Cell cycle dependence of gene transfer by lipoplex, polyplex ... - Nature

Gene Therapy (2000) 7, 401–407  2000 Macmillan Publishers Ltd All rights reserved 0969-7128/00 $15.00 www.nature.com/gt

NONVIRAL TRANSFER TECHNOLOGY

RESEARCH ARTICLE

Cell cycle dependence of gene transfer by lipoplex, polyplex and recombinant adenovirus S Brunner1, T Sauer2, S Carotta1, M Cotten3, M Saltik3 and E Wagner1 1

Institute of Biochemistry and 2Institute of Molecular Biology, Unversity of Vienna, Vienna; and 3Institute for Molecular Pathology, Vienna, Austria

The aim of this study was to investigate the influence of cell cycle on transfection efficiency. Counterflow centrifugal elutriation was used which avoids possible side-effects from chemical treatment of cells. With this method, cell populations were fractionated by means of size and density, and fractions corresponding to discrete cell cycle phase-specific populations were transfected with various nonviral methods (Lipofectamine, TfpLys and TfPEI), adenovirus-enhanced transferrinfection (AVET system) and recombinant adenovirus. Transfection efficiency was found to be strongly dependent on the cell cycle stage at the time of transfection. Luciferase activity from cells transfected with polycation- or lipid-based transfection systems was 30- to more than 500fold higher when transfection was performed during S or G2

phase compared with cells in G1 phase which have the lowest expression levels. In contrast, this effect was not observed with recombinant adenovirus which varied only four-fold. Our results indicate that mitotic activity enhances transfection not only by lipoplexes but also by polyplexes, but not a viral system which has an efficient nuclear entry machinery, suggesting that transfection close to M phase is facilitated perhaps by nuclear membrane breakdown. Furthermore, low transfection success into G1 cells indicates that DNA complexes deposited in G1 cells are probably not retained long enough to take advantage of mitosis effects or that passage of transfected cells through S phase is inhibitory. Gene Therapy (2000) 7, 401–407.

Keywords: gene transfer; cell cycle; lipoplex; polyplex; adenovirus; elutriation

Introduction Gene delivery in vitro and in vivo requires uptake of DNA into target cells, release from intracellular vesicles and transport to the nucleus where transcription takes place. Stability of the transfected DNA in the various cellular compartments is another factor that influences transfection efficiency. For in vitro experiments a variety of gene delivery systems has been developed. However, recent publications1–5 suggest that for lipofection6 high levels of reporter gene expression are only obtained in actively dividing cells. This is also an important consideration for in vivo experiments because a large fraction of the target cells of the organism are nondividing. Studies with cationic lipids1 showed that it was not the uptake of complexes but subsequent steps that limit transfection efficiency in growing cells. Wilke et al3 investigated transfection with lipopeptide DNA complexes of growth arrested cells and of cells synchronised by thymidine block; it was shown that all cells had taken up DNA but gene expression was limited by mitotic activity. Mortimer et al2 studied the influence of cell cycle on transfection by comparing logarithmically growing cell populations versus aphidicoline-treated cells that are arrested in G1 phase. They also found that uptake into cells was

Correspondence: S Brunner, Institute of Biochemistry, University of Vienna, Dr Bohrgasse 9/3, A-1030 Vienna, Austria Received 13 September 1999; accepted 26 October 1999

not limiting but transfection efficiency with cationic lipid was dependent on cell cycle. Similar conclusions were made by Tseng et al5 who suggested that entry of plasmid into the nucleus is the limiting step in lipofection. For other gene transfer systems such as polymer-based systems (polyfection6) the dependence on cell cycle is far less clear. For the further optimisation of these gene transfer systems it will be important to know more about their nuclear entry capability. The work presented here investigates, for the first time, the influence of cell cycle on the efficiency of reporter gene expression after transfection of cultured cells with very different transfection systems. We compared lipofection (Lipofectamine) to polyfection (PEI7 and pLys8 based), adenovirus-enhanced transferrinfection9 and transduction with recombinant adenovirus.10,11 In order to test precisely the effect of cell cycle status on transfection, it is necessary to isolate a sufficient number of cells in different phases with high purity. There are several methods available including nutritional deprivation protocols or the use of chemical agents such as nocodazole or aphidicoline to block cells at specific stages in the cell cycle.12 We chose to use counterflow centrifugal elutriation to prepare pure cell cycle stage populations thus avoiding the possible side-effects from treatment of cells with chemicals. This method is known to be an excellent tool to isolate cells in specific phases of the cell cycle.13,14 The separation of cells is based on size and density. Starting from one cell population it is possible to obtain several pure fractions of cells in different

Influence of cell cycle on transfection efficiency S Brunner et al

402

phases simultaneously, not only in one specific phase as is the case with other methods. Therefore, this method provides an important tool for examining cell cycle effects on transfection and offers a useful complementary technique to drug methods. We chose the suspension cell line K562 cells as our model system. Compared with adherent cells, these suspension cells can be handled more easily and grown in larger quantities (a minimum of 108 cells are needed per elutriation) and are known to be an excellent cell line for elutriation resulting in very pure fractions.13 K562 cells have also previously been used for the characterisation of transferrin receptor-mediated gene transfer.7,9,15 Control experiments using HeLa cells demonstrated that our findings are not unique to the use of suspension cells. It is shown here that transfection efficiency with all the tested synthetic gene delivery systems, surprisingly, is strongly dependent on the cell cycle phase with S/G2 cells giving 30- to 500-fold greater levels of transfection than G1 cells. In sharp contrast, recombinant adenovirus transduction was not strongly influenced by cell cycle.

Results Elutriation Our elutriation conditions result in 9–12 fractions of K562 cells differing in size and DNA content. Figure 1a shows DNA profiles of logarithmically growing cells and of representative fractions after elutriation. The first fraction of cells is pure G1 cells. The next fractions are early, middle and late S phase cells, followed by cells in G2. The last two fractions are mixtures of G2 and M phase cells as it is not possible to obtain pure M phase fractions. The quality of fractionation was controlled by FACS analysis of the DNA content of the cells and also by measurement of the cells in a multichannel cell analyser. A continuous increase in cell size was observed from around 12.5 ␮m in G1 cells to 16.5 ␮m in the largest cells, corresponding to an increase of mean cell volumes from 1112 fl to 2550 fl. To monitor synchronous progression through the cell cycle, samples of recultivated cells were taken after 4, 9 and 20 h of recultivation. Under optimal growth conditions, K562 cells double after approximately 18 h. As can be seen in the first line of Figure 1a, almost all cells derived from a pure G1 fraction (DNA content 2n, peak at the left side) underwent mitosis by 20 h (part of the cells are still in G2, right peak; the others are already in G1 again, left peak) demonstrating that the elutriation procedure did not significantly affect the length of the cell cycle. Reporter gene expression after transfection of elutriated K562 cells After elutriation, equal numbers of logarithmically growing cells (as a control for transfection efficiency) and of cells from the elutriated fractions were collected by centrifugation, resuspended in the appropriate medium and transfected with two different polycation-based transfection systems: transferrin–polylysine (TfpLys8) and transferrin–polyethylenimine (TfPEI7; large DNA complexes, mixed in HBS16) complexed with luciferase-encoding vector pCMVLuc.17 Figure 1b and 1c show relative luciferase light units (RLU) 20 h after transfection of 200 000 cells Gene Therapy

with TfpLys complexes or with TfPEI complexes, respectively. Mean luciferase activity from cells with the highest expression (transfection during S or G2 phase) compared with cells transfected during G1 phase is 130-fold higher for TfpLys complexes (Figure 1b) and 50-fold higher for TfPEI complexes (Figure 1c).

Luciferase expression at different time-points after transfection of elutriated cells The differences in reporter gene expression between the elutriated fractions as a function of the time between transfection and harvest were examined. Logarithmically growing cells and cells from all elutriated fractions were transfected with TfPEI (Figure 2a; small complexes, mixed in HBG16) or Lipofectamine (Figure 2b) and measurements of luciferase activity were performed 20 h and 45 h after transfection. Values from a representative elutriation are presented. The difference in luciferase expression after 20 h between the fractions with lowest expression (G1 during transfection) and highest expression (S or G2) is 290-fold for TfPEI and 150-fold for Lipofectamine. After 45 h, in cells that were transfected in G1, complexes theoretically had a chance to enter the nucleus during the next mitosis; nevertheless, the difference in luciferase expression is still approximately 40-fold for both systems. Instability of DNA could explain these effects (see Discussion). Transduction of cells with recombinant adenovirus and transfection by adenovirus-enhanced transferrinfection In recombinant adenovirus, the reporter gene is packed in the viral genome. In contrast, for adenovirus-enhanced transferrinfection (AVET),9 where psoralen-inactivated adenovirus as an effective endosome-disrupting agent is included in TfpLys/DNA complexes, adenovirus only serves as a carrier. We thus compared a recombinant adenoviral vector and adenovirus-enhanced transferrinfection. Figure 3a shows the gene expression levels of K562 cells after transduction with recombinant E1-negative type 5 adenovirus bearing the same CMV luciferase cassette used in the DNA of the nonviral transfection systems (AdLuc, described in Ref. 18). Only small differences in luciferase expression were found between the different elutriated fractions (maximum a factor of 4 at 24 h). The result of adenovirus-enhanced transferrinfection is shown in Figure 3b. Interestingly, in contrast to infection with recombinant adenovirus, transfection with the AVET complexes results in a transfection pattern similar to the nonviral gene transfer systems presented before, suggesting that the G1 limitation may be a property of the plasmid DNA complexes. After 18 h, luciferase activity in the fraction with highest expression was 170fold higher compared with cells that were in G1 phase during transfection. Transfection of elutriated cells with green fluorescent protein (GFP) as reporter gene We then asked whether different gene expression levels of fractions reflect a different number of cells expressing the reporter gene or whether the expression level per cell is higher in cells transfected in S/G2 phase compared with cells in G1 phase. To answer this, K562 cells were transfected with pEGFP-C1/TfPEI complexes. Figure 4a shows the percentage of cells expressing GFP, Figure 4b


403

Figure 1 Cell cycle dependence of transfection. K562 cells were elutriated as described in Materials and methods. Logarithmically growing cells and cells from the elutriated fractions were stained with 6 ␮m 4,6-diamino-2-phenylindole dihydrochloride (DAPI) and analysed in a flow cytometer. (a) Shows DNA profiles of logarithmically growing cells and of five representative fractions after elutriation (cell number plotted versus DNA content). The left peak in the histogram corresponds to the 2n DNA content (cells in G1), the right peak indicates that the cells had already doubled their chromosomes (4n, cells in G2). After 4, 9 and 20 h, respectively, aliquots of the recultivated cells were analysed again to monitor progression through cell cycle. The DNA profiles represent nontransfected cells. Logarithmically growing cells (as a control) and cells from fractions after elutriation were transfected with two different polycation-based transfection systems: luciferase-encoding vector pCMVLuc complexed with TfpL, (b) or TfPEI mixed in HBS (c). For each time-point, 200 000 cells were transfected with 1.3 ␮g complexed DNA. Cells were harvested after 20 h. Mean relative light units (RLU) after transfection of logarithmically growing cells and of five fractions with similar DNA profiles from two independent elutriations were chosen to illustrate the dramatic increase in luciferase expression.

the mean fluorescence intensities (MFI) of the GFP-positive cells. Both a difference in the percentage of GFP-positive cells and a difference in the mean fluorescence of GFP-positive cells contribute to the cell cycle-dependent expression.

Discussion In the present work we analysed the cell cycle dependence of six very different gene transfer systems. Lipofectamine was used as an example for lipid-based transfection (lipofection), TfpLys8,19 and two different TfPEI7 complexes (small complexes mixed in HBG, large complexes mixed in HBS16) served as a model for polymerbased transfection (polyfection). In addition, adenovirusenhanced transferrinfection (AVET)9 and infection with recombinant adenovirus18 were tested. We find that with all systems except recombinant adenovirus there is a

clear and comparable difference in transfection efficiency between cells transfected shortly before their next cell division (cells in late S/G2) to cells that are just at the beginning of cell cycle at the time of transfection (G1 cells). Transfection of elutriated cells showed that for all nonviral transfection systems we analysed and for the AVET system the difference in luciferase expression between cells transfected in G1 phase to cells with the highest expression (S or G2 phase) was at least 30- to more than 500-fold. Mean values were approximately 140-fold for Lipofectamine, 160-fold for TfPEI complexes and more than 300-fold for TfpLys. We hypothesise that transport of DNA into the nucleus is very inefficient so that only breakdown of the nuclear membrane allows inclusion of sufficient amounts of DNA for high reporter gene expression. By 20 h after transfection, cells transfected in early G1 had little or no time after the following cell division and breakdown of the nuclear membrane to Gene Therapy


404

Figure 2 Luciferase expression 1 and 2 days after transfection of elutriated cells. Relative light units (RLU) of all fractions after elutriation and transfection are shown. K562 cells were transfected with (a) TfPEI complexes mixed in HBG or (b) Lipofectamine, a lipid-based transfection system. For each transfection, 200 000 cells were transfected with 1.3 ␮g DNA. Measurements of luciferase activity were done after 20 h and 45 h, respectively. The RLU presented correspond to values of a representative elutriation. Similar results were obtained after all elutriations.

express the reporter gene. Therefore the difference in reporter gene expression between these cells and cells from other fractions is quite high. The size of transfection particles did not have a strong influence; small TfPEI complexes (approximately 50 nm, mixed in HBG under low ionic strength, Figure 2a) behaved similarly to large TfPEI complexes that are several 100 nm in size (complexes mixed in HBS under high ionic strength, Figure 1c). Surprisingly, adenovirusenhanced transferrinfection gave results comparable with the nonviral systems that were tested suggesting that it was the polycation/plasmid component that was limiting rather than the adenovirus entry pathway. In sharp contrast, gene transfer with recombinant adenovirus resulted in only minor differences in reporter gene expression between the elutriated fractions, clearly demonstrating that adenovirus can deliver DNA to the nucleus independent of the cell cycle. To learn whether this elutriation/recultivation/ transfection procedure gives similar results for other cell types, we evaluated an adherent HeLa cell line. After elutriation, the cells were immediately transfected in suspenGene Therapy

Figure 3 Luciferase expression after transduction with recombinant adenovirus and transfection with the AVET system. Elutriated K562 cells as well as logarithmically growing cells were (a) transduced with recombinant E1-negative adenovirus expressing luciferase (AdLuc), or (b) transfected by adenovirus-enhanced transferrinfection (AVET system, psoralen-inactivated adenovirus in DNA complexes with TfpLys). Results of luciferase measurements after 24 h (adenovirus infection) or 18 h (AVET system) are shown. The RLU correspond to values of a representative elutriation.

sion before they became adherent again. Transfection with Lipofectamine resulted in a 170-fold induction (G1 cells compared with late S/G2 cells) which is comparable with the findings with K562 cells (data not shown). Thymidine block of HeLa cells, release from block for different periods of time and subsequent transfection with Lipofectamine resulted in an approximately 30-fold difference in luciferase light units between cells transfected at the G1/S boundary compared with cells in S/G2 (data not shown). Thus, the conclusions made with elutriation of K562 cells are also valid in an adherent cell type. In our experiments we concentrated mainly on transfection with a plasmid coding for luciferase (pCMVLuc). When elutriated cells were transfected with a plasmid containing the cDNA for green fluorescent protein (pEGFP-C1) also a clear increase in the number of transfected cells as well as in the mean fluorescence intensities of transfected cells was observed in S/G2 cells compared


Figure 4 Transfection of elutriated K562 cells with green fluorescent protein as reporter gene. Elutriated cells and logarithmically growing cells were transfected with pEGFP-C1/TfPEI complexes (complexes mixed in HBS). Data are taken from one representative elutriation. (a) Shows the percentage of cells expressing GFP, (b) mean fluorescence intensities of GFP-positive cells.

with cells in G1 at the beginning of transfection. To exclude the possibility that variations in the CMV promoter activity are responsible for the observed cell cycle effect we elutriated HeLa cells stably transfected with GFP driven by the CMV promoter (data not shown). Mean fluorescence intensity of positive cells varies less than a factor of 2 in the fractions corresponding to the different phases. It was also shown by Western blot analysis that GFP protein levels in all elutriated fractions are the same (personal communication, J Gotzmann and W Mikulits, data not shown). A gene transfer system that allows efficient transfection not only of rapidly growing but also of resting cells needs to be capable of importing foreign DNA into the nucleus independent of cell cycle. Trafficking of molecules between cytoplasm and nucleus is size dependent and can occur by passive diffusion or as a controlled process. Import of molecules larger than approximately 40– 60 kDa is an energy- and signal-dependent active mechanism which can be inhibited (eg in contrast to diffusion

of oligonucleotides into the nucleus) by cold treatment or depletion of energy. These molecules need a nuclear localisation signal in order to be actively transported through nuclear pores with 9–10 nm diameter. Hagstrom et al20 used digitonin permeabilized HeLa cells as a model system to study nuclear import. They found that 1 kb double-stranded transfected DNA enters the nuclei in these cells and that import can be inhibited by energy depletion or cold treatment. DNA larger than 2 kb remained in the cytoplasm. On the other hand, DNA (1 kb) microinjected into the cytoplasm did not efficiently enter the nucleus of HeLa cells. Dowty et al21 studied transport of plasmid DNA into the nucleus of postmitotic primary rat myotubes. They found that DNA microinjected into the cytoplasm can enter postmitotic intact nuclei. Inhibition of import by wheat germ agglutinin and at 4°C suggests import through nuclear pores. However, when DNA is injected far from nuclei, less positive cells are obtained as the microinjected DNA is not efficiently transported to nucleus. Cytoplasmic sequestration of plasmid DNA could be responsible for this. Recent work has demonstrated nuclear uptake as a major barrier for lipofection and related lipid-based gene transfer.1–5 The situation is less clear for polymer-based gene transfer systems (polyplexes). It has been proposed that polylysine or polyethylenimine, in contrast to cationic lipids, can promote gene delivery from the cytoplasm to the nucleus.21–23 However, Chan et al24 found that pLys itself in contrast to a pLys/NLS conjugate does not increase nuclear import. Sebestyen et al25 reported covalent attachment of NLS peptides to DNA as a method to increase uptake into the nucleus. However, when this DNA was microinjected cytoplasmically into HeLa cells, no nuclear transport was found. Significant uptake only occurred in digitonin-permeabilized cells. The authors speculate that avoidance of cytoplasmic sequestration is also necessary for efficient nuclear entry. Alternatively, many NLS peptides were coupled to each DNA complex which might result in inhibition of translocation as the linked DNA attempts to enter by too many pores. Exciting work was recently described by Zanta et al.26 Using a capped linear DNA fragment where on one side a single NLS peptide was covalently coupled they found that small amounts of DNA were sufficient for transfection of cells. Enhancement of reporter gene expression as a result of this NLS peptide was 10- to 1000-fold. Moreover, transfection kinetics was different to other systems in that reporter gene expression reached its maximum very early after transfection (12 h with NLS compared with 24 h without NLS). Similar data were also obtained by Branden et al27 using a PNA-NLS construct. Another effort to overcome the barrier of the nuclear membrane was described with sequence specific import of DNA into the nucleus achieved by introducing DNA elements containing binding sites for transcription factors.28–30 These transcription factors then complex the transfected DNA and due to their NLS signal transport it through the nuclear pore complex into the nucleus of all cells or even in a cell-type specific manner. We also investigated whether uptake of transfection complexes is a limit for gene transfer. The relative uptake of DNA/TfPEI complexes into elutriated K562 cells was determined by comparing the total amount of DNA asso-

405

Gene Therapy


406

ciated with the cells with the amount of DNA adhering to the outside of cells. We used a recently developed FACS technique (Ogris et al, manuscript submitted) and found that receptor-mediated uptake of TfPEI complexes is constant in all phases but is actually slightly diminished in late G2/M cell fractions (data not shown). This is consistent with the fact that during mitosis endocytic events are generally reduced.31 As the DNA complexes are normally applied to cells for 3–4 h during transfection, which is longer than M-phase (which lasts approximately 1 h), this reduced uptake is not reflected by the level of reporter gene expression. The stability of transfected DNA can also be a limit to successful gene transfer. Ross et al32 showed that, dependent on the transfection method, transfection in the presence of nuclease inhibitor can increase reporter gene expression. Metabolic instability of delivered DNA in the cytosol was investigated by Lechardeur et al33 with naked DNA (uncomplexed plasmid) found to have a half-life of approximately 1–2 h in HeLa and COS cells, whereas for encapsulated plasmid (stable plasmid-lipid particles, SPLP) degradation was much slower. Nevertheless, after 4 h more than 25% of encapsulated plasmid was degraded. Our observations support the idea that inactivation or degradation of DNA may play an important role in limiting gene transfer. As a result, during the following mitosis only a small amount of intact plasmid would be included in the nucleus and contribute to reporter gene expression. This is consistent with our finding that the gene transfer efficiency of cells transfected in G1 is still poor after 45 h, when cells had a chance to divide. Taken together, our data suggest that synthetic gene transfer systems are inefficient in nondividing cells. Interestingly, adenovirus-enhanced transferrinfection behaves similarly to the other synthetic gene transfer systems. We hypothesise that even with the AVET system no efficient import of transfected DNA into the nucleus can be achieved. Only recombinant adenovirus, in contrast to all synthetic systems, can transduce cells independent of their cell cycle phase, suggesting that a large difference lies in the viral packaging of the delivered gene. These findings support the notion that it may be critical to develop nuclear targeting strategies for transfection systems to be used not only in vitro but also in vivo, as most cells in a body are postmitotic cells.

Materials and methods Chemicals A TfPEI conjugate with a molar ratio of approximately three transferrin molecules linked to PEI 800 kDa was used. The synthesis of TfPEI is described in Kircheis et al.7 TfpLys and StAVpL conjugates are described in Refs 8,9 and 19. Lipofectamine was obtained from Life Technologies (Gaithersburg, MD, USA). Plasmids were purified with the EndoFree Plasmid Kit from Qiagen (Hilden, Germany). Plasmid pCMVLuc (Photinus pyralis luciferase under control of the CMV enhancer/promoter) is described in Plank et al.17 Plasmid pEGFP-C1 was obtained from Clontech (Palo Alto, CA, USA). Cell culture Media, antibiotics and FCS were purchased from Life Technologies. K562 cells (ATCC CCL-243) were cultured Gene Therapy

in RPMI medium 1640/10% fetal calf serum (FCS)/2 mm glutamine/antibiotics. Chloroquine was obtained from Sigma (St Louis, MO, USA).

Elutriation Centrifugal elutriation was performed with a Beckman J2–21M centrifuge and a JE-6B rotor. 1.5 to 3 × 108 K562 cells were introduced in the separation chamber. The elutriation medium was PBS with 0.9 mm CaCl2 and 0.5 mm MgCl2 containing 5% FCS. A constant rotor speed of 2000 r.p.m. was applied; temperature was kept at 15°C during elutriation. Separation of the cells was accomplished by stepwise rising the pump speed (ColeParmer Masterflex pump). With a multichannel cell analyser (CASY-1; Scha¨rfe Systems, Reutlingen, Germany) the number and volume of the cells in the different fractions was determined immediately after elutriation. Cell fractions were kept at 4°C until they were recultivated and transfected. An aliquot of each fraction was used for cell cycle analysis. Cell cycle analysis To assess the homogeneity of each cell fraction after elutriation, approximately 5 × 105 cells were used for analysis without prior fixation. For analysis of fractions at different time-points after elutriation, cells were fixed as follows. Approximately 5 × 105 cells were centrifuged, resuspended in 0.5 ml PBS, added dropwise to 5 ml of ice-cold 80% ethanol and kept at −20°C (at least over night). Cells were stained with 6 ␮m 4,6-diamino-2-phenylindole dihydrochloride (DAPI) and analysed in a PAS-III flow cytometer (Partec, Mu¨nster, Germany). Complex formation and transfection TfPEI/DNA complexes for analysis of receptor-mediated uptake and for transfections, where indicated, were mixed salt-free in HBG (20 mm Hepes pH 7.4, 5% glucose). All other complexes were mixed in HBS (20 mm Hepes pH 7.4, 140 mm NaCl). TfpLys/DNA complexes were mixed at a charge ratio (lysine to phosphate6) of 1.5, TfPEI complexes at N/P ratio7 of 6.7. For lipofection, 2 ␮l of Lipofectamine were used for 1 ␮g DNA. AVET system9: for transfection of 200 000 cells, 1 ␮l biotinylated psoralen-inactivated adenovirus (dl 1014; titer 3 × 1012) was diluted in 16 ␮l of HBS and mixed with 0.2 ␮g streptavidine-polylysine in 16 ␮l of HBS. After 20 min incubation at room temperature, this was mixed with 2 ␮g DNA in 24 ␮l HBS. After a further 20 min 2 ␮l TfpLys (0.91 mg/ml pLys) in 24 ␮l HBS was added. Following a 20 min incubation, transfection complexes were added to the cells. For transfection, 200 000 cells per fraction and timepoint were collected by centrifugation, resuspended in 200 ␮l medium and transfected with 1.3 ␮g DNA in 75 ␮l complexation medium. Lipofection was done in OptiMEM, transduction with AdLuc18 in RPMI/5% FCS; RPMI/10% FCS was used for all other transfections. After 3–4 h, transfection medium was replaced by fresh RPMI/10% FCS. Cells transfected with TfpLys complexes and with TfPEI complexes mixed in HBG were treated with 200 ␮m chloroquine during transfection to enhance endosomal release. Transduction with recombinant adenovirus: 1.5 ␮l AdLuc (titer 1.3 × 1012/ml) was added to 200 000 cells (total volume 400 ␮l).


Determination of luciferase activity Cells were harvested at the indicated time-points by centrifugation and lysis in 0.25 m Tris pH 7.4, 0.1% Triton X-100 (two freeze–thaw cycles). Luciferase activity was measured from an aliquot of the lysate (using a Lumat LB9507 instrument from Berthold, Bad Wildbad, Germany). Values are given as light units per 200 000 cells transfected. Analysis of GFP-positive cells Cells were collected by centrifugation and resuspended in PBS. An equal volume of 4% paraformaldehyde (in 20 mm Hepes pH 7.4, 140 mm NaCl) was added. Cells were fixed at room temperature for 30–60 min; subsequently, they were resuspended in 500 ␮l of PBS and added dropwise to 5 ml ice-cold 80% ethanol. Cells were kept at −20°C at least overnight before flow cytometric analysis (FACScalibur; Becton Dickinson, San Jose, CA, USA).

Acknowledgements We would like to thank Josef Gotzmann and Wolfgang Mikulits (Institut fu¨r Tumorbiologie und Krebsforschung) for the stably transfected HeLa cells. We thank Ingrid Mudrak for the HeLa cells that were elutriated and Karin Paiha and Peter Steinlein for help with FACS analysis. Thanks to Bettina Grosse, Helga Vetr, Thomas Blessing and Peter Wallner for inspiring discussions and for critically reading the manuscript. This work was supported by grant S07405 from the Austrian Fonds zur Fo¨rderung der Wissenschaftlichen Forschung.

References 1 Zabner J et al. Cellular and molecular barriers to gene transfer by a cationic lipid. J Biol Chem 1995; 270: 18997–19007. 2 Mortimer I et al. Cationic lipid-mediated transfection of cells in culture requires mitotic activity. Gene Therapy 1999; 6: 403–411. 3 Wilke M et al. Efficacy of a peptide-based gene delivery system depends on mitotic activity. Gene Therapy 1996; 3: 1133–1142. 4 Brisson M, Huang L. Liposomes: conquering the nuclear barrier. Curr Opin Mol Ther 1999; 1: 140–146. 5 Tseng WC, Haselton FR, Giorgio TD. Mitosis enhances transgene expression of plasmid delivered by cationic liposomes. Biochim Biophys Acta 1999; 1445: 53–64. 6 Felgner PL et al. Nomenclature for synthetic gene delivery systems. Hum Gene Ther 1997; 8: 511–512. 7 Kircheis R et al. Coupling of cell-binding ligands to polyethylenimine for targeted gene delivery. Gene Therapy 1997; 4: 409–418. 8 Wagner E et al. Transferrin–polycation conjugates as carriers for DNA uptake into cells. Proc Natl Acad Sci USA 1990; 87: 3410– 3414. 9 Wagner E et al. Coupling of adenovirus to transferrinpolylysine/DNA complexes greatly enhances receptormediated gene delivery and expression of transfected genes. Proc Natl Acad Sci USA 1992; 89: 6099–6103. 10 Shenk T. Group C adenovirus as vectors for gene therapy. In: Kaplitt MG, Loewy AD (eds). Viral Vectors. Academic Press: San Diego, CA, 1995, pp 43–54. 11 Shenk T. Adenoviridae: the viruses and their replication. In: Fields BN, Knipe DM et al (eds). Virology, 3rd edn. LippincottRaven Publishers: Philadelphia, PA 1996, pp 2111–2148.

12 Krek W, DeCaprio JA. Cell synchronization. Meth Enzymol 1995; 254: 114–124. 13 Kauffman MG, Noga SJ, Kelly TJ, Donnenberg AD. Isolation of cell cycle fractions by counterflow centrifugal elutriation. Anal Biochem 1990; 191: 41–46. 14 Mikulits W et al. Dynamics of cell cycle regulators: artifact-free analysis by recultivation of cells synchronized by centrifugal elutriation. DNA Cell Biol 1997; 16: 849–859. 15 Cotten M et al. Transferrin-polycation-mediated introduction of DNA into human leukemic cells: stimulation by agents that affect the survival of transfected DNA or modulate transferrin receptor levels. Proc Natl Acad Sci USA 1990; 87: 4033–4037. 16 Ogris M et al. The size of DNA/transferrin-PEI complexes is an important factor for gene expression in cultured cells. Gene Therapy 1998; 5: 1425–1433. 17 Plank C et al. Gene transfer into hepatocytes using asialoglycoprotein receptor mediated endocytosis of DNA complexed with an artificial tetra-antennary galactose ligand. Bioconjug Chem 1992; 3: 533–539. 18 Michou AI, Lehrmann H, Saltik M, Cotten M. Mutational analysis of the avian adenovirus CELO, which provides a basis for gene delivery vectors. J Virol 1999; 73: 1399–1410. 19 Zauner W, Ogris M, Wagner E. Polylysine-based transfection systems utilizing receptor-mediated delivery. Adv Drug Deliv Rev 1998; 30: 97–113. 20 Hagstrom JE et al. Nuclear import of DNA in digitonin-permeabilized cells. J Cell Sci 1997; 110: 2323–2331. 21 Dowty ME et al. Plasmid DNA entry into postmitotic nuclei of primary rat myotubes. Proc Natl Acad Sci USA 1995; 92: 4572– 4576. 22 Pollard H et al. Polyethylenimine but not cationic lipids promotes transgene delivery to the nucleus in mammalian cells. J Biol Chem 1998; 273: 7507–7511. 23 Zauner W et al. Differential behaviour of lipid based and polycation based gene transfer systems in transfecting primary human fibroblasts: a potential role of polylysine in nuclear transport. Biochim Biophys Acta 1999; 1428: 57–67. 24 Chan CK, Jans DA. Enhancement of polylysine-mediated transferrinfection by nuclear localization sequences: polylysine does not function as a nuclear localization sequence. Hum Gene Ther 1999; 10: 1695–1702. 25 Sebestyen MG et al. DNA vector chemistry: the covalent attachment of signal peptides to plasmid DNA. Nat Biotechnol 1998; 16: 80–85. 26 Zanta MA, Belguise Valladier P, Behr JP. Gene delivery: a single nuclear localization signal peptide is sufficient to carry DNA to the cell nucleus. Proc Natl Acad Sci USA 1999; 96: 91–96. 27 Branden LJ, Mohamed AJ, Smith CIE. A peptide nucleic acidnuclear localization signal fusion that mediates nuclear transport of DNA. Nat Biotechnol 1999; 17: 784–787. 28 Langle Rouault F et al. Up to 100-fold increase of apparent gene expression in the presence of Epstein–Barr virus oriP sequences and EBNA1: implications of the nuclear import of plasmids. J Virol 1998; 72: 6181–6185. 29 Dean DA. Import of plasmid DNA into the nucleus is sequence specific. Exp Cell Res 1997; 230: 293–302. 30 Vacik J, Dean BS, Zimmer WE, Dean DA. Cell-specific nuclear import of plasmid DNA. Gene Therapy 1999; 6: 1006–1014. 31 Raucher D, Sheetz MP. Membrane expansion increases endocytosis rate during mitosis. J Cell Biol 1999; 144: 497–506. 32 Ross GF et al. Enhanced reporter gene expression in cells transfected in the presence of DMI-2, an acid nuclease inhibitor. Gene Therapy 1998; 5: 1244–1250. 33 Lechardeur D et al. Metabolic instability of plasmid DNA in the cytosol: a potential barrier to gene transfer. Gene Therapy 1999; 6: 482–497.

407

Gene Therapy