Viral vector systems are the most commonly used gene transfer tools for clinical gene therapy. However, lipofection systems are potential alternatives because of lower immunogenicity and easier cGMP production, but in vivo stability and transduction efficacy need to be improved. Therefore, we investigated gene transduction efficiency of our novel cGMP cationic lipids, CCQ22 and CCQ32, by FACS analysis. Toxicity analysis was performed to determine the cytotoxic side effects of the novel lipids. To evaluate the stability of the compounds in the context of local delivery to patients with intraperitoneally metastatic ovarian cancer, gene transfer was also tested in the presence of malignant ascites. Our novel cGMP standard lipids mediated gene transfer rates of more than 50%. However, for most cell lines cytotoxic side effects were similar to our reference lipofection system. In general, ascites had no major influence on gene transduction rates with the novel lipids. Our results suggest that CCQs may compare favorably with commercially available lipofection systems. These promising results facilitate further analysis of the compounds.
Recent clinical results have revealed gene transfer efficacy to be a crucial obstacle to a successful gene therapy. Many viral vector systems display superior gene transfer efficacy. However, their complex biological character results in complicated pharmacodynamics and pharmacokinetics, which could lead to surprising side effects. Further, many viral agents are immunogenic, which may cause dose-limiting toxicity and reduced efficacy upon readministration. In contrast, nonviral vectors show lower immunogenicity and easier cGMP-production.1 Therefore, an attractive concept is the development of highly efficient nonviral, liposomal vector systems with acceptable toxicity profiles.
Some of the existing commercially available lipofection systems show efficient gene transfer rates and tolerable cytotoxic effects.2 However, clinical applications require cGMP-grade reagents, which is not attained by all commercially available lipofection systems, such as Fugene6™ (Personal communication by the manufacturer). We designed novel liposomal vectors consisting of cationic lipids, CCQ22 or CCQ32, and a helper lipid (DOPE), which can be produced at cGMP-quality.3
All transduction techniques including lipofection result in cytotoxic side effects.4 Different ratios of lipids (DOPE, CCQ) were used to attempt the balance of maximal gene transfer efficiency and minimal cytotoxicity.
Some existing cationic lipids, for example, DC-Chol have been used in clinical trials for the treatment of cancer and cystic fibrosis. However, the lipids showed low efficiency of in vivo gene transfer and transient gene expression.5
The aim of this study was to show the in vitro applicability of CCQ22 and CCQ32 to breast and ovarian cancer cells. We demonstrate that both novel lipids mediate gene transfer rates at least comparable to commercially available lipofection systems. In addition, we investigated effects of ascites on gene transfer efficiencies.
Materials and methods
For X-Gal staining experiments we used pRc/CMVlacZ.3 For FACS analysis we used pEGFP-C1 (Clontech, Heidelberg, Germany). The control plasmid pRc/CMV was purchased from Invitrogen, Groningen, The Netherlands.
Cell lines, cell culture and patients material
The breast cancer cell lines MCF7, BT-20, ZR-75-1, SK-BR-3 and the ovarian cancer cell lines SK-OV-3 and MDAH 2774 were purchased from ATCC (Rockville, MD). The ovarian cancer cell line Hey was a generous gift from Dr Pfleiderer (University of Freiburg, Germany). MCF7 and BT-20 were cultivated in DMEM, SK-OV-3 and Hey in RPMI 1640, ZR-75-1 in MEM and MDAH 2774 in DMEM (high glucose). All media were supplemented with 10% FCS, vitamins, I-glutamine and antibiotics. Cells were maintained in a humidified incubator at 37°C and 5% CO2. In experiments using ascites, FCS was replaced by ascites. Cells were grown in medium containing ascites 1 day prior to lipofection.
Ascites was collected from a patient suffering from intraperitoneal ovarian carcinoma with overproduction of ascites. It was centrifuged five times at 2000 rpm for 10 minutes to eliminate cells or cellular components.
Lipids of CCQ22, CCQ32 and DC-Chol were synthesized as described elsewhere.3 Designations of CCQs give the molar percentage (mol%) of cationic lipid (e.g. CCQ32-20 denotes 20 mol% of cationic lipid and 80 mol% helper lipid). The liposomes CCQ22, CCQ32, DOPE and the lipids DC-Chol or Fugene6™ (Roche, Mannheim, Germany) were incubated in serum-free medium for 5 minutes at room temperature. These diluted liposomes were added to 1 μg of plasmid DNA. After incubation for 15 minutes at room temperature, the resulting lipoplexes were added to the cells, which were maintained in normal cultivation medium. Further details on CCQs chemistry and lipofection conditions are available on request.
In 96-well plates, each well was seeded with 1.2 × 104. cells and lipofection was performed with varying DNA/liposome ratios. Cytotoxicity was examined for different molar ratios of CCQ22 or CCQ32 and the helper lipid DOPE. At 2 days after lipofection of the cells with pRc/CMVlacZ, the dual assay was carried out as described elsewhere.3
For detailed analysis of transfection efficiency, we used pEGFP-C1. We seeded 0.4–1 × 105 cells/well in 24-well plates and lipofection was performed 24 hours later using 1 μg of plasmid DNA per well. At 48 h after lipofection, cells were harvested, resuspended in 500 μl PBS and stored on ice. For FACS analysis, 5000 cells were counted in a cytofluorometer (Becton-Dickinson).
Data are expressed as mean ±SD of separate experiments. For cytotoxicity assays, experiments were done twice and in FACS analysis all values consist of three single experiments. The Student's t-test (pairwise, double sided) was used to determine the significance of differences between groups. P-values less than 0.05 were considered significant.
Cytotoxicity of CCQ22 and CCQ32
To reduce cytotoxic effects of lipofection, the ratio of cationic and helper lipid was optimized for each cell line. Depending on the cell line, liposomes consisting of 20–50 mol% of cationic lipid displayed favorable results (Table 1A).
Some cell lines (e.g. BT-20) were more susceptible to cytotoxicity of the lipofection procedure than other cell lines (e.g. MDAH 2774). BT-20 displayed 28% surviving cells after lipofection using Fugene6™ and 19.5% using CCQ32-30. In MDAH 2774 cells, we did not see any cytotoxic effect using Fugene6™ or CCQ32-30.
In most cell lines, CCQ22 was less toxic than CCQ32 and toxicity increased with raising percentages of cationic lipid. In relation to our reference reagent, Fugene6™, which was the best commercially available lipofection system we tested in our cell lines, CCQ22 showed moderate to low toxicity with all lipid ratios investigated, whereas CCQ32 resulted in similar or higher toxicity (Fig 1).
Optimization of transfection efficiency using CCQs
Different ratios of cationic lipid and helper lipid may have distinct gene transfer properties. Gene transfer efficiency depended on the cell line and lipofection conditions (Table 1B). Gene transfer rates in MDAH 2774 were relatively low (7.1%) using Fugene6™, but could be increased with CCQ22–30 (10.2%) and significantly enhanced with CCQ22–50 (17.2%, P=0.015).
In Hey cells, the highest gene transduction efficiency was obtained using CCQ32-30 (Fig 2b). The result of 51.9% GFP-positive cells was significantly higher as compared to Fugene6™ (14.5%, P=0.042).
ZR-75-1 cells, which were inefficiently transfected with lipofection so far, displayed a gene transfer rate of 36.5% using CCQ32–50 (Fig 2b). Compared with Fugene6™ (5.9%) and DC-Chol (5.4%), this increase revealed to be significant (P=0.0043 and 0.0057, respectively).
Summarizing our transduction experiments, we found variation of gene transfer efficiencies dependent on cell lines and lipofection conditions. CCQ32 was the favorable gene transduction system in four cell lines (Hey, MCF7, SK-OV-3, ZR-75-1), CCQ22 in MDAH 2774 cells, and Fugene6™ in two cell lines (BT-20, SK-BR-3). In comparison with Fugene6™, we observed in MDAH 2774, Hey, and ZR-75-1 cells significantly higher transduction rates after lipofection with CCQs.
Influence of ascites on transduction efficiency
Ovarian cancer patients often suffer from intraperitoneal metastases and ascites. Ascites in many ways resembles serum, which inhibits lipofection by affecting complex formation of DNA and lipids. Regarding possible clinical use of the new lipids, we were interested in examining the influence of ascites on gene transfer rates.
First, we evaluated the influence of various ascites concentrations on lipofection of SK-OV-3 cells, and found a moderate decrease in transduction efficiency using CCQ32-40 (Table 1C, Fig 3a). Interestingly, a 14.5% reduction of gene transfer efficiency was obtained with 5% ascites, whereas this effect was smaller (7.5%) with 40% ascites. Using Fugene6™ (Table 1C), gene transfer rates decreased dramatically from 59.2% (10% FCS) to 32.2% (5% ascites) and 1.0% (40% ascites). We tried to compensate for these effects by variation of liposomal formulations, but the decrease of gene transfer efficiency was constant in SK-OV-3 (10–15%), Hey (5–10%), and MDAH 2774 (<3%) (Table 1D,Fig 3b). Therefore, MDAH 2774 cells seemed to be less sensitive to the negative influence of ascites on gene transduction.
Since successful in vitro transfection with cationic lipids was first described, cationic lipids have been extensively studied. Although gene transfer rates have been recently improved, they typically remain lower as compared to viral vector systems.5 Many companies have developed lipofection systems with improved transduction rates.1,2 Some of these lipofection systems, especially Fugene6™, however, are not cGMP-conform.3 For these reasons, we tested our cGMP-quality lipids, CCQ22 and CCQ32,3 in breast and ovarian cancer substrates.
In general, all gene transfer techniques are associated with cytotoxic effects.4,5 Compared to our reference system, Fugene6™, the novel lipids displayed similar or slightly higher cytotoxicity in most cell lines tested in this study. In recent reports,4 modified liposomes yielded 15–60% viable cells after lipofection. In CCQ-mediated gene transfer experiments, viable cells varied between 19.5 and 95.5% under optimal gene transfer conditions. Thus, CCQs caused cytotoxicity comparable to other liposomal vector systems.
It is well known that certain cell lines are more susceptible to lipofection than to others.6 Systematic analysis in the different cell lines revealed that cell line-specific lipofection conditions for optimal gene transfer rates have to be evaluated. In ZR-75-1, variation of the molar ratio of lipids (CCQ32 and DOPE) resulted in a significant increase of gene transduction rates from below 1% to more than 36%.
Transduction rates below 20% in some cell lines demonstrate that there is still need for improved lipofection systems. Boussif et al7 demonstrated that modified analogs of DOPE may enhance gene delivery with cationic lipids. Combinations of this modifications might further enhance gene transfer with our novel cGMP liposomes.
In recent reports, ascites fluid has been shown to inhibit gene transfer with viral and nonviral vector systems.8,9 The reasons are not completely clear so far, but lipoplex disintegration and DNA degradation seems to be involved.5 Using ascites in lipofection experiments with CCQs, we found a reduction of gene transfer efficiency of 3–15%. This effect does not probably result from ascites interference in lipoplex stability or gene transfer, because higher ascites concentrations did not increase reduction of gene transfer mediated by CCQs as seen in experiments using Fugene6™. More likely the ascites effect is caused by the reduction of the cell proliferation rate observed using ascites instead of FCS, because mitotic activity is known to be a critical factor for successful lipofection.10 These differences in mitotic activity should be further investigated, for example, by cell cycle analysis.
In summary, our data show that the novel cGMP-grade lipids mediate efficient gene transfer in different breast and ovarian cancer cell lines. In three cell lines (Hey, MDAH 2774, ZR-75-1) we got significantly higher transduction rates using CCQs instead of Fugene6™. Cytotoxic side effects were similar to those reported for other lipofection systems. However, lipofection conditions required optimization for each cell line in order to achieve optimal gene transfer rates. With regard to the in vivo application of CCQs, ascites seems to have no major influence on gene transfer rates. These in vitro results suggest that our novel GMP-quality cationic lipids should be studied for in vivo efficacy.
Haines AMR, Irvine AS, Mountain A, et al. CL22 — a novel cationic peptide for efficient transfection of mammalian cells. Gene Ther. 2001;8:99–110.
Madry H, Trippel SB . Efficient lipid-mediated gene transfer to articular chondrocytes. Gene Ther. 2000;7:286–291.
Keil O, Bojar H, Prisack HB, Dall P . Novel lipophilic chloroquine analogues for a highly efficient gene transfer into gynecological tumors. Bioorg Med Chem Lett. 2001;11:2611–2613.
Yamazaki Y, Nango M, Matsuura M, et al. Polycation liposomes, a novel nonviral gene transfer system, constructed from cetylated polyethylenimine. Gene Ther. 2000;7:1148–1155.
Li S, Huang L . Nonviral gene therapy: promises and challenges. Millenium Review. Gene Ther 2000;7:31–34.
Hernandez A, Zöllner K, Enczmann J, et al. Differential transfection efficiency of the GM-CSF gene into human renal cell carcinoma lines by lipofection. Cancer Gene Ther. 1997;4:59–65.
Boussif O, Gaucheron J, Boulanger C, et al. Enhanced in vitro and in vivo cationic lipid-mediated gene delivery with a fluorinated glycerophosphoethanolamine helper lipid. J Gene Med. 2001;3:109–114.
Audouy S, Molema G, de Leij L, Hoekstra D . Serum as a modulator of lipoplex-mediated gene transfection: dependence of amphiphile, cell type and complex stability. J Gene Med. 2000;2:465–476.
Blackwell JL, Li H, Gomez-Navarro J, et al. Using a tropism-modified adenoviral vector to circumvent inhibitory factors in ascites fluid. Hum Gene Ther. 2000;11:1657–1669.
Brunner S, Sauer T, Carotta S, et al. Cell cycle dependence of gene transfer by lipoplex polyplex and recombinant adenovirus. Gene Therapy 2000;7:401–407.
We thank B Hanzen for excellent technical assistance. G R is a member of the Duesseldorf Entrepreneurs Foundation. P D was supported by the Ministry of Science and Research (MSWF), NRW, Germany (Grant 9772125).
About this article
Technology in Cancer Research & Treatment (2003)