiency of electropermeabilization, if bipolar rectangular pulses of the same amplitude and duration are used instead of the
commonly applied unipolar pulses. q 2001 Elsevier Science B.V. All rights reserved.

Keywords: Electropermeabilization; Electroporation; Bipolar pulses; Electrolytic contamination; Metal ions

1. Introduction tropermeabilization in vitro also depends on the composi-

tion of the extracellular medium in which the cells are
Electropermeabilization, also referred to as electropora-
suspended w2x. Electrolytic action of the pulses alters this
tion, is an effective method for the internalization of
composition, both by release of metal ions from the elec-
various molecules into biological cells, with an increasing
trodes and by additional ionization and dissociation of the
number of applications in oncology, genetics, immunol-
medium constituents.
ogy, and cell biology. A short list of bibliographical
To date, several reports have been published on the
sources, in which this phenomenon and its diverse applica-
detrimental effects of metal ions released from the elec-
tions are described in detail, can be found in the Introduc-
trodes w3–5x. Loomis-Husselbee et al. w3x have reported
tion of Part I of this study w1x.
that Al 3q released from aluminum electroporation cuvettes
Besides the parameters of electric pulses and various
can significantly affect biochemical processes involving
physical and chemical parameters, the efficiency of elec-
inositol phosphates. Friedrich et al. w4x have measured the
release of Al 3q in concentrations up to 1 mM, and have
q
Part of this paper was presented by Tadej Kotnik as the Luigi observed that this release also affects the pH of the suspen-
Galvani prize lecture 2001 at the XVIth International Symposium on sion. Similarly, Stapulionis w5x has measured the release of
Bioelectrochemistry and Bioenergetics in Bratislava, Slovakia. Fe 2qrFe 3q from stainless steel electrodes in concentra-
)
Corresponding author. Faculty of Electrical Engineering, University
of Ljubljana, Trzaska 25, SI-1000 Ljubljana, Slovenia. Tel.: q38-614-76-
tions up to 1.2 mM, as well as the release of Al 3q from
8264; fax: q38-614-26-4658. aluminum electrodes and of Cu2q from copper electrodes,
E-mail address: tadej@svarun.fe.uni-lj.si ŽT. Kotnik.. and has reported that the release of each of these ions can

1567-5394r01r$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.
PII: S 1 5 6 7 - 5 3 9 4 Ž 0 1 . 0 0 1 1 5 - 3
92 T. Kotnik et al.r Bioelectrochemistry 54 (2001) 91–95

cause substantial precipitation of nucleic acids and proteins 2.3. Cell culture
from the solution. The release of Fe 2qrFe 3q from stain-
less steel electrodes has also been confirmed by Tomov DC-3F cells, a line of spontaneously transformed Chi-
and Tsoneva w6x. nese hamster fibroblasts w7x, were grown in monolayers at
In our study, we have measured concentrations of metal 37 8C and 5% CO 2 ŽUniversal Water Jacketed Incubator,
ions released into the cell suspension from aluminum Forma Scientific, Marietta, OH, USA.. The cells were
electroporation cuvettes Žrelease of Al 3q . and stainless cultivated in 150-cm3 flasks, and 60-mm petri dishes were
steel electrodes Žrelease of Fe 2qrFe 3q . during cell mem- used for cloning efficiency assays Žboth from TPP,
brane electropermeabilization by unipolar and symmetrical Trasadingen, Switzerland.. The culture medium consisted
bipolar rectangular electric pulses. We show that in both of Eagle minimum essential medium EMEM 41090 sup-
cases, the concentrations of ions released by bipolar pulses plemented with 10% fetal bovine serum Žboth from Life
are more than an order of magnitude lower than those Technologies., 100 Urml of penicillin and 125 mgrml of
released by unipolar pulses of the same amplitude and streptomycin Žboth from SarbachrSolvay Pharma, Brus-
duration, and thus, the detrimental effects of electrolytic sels, Belgium..
contamination on cells can be largely reduced by the use
of bipolar instead of unipolar pulses. 2.4. Determination of cell Õiability

After trypsination with trypsin-EDTA ŽLife Technolo-

2. Materials and methods gies., cells were centrifuged for 5 min at 1000 rpm in a
C312 centrifuge ŽJouan., and suspended at 2 = 10 7
cellsrml in SMEM to which AlCl 3 Žresp. FeCl 3 . was
2.1. Electric pulses and the medium
previously added in quantities giving concentrations of
Al 3q Žresp. Fe 2qrFe 3q . in the range from 0 to 2.5 mM.
Unipolar electric pulses were generated by a PS 15r One milliliter of cell suspension was incubated for 1 h at
GHT 1287 B electropulsator ŽJouan, St. Herblain, France., room temperature, then diluted in the culture medium Žsee
while bipolar pulses were generated by an AFG 310
programmable function generator ŽTektronix, Wilsonville,
OR, USA. and were amplified to the required voltages in
the range from 0 to 280 V using a bipolar amplifier built in
the Laboratory of Biocybernetics at the Faculty of Electri-
cal Engineering of the University of Ljubljana. A 50-ml
droplet of Spinner minimum essential medium SMEM
21385 ŽLife Technologies, Rockville, MD, USA., which
contains salts of neither aluminum nor iron, was placed
either into an aluminum electroporation cuvette ŽEppen-
dorf, Hamburg, Germany. or between a pair of flat stain-
less steel electrodes Žin both cases, the distance between
the electrodes was 2 mm. to which the pulses were then
delivered. We have used a single pulse and a train of eight
pulses, with 100-ms and 1-ms pulse durations, and ampli-
tudes ranging from 0 to 400 V for unipolar, and from 0 to
280 V for bipolar pulses. Within seconds after the delivery
of pulses, the droplet was removed from the electrodes. To
allow for comparability of our results with experiments
performed with electrodes at various distances, the pulse
amplitude was given as the ratio between the voltage and
the distance between the electrodes.

2.2. Measurements of ionic concentrations

The concentrations of Al 3q were determined using an

ICP-MS 4500 Series inductively coupled plasma-mass Fig. 1. The concentration of Al 3q ions in the suspension as the function
spectrometer ŽHewlett-Packard, Wilmington, DE, USA., of pulse amplitude Žthe voltage-to-distance ratio. for eight unipolar ŽB.
and symmetrical bipolar ŽI. rectangular electric pulses. Each pulse was
while the concentrations of Fe 2qrFe 3q were measured on of 1 ms total duration Žfor bipolar pulses, 500 ms of one polarity followed
a CX7 clinical system for chemical analysis ŽBeckman by 500 ms of the opposite polarity., and the pulses were delivered in 1 s
Coulter, Fullerton, CA, USA.. intervals.
 T. Kotnik et al.r Bioelectrochemistry 54 (2001) 91–95 93

above. to 100 cellsrml, and 4 ml of suspension was

transferred into each 60-mm petri dish. The cells were
grown for 5 days, then fixed by a 15-min exposure to
100% ethanol ŽCarlo Erba Reagenti, Milan, Italy. and
stained for 15 min with 1% crystal violet ŽSigma, St.
Louis, MO, USA.. Colonies were counted under a light
microscope ŽLeica, Wetzlar, Germany. and normalized to
the control Žcells incubated in pure SMEM, in which 1 h
incubation does not cause any detectable reduction in the
cloning efficiency of these cells. to obtain the percentage
of cells surviving a 1-h exposure to different concentra-
tions of Al 3q or Fe 2qrFe 3q. Each experiment was re-
peated three times at intervals of at least several days.

3. Results and discussion

Fig. 1 shows the concentration of Al 3q ions released

from aluminum electroporation cuvettes by an application
of eight unipolar, and of eight bipolar rectangular pulses.
Pulses, each of 1 ms duration, were delivered at intervals
of 1 s. In bipolar pulses, the polarity was reversed after
500 ms. Fig. 2 shows the results of an analogous study
performed for Fe 2qrFe 3q ions released from the stainless Fig. 3. Cell survival Žmean"S.D.. without Ž`. and with electroperme-
steel electrodes. abilization Žv . as the function of Al 3q concentration in the suspension.
The cells were incubated for 1 h at room temperature. Electropermeabi-
lization was performed at the beginning of the incubation using a train of
eight unipolar rectangular pulses, each of 100 ms duration and 240 V
amplitude Ž1200 Vrcm voltage-to-distance ratio., delivered in 1 s inter-
vals.

These results show that with bipolar pulses, the released

concentrations of both Al 3q and Fe 2qrFe 3q are more than
one order of magnitude lower than with unipolar pulses of
the same amplitude and duration. For unipolar pulses, the
released concentrations are also roughly proportional to the
number, duration, and amplitude of pulses.
To date, there are no commercially available generators
of bipolar rectangular pulses with adequate amplitude and
duration for electropermeabilization. While Figs. 1 and 2
testify that electrolytic erosion of the electrodes is de-
creased if bipolar pulses are applied instead of unipolar
ones, it is unlikely that this improvement alone would
provide sufficient motivation for the development of bipo-
lar pulse generators for electropermeabilization. On the
other hand, if electrolytic contamination of the suspension
by unipolar pulses leads to detrimental effects on the
suspended cells, this can provide a considerable impetus
for such a development.
The most obvious detrimental effect of the released
metal ions would be the loss of cell viability, and to
Fig. 2. The concentration of Fe 2q and Fe 3q ions in the suspension as the investigate whether such an effect actually exists, we have
function of pulse amplitude Žthe voltage-to-distance ratio. for eight determined the viability of DC-3F cells after 1 h of
unipolar ŽB. and symmetrical bipolar ŽI. rectangular electric pulses.
Each pulse was of 1 ms total duration Žfor bipolar pulses, 500 ms of one
incubation in the medium containing concentrations of
polarity followed by 500 ms of the opposite polarity., and the pulses were Al 3q and Fe 2qrFe 3q, respectively, ranging from 0 to 2.5
delivered in 1 s intervals. mM ŽFigs. 3 and 4, hollow circles.. As can be seen from
94 T. Kotnik et al.r Bioelectrochemistry 54 (2001) 91–95

improvement of efficiency of electropermeabilization, since

in most of the applications, only the cells that remain
viable after the treatment are of interest. In this respect, the
study of cell survival alone offers no evidence for a need
to decrease the electrolytic contamination of the medium
in the case of aluminum cuvettes. However, the reports on
the effects of released Al 3q on biochemistry of inositol
phosphates w14x, pH of the suspension w15x, and preci-
pitation of macromolecules w16x also suggest that with
aluminum electrodes, electrolytic contamination can be
detrimental, as well as biasing for the outcome of the
experiments.
As our study shows, electrolytic contamination from
both stainless steel electrodes and aluminum cuvettes is
largely reduced if unipolar rectangular pulses are replaced
by symmetrical bipolar pulses of the same amplitude and
duration. It is important to stress that this substitution does
not lead to any loss in the efficiency of electropermeabi-
lization. On the contrary, as we have demonstrated in Part
I of this study w1x, replacing unipolar pulses by bipolar
pulses of the same amplitude and duration can even lead to
a noticeable increase of both the percentage of electroper-
meabilized cells and the uptake of molecules into these
Fig. 4. Cell survival Žmean"S.D.. without Ž`. and with electroperme- cells. Together with the results presented in this paper, this
abilization Žv . as the function of Fe 2qrFe 3q concentration in the offers a strong motivation for the use of bipolar instead of
suspension. The cells were incubated for 1 h at room temperature. unipolar rectangular pulses, and consequently for the de-
Electropermeabilization was performed at the beginning of the incubation
velopment of commercial bipolar pulse generators which
using a train of eight unipolar rectangular pulses, each of 100 ms duration
and 240 V amplitude Ž1200 Vrcm voltage-to-distance ratio., delivered in could, in due course, gain a wide use in cell membrane
1 s intervals. electropermeabilization.

these figures, no effect on cell survival was detected within Acknowledgements

the investigated range for Al 3q ions, while for Fe 2qrFe 3q
ions, loss of cell viability became significant at concentra-
tions above 1.5 mM. This study reveals that with stainless This work was supported in part by CNRS, Institute
steel electrodes, even nonpermeabilized cells can be af- Gustave-Roussy, University of Paris XI, and Ministry of
fected by the released concentrations of Fe 2qrFe 3q ions. Science and Technology of the Republic of Slovenia. The
To investigate if electropermeabilization increases the vul- high-voltage amplifier has been developed within the frame
nerability of the cells to Al 3q and Fe 2qrFe 3q ions in the of the Cliniporator project ŽGrant QLK3-1999-00484. un-
medium, an additional experiment was performed, in which der the framework of the 5th PCRD of the European
at the beginning of incubation, the cells were permeabi- Commission. The authors would like to thank Dr. M.
lized by eight unipolar pulses of 100-ms duration and Bonnay from the Department of Clinical Biology of Insti-
240-V amplitude Ž1200 Vrcm voltage-to-distance ratio.. tute Gustave-Roussy, who kindly provided the equipment
Under our experimental conditions, this established proto- for the measurements of Fe 2qrFe 3q concentrations. T.K.
col caused permeabilization of f 94% of the cells accom- was also a recipient of a part-time PhD Scholarship of the
panied by only f 4% cell death Ždata not shown., but if French Government ŽCNOUS..
the cells were electropermeabilized in a medium contain-
ing Fe 2qrFe 3q ions, the cytotoxic effect was more pro-
nounced ŽFig. 4, filled circles. than for nonpermeabilized References
cells ŽFig. 4, hollow circles.. For Al 3q ions, no cytotoxic
effect was observed even when the cells were electroper- w1x T. Kotnik, L.M. Mir, K. Flisar, M. Puc, D. Miklavcic,
ˇ ˇ Cell membrane
meabilized ŽFig. 3, filled circles.. electropermeabilization by symmetrical bipolar rectangular pulses:
These results, which focus on the most obvious detri- Part I. Increased efficiency of permabilization. Bioelectrochemistry
54 Ž2001. 83–90.
mental effect of the released ions—loss of cell viability— w2x S. Orlowski, L.M. Mir, Cell electropermeabilization: a new tool for
show that in the case of stainless steel electrodes, decrease biochemical and pharmacological studies. Biochim. Biophys. Acta
of electrolytic contamination of the medium can lead to an 1154 Ž1993. 51–63.
 T. Kotnik et al.r Bioelectrochemistry 54 (2001) 91–95 95

w3x J.W. Loomis-Husselbee, P.J. Cullen, R.F. Irvine, A.P. Dawson, Elec- w5x R. Stapulionis, Electric pulse-induced precipitation of biological
troporation can cause artifacts due to solubilization of cations from macromolecules in electroporation. Bioelectrochem. Bioenerg. 48
the electrode plates. Biochem. J. 277 Ž1991. 883–885. Ž1999. 249–254.
w4x U. Friedrich, N. Stachowicz, A. Simm, G. Fuhr, K. Lucas, U. w6x T. Tomov, I. Tsoneva, Are the stainless steel electrodes inert? Bio-
Zimmermann, High efficiency electrotransfection with aluminum electrochemistry 51 Ž2000. 207–209.
electrodes using microsecond-controlled pulses. Bioelectrochem. w7x J.L. Biedler, H. Riehm, Cellular resistance to actinomycin D in
Bioenerg. 47 Ž1998. 103–111. Chinese hamster cells in vitro. Cancer Res. 30 Ž1970. 1174–1184.