Bioelectrochemistry 64 (2004) 113 – 124

www.elsevier.com/locate/bioelechem

Review
Techniques of signal generation required for electropermeabilization.
Survey of electropermeabilization devices
Marko Puc, Selma Čorović, Karel Flisar, Marko Petkovšek, Janez Nastran, Damijan Miklavčič *
Damijan Miklavčič, University of Ljubljana, Faculty of Electrical Engineering, Tržaška 25, SI-1000 Ljubljana, Slovenia
Received 27 November 2003; received in revised form 23 January 2004; accepted 8 April 2004
Available online 15 June 2004

Abstract

Electropermeabilization is a phenomenon that transiently increases permeability of the cell plasma membrane. In the state of high
permeability, the plasma membrane allows ions, small and large molecules to be introduced into the cytoplasm, although the cell plasma
membrane represents a considerable barrier for them in its normal state. Besides introduction of various substances to cell cytoplasm,
permeabilized cell membrane allows cell fusion or insertion of proteins to the cell membrane. Efficiency of all these applications strongly
depends on parameters of electric pulses that are delivered to the treated object using specially developed electrodes and electronic devices—
electroporators. In this paper we present and compare most commonly used techniques of signal generation required for
electropermeabilization. In addition, we present an overview of commercially available electroporators and electroporation systems that
were described in accessible literature.
D 2004 Elsevier B.V. All rights reserved.

Keywords: Electroporation; Electropermeabilization; Instrumentation; Electrochemotherapy; Gene transfection

1. Introduction cells, their surroundings and cell geometry (i.e. temperature,

osmotic pressure, cell size and shape, etc.) [7,14]. With
The use of high voltage electric pulse technology, elec- properly chosen values of the electric field parameters, the
tropermeabilization, in cell biology, biotechnology and process of electropermeabilization is reversible and cells
medicine has attracted significant interest ever since first return into their normal physiological state. If these param-
reports were published several decades ago [1– 3]. Electro- eters exceed certain values (e.g. amplitude of pulses is too
permeabilization is a transient phenomenon that increases high or duration of pulses is too long), cells are irreversibly
permeability of the cell plasma membrane. In the state of permeabilized and lose their viability (Fig. 1) [5 –7].
high permeability, the plasma membrane allows ions, small Permeabilization of cell plasma membrane is achieved by
and large molecules to be introduced into the cytoplasm, exposure of the cell to a short but intense electric field. The
although the cell plasma membrane in its normal state basic quantity underlying this process is presumably the
represents a considerable barrier for them. Besides intro- induced transmembrane potential difference, which is in the
duction of different substances to the cytoplasm, the per- first approximation proportional to the product of the applied
meabilized cell membrane allows cell fusion or insertion of electric field strength E and cell radius R [7,16]. Furthermore,
proteins into cell membrane (Fig. 1) [4– 7]. Efficacy of it has been shown that electric field controls the permeabi-
electropermeabilization and its applications strongly lization of cell membrane in two ways. (1) Electric field
depends on many parameters that can be divided into initiates permeabilization of cell membrane in the regions
parameters of the electric field (i.e. pulse amplitude, pulse where transmembrane potential difference exceeds the
duration, pulse repetition frequency, number of pulses and threshold value (between 200 and 300 mV) [7,9]. (2) Electric
pulse shape) [8– 13], and parameters that define the state of field strength defines the size of permeabilized area of cell
membrane [7,9,11]. This means that permeabilization of cell
* Corresponding author. Tel.: +386-1-4768-456; fax: +386-1-4264- membrane will occur only if the applied electric field is larger
658. than the threshold value. Since the induced transmembrane
E-mail address: damijan@svarun.fe.uni-lj.si (D. Miklavčič). potential difference is also proportional to the cell radius, it is

1567-5394/$ - see front matter D 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.bioelechem.2004.04.001
114 M. Puc et al. / Bioelectrochemistry 64 (2004) 113–124

Fig. 1. Exposure of a cell to an electric field may result either in permeabilization of cell membrane or its destruction. In this process the electric field
parameters play a major role. If these parameters are within certain range, the permeabilization is reversible, therefore it can be used in applications such as
introduction of small or large molecules into the cytoplasm, insertion of proteins into cell membrane or cell fusion.

evident that threshold value of electric field varies with cell the required signal to its output loaded by impedance of
size. This means that large cells are more sensitive to lower sample between electrodes.
electric field strengths than small cells [7,9]. Moreover, it has Probably the major problem that every engineer faces
been shown that induced transmembrane potential difference during the design of electroporator is characterization of the
also depends on cell density, arrangement and cell position load, which in principle has resistive and capacitive com-
[17 –19]. Considering this, it is very difficult to generalize the ponent. The value of each component is defined by geom-
electric field parameters for different experimental conditions etry and material of electrodes and by electrical and
(i.e. single cell permeabilization, in vitro, in vivo, etc.), or for chemical properties of the treated sample. In in vitro
different cell types (i.e. animal, plant, fungi, prokaryotic). In conditions these parameters that influence on impedance
addition, different applications require different time varia- of load can be well controlled since size and geometry of
tion of electric fields (i.e. exponentially decaying, square sample is known especially if cuvettes are used, furthermore
wave, etc.) and different exposure times. by using specially prepared cell mediums electrical and
It is not an aim of this paper to focus on further chemical properties are defined or can be measured [27 –
description of electric field parameters that are required in 30]. On the other hand, in in vivo or clinical conditions, size
different applications of electropermeabilization. Instead, and geometry can still be controlled to a certain extent but
we would like to present and compare advantages and electrical and chemical properties can only be estimated.
drawbacks of the existing and most commonly used con- But what is practically impossible to predict during the
cepts of electric signal generation and available devices that development of the device are changes in the electrical and
fulfill electrical requirements of applications such as: elec- chemical properties of the sample due to exposure to high-
trochemotherapy, electrotransfection, insertion of proteins voltage electric pulses. Besides electropermeabilization of
into cell membrane, cell fusion and transdermal drug deliv- cell membranes which increases electrical conductivity of
ery [5– 7,15,20 –23]. the sample [31 – 33,38,39], electric pulses also cause at least
two known side effects: heating and electrolytic contami-
nation of the sample [10,34 –37]. Furthermore, there are
2. Techniques of signal generation required for several other side effects that evolve from interactions
electropermeabilization between electrodes and treated sample, but we will not
explain their influence on electrical and chemical properties
Effectiveness of electropermeabilization in either in vitro, of the sample because this is beyond the scope of this paper.
in vivo or clinical environment depends on the distribution When most of the electrical parameters that electropo-
of electric field inside the treated sample [24 – 26]. To rator should provide are determined, engineer has to choose
achieve this, we have to use an appropriate set of electrodes the type of electroporator he is going to design. In principle,
(e.g. needle, parallel plates, cuvettes, etc.) and an electro- electroporators can be divided in several groups depending
permeabilization device—electroporator that generates re- on the biological applications, but from the electrical point
quired voltage or current signals. Although both parts of the of view only two types of electroporators exist: devices with
mentioned equipment are equally important for effective- voltage output (output is voltage signal U(t)) and devices
ness of electropermeabilization, electroporator has substan- with current output (output is current signal I(t)). Both types
tially more important role since it has to be able to deliver of devices have their advantages and disadvantages, but one
 M. Puc et al. / Bioelectrochemistry 64 (2004) 113–124 115

point definitely speaks in favor of devices with voltage

output. For example, if we perform in vitro experiments
with stainless steel parallel plate electrodes with plate sides
substantially larger than the distance between them, the
electric field strength E that is applied to the sample can
be approximated by the voltage-to-distance ratio U/d, where
d is the electrode distance and U the amplitude of applied
signal obtained from an electroporator with voltage output.
On the other hand, if an electroporator with current output is
Fig. 3. Switching high voltage power supply with power transistors
used, the same approximation could be used only if addi- (generator of square wave pulses). The device consists of power supply part
tional measurement of voltage difference between electrodes and pulse generator. The variable high-voltage power supply (V)
is performed or if the impedance Z of the sample is known, continuously charges the capacitor (C) that stores energy required during
measured or approximated and voltage difference between the pulse. To deliver the pulse to the electrodes, the triggering circuit
electrodes is estimated using Ohm’s law U = IZ. This exam- generates low-voltage pulse, usually around 10 V, that opens transistor (Q)
(e.g. fast power MOSFET or IGBT) for the duration of the low-voltage
ple shows that if an electroporator with voltage output is pulse.
used, estimation of applied electric field strength can be
made without additional measurements or knowledge of
samples passive electrical properties. constant of discharge, since the impedance of load (e.g.
Since electroporators with voltage output are much more cell suspension) varies [38 –40]. If additional resistors are
widespread than the electroporators with current output, we connected in parallel to the output the time constant of
will concentrate on most commonly used techniques to discharge is defined by: (RjjZL) C, where R is resistance of
generate voltage signals required for electropermeabilization. the internal resistor. If absolute value of the impedance of
load ZL is at least 10 times larger than the resistance R
2.1. Discharge of a capacitor (ZL z 10R), the time constant can be approximated by the
RC product.
This is the oldest technique used to generate signals for The presented concept is very simple and the generated
electropermeabilization primarily in in vitro environment. pulse could be used even for gene transfection since it
The device consists of: high voltage power supply, capac- includes the high voltage part for permeabilization and low
itor, switch, and optionally resistance (Fig. 2). The device voltage electrophoretic part [54]. Although the transition
operates in two phases, charge and discharge, and gener- from high voltage to low voltage is smooth, the respective
ates exponentially decaying pulses. During the charge lengths of each part is ill-defined. Definition of electric field
phase, the switch (S) is in the position 1 and variable parameters is probably the major drawback of the presented
high voltage power supply (V) charges the capacitor (C) to technique. Moreover, repetition frequency of signal delivery
the preset voltage. In the discharge phase, the switch is in is low due to a long charge phase, and the flexibility of
the position 2, and the capacitor discharges through the electric field parameters is in general poor. Besides this, the
load connected to the output. Time constant of discharge s presented technique usually requires additional circuits to
can be approximated by product ZLC, where C is the prevent sparking that might be caused during the change of
capacitance of capacitor and ZL is the absolute value of the switch position.
load impedance. But most commercially available devices
have built-in resistances that are connected in parallel to 2.2. Square wave generators
the load. Their main purpose is to define exactly the time
For better control of electric field parameters, square
wave pulse generator has been introduced. The device still
comprises a variable high voltage power supply (V) and a
capacitor (C) for energy storage, yet the switch is replaced
with a fast power MOSFET (metal oxide silicon field effect
transistor) or IGBT (insulated gate bipolar transistor) (Q)
and a triggering circuit (Fig. 3). In principle, such a device
can continuously deliver square wave pulses to the output,
provided that the high voltage power supply is able to
recharge the capacitor during the delay between two con-
Fig. 2. Discharge of a capacitor (generator of exponentially decaying secutive pulses. The output amplitude of pulses is defined
pulses). The basic setup comprises: variable high-voltage power supply (V), by amplitude of variable power supply, while pulse dura-
capacitor (C), switch (S), and optionally resistance (R). The device operates
in two phases: charge (switch is in position 1 and capacitor charges to the
tion, pulse repetition frequency and possibly number of
preset voltage) and discharge (switch is in position 2 capacitor discharges pulses are programmed by a computer that also comprises
through the load connected to the electrodes). triggering circuit.
116 M. Puc et al. / Bioelectrochemistry 64 (2004) 113–124

Despite improved control over the electric field param-

eters, this technique still has drawbacks that limit flexibil-
ity and accuracy of pulse parameters available to the user.
The main problem lies in limited power capabilities of
high voltage power supply. The charging current of ca-
pacitor that is delivered from power supply is usually
much smaller than the discharging current that flows
through the load during the pulse. Since more charge is
taken from the capacitor than delivered, the voltage on
capacitor decreases, which results in a decrease of pulse
amplitude. The decrease of voltage can be limited by
increasing capacitance of the capacitor, or it can be totally
eliminated by using power supply that meets power
requirements of the load. Because the first solution to Fig. 5. Simplified circuit of an analogue generator of arbitrary signal. The
the problem is more common, the accuracy of pulse signal generated by function generator FG is delivered through the unity
gain amplifier to the voltage stage, where the amplitude of signal is
amplitude of delivered pulses is within the range of few
increased. The amplitude of signal delivered from driving stage (i.e.
percent of the maximum value. In addition, limited power function generator and unity gain amplifier) defines the output amplitude of
supply also influences the limitation of pulse duration and voltage stage. The signal then enters the current stage, which ensures the
pulse repetition frequency. If consecutive pulses are gen- power required by the load ZL.
erated, it is usually required that each pulse has the same
amplitude as the first one that was generated. Due to the
decrease of voltage on the capacitor during the pulse, next complex due to nonlinear relationship between magnetic
pulse can be delivered only after the capacitor is recharged field density (B) and magnetic field strength (H) in the
to the preset voltage. core of transformer. Beside this, additional output circuits
Despite these drawbacks, square wave pulse generators are usually necessary to demagnetize the transformer after
are still very often used to generate pulses especially in the end of the pulse. With no additional circuit at the
combination with pulse transformers (Fig. 4). This tech- output, demagnetization is carried out through the load,
nique requires a square wave generator that generates low and consequently the shape of the signal is distorted (i.e.
voltage pulses, while pulse transformer (T) outputs a high quasi bipolar pulses are produced).
voltage pulse due to translation function that is defined by
its properties. Furthermore, this configuration provides great 2.3. Analogue generator of unipolar arbitrary signals
safety margin because by using pulse transformer, the
output floats and pulse transformer can be built to saturate Although square wave and exponentially decaying pulses
if the pulse length exceeds the maximum pulse length were and probably still are most frequently used signals for
[41,42]. electropermeabilization, in some experiments pulses of
Improved safety reduces the flexibility of pulse param- different shape (e.g. trapezoidal pulses with possibility of
eters, and while amplitude of pulses can be as high as 3 control of rise and fall time or square wave pulses modu-
kV, pulse duration and pulse repetition frequency are lated with high-frequency sinusoidal signals) have been
limited by the characteristics of the pulse transformer. used [13,43].
Despite the safety feature of the pulse transformer, it has For generation of arbitrary unipolar signals, technique
to be stressed that development of such a transformer is requires at least two amplification stages (voltage and
current) and appropriate driving stage (Fig. 5) [44]. The
driving stage consists of a signal generator (FG), which is
usually a computer with a digital-to-analog converter, and a
unity-gain amplifier (AD) that meets power and impedance
requirements at the input of a voltage stage. The voltage
stage in the presented case is composed of a MOSFET (QV)
and a resistor network connected to the source of the
transistor. The signal delivered to the input of the voltage
stage opens the transistor according to the transfer function,
thus the output voltage changes (e.g. input of 4 V results in
200 V at the output). The major drawback of such voltage
stage is that the ground of voltage stage must be electrically
Fig. 4. Square wave pulse generator with pulse transformer. Similarly to the
previous technique (see Fig. 3) the device comprises power supply and
isolated from the ground of the driving stage. The signal is
pulse generator, but between the load ZL and pulse generator there is also then delivered to the current stage, which is a classical
pulse transformer (T) that additionally increases the amplitude of pulses. source follower made of power MOSFETs connected in
 M. Puc et al. / Bioelectrochemistry 64 (2004) 113–124 117

This design allows wide flexibility of all electrical

parameters, yet some drawbacks still exist. The driving
stage is much more complex than in previously described
techniques, and besides this, it must have electrically
isolated power supplies. With this design it is possible
to generate signals with maximal amplitudes that are
approximately 20 to 30 V lower than supply voltage
( + UCC). Probably the major problem remains general
limitation of output voltage and current due to limitations
of semiconductor technology (SOA-safe operation area of
transistors).

2.4. Analogue generator of bipolar arbitrary signals

Until now we presented techniques that are only able to

deliver unipolar signals. But some researchers in the field
of electropermeabilization tend to utilize bipolar signals
[9,10,46]. Today probably one of the best techniques that
have been evaluated is a class AB bipolar amplifier, in
Fig. 6. Simplified circuit of an analogue generator of bipolar arbitrary
signal. The signal generated by an arbitrary signal source (FG) is delivered other words the closed-loop push – pull amplifier (Fig. 6)
to the input stage where the signal is subtracted from attenuated output [47].
signal delivered through the feedback network. The differential signal is The signal generated by an arbitrary signal source (FG) is
delivered to the inputs of two transconductance stages that increase voltage delivered to the input stage where the signal is subtracted
of signal (upper stage for positive signal and lower stage for negative
from appropriately reduced output signal delivered through
signal). The two signals from each transconductance stage are then
delivered to two output stages, where signals are recombined and amplified the feedback network. The difference of the two signals is
to meet power requirements required by load ZL [47]. delivered to the input of a bipolar voltage amplifier that
comprises two transconductance stages, one for the positive
parallel. This last stage meets the power requirements and one for the negative period of the signal. Each ampli-
determined by the impedance of load (ZL) between the fying stage is composed of two bipolar transistors (PNP-
electrodes [45].

Table 1
Comparison of presented techniques of signal generation for electro-
permeabilization
Technique Advantages Disadvantages
Discharge of capacitor – simple and – poor flexibility of
inexpensive parameters
construction
Square wave – simple construction – limitation of output
generator (power – better control of parameters due to
transistors) pulse parameters semiconductor
technology
Square wave – very safe – limitations of pulse
generator (pulse (possibility to use in duration and
transformer) clinical environment) repetition frequency
– very high pulse – complex design of
amplitudes pulse transformer
Analogue generator of – wide flexibility of – limitation of output
unipolar arbitrary pulse parameters current and voltage
signals – arbitrary signal due to semiconductor
shape technology
Analogue generator of – genuine bipolar – limitation of
Fig. 7. Simplified circuit of modular high-voltage source. Operation of the bipolar arbitrary arbitrary signals bandwidth, output
device is based on a principle of digital-to-analogue converter, thus the signals – arbitrary signal current and voltage
device comprises several (N) individually controlled electrically isolated DC shape due to semiconductor
voltage modules, where the amplitude of the particular voltage source VN is technology
twice as high as in the preceding module. With an appropriate control of Modular high voltage – high dynamics – price
output transistors Q1 – QN the modules are connected in series and a total of source – high currents and
2N different output voltage levels with the resolution of V1 are obtained [48]. voltages
118 M. Puc et al. / Bioelectrochemistry 64 (2004) 113–124

Table 2
List of commercially available electroporators with their parameters, biological applications and possible signal generation technique
Company/product Output Voltage range Time constant Charge time (tc)/pulse Biological application Possible signal
characteristics (s)/pulse length (T) repetition frequency generation technique
( f)
ADITUS MEDICAL http://www.aditusmedical.com
CythorLab Arbitrary LV: 0 V – 600 Vpp, LV: T = 400 ms NA in vitro, in vivo NA
HV: 0 V – 3000 Vpp HV: T = 5 ms

AMAXA Biosystems http://www.amaxa.com

Nucleofector NA NA NA NA in vitro transfection NA

BIORAD http://www.biorad.com
Micro Pulser Exponential 200 – 3000 V s = 1 – 4 ms tc = 5 s bacterial, yeast Capacitor discharge
Gene Pulser Xcell Exponential 10 – 3000 V s = 0.5 ms – 3.3 s tc = 5 s all cell type, eukaryotic Capacitor discharge
Square wave T = 0.05 – 10 ms f = 0.1 – 10 Hz and prokaryotic cells

BTX http://www.btxonline.com
ECM 399 Exponential LV: 2 – 500 V LV: s = 157 ms tc < 5 s bacterial, yeast, Capacitor discharge
HV: 10 – 2500 V HV: s = 5.4 ms mammalian
ECM 630 Exponential LV: 10 – 500 V LV: s = 25 As – 5 s tc < 5 s bacterial, yeast, Capacitor discharge
HV: 50 – 2500 V HV:s = 625 As – 78 ms mammalian, plant,
in vivo
ECM 830 Square wave LV: 5 – 500 V LV: T = 10 As – 10 s f = 0.1 – 10 Hz bacterial, yeast, LV: square wave
HV: 30 – 3000 V HV: T = 10 – 600 As mammalian, plant, generator HV: pulse
in vivo, in ovo transformer
ECM 2001 Square wave LV: 10 – 500 V LV: T = 10 As – 99 ms NA mammalian, plant, LV: square wave
HV: 10 – 3000 V HV: T = 1 – 99 As NA electrofusion generator
Sinus (AC) 0 V – 150 Vpp fAC = 1 MHz HV: pulse transformer
AC: NA
HT 3000 Square wave LV: 0 – 500 V LV: T = 10 ms – 1 s f = 0.1 – 10 Hz in vitro LV: square wave
HV: 0 – 3000 V HV: T = 10 – 600 ms generator HV: pulse
transformer

CLONAID http://www.clonaid.com
RMX2010 Square wave 5 – 200 V T = 10 As – 990 ms f = 1 – 10 Hz gene transfection Square wave generator

CYTO PULSE SCIENCES http://www.cytopulse.com

PA-2000 Square wave 5 – 1000 V T = 1 As – 2 ms f < 8 Hz in vitro, in vivo, Square wave generator
ex vivo
PA-4000 Square wave 5 – 1100 V T = 1 As – 2 ms f < 8 Hz in vitro, in vivo, Square wave generator
ex vivo
PA-101 Sinus (AC) 10 – 150 Vpp fAC = 0.2 – 2 MHz dielectrophoresis AC: NA

EPPENDORF SCIENTIFIC http://www.eppendorf.com

Electroporator Exponential 200 – 2500 V s = 5 ms tc < 8s bacterial, yeast Capacitor discharge
2510
Multiporator:
Eukaryotic Exponential 20 – 1200 V s = 15 – 500 As tc < 30s mammalian, plant, Capacitor discharge
module oocytes
Bacterial module Exponential 200 – 2500 V s = 5 ms tc < 30 s bacterial, yeast Capacitor discharge
Fusion module Square wave 0 – 300 V T = 5 – 300 As f = 1 Hz mammalian, plant Square wave generator
sinus (AC) 2 – 20 Vpp fAC = 2 MHz AC: NA

EQUIBIO http://www.equibio.com
Easyjec T Plus Exponential 100 – 3500 V s = 10 As – 7 s NA all cell types Capacitor discharge
Easyjec T Optima Exponential 20 – 2500 V s = 1.5 ms – 7 s NA all cell types Capacitor discharge
Easyjec T Prima Exponential 1800 – 2500 V s = 5 ms NA bacterial Capacitor discharge

GENETRONICS http://www.genetronics.com
MEDPULSER Square wave NA NA NA electrochemotherapy, NA
clinical device

IGEA http://www.igea.it
Cliniporator Square wave LV: 20 – 200 V LV: T = 10 As – 20 ms f = 1 Hz – 10 kHz electrochemotherapy, unipolar arbitrary
HV: 50 – 1000 V HV: T = 30 – 200 As gene therapy, clinical generator
device
 M. Puc et al. / Bioelectrochemistry 64 (2004) 113–124 119

Table 2 (continued)
Company/product Output Voltage range Time constant Charge time (tc)/pulse Biological application Possible signal
characteristics (s)/pulse length (T) repetition frequency generation technique
( f)
JOUAN
Electropulsator Square wave 0 – 1000 V T = 5 As – 24 ms f = 1 – 10 Hz bacterial, yeast, Square wave generator
PS10 mammalian, plant
Electropulsator Square wave 0 – 1500 V T = 5 As – 24 ms f = 1 – 10 Hz bacterial, yeast, Square wave generator
PS15 mammalian, plant

PROTECH INTERNATIONAL http://www.protechinternational.com

CUY-21 Square wave LV: 0.1 – 199 V LV: T = 0.1 – 999 ms f = 0.1 – 10 Hz in vitro, in vivo, Square wave generator
HV: 200 – 500 V HV: T = 0.1 – 100 ms in ovo, in utero
LF101 Square wave 0 – 999 V T = 5 – 99 ms f = 0.1 – 10 Hz mammalian, plant, Square wave generator
electrofusion

TRITECH RESEARCH http://www.tritechresearch.com

Mammo Zapper Exponential NA NA tc = 15 s mammalian Capacitor discharge
Bacto Zapper Exponential < 2000 V s < 10 ms tc = 5 s Bacterial Capacitor discharge

THERMO ELECTRON CORPORATION http://www.savec.com

CelljecT Uno Exponential 1800 or 2500 V s = 5 ms NA bacterial, yeast Capacitor discharge
CelljecT Duo Exponential 20 – 2500 V s = 1.5 ms – 7 s NA all cell type, Capacitor discharge
eukaryotic and
prokaryotic cells
CelljecT Pro Exponential 20V – 3500V s = 10 As – 7 s tc < 30 s bacterial, yeast, Capacitor discharge
mammalian, plant
Signal generation techniques that are given for each device were anticipated according to the output characteristic. During our investigation we did not have
access to the electrical schemas of the devices nor had we any of the listed device in our hands.
NA stands for not available.

type for positive and NPN-type for negative period) modules (Fig. 7). Its operation is based on a principle of a
connected in cascade and a resistor network necessary for digital-to-analog converter, thus the amplitude of the partic-
normal operation. At this point it has to be stressed that ular source V N is twice as high as the predecessor
complementary transistors have to be used (i.e. NPN and (VN = 2VN  1). The voltage of the individual source is con-
PNP type which are close match) otherwise symmetry stant and can participate in a generation of a common output
between positive and negative part of amplifier cannot be pulse at any time. With an appropriate control of output
achieved. The two signals amplified in each transconduc- transistors Q1 –QN that operate as switches and connect the
tance stage are delivered to two output stages, again one for modules in series, a total of 2N different output voltage levels
positive and one for negative period of signal. The output with the resolution of V1 are obtained [48]. Although the
stages are composed of power MOSFETs, if possible design of each individual source is similar to the design of
complementary (N-type for positive and P-type for negative previously described square wave pulse generator, the indi-
period), that are connected in cascade as source followers. vidual source used in this concept has no problems with the
These last two stages recombine two signals from voltage shortage of power. For correct operation, each source (even
amplifier and meet the power requirements defined by the the smallest one) must be able to produce and sustain the
impedance of the load between electrodes [47]. maximum possible current during the pulse generation. If this
Although, the design by itself has no problems and is is not ensured, the pulse amplitude will decrease.
given as an example in any electronic design book, the major The presented modular topology has many advantages
problem originates in poor availability of semiconductor due to very high dynamics and high power that can be
components (i.e. high voltage and high power complemen- delivered to its output. Furthermore, with a supplemented
tary transistors) necessary to build each of the amplifying single-phase transistor bridge on the output, bipolar pulses
stage. Since those transistors exist only up to 250 V, undesired can be generated as well. Besides the electrode polarity
cascades that gradually reduce dynamics have to be used to change, the transistor bridge also increases the incorporated
generate signals required for electropermeabilization. safety measures of the device in case of malfunction, which
could result in a delivery of huge power to the output.
2.5. Modular high voltage source Namely, for any given pulse amplitude at least three power
transistor switches have to be turned ON (two for the
Another possible improvement of a square wave generator selection of the pulse polarity and at least one for the
is a modular high voltage source that consists of several (N) selection of the desired output pulse amplitude). The mod-
individually controlled and electrically isolated DC voltage ular solution and consequently the increased number of
120 M. Puc et al. / Bioelectrochemistry 64 (2004) 113–124

Table 3
List of commercially available electrodes with their properties and biological applications suggested by manufacturer
Company/product Number of Electrode Needle length (L)/electrode Electrode Biological
electrodes distance size (shape)/Volume (V) material application
BIORAD http://www.biorad.com
CUVETTES: Compatible with Micro pulser, Gene Pulser Xcell
2 1 mm V = 100 Al Aluminum in vitro
2 2 mm V = 400 Al Aluminum in vitro
2 4 mm V = 800 Al Aluminum in vitro

BIOSMITH http://www.biosmith.com
CUVETTES: Compatible with electroporation devices from all major manufacturers
72001 2 1 mm V = 100 Al Aluminum in vitro
72002 2 2 mm V = 400 Al Aluminum in vitro
72004 2 4 mm V = 800 Al Aluminum in vitro

BTX http://www.btxonline.com
2-NEEDLE ARRAY: Compatible with: ECM 830, 630, 395, 399, 600, T820
Model 531 2 10 mm L = 200 mm Stainless in vivo
steel
Model 532 2 5 mm L = 200 mm Stainless in vivo
steel
GENETRODES: Compatible with: ECM 630, 830, 2001, 600, T820
Model 508 2 1 – 10 mm L = 5 mm Gold plating in vivo
Model 510 2 1 – 10 mm L = 10 mm Gold plating in vivo
Model 512 2 0 – 13 mm L = 5 mm (L-shaped) Gold plating in ovo
Model 514 2 0 – 13 mm L = 3 mm (L-shaped) Gold plating in ovo
Model 516 2 0 – 13 mm L = 1 mm (L-shaped) Gold plating in ovo
CALIPER: Compatible with: ECM 830, 600, 630, 2001, T820
Model 384 2 1 – 130 mm 10  10 mm (square) Stainless steel in vivo
Model 384L 2 1 – 130 mm 20  20 mm (square) Stainless steel in vivo
TWEZERTRODES: Compatible with: ECM T820, 630, 830, 2001
Model 520 2 1 – 20 mm 7 mm diameter (disk) Stainless steel in vivo
Model 522 2 1 – 20 mm 10 mm diameter (disk) Stainless steel in vivo
GENEPADDLES: Compatible with: ECM 830, 2001, 630, 600, T820
Model 542 2 1 – 10 mm 3  5 mm (rectangle) Gold plating in vitro, in vivo
Model 543 2 1 – 10 mm 5  7 mm (rectangle) Gold plating in vitro, in vivo
PETRI PULSER: Compatible with: ECM 830, 630, 600, 399, 395, T820
PP35-2P 13 2 mm V = 0.5 – 30 ml Gold plating in vitro
PETRI DISH ELECTRODES: Compatible with:ECM 830, 630, 2001, 600, T 820
24 2 mm V = 10 – 50 ml Stainless steel in vitro

BTX http://www.btxonline.com
MICROSLIDE: Compatible with: ECM 630, 830, 395, 399, 2001, 600, T820
Model 450 2 0.5 mm V = 20 Al Stainless steel in vitro, fusion
Model 450-1 2 1 mm V = 40 Al Stainless steel in vitro, fusion
Model 453 2 3.2 mm V = 0.7 ml Stainless steel in vitro, fusion
Model 453-10 2 10 mm V = 2.2 ml Stainless steel in vitro, fusion
FLAT ELECTRODE CHAMBER: Compatible with: ECM 630, 830, 2001, 600, T820
Model 484 2 1 mm V = 0.5 ml Stainless steel in vitro, fusion
Model 482 2 2 mm V = 1 ml Stainless steel in vitro, fusion
MEANDER FUSION CHAMBER: Compatible with: ECM 630, 830, 2001, 200, 600, T820
2 0.2 mm – Silver in vitro, fusion
Electroporation plates:
Model HT-P96-2B/W 96 2 mm V = 150 Al Gold plating in vitro
Model HT-P96-4B/W 96 4 mm V = 300 Al Gold plating in vitro
Model HT-P384-2B/W 384 2 mm V = 700 Al Gold plating in vitro
MULTI-WELL COAXIAL ELECTRODES: Compatible with: ECM 630, 830, 2001, 600, T820
Model 491-1 1 1.6 mm V = 0.3 ml (circular) Gold plating in vitro
Model 747 8 1.6 mm V = 0.3 ml (circular) Gold plating in vitro
Model 840 96 1.6 mm V = 0.3 ml (circular) Gold plating in vitro
Flatpack chambers:
Model 485 2 1.83 mm V = 1.5 ml Stainless steel in vitro
 M. Puc et al. / Bioelectrochemistry 64 (2004) 113–124 121

Table 3 (continued)

Company/product Number of Electrode Needle length (L)/electrode Electrode Biological

electrodes distance size (shape)/Volume (V) material application
Model 486 2 0.56 mm V = 85 Al Stainless steel in vitro
Cuvettes:
Model 610 2 1 mm V = 20 – 90 Al Aluminum in vitro
Model 620 2 2 mm V = 40 – 400 Al Aluminum in vitro
Model 640 2 4 mm V = 80 – 800 Al Aluminum in vitro

EPPENDORF http://www.eppendorf.com
CUVETTES: Compatible with Multiporator, Electroporator 2510
2 1 mm V = 100 Al Aluminum in vitro
2 2 mm V = 400 Al Aluminum in vitro
2 4 mm V = 800 Al Aluminum in vitro

CYTOPULSE http://www.cytopulse.com
COAXIAL ELECTRODES: Compatible with PA-101
Model FE-C25/400 2 2.5 mm V = 350 ml NA in vitro, fusion
Model FE-C25/800 2 2.5 mm V = 750 ml NA in vitro, fusion
Model FE-C20/1000 2 2 mm V = 1000 ml NA in vitro, fusion
Tweezers:
TE-5-10 2 Adjustable 5  10 mm (rectangular) NA in vivo
TE-5R 2 Adjustable 5 mm diameter (circular) NA in vivo
2-row needle array:
NE-4-4 8 4 mm NA in vivo
NE-4-6 12 4 mm NA in vivo
NE-6-4 8 6 mm NA in vivo
NE-6-6 12 6 mm NA in vivo
Cuvettes:
CUV-01 2 1 mm V = 100 Al NA ex vivo
CUV-02 2 2 mm V = 400 Al NA ex vivo
CUV-04 2 4 mm V = 800 Al NA ex vivo
Electrode array:
96W-A 96 wells 5.5 mm V = 300 Al/well NA ex vivo

EUROGENTEC http://www.eurogentec.com
CUVETTES: Compatible with most existing electroporation systems
2 1 mm NA Aluminum in vitro
2 2 mm NA Aluminum in vitro
2 4 mm Aluminum in vitro

ICHOR http://www.ichorms.com
Trigrid
multiple NA NA NA in vivo

IGEA http://www.igea.it.
TYPE I: Compatible with Cliniporator
Plate electrodes 2 6 – 8 mm 10  30 mm (rectangular) Stainless steel clinical applications
TYPE II: compatible with cliniporator
Needle rows 8 4 mm L = 20 – 30 mm Stainless steel clinical applications
TYPE III: Compatible with Cliniporator
Hexagonal needle array 7 8 mm L = 20 – 30 mm Stainless steel clinical applications
NA stands for not available.

assembly parts (isolated DC modules, IGBT driver circuitry, such as electrochemotherapy, gene transfer, electroinsertion
etc.), on the other hand, increase the costs of the device, of proteins into cell plasma membrane, electrofusion of cells,
which is a subject of optimization during the design stage. transdermal drug delivery, water treatment and food preser-
vation [5– 7,15,20 –23,55 – 57]. Efficiency of all these appli-
cations strongly depends on parameters of electric pulses,
3. Discussion which are delivered to the treated object using specially
developed electrodes and electronic devices—electropora-
Nowadays electropermeabilization is widely used in var- tors. Both parts of equipment play equally important role in
ious biological, medical, and biotechnological applications process of electropermeabilization, but in this paper we
122 M. Puc et al. / Bioelectrochemistry 64 (2004) 113–124

have focused exclusively on electroporators and advantages addition, researcher must also be able to set the oscilloscope
and disadvantages of techniques used for generating re- before the experiments, which requires additional training.
quired signals (Table 1). At this point we did not discuss Probably there are many more drawbacks (e.g. expensive
how each of the presented techniques can solve different high voltage probes and current probes) of using the
problems like tissue burning, electrolytic contamination, oscilloscope that could be overcome by built in current
etc., since this would require additional analysis of electrode and voltage monitors.
designs and materials. Although today we can find several new studies showing
Besides reviewing known techniques of signal genera- biological effects of nanosecond pulsed electric fields
tion, we also investigated the world market of electropora- [51,52], we did not review the parameters and technologies
tors. A list of existing commercially available electroporators used, since this has already been done by Mankowski et al.
with their parameters, biological applications and possible [49]. In this review, they have presented several short pulse
signal generation technique are given in Table 2. Devices generator technologies such as discharge of capacitor, pulse
are grouped by manufacturer and each device is presented forming line (PFL), Marxx-generator, etc. Besides this they
with the following parameters: output characteristics, volt- also offer a list of commercially available short pulse
age range, time constant (s)/pulse length (T), and charge generators.
time (tc)/pulse repetition frequency ( f). The value of last In conclusion we can say that even though manufacturers
two parameters depends on output characteristic if the offer a brand variety of electroporators and electroporation
device produces exponentially decaying pulses, time con- systems, these devices still have specific limitations. This
stant and charge time are given as parameters. On the other was probably the main reason why many researchers have
hand, if the device generates square wave pulses, pulse developed their own custom-designed devices or systems.
length and pulse repetition frequency are given as param- Since many of these custom-built devices are poorly de-
eters. Since some of manufacturers also offer different scribed in the articles, we were unable to explore their
electrodes for different applications we have also made a parameters in details. What we offer instead is a list of
list that is given in Table 3. Electrodes are grouped by articles describing the devices (see Refs. [43,44,47 – 57]).
manufacturers and each electrode is presented with the
following parameters: electrode type, electrode distance
and biological applications. Acknowledgements
We can see that it is practically impossible to compare
the listed devices due to difference in their characteristics. This research has been supported through various grants
Even if we compare devices with identical output charac- by the Ministry of Education, Science and Sports of the
teristic (e.g. exponential, square wave, arbitrary) we see that Republic of Slovenia and in part by the European project
either their voltage range or their time constant/pulse length Cliniporator QLK3-1999-00484. The authors wish to thank
vary in incomparable range. We believe that with each of the Dr. Maša Kanudšar, Dr. Tadej Kotnik and Gorazd Pucihar
listed devices adequate experimental results can be for many useful suggestions during the preparation of the
achieved, yet some questions still remain. Do we need manuscript.
any special buffers for electropermeabilization of cells?
How can we set the required parameters for electropermea-
bilization (i.e. user friendliness of the device)? Is the device References
modular or non-modular type (i.e. with addition of new
module we extend working parameters)? From this we can [1] K. Kinosita, T.Y. Tsong, Formation and resealing of pores of con-
conclude that manufacturers of the electroporators have to trolled sizes in human erythrocyte membrane, Nature 268 (1977)
standardize electrical parameters of devices, which would 438 – 441.
[2] J. Teissie, T.Y. Tsong, Electric field induced transient pores in phos-
also include list of required buffers that have to be used for
polipid bilayer vesicles, Biochemistry 20 (1981) 1548 – 1554.
efficient electropermeabilization. This has already been [3] E. Neumann, M.S. Ridder, Y. Wang, P.H. Hofschneider, Gene transfer
done by some manufacturers (Eppendorf, BioRad, BTX, into mouse lyoma cells by electroporation in high electric fields,
etc.) who supply protocols and standardized buffers for EMBO J. 1 (1982) 841 – 845.
different procedures. [4] E. Neumann, A.E. Sowers, C.A. Jordan, Electroporation and Electro-
Besides standardization of parameters of devices, manu- fusion in Cell Biology, Plenum, New York, 1989.
[5] L.M. Mir, Therapeutic perspectives of in vivo cell electropermeabili-
facturers should also start offering a built-in module for zation, Bioelectrochemistry 53 (2000) 1 – 10.
current and voltage monitoring. It is very important that [6] E. Neumann, S. Kakorin, K. Toensing, Fundamentals of electropor-
researcher has an immediate feedback about the electro- ative delivery of drugs and genes, Bioelectrochem. Bioenerg. 48
permeabilizing signal that has been delivered to the electro- (1999) 3 – 16.
des. Monitoring of voltage and current can be performed by [7] J. Teissie, N. Eynard, B. Gabriel, M.P. Rols, Electropermeabilization
of cell membranes, Adv. Drug Deliv. Rev. 35 (1999) 3 – 19.
use of an oscilloscope, but this requires additional space for [8] M.P. Rols, J. Teissie, Electropermeabilization of mammalian cells to
another electronic device in already overstuffed laboratory macromolecules: control by pulse duration, Biophys. J. 75 (1998)
and also additional wiring for signal measurements. In 1415 – 1423.
 M. Puc et al. / Bioelectrochemistry 64 (2004) 113–124 123

[9] T. Kotnik, L.M. Mir, K. Flisar, M. Puc, D. Miklavčič, Cell membrane medium conductivity on electropermeabilization and survival of cells
electropermeabilization by symmetrical bipolar rectangular pulses: in vivo, Bioelectrochemistry 54 (2001) 107 – 115.
Part I. Increased efficiency of permeabilization, Bioelectrochemistry [31] I.G. Abidor, A.I. Barbul, D.Z. Zhev, P. Doinov, I.N. Bandarina, E.M.
54 (2001) 83 – 90. Osipova, S.I. Sukharev, Electrical properties of cell pallet and cell
[10] T. Kotnik, L.M. Mir, D. Miklavčič, Cell membrane electropermeabi- fusion in centrifuge, Biochim. Biophys. Acta 152 (1994) 207 – 218.
lization by symmetrical bipolar rectangular pulses: Part II. Reduced [32] I.G. Abidor, L.-H. Li, S.W. Hui, Studies of cell pallets: II. Osmotic
electrolytic contamination, Bioelectrochemistry 54 (2001) 91 – 95. properties, electroporation and related phenomena: membrane inter-
[11] M. Puc, T. Kotnik, L.M. Mir, D. Miklavčič, Quantitative model of actions, Biophys. J. 67 (1994) 427 – 435.
small molecules uptake after in vitro cell electropermeabilization, [33] R.V. Davalos, B. Rubinsky, D.M. Otten, A feasibility study for elec-
Bioelectrochemistry 60 (2003) 1 – 10. trical impedance tomography as a means to monitor tissue electro-
[12] G. Pucihar, L.M. Mir, D. Miklavčič, The effect of pulse repetition poration for molecular medicine, IEEE Trans. Biomed. Eng. 49
frequency on the uptake into electropermeabilized cells in vitro with (2002) 400 – 403.
possible analysis and it application, Bioelectrochem. Bioenerg. 57 [34] U. Pliquett, E.A. Gift, J.C. Weaver, Determination of the electric field
(2002) 167 – 172. and anomalous heating caused by exponential pulses with aluminum
[13] T. Kotnik, G. Pucihar, M. Reberšek, D. Miklavčič, L.M. Mir, Role of electrodes in electroporation experiments, Bioelectrochem. Bioenerg.
pulse shape in cell membrane electropermeabilization, Biochim. Bio- 39 (1996) 39 – 53.
phys. Acta 1614 (2003) 193 – 200. [35] U. Friderich, N. Stachowicz, A. Simm, G. Fuhr, K. Lucas, U. Zim-
[14] M. Golzio, M.P. Mora, C. Raynaund, C. Delteil, J. Teissie, M.P. Rols, mermann, High efficiency electrotransfection with aluminum electro-
Control by osmotic pressure of voltage-induced permeabilization and des using microsecond controlled pulses, Bioelectrochem. Bioenerg.
gene transfer in mammalian cells, Biophys. J. 74 (1998) 3015 – 3022. 47 (1998) 103 – 111.
[15] J. Teissie, Transfer of foreign receptors to living cell surfaces: the [36] F. Loste, N. Eynard, J. Teissie, Direct monitoring of the field
bioelectrochemical approach, Bioelectrochem. Bioenerg. 46 (1998) strength during electropulsation, Bioelectrochem. Bioenerg. 47
115 – 120. (1998) 119 – 127.
[16] T. Kotnik, F. Bobanović, D. Miklavčič, Sensitivity of transmembrane [37] T. Tomov, I. Tsoneva, Are the stainless steel electrodes inter? Bio-
voltage induced by applied electric fields: a theoretical analysis, Bio- electrochemistry 51 (2000) 207 – 209.
electrochem. Bioenerg. 43 (1997) 285 – 291. [38] M. Pavlin, T. Slivnik, D. Miklavčič, Effective conductivity of cell
[17] P.J. Canatella, J.F. Karr, J.A. Petros, M.R. Prausnitz, Quantitative suspensions, IEEE Trans. Biomed. Eng. 49 (2002) 77 – 80.
study of electroporation-mediated molecular uptake and cell viability, [39] M. Pavlin, D. Miklavčič, Effective conductivity of a suspension of
Biophys. J. 80 (2001) 755 – 764. permeabilized cells: a theoretical analysis, Biophys. J. 85 (2003)
[18] M. Pavlin, N. Pavšelj, D. Miklavčič, Dependence of induced 719 – 729.
transmembrane potential on cell density, arrangement, and cell [40] R.D. Fomekong, U. Pliquett, F. Pliquett, Passive electrical properties
position inside a cell system, IEEE Trans. Biomed. Eng. 49 or RBC suspensions: changes due to distribution of relaxation times
(2002) 605 – 612. in dependence on the cell volume fraction and medium conductivity,
[19] B. Valič, M. Golzio, M. Pavlin, A. Schatz, C. Faurie, B. Gabriel, J. Bioelectrochem. Bioenerg. 47 (1998) 81 – 88.
Teissie, M.P. Rols, D. Miklavčič, Effect on electric field induced trans- [41] G.A. Hofmann, S.B. Dev, G.S. Nanda, Electrochemotherapy: transi-
membrane potential on spheroidal cells: theory and experiments, Eur. tion from laboratory to clinic, IEEE Eng. Med. Biol. Mag. 15 (1996)
Biophys. J. 32 (2003) 519 – 528. 124 – 132.
[20] M.R. Prausnitz, V.G Bose, R. Langer, J.C. Weaver, Electroporation of [42] M. Okino, H. Tomie, H. Kanesada, M. Marumoto, K. Esato, H.
mammalian skin: a mechanism to enhance transdermal drug delivery, Suzuki, Optimal electric conditions in electrical impulse chemother-
P. Natl. Acad. Sci. U. S. A. 90 (1993) 10504 – 10508. apy, Jpn. J. Cancer Res. 83 (1992) 1095 – 1101.
[21] M.R. Prausnitz, A practical assessment of transdermal drug delivery [43] D.C. Chang, Cell poration and cell fusion using an oscillating electric
by skin electroporation, Adv. Drug Deliv. Rev. 35 (1999) 61 – 67. field, Biophys. J. 56 (1989) 641 – 652.
[22] S.B. Dev, D.P. Rabussay, G. Widera, G.A. Hofmann, Medical applica- [44] M. Puc, K. Flisar, S. Reberšek, D. Miklavčič, Electroporator for in
tions of electroporation, IEEE Trans. Plasma Sci. 28 (2000) 206 – 223. vitro cell permeabilization, Radiol. Oncol. 35 (2001) 203 – 207.
[23] J. Gehl, Electroporation: theory and methods, perspectives for drug [45] M. Puc, S. Reberšek, D. Miklavčič, Requirements for a clinical elec-
delivery, gene therapy and research, Acta Physiol. Scand. 177 (2003) trochemotherapy device—electroporator, Radiol. Oncol. 31 (1997)
437 – 447. 368 – 373.
[24] D. Miklavčič, K. Beravs, D. Šemrov, M. Čemažar, G. Serša, The [46] E. Tekle, R.D. Astumian, P.B. Chock, Electroporation by using
importance of electric field distribution for effective in vivo electro- bipolar oscillating electric field: an improved method foe DNA
poration of tissue, Biophys. J. 74 (1998) 2152 – 2158. transfection of NIH 3T3 cell, P. Natl. Acad. Sci. U. S. A. 88 (1991)
[25] J. Gehl, T.H. Sorensen, K. Nielsen, P. Raskmark, S.L. Nielsen, T. 4230 – 4234.
Skovsgaard, L.M. Mir, In vivo electroporation of skeletal muscle: [47] K. Flisar, M. Puc, T. Kotnik, D. Miklavčič, A system for in vitro cell
threshold, efficacy and relation to electric field distribution, Bio- membrane electropermeabilization with arbitrary pulse waveforms,
chim. Biophys. Acta 1428 (1999) 233 – 240. IEEE Eng. Med. Biol. Mag. 22 (2003) 77 – 81.
[26] G. Serša, M. Čemažar, D. Šemrov, D. Miklavčič, Changing electrode [48] M. Petkovšek, J. Nastran, D. Vončina, P. Zajec, D. Miklavčič, G.
orientation improves the efficacy of electrochemotherapy of solid Serša, High voltage pulse generation, Electron. Lett. 38 (2002)
tumors in mice, Bioelectrochem. Bioenerg. 39 (1996) 61 – 66. 680 – 682.
[27] M.P. Rols, J. Teissie, Ionic-strength modulation of electrically induced [49] J. Mankowski, M. Kristiansen, A review of short pulse generator
permeabilization and associated fusion of mammalian cells, Eur. J. technology, IEEE Trans. Plasma Sci. 28 (2000) 102 – 108.
Biochem. 179 (1989) 109 – 115. [50] Y.C. Lin, C.M. Jen, M.Y. Huang, C.Y Wu, X.Z. Lin, Electroporation
[28] E. Neumann, Membrane electroporation and direct gene transfer, Bio- microchips for continuous gene transfection, Sens. Actuators, B,
electrochem. Bioenerg. 28 (1992) 247 – 267. Chem. 79 (2001) 137 – 143.
[29] C.S. Djuzenova, U. Zimmermann, H. Frank, V.L. Sukhorukov, E. [51] S.J. Beebe, P.M. Fox, L.J. Rec, K. Somers, R.H. Stark, K.H.
Richter, G. Fuhr, Effect of medium conductivity and composition Schoenbach, Nanosecond pulsed electric field (nsPEF) effects on
on the uptake of propidium iodide into electropermeabilized mye- cells and tissues: apoptosis induced and tumor growth inhibition,
loma cells, Biochim. Biophys. Acta 1284 (1996) 143 – 152. IEEE Trans. Plasma Sci. 30 (2002) 286 – 292.
[30] G. Pucihar, T. Kotnik, M. Kandušer, D. Miklavčič, The influence of [52] K.H. Schoenbach, S. Katsuki, R.H. Stark, E.S. Buescher, S.J. Beebe,
124 M. Puc et al. / Bioelectrochemistry 64 (2004) 113–124

Bioelectrics—new applications for pulsed power technology, IEEE Escherichia coli in saline water with high electric field AC, J. Food
Trans. Plasma Sci. 30 (2002) 293 – 300. Eng. 53 (2002) 203 – 207.
[53] A.C. Durieux, R. Bonnefoy, C. Manissolle, D. Freyssenet, High-effi- [56] J. Teissie, N. Eynard, M.C. Vernhes, A. Benichou, V. Ganeva, B.
ciency gene electrotransfer into skeletal muscle: description and phys- Galutzov, P.A. Cabanes, Recent biotechnological developments of
iological applicability of a new pulse generator, Biochem. Biophys. electropulsation. A prospective review, Bioelectrochemistry 55 (2002)
Res. Commun. 296 (2002) 443 – 450. 107 – 112.
[54] S. Šatkauskas, M.F. Bureau, M. Puc, A. Mahfoudi, D. Scherman, D. [57] K. Uemura, S. Isobe, Developing a new apparatus for inactivat-
Miklavčič, L.M. Mir, Mechanisms of in vivo DNA electrotransfer: ing bacillus subtilis spore in orange juice with a high electric
respective contributions of cell electropermeabilization and DNA field AC under pressurized conditions, J. Food Eng. 56 (2003)
electrophoresis, Molec. Ther. 5 (2002) 133 – 140. 325 – 329.
[55] K. Uemura, S. Isobe, Developing a new apparatus for inactivating