Electroporation optimization guide.
Vadim A. Klenchin, University of Wisconsin
This is an attempt to aid rational optimization of
electroporation conditions. It is based on my posts
to Usenet newsgroup bionet.molbio.methds-reagnts in
response to numerous and varied questions about
electroporation. Below, I summarize my experience and
recommendations based both on practical and theoretical
A very brief summary:
Chose electroporation medium that allows highest survival
rate for your particular cell type, use highest capacitance
your electroporator can deliver, use cells at ~ 30 millions/ml,
plasmid at ~ 0.1 mg/ml in the final volume of ~ 400 ul in
0.4 cm cuvette at room temperature without preincubation of
DNA with cells. Having fixed all these parameters, "titrate"
output voltage to find optimal condition.
The detailed version:
The mechanism of the phenomenon is electrophoretically driven
DNA insertion through electropores (see Klenchin et al.,
Biophys. J., 1991 and Sukharev et al., ibid, 1992; numerous
other papers confirmed all of the conclusions of these
studies). Electric field therefore plays dual role - first it
creates electropores in lipid bilayer by a mechanism similar
to a dielecric breakdown, next it is essential for DNA
The number and size of electropores, as well as their ability
to eventually reseal, is a very complex function of electric
field intensity (E, e.g. V/cm) and pulse duration. Generally
speaking, however, the apparent "critical" E depends linearly
on cell diameter. (That is why we typically use < 1 kV/cm for
mammalian cells and > 10 kV/cm for bacterial).
DNA electrophoresis, on another hand, is a simple linear
function of the integral of E over time (E x t in case
of square pulse). There is no "critical" voltage value
below which no DNA electrophoresis takes place.
Experimentally, it is quite clear that, even with a comparable
electropermeabilization levels, cells survive better when
treated with pulses of low amplitude/long duration than those
of high amplitude/short duration. Therefore:
1a. Pulse generated by a simple capacitor discharge
(e.g., most commercial electroporators; meaning exponential
decline of E over time) should be low amplitude/long duration
exponent (under these conditions pores created are larger and
live longer; long electroporative and less damaging "tail"
works for electrophoresis). In practical terms:
Use the largest capacitor available on your electroporator
and find other optimal parameters. ~ 1000 microF capacitance
seems to be a sweet spot. Few electroporators have larger
capacitance and in my own experience a switch from
1 mF to 3 mF did not produce significant enough difference.
1b. The characteristic discharge time (tau, also
called "pulse duration"; more precisely, time it takes for
the voltage to drop by number e times) is a product of your
the circuit's electrical resistance and capacitor's electric
capacity (tau = R (Ohms) x C (Farads)). R is usually your
sample's resistance. R is inverse linear function of sample
cross-section area (i.e., suspension volume for a given
inter-electrode distance). Therefore, using lower volumes
gives longer pulses. That is, other things being equal,
0.5 ml in standard 0.4 cm cuvette would be "better" than
1c. Another way of increasing pulse duration is
to use electroporation medium with low ionic strength.
Typically, these would be buffered solutions of 250-300 mM
sucrose to provide physiological osmolarity. Certainly
other sugars that are well-tolerated by your type of cells
are also an option. In practice, for not quite known reason(s),
salt-based electroporation media seem to give better
efficiencies than sucrose-based. However, if you only have
access to 25 microF capacitors, trying a low-conducting
medium may very well give higher efficiencies due to the
longer pulse duration.
The probability for the DNA to get to the pore increases
with increasing both cells and DNA concentration. Therefore:
2. Increase both. It all depends on exact situation but
generally use final cell suspension at 3x10e7 - 4x10e7/ml
and DNA at 50-200 ug/ml. Going with lower suspension volumes
is a cheap way to achieve higher concentrations. (Although
with less than 400 ul in 0.4 cm cuvette is becoming not
terribly well reproducible because of the edge effects/shape
of meniscus, etc).
At some point, however, increasing DNA concentration
becomes inhibitory. Two reasons for this: i) increasing
concentration of "crap" contaminants that come with plasmid
preps, ii) DNA is cytotoxic during electroporation because it,
being a very large molecule forced through a very small
electropore, enlarges pores sometimes to the point that they
So the best you can do is to ensure that DNA is as
clean as possible. Cesium chloride is best but too tedious an
option, the next best is anion exchange-based purification
(e.g. Qiagen's Maxi prep kits, etc). For routine experiments,
pooled and EtOH precipitated minipreps is a valid option.
Literature suggests that some cells are much more sensitive
to DNA contaminants (LPS?) than others.
3. Room temperature works better in all cases for large
cells (mammalian, insect, etc). This is probably because
lipid bilayer is more fluid and electropores form easier and
reseal better at higher temperatures. The limitation to this
comes when working with very small cells that require very high
electric field intensities. In such cases the heat generated is
just too high, effectively boiling the cells (Joule's law,
heat production is proportional to the square of voltage).
4. It depends, but generally supercoiled and circular
plasmids work better for transient, while linearized plasmids
work better for stable transfections and homologous recombination.
Effects here could be small or large depending on cell line, the
plasmid itself and the transfection's ultimate signal.
5. No prepulse incubation of DNA is necessary. In
fact, if the electroporation medium contains Mg2+ (a good thing,
see below), it may be deleterious dure to the risk of nuclease
activity secreted by cells and degrading your plasmid becomes real.
I've seen DNA degraded pretty thoroughly in CHO suspensions in
Mg2+ containing PBS in as little as 5 min at RT.
6. I wouldn't bother with more than one pulse.
The apparent success of radiofrequency electoporation
suggests otherwise but it's a far cry from regular long
pulses. I've tried two pulses and/or second pulse with
reversing polarity and never got an an optimum that was
better than a single pulse protocol.
7. While no postpulse incubation is necessary, ~ 5 min
after pulse does not affect survival significantly. In my
experience, cells should not be diluted right away into warm
culture medium. Use RT medium and let pores reseal for ~ 20 min
before placing cells in 37C CO2 incubator.
By far the trickiest issue is electroporation medium. Effect
on the level of electricity is simple: in low salt the pulse is
much longer then in high salt (but in high [sucrose] the same
amplitude/duration gives lower frequency, probably having
something to do with lipid behavior). All other effects are
on the level of cell survival and biological/transcitptional
responses downstream. Basically, you want to find a solution
that give highest survival rate after harsh electroporation
(no need to titrate voltage here; just chose something extreme
that kills 70% cells in PBS or any other basic solution). One
spend lifetime optimizing components of the buffer. Most people
use cell culture media (DMEM, RPMI, etc. - with or without
5-10% serum). I don't doubt that it works but I don't believe
this is optimal in most cases.
8. Here is my approach to the problem. A simple buffer
that appears to work quite well for a variety of cell types
(including even Drosophila cell line and Dictyostelium cells)
and, in my hands, was consistently better or as good as most
other things described in the literature:
20 mM HEPES, 135 mM KCl, 2 mM MgCl2, 0.5% Ficoll 400,
final pH adjusted to 7.6 with NaOH. (Don't autoclave,
The idea here is that the cells after electropermeabilization
get exposed to a solution that looks more like intracellular
ionic composition, high K+ and low Na+. Good buffering and Mg2+
are essential for good cell survival, 0.5% Ficoll 400 in
my experiments quite dramatically increased survival rate
of Cos-1 cells for unknown reasons (even if it does nothing
for other cells, it more than likely can't hurt at that
concentration). A good suggestion that came from a 1992 paper
(Nuceic Acids Res, 1992, 20:2902) describing a similar
"Cytomix" solution is to add, right before electroporation,
a 1/100 volume of a stock of 200 mM ATP/500 mM glutathione.
Both of these components promote cell suirvival thus allowing
harsher (and hence more efficient) electrical conditions. It is
VERY important to pH this stock solution to pH > 7.0 as both
compounds typically come as strong acids. (Filter sterilize,
aliquote and store at -20C). Personally, I always add
ATP/glutathione and feel they help a quite a bit.
The Cytomix composition is this: 120 mM KCl, 0.15 mM CaCl2,
10 mM K2HPO4, 25 mM HEPES, 2 mM EDTA, 5 mM MgCl2.
(It looks good and definitely works well. My problem with
it is the inclusion of Ca2+ that is quite toxic during
electroporation of at least some fibroblast cells and too
high EDTA/Mg2+ concentration that I can't see the reason for;
plus, above 2 mM Mg2+ there is no benefit in terms of survival
rate. In fact, at higher concentrations Mg2+ effect of
shielding DNA's charges becomes noticeable and leads to a
slightly lower transfection efficiency).
Since Cl- is not a predominant anion in the cell, I've
always wanted to test how beneficial replacement of KCl
for K gluconate or glutamate or isethionate will be.
Never got around to back-to-back comparison...
In terms of other things to try:
- "UW solution" is a transplant storage solution that is
sold as ready to go preparation ("ViaSpan") and it seems like
a good composition well-suitable for electroporation (at least one
paper mentioned it offering great advantage);
- Addition of ~ 0.5-1% DMSO was reported to give increased
efficiency; no experience here but I can confirm that for
E.coli 7% DMSO is 2-fold better than 10% glycerol;
- Addition of ~50 mM trehalose was reported to work well.
Not sure what the effect is here. Could be just compensating
for osmolarity or might have specific effects on membranes
or even cell physiology. Same as Ficoll 400?
- Plating cells into medium containing butirate (~ 2.5 mM?
probably needs fine tuning fir each cell type). Butirate
is a non-specific transcriptional activator and it's been
reportedly shown to be beneficial in many transfection
- One might consider using cell cycle synchronized cultures.
The cell size is going to be more uniform and there are
reports that cells in G2 are more "transfection competent".
The bottom line about electroporation medium: if your needs
are casual (some number of transformants, some well-detectable
transient expression), don't bother optimizing too much. If
you are on the edge and absolutely require efficiencies as high
as could possibly be achieved, playing with electroporation
medium might pay very well. Increasing of survival rate affords
delivering higher intensity pulses, which means better
9. Assuming you've made a sound choice of most
variables, fix them all and perform a single most important
optimization: find the optimal voltage. First, "titrate"
voltage to find conditions where 30-60% of cells survive.
Ideally, this should be done by plating and counting next
day, not Trypan Blue staining test (T.B. underestimates
death dramatically). It is better to do this test *with*
DNA present since it increases cell death during
electroporation. Once you have established a reasonable range,
repeat the voltage titration within this range with narrow
increments - this time scoring the actual transfection
10. Electroporation of bacteria and yeasts is altogether
a different issue. While the basic physics is the same, small
size of these organisms requires much higher voltages that pose
overheating risk. This leads to a completely different set of
"rules" and recommendations. Also, I have absolutely no
experience with electroporation of plant cells. Apart from
media compositions (and issues with nucleases?), the same
set of basic considerations should be equally applicable to