Effective gene transfer to solid tumors using different nonviral ... - Nature

Cancer Gene Therapy (2002) 9, 399 – 406 D 2002 Nature Publishing Group All rights reserved 0929-1903 / 02 $25.00 www.nature.com / cgt

Effective gene transfer to solid tumors using different nonviral gene delivery techniques: Electroporation, liposomes, and integrin-targeted vector Maja Cemazar,1,2 Gregor Sersa,2 John Wilson,1 Gillian M Tozer,1 Stephen L Hart,3 Alenka Grosel,2 and Gabi U Dachs1 1

Tumor Microcirculation Group, Gray Cancer Institute, Mount Vernon Hospital, Northwood HA6 2JR, UK; Department of Tumor Biology, Institute of Oncology, Ljubljana SI-1000, Slovenia; and 3Molecular Immunology Unit, Institute of Child Health, University College London, London WC1N 1EH, UK.

2

In this study, we measured transfection efficiency in vitro and in vivo using the following nonviral approaches of gene delivery: injection of plasmid DNA, electroporation - assisted, liposome - enhanced, and integrin - targeted gene delivery, as well as the combination of these methods. Four histologically different tumor models were transfected with a plasmid encoding the green fluorescent protein ( GFP ) ( B16 mouse melanoma, P22 rat carcinosarcoma, SaF mouse sarcoma, and T24 human bladder carcinoma ) using adherent cells, dense cell suspensions, and solid tumors. Emphasis was placed on different electroporation conditions to optimise the duration and amplitude of the electric pulses, as well as on different DNA concentrations for effective gene delivery. In addition, transfection efficiency was correlated with cell density of the tumors. The major in vivo findings were: ( a ) electroporation assisted gene delivery with plasmid DNA, employing long electric pulses with low amplitude, yielded significantly better GFP expression than short electric pulses with high amplitude; ( b ) electroporation combined with liposome – DNA complexes yielded the highest percentage of transfected tumor area in B16F1 tumor ( 6% ); ( c ) transfection efficiency of electroporation - assisted plasmid DNA delivery was dependent on tumor type; ( d ) integrin - targeted vector, alone or combined with electroporation, was largely ineffective. In conclusion, our results demonstrate that some nonviral methods of gene delivery are feasible and efficient in transfecting solid tumors. Therefore, this makes nonviral methods attractive for further development. Cancer Gene Therapy ( 2002 ) 9, 399 – 406 DOI: 10.1038 / sj / cgt / 7700454 Keywords: electroporation; lipofectin; integrin - targeted peptide; green fluorescence protein; tumor models

T

he main stumbling block in gene therapy of various diseases still remains the effective and safe delivery of genes to the target tissue. Most of the current vectors are viral, among which adeno - and retroviruses are commonly used. Although transfection using these vectors is relatively high, several drawbacks have to be taken into consideration when using this approach of gene delivery in a clinical setting, such as eliciting an immune response and concerns about patient safety.1,2 In contrast to viral vectors, nonviral methods of gene delivery result in lower transfection efficiency, but they lack pathogenicity and can package large amounts of plasmid DNA. Specifically, it has been shown that cationic liposomes are clinically well tolerated, easy to produce, noninfectious, nonimmunogenic, and able to transfect slowly dividing cells.3,4 Receptor- mediated gene delivery can potentially improve efficiency and safety further by selectively targeting a cell population expressing the surface receptor for a ligand. The efficiency of gene transfer by receptor-mediated Received January 29, 2002. Address correspondence and reprint requests to: Prof Maja Cemazar, Institute of Oncology, Zaloska 2, Ljubljana SI - 1000, Slovenia. E - mail: mcemazar@onko - i.si

endocytosis of polycation –DNA complexes has been enhanced by the incorporation of endosomolytic agents, such as adenoviral capsids or fusogenic peptides.5 - 7 These artificial viruses have the potential to exploit the efficiency of viral vectors and the advantages of liposomes. The combination of lipofectin -based liposomes with integrin binding peptides has been shown to promote receptormediated endocytosis of DNA and also reduce endosomal degradation through lipid- mediated destabilisation of the endosomal membrane.8 Efficient transfection was demonstrated in a number of cell types in vitro and in bronchial and alveolar cells in vivo, with transgene expression sustained for 3 – 7 days.8 - 10 Electroporation (Ep ) as a gene delivery method for cells in vitro is an established method. Recently, it was employed in preclinical studies on animal models and also in clinical settings for delivery of chemotherapeutic drugs with very high antitumor efficiency and negligible side effects.11,12 In addition, some recent studies have examined the use of electroporation to deliver plasmid DNA to various types of tissues in vivo, such as skin, liver, testis, brain, skeletal muscle, and tumors.13 Various settings of electroporation were used, with no general conclusion on the optimal electrical, DNA, and environmental parameters for effective

Nonviral gene delivery to solid tumors M Cemazar et al

400 DNA transfer, especially when delivering DNA to tumors. High - voltage microsecond pulses, which were used in electrochemotherapy, and low- voltage milliseconds pulses, which were found to be the most suitable for skeletal muscle transfection, were both used in gene delivery protocols for transfecting tumors.13 - 15 However, no detailed study using different amplitudes and duration of the delivered electric pulses has been performed so far. The aim of our study was therefore to measure transfection efficiency using a plasmid encoding the green fluorescent protein (GFP ) in histologically different tumor models both in vitro and in vivo. Different nonviral approaches of gene delivery were compared and combined to establish the most effective method. Compared were naked DNA injection, electroporation -assisted, liposomes- enhanced, and receptormediated gene delivery method, including the combination of these methods. Emphasis was on different electroporation conditions, in order to optimise the duration and amplitude of the electric pulses, as well as on different DNA concentrations, for effective gene expression.

Materials and methods Cell lines and tumor models

B16F1 mouse melanoma ( CRL6323; American Type Culture Collection, Manassas, VA ), P22 rat carcinosarcoma,16 SaF mouse sarcoma,17 and T24 human bladder carcinoma ( European Collection of Cell Cultures, Ref. no. 92091712 , Salisbury, UK) were used in this study. All cells were maintained in Dulbecco’s modified Eagle’s medium ( Life Technologies, Paisley, UK ) with 10% fetal calf serum and L -glutamine in a 5% CO2 humidified incubator at 378C. Tumors were implanted subcutaneously in the flank of C57Bl/ 6 mice ( B16F1 ), SCID mice ( P22 and T24 ), and CBA mice ( SaF ) and grown to a size of 6– 7 mm in diameter. Procedures were performed with approved protocols, in accordance with the UK Animals (Scientific Procedures ) Act 1986 and with the approval from the Ethical Review Committee of the Gray Cancer Institute. Gene transfection in vitro

The commercial plasmid, pEGFP - N1 ( Clontech, Basingstoke, UK ), encoding the enhanced GFP controlled by the cytomegalovirus promoter, was used throughout to assess transfection efficiency. Cells from exponential growth phase were used in all experiments. Two different conditions of in vitro transfection were tested — either adherent cells ( optimised in vitro) or dense cell suspension ( simulated in vivo conditions ).8,9 The in vitro conditions for adherent cells were as follows: 1.5 L of lipofectin ( Life Technologies ) and 2.0 g of DNA ( LD ), or 1.5 L of lipofectin, 3.4 g of integrin -targeted peptide, and 2.0 g of DNA (LPD ) in 1 mL of OptiMEM (Life Technologies ) per 1– 3105 preplated adherent cells.8 Transfection was performed for 5 hours at 378C. After that, 5 mL of OptiMEM was added to the cells and left overnight in a 5% CO2 humidified incubator at 378C. The in vitro conditions for dense cell suspensions were as follows: 10 g of DNA ( D10 ); or 50 g of DNA (D50 ); or

Cancer Gene Therapy

7.5 L of lipofectin and 10 g of DNA (LD ); or 7.5 L of lipofectin, 20 g of peptide, and 10 g of DNA (LPD ) in 75 L of Hank’s balanced salt buffer, with or without subsequent electroporation. The concentrations used for LD and LPD were about 5 -fold of those used for adherent cells, and equivalent to those used in vivo (see below ). Electroporation was performed by placing the cells ( 1106 cells in 50 L of suspension ) between two flat parallel stainless steel electrodes with a 2- mm gap connected to the electroporator ( built in -house ) and subjected to eight square wave electric pulses ( pulse width 5 milliseconds, repetition frequency 1 Hz, electric field intensity 600 V / cm ). After exposure to electric pulses, the cells were incubated for 5 minutes at room temperature. After that, 5 mL of OptiMEM was added to the cell suspension and left overnight in a 5% CO2 humidified incubator at 378C. Transfection efficiency was analysed 24 hours after transfection by fluorescent activated cell sorting ( Becton Dickinson, Oxford, UK ) and defined as the percentage of cells with increased fluorescence over untransfected controls. Experiments were performed at least three times using duplicate samples. Gene transfection in vivo

DNA or DNA complexes were injected intratumorally (i.t. ) with or without subsequent electroporation. Electric pulses were delivered to anesthesized animals ( Metofane; C -Vet, Edmonds, UK ), as previously described.18 Briefly, the applicator consisted of two flat parallel electrodes 7 mm apart ( two stainless steel strips, width 7 mm, with rounded corners ). Electrodes were placed percutaneously at the opposite margins of the tumor. Good contact between the electrodes and the overlying skin was assured by means of a conductive gel (Parker Laboratories, Fairfield, NY ). Eight square wave pulses of different amplitudes and pulse duration at repetition frequency of 1 Hz were generated by a Jouan GHT 1287 electroporator ( Jouan, St. Herblaine, France ) or an electroporator built in - house. Different electric pulse amplitude, pulse duration, and DNA concentrations were further tested on the SaF tumor model to determine the optimal protocol for in vivo transfection of solid tumors using electroporation. Plasmid DNA was injected at 50 g/tumor ( D50 ), or 50 g of DNA with Ep1: voltage /electrode distance ratio 600 V /cm, pulse length 5 milliseconds (D50Ep1 ); or 50 g of DNA with Ep2: voltage /electrode distance ratio 1200 V /cm, pulse length 0.1 millisecond ( D50Ep2 ). Initial experiments using DNA lipofectin or DNA lipofectin peptide mixtures were carried out in P22 and SaF tumors ( two to four tumors each ). These initial conditions were, for LD: ( a) 15 L of lipofectin, 20 g of DNA; and (b ) 30 L of lipofectin, 40 g of DNA per tumor. The more dilute mixture was found to be superior. The initial conditions for LPD: ( a) 15 L of lipofectin, 40 g of peptide, 20 g of DNA per tumor, which results in a charge ratio of 3:1; (b ) 15 L of lipofectin, 80 g of peptide, 20 g of DNA, charge ratio 7:1; ( c ) 7.5 L of lipofectin, 20 g of peptide, 10 g of DNA, charge ratio 3:1, but half the concentration of (a ) above. Although little difference was seen among these conditions, ( c ) was slightly superior and less viscous than (a ). The


401 resulting in approximately 30% and 15% cells transfected, respectively. Electroporation of dense cell suspensions ( simulated in vivo conditions) ( Fig 1b ), with or without LD, was significantly more effective in B16 melanoma compared to other cell lines used in the experiments. In addition, under simulated in vivo conditions, LPD transfection with or without electroporation was more effective than transfection with DNA alone, but much less compared to DNA or LD combined with electroporation. Transfection with LD alone or combined with electroporation was generally much more effective than transfection with DNA alone. In P22 cells,

A 100

Transfection (%)

following conditions were therefore carried forward for detailed analysis in all four tumor models: 15 L of lipofectin and 20 g of DNA ( LD ); or with Ep1 (LDEp1); 7.5 L of lipofectin, 20 g of peptide, and 10 g of DNA ( LPD ); or with Ep1 ( LPDEp1 ), in 50– 70 L of PBS per tumor. No side effects were observed after the performed procedures. The tumors were excised 24 hours after treatment and were either rapidly frozen in liquid nitrogen and stored at 708C for subsequent analysis of reporter gene expression, or fixed with formaldehyde, paraffin- embedded, and stained with haematoxylin /eosin (H&E ) for determination of necrosis and cell density. To visualise GFP fluorescence, frozen sections (20 m ) of the tumors were cut at different depths throughout the tumor. On average, 20 sections were cut per tumor. An epifluorescence microscope ( Nikon type TE200 ) with a narrow band filter ( 500 –510 nm; Glen Spectra, Stanmore, UK ), equipped with a custom -made imaging system, was used to visualise GFP fluorescence and to quantify fluorescent area. Transfection efficiency was defined as the percentage of tumor area expressing GFP with regard to the total tumor area.

*

LD LPD

* 10

Cell survival after treatment

Cell survival after in vitro transfection was estimated using the MTS ( Promega, Southampton, UK ) proliferation assay according to manufacturer’s instructions. Cell survival in vivo was estimated by determining the percentage of necrotic area from H&E -stained sections using image analysis. Six fields were scored per section and six tumor sections were used.

1

B16

P22

Significance tests were carried out on the data groups using analysis of variance followed by the Student’s t test for individual pairwise comparisons, with values of P < .05 considered as significant.

D10

100

Transfection (%)

Statistical analysis

T24

B

Cell density assessment

In H&E - stained tumor sections, the number of cells in the microscopic fields was determined by image analysis. Six fields were scored per section and three to six tumors were used per tumor model. Cell density was expressed as average number of cells in the fields.

SaF

* * *

10

D50Ep

LD

LDEp

LPD

*

D10Ep

D50

LPDEp

*

** *

P22

SaF

*

*

*

*

1

0.1

0.01

B16

Results In vitro transfection efficiencies

Transfection efficiencies in vitro varied depending on the protocol and the cell line used (Fig 1, a and b ). Using preplated adherent cells (optimised in vitro conditions) ( Fig 1a ), integrin -targeted transfection (LPD ) was very effective for T24 tumor cells with 66% of the cells transfected, which was significantly better than using lipofectin ( LD ). LPD was also more effective for P22 cells; whereas for SaF and B16 cells, there were no differences between the two methods,

T24

Figure 1 A: In vitro transfection efficiencies in rodent and human cell lines using lipofectin - enhanced ( LD ) and integrin - targeted ( LPD ) methods; optimised for in vitro conditions. Experiments were performed at least three times using duplicate samples. Bars are arithmetic means ( AM ) ± standard error of the mean ( SE ). *P < .05 compared to LD group. B: In vitro transfection efficiencies using different DNA concentration ( D10, 10 g of DNA; D50, 50 g of DNA ), lipofectin - enhanced ( LD ), and integrin - targeted ( LPD ) methods with or without subsequent electroporation ( D10Ep, D50Ep, LDEp, LPDEp ) in simulated in vivo conditions. Experiments were performed at least three times using duplicate samples. Bars are AM ± SE. *P < .05 compared to specific group without Ep.

Cancer Gene Therapy


*

Transfection (%)

10

D50

D50Ep1

LD

LDEp1

LPD

LPDEp1

D50Ep2

* *

*

1

*

* *

*

0.1

* 0.01

B16

P22

SaF

T24

Figure 2 In vivo transfection efficiencies in rodent and human tumor xenografts using naked DNA injection ( D50, 50 g of DNA per tumor ), lipofectin - enhanced ( LD ), integrin - targeted ( LPD ) methods with or without subsequent electroporation [ D50Ep1 ( 600 V / cm, 5 milliseconds ), D50Ep2 ( 1200 V / cm, 0.1 millisecond ), LDEp1, LPDEp1 ]. Bars are AM ± SE of three to eight tumors per bar. In all tumor models, Ep1 ( 600 V / cm, 5 milliseconds, 1 Hz ) yielded significantly better transfection efficiency compared to Ep2 ( 1300 V / cm, 0.1 millisecond, 1 Hz ). *P < .05 compared to specific group without Ep.

transfection efficiency using 10 g of DNA was better than transfection efficiency using 50 g of DNA, possibly due to the toxicity of high concentrations of DNA for this particular cell line. An MTS proliferation assay showed that electroporation generally reduced survival by 20%, whereas survival of SaF cells was reduced by 50%. LD and LPD methods alone did not reduce cell survival. Interestingly, in combination with electroporation, LPD transfection tended to increase cell survival compared to DNA or LD. Hence, LPD transfection combined with electroporation resulted in lower transfection efficiency, but higher cell survival ( data not shown ).

efficiencies varied depending on the protocol and the tumors used. Parameters tested were: plasmid DNA injection (D50 ), electric pulse (Ep ) duration and amplitude, lipofectin –DNA complex ( LD ), lipofectin –peptide– DNA mixtures (LPD ), and combinations of these. LD transfection was only measurable in P22 and SaF tumors but was significantly improved by Ep in B16F1 and P22 tumors. In general, the most effective transfection method was electroporation assisted transfection using plasmid DNA or lipofectin –DNA complexes, which were superior to lipofectin – DNA complexes alone, and the integrin -targeted peptide transfection method. Transfection using simple i.t. DNA injection was detectable only in P22 and SaF tumors with 0.013% and 0.03% of transfected area, respectively. Lipofectin - enhanced transfection using lower DNA concentration than plasmid DNA injection was more effective than DNA injection alone, except in the T24 tumors where lipofectin -enhanced method with or without subsequent electroporation did not yield GFP expression. In all tumors tested, integrin - targeted transfection alone or combined with electroporation resulted in negligible GFP expression. These results were similar to the

10

A Transfection (%)

402

SaF

1

0.1

0.01 600

In vivo transfection efficiencies 10

B Transfection (%)

Transfection in vivo was far less efficient than under in vitro conditions ( Fig 2 ). Similar to the in vitro data, transfection 50

SaF Necrosis (%)

40 30 20

1000

1200

1400

SaF

1

0.1

0.01 200

10 0

800

Amplitude/electrode distance (V/cm)

300

400

500

600

Amplitude/electrode distance (V/cm) Control

D50

Ep

D50Ep

LD

LDEp

LPD

LPDEp

Figure 3 Extent of necrosis in SaF tumors after transfection using naked DNA injection ( D50 ), lipofectin - enhanced ( LD ), and integrin targeted ( LPD ) methods with or without subsequent electroporation ( D50Ep1, LDEp, LPDEp ). Bars are AM ± SE of six tumors per bar.

Cancer Gene Therapy

Figure 4 A: In vivo transfection efficiencies in SaF using different electric pulse amplitudes ( voltage / electrode distance ratio ) at 0.1 millisecond pulse duration. Data points are AM ± SE of three to seven tumors per point. B: In vivo transfection efficiencies in SaF using different electric pulse amplitudes ( voltage / electrode distance ratio ) at 5 milliseconds pulse duration. Data points are AM ± SE of four to eight tumors per point.


403

10

Transfection (%)

SaF 1

0.1

0.01 0.0

2.5

5.0

7.5

10.0

Pulse length (ms) Figure 5 In vivo transfection efficiencies in SaF using different electric pulse durations at 600 V / cm amplitude / electrode distance ratio. Data points are AM ± SE of four to eight tumors per point.

data obtained under simulated in vivo conditions, where this transfection method was less effective than lipofectinenhanced or electroporation -assisted gene delivery methods. Electroporation using two different electrical parameters increased GFP expression in all tumors. In all tumor models, Ep1 (600 V /cm, 5 milliseconds, 1 Hz ) yielded significantly better transfection efficiency compared to Ep2 ( 1300 V /cm, 0.1 millisecond, 1 Hz ). This was particularly evident in T24 tumor model, where Ep1 resulted in almost 20 times higher transfection efficiency compared to Ep2. In P22, SaF, and T24 tumors, the transfection efficiency using Ep1 was around 1%, whereas in B16F1 melanoma tumor it increased to 3%. Electroporation also significantly increased transfection efficiency of lipofectin -enhanced transfection, which was the most pronounced in B16F1 tumor with 6.4% of tumor area transfected. This was the highest transfection efficiency obtained in our experimental setup (Fig 2). Cell survival estimated by the extent of necrosis in SaF tumors demonstrated that electroporation combined with

DNA alone or lipofectin– DNA complexes tended to increase extent of necrosis in tumors, but not significantly compared to untreated control (Fig 3 ). Addition of the integrin - targeted peptide to the mixture, which appeared to abolish transfection in all tumor models, on the other hand, protected cells from dying, similarly to the simulated in vivo conditions. Further studies were conducted in SaF tumors to improve electroporation -assisted gene delivery by changing pulse amplitude and duration, as well as DNA concentration. Amplitude dependence was tested at two different pulse lengths, 0.1 and 5 milliseconds (Fig 4, a and b ). High voltage electric pulses at short pulse duration were ineffective in transfecting tumors yielding less than 0.1% transfected tumor area (Fig 4a ). Electric pulses using longer pulse duration and lower amplitude proved to be significantly more effective compared to high pulse amplitude with short duration. Furthermore, with increasing electric pulse amplitude, transfection efficiency increased, with up to 1% tumor area transfected at 600 V / cm, 5 milliseconds (Fig 4b). This increase was, however, not statistically significant compared to 200 V / cm. Transfection efficiency was also dependent on the pulse duration at 600 V /cm ( Fig 5 ). Transfection was obtained above the pulse duration of 1 millisecond with transfection efficiency at around 1%, therefore demonstrating a threshold phenomenon. Although not statistically significant, longer pulse duration appeared to yield better transfection efficiency compared to shorter pulses. Transfection efficiency in SaF tumors was not affected by increasing the DNA amount injected in tumors in the same volume from 10 to 150 g combined with Ep1 (Fig 6 ). Although a trend existed for increased transfection efficiency with higher DNA amounts, this was not statistically significant. Cell size and cell density were proposed to influence electroporation -assisted transfection. 19,20 This was investigated for the two most effective transfection regimes, namely plasmid DNA with Ep1 and Ep2. A negative

10

10

B16

Transfection (%)

Transfection (%)

SaF 1

0.1

0.01 0

25

50

75

100

125

150

DNA concentration (µg/tumour) Figure 6 In vivo transfection efficiencies in SaF tumors using different DNA amounts injected per tumor at electric pulse amplitude / electrode distance ratio 600 V / cm and 5 milliseconds pulse duration. Data points are AM ± SE of 4 – 10 tumors per point.

D50Ep1 r2=0.64 D50Ep2 r2=0.49 T24

1

SaF

P22

0.1

0.01 400

450

500

550

600

2

Cell density (# of cells/mm ) Figure 7 Correlation between cell density and transfection efficiency using two different electroporation protocols ( Ep1: 600 V / cm, 5 milliseconds; Ep2: 1200 V / cm, 0.1 millisecond ). Data points are AM ± SE of four to eight tumors per point.

Cancer Gene Therapy


404 correlation between cell density and transfection efficiency was found (r =0.64 and 0.49 for Ep1 and Ep2, respectively ), which indicates that transfection efficiency is better at lower cell density, with B16F1 melanoma having the lowest density of the four tumors tested (Fig 7). Discussion

In this study, the transfection efficiency of four nonviral methods of gene delivery and their combinations were compared. A plasmid encoding GFP was delivered to four histologically different tumor models in vitro and in vivo using either naked DNA injection, electroporation- assisted, liposome -enhanced, integrin - targeted gene delivery methods, or combinations of these methods. The results show that in vitro models are generally not predictive for the more complex in vivo situation, especially when using electroporation for gene delivery. The in vivo results demonstrate that electroporation - assisted, as well as lipofectinenhanced, gene delivery methods to solid tumors in vivo are feasible with relatively good transfection efficiency in some tumors. Our results demonstrated that electric pulses of lower amplitude and longer duration (600 V /cm, 5 milliseconds ) yielded significantly better transfection efficiency of plasmid DNA, compared to electric pulses of higher amplitude and shorter duration ( 1200 V /cm, 0.1 millisecond ). Therefore, the electroporation protocol employing longer pulses was used in combination with the liposome –DNA complex and with integrin -targeted vectors. Due to the limitation of the injection volume into the tumors (50 L ), DNA amounts in the lipofectin – DNA complex and in integrin - targeted vector had to be reduced to 20 and 10 g per tumor, respectively, compared to 50 g per tumor in the case of plasmid DNA injection. Liposome -enhanced gene delivery, by itself, even at less than half the DNA amount, was better than plasmid DNA injection alone. These results are in accordance with several preclinical and clinical studies demonstrating efficient gene transfer to solid tumors using i.t. liposome – DNA complexes injection.21 – 23 Furthermore, if we take into account lower DNA concentration, liposome -enhanced method combined with electroporation yielded better transfection than electroporation combined with plasmid DNA ( except in the case of T24 human bladder carcinoma ). Our results on T24 carcinoma are in agreement with the observation made by Wells et al.24 They found that electroporation increased transfection efficiency of plasmid DNA in MC2 mammary tumors, but not of a liposome – DNA complex.24 The reason for that is currently not known, but it might be due to the tumor type because other tumor types in our study responded well to this combination with evident transfection. In contrast to expectations, the integrin - targeted method was ineffective in vivo in solid tumors. In vitro data using this vector were promising, showing high transfection efficiency, and a recent in vivo report demonstrated delivery of the lacZ reporter gene to the lung, transfecting bronchial epithelium and parenchymal cells with similar efficiency to an adenoviral vector and with greater efficiency compared to

Cancer Gene Therapy

cationic liposomes.8 - 10 A possible reason for the lack of GFP expression in tumors is the very low amount of DNA used in our experiments. In the study of Jenkins et al,10 five times-higher DNA amounts were used to transfect the lung. In our study, the DNA amount was reduced due to tumor size and consequently limited DNA injection volume. On the other hand, the integrin -targeted method protected cells from dying from electroporation. This was shown in dense cell suspensions in vitro ( simulated in vivo conditions ) and in transfection of solid tumors in vivo. The reason for this observation is currently unknown, and deserves further studies. It could be due to the fact that electroporation induced damage to the cell membrane is counterbalanced by binding of the integrin -targeted peptide to the receptors, leading to a signalling cascade. Several studies have already reported the use of electroporation for gene delivery in vivo to different tissues.13 In the case of tumors, many different electrical parameters for electroporation were employed for delivering plasmid DNA encoding either reporter (luciferase, -galactosidase, GFP ) or therapeutic genes (IL - 12, Stat3, GM -CSF, IL - 2).25 – 35 Our study focused on optimisation of plasmid DNA delivery to tumors by using different electrical parameters and DNA concentrations in the SaF tumor model. Transfection was achieved with all electrical conditions tested; however, pulses of lower amplitude and longer duration yielded significantly better transfection efficiency. These results are in agreement with the results of Rols et al,29 showing that approximately 4% of cells in B16 melanoma tumors was transfected using 800 V /cm amplitude / electrode distance ratio and 5 milliseconds pulse duration. Furthermore, these results also support the results presented by Bettan et al,33 although in that study only long duration pulses with varying pulse amplitude were used. Based on the current results, further studies on pulse amplitudes and duration in different tumor models are needed to make predictions of optimal electroporation conditions for transfection of different tumor types. Also, these studies, as well as a computer modelling of the response of different tissues to the different electroporation conditions, are needed to determine whether the use of longer pulses than used in our study, with low amplitude, may yield better transfection efficiency. Our study showed that transfection efficiency varied between the tumor models used. Of the four histologically different tumors used, melanoma was easier to transfect than carcinoma, followed by carcinosarcoma and sarcoma. Therefore, it is apparent that tumor type is an important determinant for successful transfection. It was already proposed that this observation could be due to the specific properties of the tumor tissue, such as differences in tissue organisation, in extracellular matrix, presence or absence of necrosis, overall tissue conductivity, the ability of cells to express transfected genes, cell density, and cell size.13,19,33 A correlation between cell density and transfection efficiency by electroporation demonstrated that lower cell density yielded better transfection (Fig 7 ), which is in agreement with in vitro data obtained on adherent cells.36 The correlation was not very high, as only four tumor types


405 were included and should be expanded in a further study to prove the hypothesis. As we and others have shown that different intrinsic sensitivities to electroporation exist between different cell types in vitro, this might also be the case in solid tumors in vivo.37,38 It has already been shown that electroporation, in addition to the direct effect on cell membranes, also affects the cytoskeleton and chromosomal DNA, and therefore might trigger different cellular processes leading to apoptosis or necrosis.39,40 Another factor, which should be taken into account for gene delivery in vivo, is the amount of hypoxia (low oxygen ) present in the solid tumors. In vitro nonviral transfection was shown not to be reduced by hypoxia.9 However, GFP fluorescence in T24 tumor cells in vitro and in vivo was shown to be reduced by low oxygen tension.41 In that report, a heterogeneity of GFP expression was detected in solid tumors grown from stably transfected T24 cells.41 It is, therefore, possible that transfection efficiencies in solid tumors were underestimated in this report. It appears that an immune response did not influence the efficacy of the nonviral gene delivery methods because two tumors ( SaF, B16F1 ) were implanted in immunocompetent and two (P22, T24 ) in immunocompromised animals. Therefore, in contrast to viral gene delivery, our study with nonviral gene delivery further supports the notion that these methods are not eliciting an immune response and are feasible regardless of the immune status. In conclusion, our results demonstrate that nonviral methods of gene delivery are feasible and show promise to replace the viral methods of gene delivery. However, further studies are needed to optimise these methods for clinical applications.

Acknowledgments

This work was supported by the Cancer Research Campaign, Grant SP2292 /0102, and the Ministry of Education, Science, and Sport of the Republic of Slovenia. We would like to thank the Advanced Technology Development Group (Gray Cancer Institute ) for the production of the custom - made square wave electroporator and imaging system.

References 1. Harvey BG, Worgall S, Ely S, et al. Cellular immune responses of healthy individuals to intradermal administration of an E1 – E3 – adenovirus gene transfer vector. Hum Gene Ther. 1999;10:2823 – 2837. 2. Marshall E. Gene therapy death prompts review of adenovirus vector. Science. 1999;286:2244 – 2245. 3. Cotten M, Wagner E. Non - viral approaches to gene therapy. Curr Opin Biotechnol. 1993;4:705 – 710. 4. Kikuchi A, Aoki Y, Sugaya S, et al. Development of novel cationic liposomes for efficient gene transfer into peritoneal disseminated tumor. Hum Gene Ther. 1999;10:947 – 955. 5. Curiel DT, Agarwal S, Wagner E, et al. Adenovirus enhancement of transferrin – polylysine – mediated gene delivery. Proc Natl Acad Sci USA. 1991;88:8850 – 8854.

6. Wagner E, Plank C, Zatloukal K, et al. Influenza virus hemagglutinin HA - 2 N - terminal fusogenic peptides augment gene transfer by transferrin – polylysine – DNA complexes: toward a synthetic virus - like gene - transfer vehicle. Proc Natl Acad Sci USA. 1992;89:7934 – 7938. 7. Simoes S, Slepushkin V, Gaspar R, et al. Gene delivery by negatively charged ternary complexes of DNA, cationic liposomes and transferrin or fusogenic peptides. Gene Ther. 1998;5:955 – 964. 8. Hart SL, Arancibia - Carcamo CV, Wolfert MA, et al. Lipid - mediated enhancement of transfection by a nonviral integrin - targeting vector. Hum Gene Ther. 1998;9:575 – 585. 9. Dachs GU, Coralli C, Hart SL, et al. Gene delivery to hypoxic cells in vitro. Br J Cancer. 2000;83:662 – 667. 10. Jenkins RG, Herrick SE, Meng QH, et al. An integrin - targeted non - viral vector for pulmonary gene therapy. Gene Ther. 2000;7:393 – 400. 11. Mir LM, Glass LF, Sersa G, et al. Effective treatment of cutaneous and subcutaneous malignant tumors by electrochemotherapy. Br J Cancer. 1998;77:2336 – 2342. 12. Sersa G, Stabuc B, Cemazar M, et al. Electrochemotherapy with cisplatin: clinical experience in malignant melanoma patients. Clin Cancer Res. 2000;6:863 – 867. 13. Somiari S, Glasspool - Malone J, Drabick JJ, et al. Theory and in vivo application of electroporative gene delivery. Mol Ther. 2000;2:178 – 187. 14. Mir LM, Bureau MF, Gehl J, et al. High - efficiency gene transfer into skeletal muscle mediated by electric pulses. Proc Natl Acad Sci USA. 1999;96:4262 – 4267. 15. Heller L, Jaroszeski MJ, Coppola D, et al. Electrically mediated plasmid DNA delivery to hepatocellular carcinomas in vivo. Gene Ther. 2000;7:826 – 829. 16. Tozer GM, Shaffi KM. Modification of tumour blood flow using the hypertensive agent, angiotensin II. Br J Cancer. 1993;67:981 – 988. 17. Hill SA, Collinrigde DR, Vojnovic B, et al. Tumour radiosensitization by high - oxygen - content gases: influence of the carbon dioxide, content of the inspired gas on pO 2, microcirculatory function and radiosensitivity. Int J Radiat Oncol Biol Phys. 1998;40:943 – 951. 18. Cemazar M, Milacic R, Miklavcic D, et al. Intratumoral cisplatin administration in electrochemotherapy: antitumor effectiveness, sequence dependence and platinum content. Anticancer Drugs. 1998;9:525 – 530. 19. Miklavcic D, Beravs K, Semrov D, et al. The importance of electric field distribution for effective in vivo electroporation of tissues. Biophys J. 1998;74:2152 – 2158. 20. Cemazar M, Miklavcic D, Vodovnik L, et al. Improved therapeutic effect of electrochemotherapy by intratumoral drug administration and changing of electrode orientation for electropermeabilization on EAT tumor model in mice. Radiol Oncol. 1995;29:121 – 127. 21. Nabel GJ, Nabel EG, Jang Z, et al. Direct gene transfer with DNA – liposome complexes in melanoma: expression, biologic activity, and lack of toxicity in humans. Proc Natl Acad Sci USA. 1993;90:11307 – 11311. 22. Larchian WA, Horiguchi J, Nair SK, et al. Effectiveness of combines interleukin - 2 and B7.1 vaccination strategy is dependent on the sequence and order: a liposome - mediated gene therapy treatment for bladder cancer. Clin Cancer Res. 2000;6:2913 – 2920. 23. Yoo GH, Hung MC, Lopez - Berenstein G, et al. Phase I trial of intratumoral liposome EiA gene therapy in patients with recurrent breast and head and neck cancer. Clin Cancer Res. 2001;7:1237 – 1245.

Cancer Gene Therapy


406 24. Wells JM, Li LH, Sen A, et al. Electroporation - enhanced gene delivery in mammary tumors. Gene Ther. 2000;7:541 – 547. 25. Heller R, Schultz J, Lucas ML, et al. Intradermal delivery of interleukin - 12 plasmid DNA by in vivo electroporation. DNA Cell Biol. 2001;20:21 – 26. 26. Heller L, Pottinger C, Jaroszeski MJ, et al. In vivo electroporation of plasmids encoding GM - CSF or interleukin - 2 into existing B16 melanomas combined with electrochemotherapy induces long - term antitumour immunity. Melanoma Res. 2000;10:577 – 583. 27. Niu G, Heller R, Catlett - Falcone R, et al. Gene therapy with dominant - negative Stat3 suppresses growth of the murine melanoma B16 tumor in vivo. Cancer Res. 1999;59:5059 – 5063. 28. Yamashita YI, Shimada M, Hasegawa H, et al. Electroporation - mediated interleukin - 12 gene therapy for hepatocellular carcinoma in the mice model. Cancer Res. 2001;61:1005 – 1012. 29. Rols MP, Delteil C, Golzio M, et al. In vivo electrically mediated protein and gene transfer in murine melanoma. Nat Biotechnol. 1998;16:168 – 171. 30. Nishi T, Yoshizato K, Yamashiro S, et al. High - efficiency in vivo gene transfer using intraarterial plasmid DNA injection following in vivo electroporation. Cancer Res. 1996;56:1050 – 1055. 31. Goto T, Nishi T, Tamura T, et al. Highly efficient electro - gene therapy of solid tumor by using an expression plasmid for the herpes simplex virus thymidine kinase gene. Proc Natl Acad Sci USA. 2000;97:354 – 359. 32. Baba M, Iishi H, Tatsuta M. In vivo electrophoretic transfer of bcl - 2 antisense oligonucleotide inhibits the development of

Cancer Gene Therapy

hepatocellular carcinoma in rats. Int J Cancer. 2000;85:260 – 266. 33. Bettan M, Ivanov MA, Mir LM, et al. Efficient DNA electrotransfer into tumors. Bioelectrochemistry. 2000;52:83 – 90. 34. Lohr F, Lo DY, Zaharoff DA, et al. Effective tumor therapy with plasmid - encoded cytokines combined with in vivo electroporation. Cancer Res. 2001;61:3281 – 3284. 35. Yoshizato K, Nishi T, Goto T, et al. Gene delivery with optimized electroporation parameters shows potential for treatment of gliomas. Int J Oncol. 2000;16:899 – 905. 36. Susil R, Semrov D, Miklavcic D. Electric field induced transmembrane potential depends on cell density and organization. Electro - Magnetobiol. 1998;17:391 – 399. 37. O’Hare MJ, Ormerod MG, Imrie PR, et al. Electropermeabilization and electrosensitivity of different types of mammalian cells. In: Neumann E, Sowers AE, Jordan CA, eds. Electroporation and Electrofusion in Cell Biology. New York: Plenum, 1989:319 – 330. 38. Cemazar M, Jarm T, Miklavcic D, et al. Effect of electric - field intensity on electropermeabilization and electrosensitivity of various tumor - cell lines in vitro. Electro - Magnetobiol. 1998;17:261 – 270. 39. Teissie J, Rols MP. Manipulation of cell cytoskeleton affects the lifetime of cell membrane electropermeabilization. Ann NY Acad Sci. 1994;720:98 – 110. 40. Meaking WS, Edgerton J, Harton CW, et al. Electroporation induced damage in mammalian cell DNA. Biochim Biophys Acta. 1995;1264:357 – 362. 41. Coralli C, Cemazar M, Kanthou C, et al. Limitations of the reporter green fluorescent protein under simulated tumor conditions. Cancer Res. 2001;61:4784 – 4790.