Improving Electrotransfection Efficiency by Post-Pulse Centrifugation

LH Li1,2, P Ross1 and SW Hui1 1Membrane Biophysics Laboratory, Roswell Park Cancer Institute, Buffalo, NY, USA

We have demonstrated that the viability of electro- dorf desktop centrifuge. Pelleting improves the cell viability transfected adherent CHO and suspended NK-L, K-562, over the whole range of the NK-L, K-562, L1210 and MC2 L1210 and MC2 cells is improved if pelleting by centrifug- cell concentrations studied. When this pelleting method is ation is performed immediately after pulsing. The protec- applied to load CHO cells with FITC-dextran (41 000 MW), tion effect on cell viability is cell line- and pellet thickness- not only is the success rate close to 100%, but the growth dependent. For forming CHO cell pellets, centrifugation rate is similar to the control, which is far better than the force (300–13 000 g) and duration are not crucial; about conventional electroporation method. Furthermore, the five to 10 cell layers in the pellet provide the optimal protec- transfection efficiency of the five cell lines in pellet is sig- tion effect. NK-L, K-562, L1210 and MC2 cell pellets are nificantly higher than that in suspension. optimally formed by centrifugation at 13 000 g in an Eppen-

Keywords: centrifugation pellet; viability; electroporation; electroloading; electrotransfection

Introduction method that can improve cell viability is likely to improve electrotransfection efficiency. Electroporation has become a popular method in biotech- Pelleting by centrifugation effectively reduces extra- nology for transferring genetic materials, as well as other cellular volume between cells and leads to reduced biochemicals, such as foreign proteins and drugs, into erythrocyte electrohemolysis by restricting post-pulse cell 1–3 cells. The advantage of this method is that it can be swelling.10,15 This feature may be used to reduce post- 4,5 applied to almost all cell types. Also, the parameters in pulse cell mortality in cultured cells. Furthermore, post- electroporation are easier to control and the efficiency in pulse pelleting does not interfere with the pulsing pro- 6–9 some cases is higher than those by other methods. cedure. The combination of the post-pulse pelleting tech- Moreover, it avoids biological contamination that poten- nique and the conventional electroporation method may tially exists in virus-induced methods. It offers an attract- improve cell viability, thereby increase electrotransfec- ive way for ex vivo gene delivery if the transfection tion and electroloading efficiency. efficiency can be significantly improved. In this article, we demonstrated that the post-pulse pel- Electroporation includes two major steps. The first step leting technique could improve cell viability of pulsed is the temporary breakdown of the membrane barrier to cultured cells. Five cell lines were studied, an adherent allow uptake of exogenous molecules, such as DNA or cultured cell line CHO cells and four suspension cultured 10,11 other materials, by diffusion or electrophoresis. The cell line, NK-L, K-562, L1210 and MC2 cells. We deter- second step is to allow cell membranes to recover to their mined that the post-pulse pelleting technique improves 7 original impermeable state. Initially, in the reversible the electroloading and electrotransfection efficiency by breakdown step, the higher the pulse energy (including improving cell viability in these five cell lines. pulse duration and pulse strength), the higher the intake by electroloading.12 If the pulse electrical field is too strong, an irreversible breakdown occurs, and the Results efficiency of electrotransfection and electroloading suffers due to the drop of cell viability.13,14 It is found that the Effects of electroporation and pelleting on post-pulse best electrotransfection and electroloading occur when cell viability 8,12 the cell viability is approximately 50%. This viability Our goal is two-fold. The first is to determine if post- problem is especially serious in the application of elec- pulse pelleting can improve the viability of pulsed cul- troporation for gene delivery in some cell lines, for tured cells; if the result is positive, the second goal is to example, natural killer cells and lymphomas. Any apply this technique to improve the efficiency of electrotransfection and electroloading. It is known that post- pulse pelleting can decrease the electrolysis of pulsed Correspondence: SW Hui, Roswell Park Cancer Institute, Elm & Carlton erythrocytes due to the inhibition of colloidal osmotic Streets, Buffalo, NY 14263, USA 10,15 2On leave from: Biomedical Engineering Department, Hunan Medical swelling. The thicker the post-pulse pellet, the lower University, Changsha, PR China the electrolysis. However, the thick pellet also poses a Received 13 March 1998; accepted 18 September 1998 hindrance to the supply of nutrients. Therefore, one may Improving electroporation efficiency by pellet LH Li et al 365 expect that there is an optimal thickness of the pellet for post-pulse viability of cultured cells. To check this idea, the dependence of the viability of the pulsed CHO cells on cell density was examined, as shown in Figure 1a. For cells incubated in pellet, the optimal viability occurs in the concentration range of 8 × 106/ml–16 × 106/ml (120 000–240 000 total cells). The viability is very low at cell densities less than 8 × 106/ml, and decreases gradually when cell density exceeds 16 × 106/ml. The decrease of cell viability in pellet with the increase of cell density shows the adverse effect of pellet thickness on cell viability once it exceeds the optimal thickness. For cells incubated in suspension, the viability increases with the increase of the cell density. When the cell density reaches 40 × 106/ml (corresponding to 600 000 total cells), there is little difference in viability between cells incubated in pellet or in suspension. When cell density is low (Ͻ16 × 106/ml), the viability of cells incubated in pellet is much higher than that incubated in suspension. To illustrate the advantage of pelleting at low cell density further, we pulsed CHO cells at the highest cell density (80 × 106/ml, corresponding to 1 200 000 cells) and resuspended the pulsed cells in 115 ␮l of culture medium (about eight times dilution, 10 × 106/ml). The diluted sample was divided into two parts, one with 120 000 cells and pelleted, the other with the remaining cells in suspension. The resultant viability is shown in Figure 1b. Again, the viability of cells incubated in the post-pulse pellet is much higher than that incubated in suspension. There is no pulse treatment for cells in the control sample. The effect of the pellet thickness on post-pulse cell viability is studied by changing the bottom surface area of the centrifuge chambers. Cylindrical chambers with different flattened bottom surface areas were used to allow the estimation of relative pellet thickness, using the same number of cells. Cells were centrifuged by a table- top centrifuge at 2200 g for 0.5 min (IEC HN-S Centrifuge, Needham Heights, MA, USA). In Figure 1c, the dependence of cell viability on relative pellet thickness, defined as cell number per unit bottom surface area, is shown. The shape of the curves in Figure 1c is similar to that in Figure 1a (solid circles), though a different centrifugation force was used. Also, the thickness of the pellet for optimal viability occurs at the same value for three chambers with different bottom diameters, 2, 3 and 4.3 mm, respectively. The effect of post-pulse manipulation on cell viability Figure 1 (a) The dependence of CHO cell viability on cell density. CHO was examined. Figure 2 shows the dependence of the cells were pulsed by four 2.5 kV/cm, 400 ␮s pulses. Solid circles represent pulsed cell viability on the delay time between electro- the viability of cells incubated in pellet, while the empty circles represent poration and centrifugation. When cells were centrifuged those incubated in suspension. (b) The dependence of CHO cell viability at a delay time of 5 min or more after pulses, the cell on incubation methods. CHO cells were pulsed by four 2.5 kV/cm, 400 ␮s × 6 viability decreases to almost zero. It is clear that cells pulses in a cell density of 80 10 /ml (1 200 000 total cells), diluted eight times and divided into two parts. Pellet column represents the viability have to be pelleted immediately (usually within 0.5 min) of one part with 120 000 CHO cells incubated in pellet; suspension column after pulses to achieve high cell viability. represents the viability of the other part with 1 000 000 CHO cells incu- Centrifugation time is another concern of how post- bated in suspension. Control column is the viability without pulse appli- pulse centrifugation affects cell viability. As shown in cation. (c) The dependence of CHO cell viability on relative pellet thickness Figure 3, the centrifugation (13 000 g) times, within a 3 to (cell number per unit bottom surface area). CHO cells were pulsed by ␮ 60 s range, does not affect the CHO cell viability if cells four 2.5 kV/ml, 400 s pulses. Circles, squares and triangles represent the cell viability incubated as pellets in cylindrical chambers with 4.3, 3 and are centrifuged immediately after pulses. 2 mm bottom diameter, respectively. Suspension NK-L, K-562, L1210 and MC2 cells were used to test how generally this pelleting method could be applied. Similar advantages in the viability were found for cells incubated in pellet as against those incu- Improving electroporation efficiency by pellet LH Li et al 366

Figure 2 The dependence of CHO cell viability on the delay time between electroporation and centrifugation. Cells were pulsed by four 2.5 kV/cm, 400 ␮s pulses.

Figure 3 The dependence of CHO cell viability on the centrifugation time (13 000 g). Cells were pulsed by four 2.5 kV/cm, 400 ␮s pulses.

bated in suspension, as shown in Figure 4a, b and c. It has to be mentioned that Figure 4c was done in a more stringent way in testing the effectiveness of the pellet method. The pulsed L1210 cells were resuspended and split into two equal parts, one for incubation in pellet and the other in suspension.

Improvement of cell loading and transfection by post- pulse pelleting The knowledge derived from cell viability measurements is applied to improve cell loading and transfection. Fig- Figure 4 (a) The dependence of NK-L cell viability on cell density. Cells ure 5 demonstrates that CHO cells are successfully were pulsed by four 2.1 kV/cm, 400 ␮s pulses. Solid circles and empty loaded with FITC-dextran (41 000 Mr) by the post-pulse circles represent the viability of cells incubated in pellet and suspension, pelleting technique. Phase-contrast micrographs of respectively. (b) The dependence of K-562 cell viability on cell density. pulsed-pelleted (Figure 5c) and unpulsed control (Figure Cells were pulsed by four 2.3 kV/cm, 400 ␮s pulses. Solid squares and 5a) cells indicate that the growth rate of pulsed-pelleted empty squares represent the viability of cells incubated in pellet and suspension, respectively. (c) The dependence of L1210 cell viability on cell cells is not affected by pulses. Fluorescent micrographs density. Cells were pulsed by up to four 1.7 kV/cm, 400 ␮s pulses. After of pulsed-pelleted cells (Figure 5d) and unpulsed control pulses, the cells were split into two equal parts, one for pellet incubation cells (Figure 5b) demonstrate that the FITC-dextran was (᭿, ̆, ᭹), the other for suspension incubation (,̅,᭺). successfully loaded into the pulsed-pelleted cells, but not into control cells. The phase-contrast micrograph of improve transfection. Phase-contrast micrographs of pulsed but unpelleted cells (Figure 5e) shows the Figure 6a and b show the viability and the transfection decrease of a growth rate, but with successful loading of of CHO cells by pulsed-pelleted and pulsed-unpelleted FITC-dextran into those living cells (Figure 5f). methods, with all other conditions the same. Similarly, The post-pulse pelleting technique was applied to Figure 6c and d correspond to the transfection of pulsed- Improving electroporation efﬁciency by pellet LH Li et al 367

␮ Figure 5 Micrographs of electroloading of CHO cells with 20 mg/ml FITC-dextran (41 000 Mr) by four 2.5 kV/cm, 400 s pulses. Micrographs a, c and e represent the control, pulsed-pelleted, and pulsed-nonpelleted cells observed by phase contrast microscope respectively; b, d and f represent the control, pulsed-pelleted, and pulsed-nonpelleted cells observed by ﬂuorescent microscope. Bar is 30 ␮m.

pelleted and pulsed-unpelleted NK-L cells (1.6 × 107/ml), Furthermore, three more cell lines, K-562, L1210 and respectively. The effectiveness of the post-pulse pelleting MC2, were used to check the effectiveness of the pellet method was additionally demonstrated by comparing the method. The expressed luciferase is always more in pel- transfection of NK-L cells, by using plasmids coding for leted than in suspension cells, as shown in Figures 7b, c green fluorescent protein. As shown in Figure 7a, when and d. It needs to note that L1210 and MC2 cells, in Fig- three cell densities were compared, the TE (transfection ure 7c and d respectively, were done in a slightly differ- efficiency) by the conventional method is significantly ent and, in a sense, more stringent manner than for CHO lower than that by the post-pulse pellet technique. and NK-L cells. Instead of pulsing and incubating in Improving electroporation efficiency by pellet LH Li et al 368

Figure 6 Micrographs of electrotransfection of CHO and NK-L cells with (250 ␮g/ml pSV-␤-gal). Cells were pulsed by four 400 ␮s pulses with 2.5 kV/cm for CHO cells and 2.1 kV/cm for NK-L cells. Micrographs a and b represent pulsed-pelleted and pulsed-nonpelleted CHO cells, respectively. Micrographs c and d represent pulsed-pelleted and pulsed-nonpelleted NK-L cells, respectively. Bar is 45 ␮m.

sequential lots, the pulsed cells were immediately resus- Discussion pended and split into two equal parts, one for incubation in the pellet form, the other in suspension. The events occurred after electroporation include the In order to see if the advantage of the pelleting method natural resealing process of the electropores on cell mem- applies also to cuvette chambers and exponential decay branes, and the post-pulse colloidal osmotic swelling.16,17 pulses used in most commercial apparatus, L1210 cells The relative time rates of these two processes determines were used in experiments with this setup. The transfec- if cells die or survive.3,10 There are several ways to reduce tion efﬁciency in pellet is much higher than that in sus- post-pulse cell swelling.10,17 Forming a pellet immediately pension, for both cell concentrations, as shown in Figure after pulse would reduce erythrocyte swelling,10,15 and 8. It indicates that the advantage of the pelleting method therefore give more time for cells to reseal and reduce is not restricted only to using our mini-chamber and cell lysis. In this study, we set to test if what we learned square pulses. from erythrocytes10 and hybrid cells3 can be applied to Improving electroporation efﬁciency by pellet LH Li et al 369

Figure 7 (a) The fluorescence intensity (per 750 000 NK-L cells) of green fluorescent protein after being electrotransfected (250 ␮g/ml pEGFP-N1 plasmid). Cells of three densities were pulsed by four 2.1 kV/cm, 400 ␮s pulses and incubated in pellet (empty columns) or nonpellet (filled columns) after pulses. (b) The dependence of transfection efficiency of K-562 cells on cell density. Cells were pulsed by four 2.3 kV/cm, 400 ␮s pulses. Solid squares and empty squares represent the transfection efficiency of cells incubated in pellet and suspension, respectively. (c) The dependence of transfection efficiency of L1210 cells on cell density. Cells were pulsed by one, two and four 1.7 kV/cm, 400 ␮s pulses, respectively. After pulses, cells were splited into two equal parts, one for pellet incubation (í, ̆, b), the other for suspension incubation (, ̅, ᭺). (d) The dependence of transfection efficiency of MC2 cells on cell density. Cells were pulsed by two and four 2.3 kV/cm, 400 ␮s pulses, respectively. After pulses, cells were split into two equal parts, one for pellet incubation (í, ̆), the other for suspension incubation (, ᭝). improve the transfection of cultured cells in general. If incubated in pellet (Figure 1a). Cells suspended in high the results are proven positive, this method can be very density (50 × 106/ml CHO cells) are like those in a loose useful for in vitro and ex vivo gene delivery. pellet, even more so if cells settle under gravitation dur- Certain limitations apply to the pellet thickness, as ing incubation. It is possible that this loose pellet was indicated from Figure 1a. The post-pulse pellet formed responsible for the high cell viability when cells were by the cell density of 8–16 × 106/ml gives the highest pulsed and incubated in suspension at high cell density. protection effect for CHO cells. At lower cell densities, This phenomenon was also observed in NK-L and K-562 the pellets become too thin and behave more like cells in cells (Figure 4a and b). Figure 1b supports this expla- suspension. The viability falls to the level of cells incu- nation, since eight times dilution of pulsed cells at 80 × bated in suspension. At high cell densities, the gradual 106/ml cell density (1 200 000 total cells) caused the cell decrease of viability is most likely because of the restric- viability to decrease significantly when cells were incu- tion of nutrient supply in thick pellets, therefore slowing bated in suspension. Furthermore, high cell density also cell recovery. Figure 1c further supports this hypothesis. reduces the electric conductivity of the suspension and The optimal pellet thickness for cell viability coincides, may be partially responsible for the enhancement of the regardless of chamber bottom surface area. The common cell viability by reducing the pulse current.18 optimal pellet thickness is estimated to be about five to To generalize the post-pulse pelleting technique, we 10 layers of cells. tested four more cell lines, NK-L, K-562, L1210 and MC2 When incubated in suspension, cell viability increases cells. We found the same protection effect of pellet on with cell density and reaches a level comparable to those viability (Figure 4a, b and c). However, there is no Improving electroporation efficiency by pellet LH Li et al 370 sion method is that pelleting results in higher cell viability, therefore many more loaded cells could survive the electrical trauma. Electroloading of neutral molecules, such as dextran, could not utilize the whole advantage of the post-pulse pelleting technique, since the loading efficiency is mostly dependent on molecular diffusion through pores after pulses.3,6,13 In the pellet system, immediate centrifugation is the key to improve cell viability. This process limits the time range during which molecules can diffuse into electroporated cells. On the other hand, the post-pulse pelleting technique should be good for loading charged molecules, such as DNA, since it is known that electrophoresis of DNA during pulses is the major mechanism for electrotransfection of cells.7,11 Centrifugation immediately after pulses is not expected to significantly reduce the efficiency of DNA internalization.20 With improved cell viability in the post-pulse centrifugation pellet, the Figure 8 Transfection efficiency of L1210 cells pulsed by exponential pulses in 1 ml cuvette chamber (1.7 kV/cm peak field strength, 200 ␮s half- pelleting technique should greatly enhance the electro- time). After pulses, cells were split into two equal parts, one for incubation transfection efficiency. This is indeed the case as shown in pellet (í, ̆), the other for suspension (, ᭝). Pellet 1 (í) and suspen- in Figure 6a, b, c and d for CHO and NK-L cells. From sion 1 () were pulsed at cell concentration of 2 × 106/ml; pellet 2 (̆) the micrographs, one can estimate that about 30–50% of ᭝ × 5 and suspension 2 ( ) were pulsed at cell concentration of 6 10 /ml. CHO cells and 10–15% of NK-L cells could be transfected as assayed by in situ ␤-gal staining. When the plasmid pEGFP-N1 is used, about 30–50% of NK-L cells could be sharply defined optimal pellet thickness effect on cell transfected (data not shown). The quantitative data in viability of pulsed NK-L, K-562 and L1210 cells. Figures Figure 7a, b, c and d, and Figure 8 gave further evidence 1a and 4a, b and c suggest that, although the general that post-pulse pellet technique is better than conven- trend, ie the post-pulse pelleting could improve pulsed tional electroporation method. It is interesting to note cell viability, one still has to find the exact optimal con- that the pellet method demonstrates a stronger effect ditions, eg cell density, the electrical pulse parameters, when high pulse conditions are used, as shown in Figure and incubation time both before and after pulse. 4c, 7c and 8. Because cells are believed to be susceptible to centri- The pellet method improves cell viability only if the fugation disruption immediately after electroporation,19 colloidal osmotic effect is the major factor in cell killing. special attention was paid to the centrifugation effect on Since different cell lines may behave differently in pellets, pulsed cells. As shown in Figures 1a and b, 2, 4a, b and it is not surprising that the pellet effect is cell-line depen- c and 5, it is clear that immediate centrifugation is neces- dent. In fact, 3.3 lymphomas do not show significant sary in order to obtain the protection effect of pelleting improvement in both viability and transfection efficiency on cell viability. If the pulsed cells stay in suspension for in pellets because of electrotransfection-induced large 5 min and more, the centrifugation effect diminishes scale apoptosis (data not shown). (Figure 2). Surprisingly, centrifugation time (Figure 3) The two-phase method6 also improves cell viability by and centrifugation force from 13 000 g to 170 g (data not restricting post-pulse colloidal osmotic swelling. How- shown) did not affect CHO cell viability, but only 13 000 g ever, its major advantage is that it can also concentrate centrifugation can protect the pulsed NK-L cells from DNA and desired cells into one phase by adjusting poly- lysing (data not shown). The difference may be because mer composition or ratio, therefore it can transfect cells different cell lines (CHO and NK-L cells) pack differently selectively. Both methods can transfect certain cell lines in pellet, have differences in surface charge density and more successfully than conventional electroporation membrane speculation, and have different swelling lim- methods. If CHO cells were used to compare these two its. Understanding the protection effect of immediate cen- methods, the transfection efficiency is comparable. How- trifugation on cell viability allows us to define optimal ever, the pellet method is easier to perform, and only one conditions for pulsing and incubating processes separ- more parameter of cell concentration needs to be optim- ately, and gives us independent control of pulse and ized, in addition to those applied to the conventional incubation conditions to achieve optimal electro- electroporation method. The advantage of pelleting in poration efficiency. electrotransfection and electroloading is particularly The understanding of the benefit of pelleting on cell attractive where cell sources are limited, as in many ex viability helps us to enhance electroporation efficiency. vivo situations. The electroloading of macromolecules to CHO cells was first tried. As shown in Figure 5c and d, the loading Materials and methods efficiency of 40 000 Mr FITC-dextran was almost 100%. More importantly, the growth of loaded cells is as high Adherent CHO (Chinese hamster ovary) cells were cul- as the control (Figure 5a and c). As shown in Figure 5e tured in F10 medium with 15% newborn calf serum and and f, cells electroloaded by the conventional suspension 0.6% PSN antibiotic mixture (100×) (Gibco, Grand Island, method also achieved 100% loading for the surviving NY, USA) in 100 mm culture dishes (Costar, Cambridge, cells. The only difference in outcome by using the post- MA, USA). Natural killer cells (NK-L cells), erythroleuke- pulse pelleting technique and the conventional suspen- mia cells (K562 cells), L1210 lymphomas, murine Improving electroporation efficiency by pellet LH Li et al 371 mammary tumor cells (MC2) were cultured in RPMI-1640 41 000) was first dialyzed (molecular cut-off 12 000– with 0.6% PSN antibiotics mixture (100×) in 25-cm2 tissue 14 000, Spectrum Medical Industrials, Los Angeles, CA, culture flasks (Corning Glass Work, Corning, NY, USA). USA) to remove free FITC. The dialyzed FITC-dextran NK-L cells were cultured in 20% FCS (fetal calf serum) (20 mg/ml as the final concentration) was mixed with with 20 ng/ml IL-2 (Interleukin-2), all other suspension cells in culture medium. The cells were ready to be cells were cultured in 10% FCS. Cells were cultured in pulsed. Fluorescence and phase micrographs were taken 5% CO2. Only cells in the exponential growth phase, with after overnight growing. Culture medium was gently Ͼ95% viability were used, as determined by the trypan replaced by PBS (150 mm NaCl, 3 mm KCl, 5 mm NaPi, blue exclusion test. NK-L, K-562, L1210, MC2 and tryp- pH 7.4) to remove detached dead cells and to reduce sinized CHO cells were washed and suspended in their background fluorescence of the culture medium. respective culture media and adjusted to the desired con- For electrotransfection, CHO cells (8 × 106/ml or centrations. The cell concentration was determined by a 120 000 total cells), NK-L cells (1.5 × 107/ml or 225 000 hemacytometer. Cell suspension was kept at room tem- total cells), K-562, L1210 or MC2 were pulsed with the perature before pulses. presence of plasmid (final concentration 250–300 ␮g/ml) A plexiglass homemade miniature pulse chamber with coding for ␤-galactosidase (pSV-␤-galactosidase), green two rectangular stainless steel electrodes (3 × 2.5 mm, fluorescence protein (pEGFP-N1; Clontech, Palo Alto, 2.75 mm apart) was used, unless specifically mentioned. CA, USA), or luciferase (Promega, Madison, WI, USA). 15 ␮l of cells suspended in their culture media were The efficiency of ␤-galactosidase expression was placed inside the chamber and subjected to four 400 ␮s assayed by X-gal in situ staining method 40 h after pul- long quasisquare pulses at room temperature. The pre- ses.6 CHO cells were fixed by 0.25% glutaraldehyde and liminary experiments showed that 2.5 kV/cm is optimal stained in monolayer. The NK-L cells were washed by for CHO cells, 2.1 kV/cm for NK-L cells, 2.3 kV/cm for PBS and stained in PBS suspension without fixation. The K-562 and MC2 cells, and 1.7 kV/cm for L1210 cells. fluorescent intensity of green fluorescent protein of trans- These field strength conditions were used in this paper, fected NK-L cells was measured at Ex = 450 nm, and Em unless specifically mentioned. The miniature chamber = 530 nm (SLM Instruments, Urbana, IL, USA) to quanti- was used so the method may be applied when cell tatively compare the transfection efficiency of post-pulse sources are limited. Since each sample was pulsed in the pelleting technique and conventional method. The meas- same volume, the change of cell density represents the ured fluorescent intensities were normalized to express change of total cell number, hence the change of the the intensities per 750 000 cells (15 ␮l of cells with 50 × thickness of post-pulse pellets. Only one experiment was 106/ml density) before pulses. The luciferase expression performed by using exponential decay pulses and 1 ml was measured by following the method given by Pro- cuvette chamber. In this case, 0.3 ml of L1210 cells sus- mega. Briefly, the cells were pelleted and resuspended pended in culture medium were pulsed with 1.7 kV/cm in 100 ␮l lysing buffer. After one round of freezing and peak field strength and half-time ␶ = 200 ␮s at room thawing, 20 ␮l of cell lysate was mixed with 100 ␮lof temperature. luciferase assay reagent, and the luminescence was meas- After pulse application, cells were immediately trans- ured by a luminometer. ferred to an Eppendorf centrifuge tube and incubated at Because the data was quite different from day to day 37°C either in suspension (conventional method), or in a experiments, the Figures were plotted by using the pellet formed by centrifugation using an Eppendorf cen- results from the same day. Special care was taken to be trifuge (model 5414, Brinkmann Instruments, Westbury, sure that the same trends were followed by the data from NY, USA). For some specifically mentioned experiments different days, when several repeating data can not be comparing the pellet method and conventional method, finished in 1 day. the pulsed cells were resuspended first and then split into two equal parts, one for pellet incubation, the other for Acknowledgements suspension incubation. After some preliminary experiments, the optimal post-pulse incubation for CHO cells Thanks are due to the late Dr IG Abidor for his vision (pellets or suspension) was found to be 20 min, followed of the potential application of the pellet system when he by immediate resuspending for viability counting; or for worked in this laboratory. Thanks are also due to Mr cell culture. NK-L cells (pellet or suspension) were found Harold Freund for his help in making centrifuge cham- to be optimal if they were incubated for 10 min, and bers; Dr William B Carson and Ms Tamisha Palmer for resuspended for cell culture, or for continuous incubation providing NK-L cells and pEGFP-N1 plasmid. This work for 1 more hour before viability counting. K-562, L1210 is supported by grant GM30969 from National Institutes and MC2 were incubated for 20 min after pulses, resus- of Health. pended for cell culture, or for continuous incubation in suspension for another 1 h before viability counting. References The criterion for counting viable cells by phase contrast microscopy was that cells with smooth shape and bright 1 Neumann E, Schaefer-Ridder M, Wang Y, Hofschneider PH. appearance were counted as viable cells, and dark cells Gene transfer into mouse lyoma cells by electroporation in high with a completely altered phase contrast were counted electric fields. EMBO J 1982; 7: 841–845. as lysed cells. This counting method was verified by the 2 Zimmermann U, Stopper H, Jones H. Large scale transfection of mouse L-cells by electropermeabilization. Biochim Biophys trypan blue exclusion method. The difference between Acta 1987; 900: 38–44. these two methods is within 5%, in agreement with 3 Li LH, Hensen ML, Zhao YL, Hui SW. Electrofusion between 21 = other reports. The viability was calculated by V heterogeneous-size mammalian cells in a pellet: potential appli- Nviable/Ntotal. cations in drug delivery and hybridoma formation. Biophys J For loading experiment of CHO cells, FITC-dextran (Mr 1996; 71: 479–486. Improving electroporation efficiency by pellet LH Li et al 372 4 Tekle E, Astumian RD, Chock PB. Electroporation by using 1/2 fibroblasts following permeation by rectangular and bipolar oscillating electric field: an improved method for DNA exponential-decay electric field pulses. BioTechniques 1988; 6: transfection of NIH 3T3 cells. Proc Natl Acad Sci USA 1991; 88: 550–558. 4230–4234. 14 Kubiniec RT, Liang H, Hui SW. Effects of pulse length and pulse 5 Xie TD, Sun L, Tsong TY. Study of mechanisms of electric field- strength on transfection by electroporation. BioTechniques 1990; induced DNA transfection I. Biophys J 1990; 58: 13–19. 8: 1–3. 6 Hui SW, Stoicheva N, Zhao YL. High-efficiency loading, trans- 15 Abidor IG, Li LH, Hui SW. Studies of cell pellets: I. Electrical fection and fusion of cells by electroporation in two-phase poly- properties and porosity. Biophys J 1994; 67: 418–426. mer system. Biophys J 1996; 71: 1123–1130. 16 Kinosita K, Tsong TY. Voltage-induced pore formation and 7 Wolf H et al. Control by pulse parameters of electric field- hemolysis of human erythrocytes. Biochim Biophys Acta 1977; mediated gene transfer in mammalian cells. Biophys J 1994; 66: 471: 227–242. 524–531. 17 Li LH, Hui SW. Characterization of PEG-mediated electrofusion 8 Anderson GL, Evens GA. Optimization of electroporation for of human erythrocytes. Biophys J 1994; 67: 2361–2366. Analyt Biochem 180 transfection of mammalian cell lines. 1989; : 18 Ohno-Shosaku T, Okada Y. Electric pulse-induced fusion of 269–275. mouse lymphoma cells: roles of divalent cations and membrane 9 Chang DC, Gao PQ, Maxwell BL. High efficiency gene transfec- lipid domains. J Membrane Biol 1985; 85: 269–280. tion by electroporation using a radio-frequency electric field. 19 Teissie J, Rols MP. Fusion of mammalian cells in culture is Biochim Biophys Acta 1991; 1992: 153–160. obtained by creating the contact between cells after their electro- 10 Abidor IG, Li LH and Hui SW. Studies of cell pellets: II. Osmotic permeabilization. Biochem Biophys Res Commun 1986; 140: 258– properties, electroporation, and related phenomena: membrane interactions. Biophys J 1994; 67: 427–435. 266. 11 Klenchin VA et al. Electrically induced DNA uptake by cells is 20 Chernomordik LV, Sokolov AV, Budker VG. Electrostimulated a fast process involving DNA electrophoresis. Biophys J 1991; 60: uptake of DNA by liposomes. Biochim Biophys Acta 1990; 1024: 804–811. 179–183. 12 Hui SW. Effects of pulse length and strength on electroporation 21 Stenger DA et al. Optimization of electrofusion parameters for efficiency. Meth Molec Biol 1995; 2: 29–40. efficient production of murine hybridomas. Hybridoma 1988; 7: 13 Liang H et al. Uptake of fluorescence-labeled dextrans by 10T 505–518.