Successful Transfection of Lymphocytes by ... - Semantic Scholar

Bioscience Reports, Vol. 19, No. 6, 1999

Successful Transfection of Lymphocytes by Ternary Lipoplexes S. Simões,1,2,3 V. Slepushkin,1,5 R. Gaspar,2,3 M. C. Pedroso de Lima,2,4 and N. Düzgünes1,6 Received May 27, 1999, Accepted July 28, 1999 Transgene expression in lymphoid cells may be useful for modulating immune responses in, and gene therapy of, cancer and AIDS. Although cationic liposome-DNA complexes (lipoplexes) present advantages over viral vectors, they have low transfection efficiency, unfavorable features for intravenous administration, and lack of target cell specificity. The use of a targeting ligand (transferrin), or an endosome-disrupting peptide, in ternary complexes with liposomes and a luciferase plasmid, significantly promoted transgene expression in several T- and B-lymphocytic cell lines. The highest levels of luciferase activity were obtained at a lipid/DNA (±) charge ratio of 1/1, where the ternary complexes were net negatively charged. The use of such negatively charged ternary complexes may alleviate some of the drawbacks of highly positively charged plain lipoplexes for gene delivery. KEY WORDS: Cationic liposomes; luciferase; H9 cells; PM1 cells; GALA.

INTRODUCTION Lymphocytes play a major role in the immune system and represent an important target for gene transfer studies aimed at human gene therapy. Adoptive cellular immunotherapy based on the use of genetically modified T-cells represents a promising strategy to increase the immune response against viral infections and malignant diseases, as well as to correct single gene defects in T-cell immunodeficiency syndromes, such as adenosine deaminase deficiency (Hwu et al., 1993; Tran et al., 1994; Blaese et al., 1995; Hege and Roberts, 1996; Heslop et al., 1996). CD4-positive Tlymphocytes are one of the predominant cell reservoirs for HIV-1. "Intracellular immunization" of these cells, aiming at inhibiting viral replication, has been pursued by introduction of therapeutic genes whose expression would lead to suppression of 1

Department of Microbilogy, School of Dentistry, University of the Pacific, San Francisco, CA 94115, USA. 2 Laboratory of Pharmaceutical Technology, Faculty of Pharmacy. 3 Center for Neurosciences. 4 Department of Biochemistry, University of Coimbra, 3000 Coimbra, Portugal. 5 Present address: Gene Transfer Core Laboratory, 221 EMRB, University of Iowa, College of Medicine, Iowa City, IA 52242. 6 To whom correspondence should be addressed. E-mail: [email protected] 601 0144-8463/99/1200-0601S16.00/0 © 1999 Plenum Publishing Corporation

602

Simões, Slepushkin, Gaspar, Pedroso de Lima, and Düzgünes

HIV integration, inhibition of proviral gene expression (Baltimore, 1988; Woffendin, et al., 1994, Yu et al., 1994; Zhou et al., 1994; Lisziewicz et al., 1995; Konopka et al., 1998), or activation of suicide genes in virally infected cells (Harrison et al., 1992; Konopka et al., 1997). Retrovirus-mediated gene transfer is currently the method of choice for transfection of human T-lymphocytes for applications in gene therapy (Buschle et al., 1995). However, serious drawbacks are associated with the use of such vectors, namely the limited size of carried genetic material, the requirement for dividing target cells and potential safety risks related to the possibility of random integration into the host genome, oncogene activation or generation of active viral particles (Singhal and Huang, 1994; Clark and Hersh, 1999). Cationic liposome/DNA complexes ("lipoplexes"; Feigner et al., 1997) present several advantages over viral vectors for gene delivery, as they are non-infectious, appear to be non-immunogenic in vivo, can carry large pieces of DNA and are easy to produce on a large scale (Singhal and Huang, 1994; Cheng, 1996; Lasic and Templeton, 1996). However, the lower levels of transfection per particle compared to viral vectors, their unfavorable features for intravenous administration including their net positive charge, and the lack of target cell specificity, has restricted their use for gene therapy approaches in vivo. Nevertheless, recent studies with both proliferating and non-proliferating cells have shown that transfection mediated by lipid-based gene delivery systems can be drastically enhanced, either by promoting internalization via the association of targeting ligands with lipoplexes, or improving cytoplasmic delivery by the association of pHsensitive amphipathic peptides (Cheng, 1996; Simões et al., 1997, 1998, 1999a, b). Alerted by previous results obtained with adherent cell lines (Simões et al., 1998) we decided to investigate whether lipoplexes associated with the targeting ligand transferrin, or the pH-sensitive, fusogenic peptide, GALA (Subbarao et al., 1987; Parente et al., 1988), would also mediate efficient transfection of lymphoid cells. Our goal was to test if promotion of internalization of the complexes triggered by transferrin (either by receptor-mediated or non-specific endocytotic processes), or improvement of cytoplasmic delivery of DNA through destabilization of the endosomal membrane (induced by GALA), would result in higher levels of gene transfer into these cells, as compared to plain lipoplexes. Three lymphocytic cell lines were used for this purpose: H9 and PM1 cells, which are both CD4+ clonal derivatives of the Hut-78 T-cell line (Mann et al., 1989; Lusso et al., 1995), and B-lymphocytic TF228.1.16 cells that stably express functional HIV envelope proteins on their surface (Jonak et al, 1993).

MATERIALS AND METHODS Materials The cationic lipid 1,2-dioleoyl-3-(trimethylammonium) propane (DOTAP), and dioleoylphosphatidylethanolamine (DOPE), were purchased from Avanti Polar Lipids (Alabaster, AL). Iron-saturated, heat-inactivated human transferrin was obtained from Collaborative Biomedical Products (via Becton Dickinson, Bedford, MA). The pH-sensitive, amphipathic, 30-amino acid peptide, GALA, with the

Transfection of Lymphocytes

603

sequence WEAALAEALAEALAEHLAEALAEALEALAA (Subbarao et al., 1987; Parente et al, 1990), was synthetized and HPLC-purified by the UCSF Biomolecular Resource Center (San Francisco, CA). The pCMVluc plasmid (VR-1216) was kindly provided by Dr. P. L. Feigner (Vical, Inc., San Diego, CA), NaCl, and N-(2hydroxyethyl) piperazine-N'-(2-ethanesulfonic acid) (HEPES) were obtained from Sigma (St. Louis, MO).

Liposome Preparation

Cationic liposomes composed of DOTAP:DOPE (1:1 weight ratio) were prepared by first drying a chloroform solution of the lipids under a stream of argon and then in a vacuum oven at room temperature, and hydrating the lipid film with 1 ml of deionized water, at a final concentration of 5 mg/ml. The multilamellar vesicles obtained were then sonicated briefly under argon, extruded 21 times through polycarbonate filters of 50 nm pore diameter using a Liposofast device (Avestin, Toronto, Canada), diluted 5 times with deionized water and filter-sterilized utilizing Millex 0.22 mm pore-diameter filters (Millipore, Keen, NH) (Simões et al., 1998).

Cells H9 and PM1 cell lines (obtained from R. Gallo, and from P. Lusso and M. Reitz, respectively, through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH) were grown in RPMI 1640 medium (Irvine Scientific, Santa Ana, CA) supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS) (Sigma, St. Louis, MO), penicillin (100 units/ml), streptomycin (100 mg/ml) and L-glutamine (2 mM). TF228.1.16 cells (a gift from Z, L. Jonak and E. Henri at Smithkline Beecham Pharmaceuticals, King of Prussia, PA) were grown in Dulbecco's Modified Eagles's Medium-high glucose (DME-HG) (Irvine Scientific, Santa Ana, CA) supplemented with 16% (v/v) heat-inactivated fetal bovine serum (FBS), penicillin (100 units/ml), streptomycin (100 mg/ml) and L-glutamine (4 mM). Cells were maintained at 37°C, under 5% CO2, and passaged in T-25 flasks (Corning Costar, Cambridge, MA, USA) twice a week.

Preparation of the Ternary Complexes

One hundred ml of 100 mM NaCl, 20 mM Hepes, pH 7.4 (HBS), with or without 32 mg iron-saturated human transferrin (Cheng, 1996), was mixed with 2.5, 5, 10 or 20 ml liposomes, and incubated at room temperature for 15 min. One ug pCMVluc plasmid in 100 ul of HBS was then added and gently mixed. The mixture was further incubated for 15 min at room temperature. Peptide complexes were prepared in a similar manner, but initially mixing a solution of 0.6 ug of GALA in 100 ul of HBS buffer with the liposomes. These concentrations of transferrin and GALA were previously established to be optimal for transfection of epithelial cells lines (Cheng et al., 1996; Simões et al., 1998).

604


Transfection Activity

Cells were rinsed twice with serum-free medium, and 106 cells/0.3 ml of medium aliquoted into polypropylene culture tubes (Corning Costar, Cambridge, MA) before lipid/DNA complexes were added. Lipid/DNA complexes were added gently to cells in a volume of 0.2 ml per tube. After an incubation for 4 hours in 5% CO2 at 37°C, the cells were centrifuged at 180g for 5 min, the supernatant was replaced with medium containing FBS. The cells were resuspended and further incubated for 24 or 48 hours. The cells were then washed twice with phosphate-buffered saline (PBS) and 100 ul of lysis buffer (Promega, Madison, WI) were added to each tube. The level of gene expression in the lysates was evaluated by measuring light production by luciferase, using a scintillation counter protocol (Promega). The protein content of the lysates was measured by the DC Protein Assay reagent (Bio-Rad, Hercules, CA) using bovine serum albumin as the standard. The data were expressed as ng of luciferase (based on a standard curve for luciferase activity), per mg of total cell protein. RESULTS Promotion of internalization of DOTAP:DOPE/DNA complexes through the association of iron-saturated human transferrin was found to be an efficient strategy to enhance transfection of both adherent, proliferating cells (Simões et al., 1997, 1998), and non-proliferating primary human macrophages (Simões et al., 1997, 1999a). Association of the fusogenic peptide GALA, with or without transferrin, also resulted in an enhancement of transfection of these cells compared to that obtained with plain lipoplexes (Simões et al., 1997, 1998, 1999a). A similar strategy was tested in different lymphoid cell lines growing in suspension, since these cells express transferrin receptors and represent important models for studies in gene therapy applications. In this regard, it is also of interest that transferrin is required for the transformation of lymphocytes in response to cytokine stimulation (Brock and Mainou-Fowler, 1983). Results presented in Fig. 1 illustrate the effect of the association of the ligand transferrin to DOTAP:DOPE/DNA complexes on the levels of transfection activity in H9 cells, as well as the duration of luciferase gene expression. Different lipid/ DNA charge ratios were tested, since previous data indicated that the transfection activity may vary depending on the net charge of the complexes (Simões et al., 1997, 1998, 1999a). A significant enhancement of transfection was obtained with transferrin-lipoplexes, compared to control plain lipoplexes. Although this effect was observed for all the lipid/DNA (±) charge ratios tested, the highest levels of luciferase expression were obtained for the 1/1 (±) charge ratio, where an enhancement of almost 150-fold was observed compared to controls at the same charge ratio. Nevertheless, the levels of transgene expression decreased from 24 h to 48 h following transfection, indicting the short term expression of the luciferase plasmid when transferred by this method. Cell viability, evaluated by the Alamar Blue assay (Konopka et al., 1996), was approximately 100% compared to untreated cells in both cases (data not shown).


605

Fig. 1. Effect of tranferrin complexation with DOTAP:DOPE liposomes on luciferase gene expression in T-lymphocytic H9 cells, and the time-dependence of gene expression. The level of luciferase gene expression was evaluated 24 or 48 h following the 4 h incubation of the cells with the ternary complexes, as described in "Materials and Methods". The data, expressed as ng of luciferase per mg of total cell protein, indicate the mean ± standard deviation obtained from triplicate wells, and are representative of 2 independent experiments.

Parallel studies with other lymphoid cells also showed an enhancement of transfection by transferrin (Fig. 2). The 1/1 lipid/DNA (±) complexes containing transferrin were clearly superior to plain lipoplexes in mediating gene transfer to PM1 and TF228.1.16 cells. Besides illustrating the versatility of this type of complexes, successful transfection of these cell lines is of particular interest for gene therapy approaches to HIV infection. PM1 cells are characterized by a unique susceptability to a wide range of HIV-1 isolates and represent a reproducible and efficient cellular system for the in vitro propagation of primary and macrophage-tropic isolates of HIV-1. These cells may thus provide a precious tool for studies aimed at developing broadly active vaccines against HIV-1 (Lusso et al., 1995). TF228.1.16 cells are derived from a human Burkitt's lymphoma cell line (BJAB) that represents an early B cell type. They express the gp 120/gp41 protein at the cell surface, thus mimicking HIV-infected cells. TF228.1.16 cells form syncytia with human CD4+ cells, and thus provide a virus-free cell-based assay for evaluating novel agents that may prevent syncytia formation (Jonak et al., 1993). The cytoplasmic membrane is not the only obstacle for the effective intracellular delivery of genetic material mediated by non-viral vectors. In fact, recent evidence that the endocytotic pathway is the main mechanism involved in the entry of the lipid-DNA complexes into the cytoplasm, raises the importance of promoting destabilization of the endosomal membrane as an important step in enhancing gene transfer and protecting DNA degradation at the lysosomal level (Simões et al.,

606


Fig. 2. Gene delivery to T-lymphocytic PM 1 and B-lymphocytic TF228.1.16 cells by lipoplexes with and without tranferrin. In this experiment only the 1/1 lipid/ DNA (±) complex was tested and the incubation time following transfection was 24 h. The data, expressed as ng of luciferase per mg of total cell protein, indicate the mean ± standard deviation obtained from triplicate wells, and are representative of 2 independent experiments.

1999b; Clark and Hersh, 1999). The association of the synthetic fusogenic peptide, GALA, with DOTAP:DOPE/DNA complexes also led to a significant enhancement of transfection of both H9 and TF288.1.16 cells (Fig. 3). This compound is a 30amino acid, pH-sensitive peptide that undergoes a transition from a random coil at pH 7.5 to an amphipathic a-helix at pH 5.0, under which conditions it is able to strongly interact with target membranes to induce fusion, contents leakage and phospholipid flip-flop (Subbarao et al., 1987; Parente et al, 1988; Parente et al, 1990). The peptide was also shown to induce endosome destabilization (Plank et al., 1994) and to enhance gene delivery by dendrimers (Haensler and Szoka, 1993). The highest levels of luciferase expression were achieved when GALA was associated with 1/1 lipid/DNA (±) charge ratio complexes for both cell lines studied. However, the significant enhancement of transfection obtained with the complexes of 1/2 lipid/ DNA (±) charge ratio, which are a priori negatively charged, was also of particular interest. DISCUSSION Lymphocytes have been described as being cells that are difficult to transfect. Our results indicate that the association of a cell-binding ligand, or of endosomedisrupting peptides are valuable strategies to enhance intracellular delivery of cationic liposome-DNA complexes into, and transgene expression in, different lymphocyte cell lines. The highest levels of transfection appear to occur when these ternary


607

Fig. 3. Gene transfer into H9 and TF228.1.16 cells by GALA-lipoplexes. The level of luciferase gene expression was evaluated as described in "Materials and Methods". The data, expressed as ng of luciferase per mg of total cell protein, indicate the mean ± standard deviation obtained from triplicate wells, and are representative of 2 independent experiments.

complexes are net negatively charged, as determined by zeta potential measurements (Simões et al., 1997, 1998). These observations suggest that the use of the ternary complexes may alleviate the problems associated with the high positive charge necessary for transfection by plain lipoplexes, such as interaction with and neutralization by negatively charged macromolecules in serum or tissues. Thus, it may be possible to utilize such ternary complexes in both ex vivo or in vivo gene therapy protocols. The ternary complexes appear to be stable enough to facilitate enhanced transfection even in the presence of high concentrations of serum (Simões et al., 1998; C. Tros & N. Düzgünes, submitted). The combined use of both strategies, i.e., simultaneous association of transferrin and the GALA peptide with the lipoplexes, did not enhance transfection over that achieved with transferrin or GALA-lipoplexes alone (data not shown). This observation is in contrast to the results obtained with macrophages, where highest levels of transfection were observed when a combination of transferrin and GALA were employed (Simões et al., 1997, 1999a), and suggests that the rate-limiting steps for gene delivery in macrophages and lymphocytes are different. Internalization of the lipolexes appears to be the rate-limiting step of lymphocyte transfection. This is particularly evident for the case of H9 cells where transfection activity mediated by transferrin-lipoplexes (promoting internalization of the lipoplexes) is significantly

608


higher than that observed for the GALA-lipoplexes (promoting endosomal disruption and DNA cytoplasmic delivery) under the same experimental conditions (compare Figs 1 and 3 for the H9 cells). Similar results were observed for the PM1 cells (data not shown). With TF228.1.16 cells, however, no significant differences were observed between the two types of complexes in terms of their ability to mediate transfection. The very low levels of transfection observed for this type of cells makes it difficult to draw definite conclusions regarding the rate limiting step. The results obtained with H9 and PM1 cells, in terms of the difference between transferrin- and GALA-lipoplexes, are in close agreement with those observed for adherent cell lines (Simões et al., 1998). Nevertheless, it cannot be ruled out that part of the transfection enhancing effect of transferrin may be due to its ability to mediate fusion between the lipoplex and endosome membranes at low pH, as suggested by our studies utilizing inhibitors of endosome acidification (Simões et al., 1999c). Still regarding this issue, it should be noted that our conclusions on the rate-limiting step were based on the final outcome (which is gene expression) and therefore, a more accurate and definite conclusion can only be drawn upon performance of systematic studies on the kinetics of intracellular delivery of DNA mediated by the different lipoplexes. The molecular and cellular mechanisms of gene delivery by the ternary complexes are currently being investigated in our laboratories. Fluorescence studies with adherent cell lines have shown that transferrin lipoplexes are taken up to much higher levels than plain lipoplexes, suggesting that a correlation between the amount of delivered DNA and transfection activity can be established (P. Pires, S. Simões, B. Plowman, N. Dügünes, M. C. Pedroso de Lima, Abstract, NATO Advanced Studies Institute on Targeting of Drugs: Strategies for Gene Constructs and Delivery, Marathon, Greece, June 24-July 5, 1999). Studies on the competitive inhibition of uptake of transferrin-lipoplexes by free transferrin, and on the use of inhibitors of endocytosis and lysosomotropic agents, have suggested that the transferrin-lipoplexes are taken up most likely via non-specific receptor-mediated endocytosis in adherent cells, and that transferrin may act as a pH-activated fusogen (Simões et al., 1999c). Future studies will address whether similar processes take place in lymphocytes. The combination of the strategies described in this work with others involving the use of adeno-associated virus or Epstein-Barr virus plasmids aiming at sustained gene expression (Philip et al., 1994, Saeki et al., 1998) may result in promising alternatives to viral vectors for transfection of lymphocytes or stem-cells. Nevertheless, safety issues associated with the use of lipoplexes, such as the recently described problem of lymphocyte apoptosis induced by some transfection reagents (Ebert et al., 1997), still need to be addressed carefully. ACKNOWLEDGEMENTS The work was supported by the Univeristy of the Pacific School of Dentistry, the National Institutes of Health (AI 35231), PRAXIS XXI (PRAXIS/PCNA/P/ BIO/45/96) Portugal and Grant BIO4-CT97-2191 from the European Union. We thank Z. L. Jonak and E. Henri (SmithKline Beecham) for the gift of TF228.1.16 cells, and P. L. Feigner (Vical) for the pCMVluc plasmid.


609

REFERENCES Baltimore, D. (1988) Nature 335:395-396. Blaese, R. M. et al. (1995) Science 270:475-480. Brock, J. H., and Mainou-Fowler, T. (1983) Immunol. Today 4:347-351. Buschle, M. et al. (1995) Hum. Gene Ther. 6:753-761. Cheng, P. (1996) Hum. Gene Ther. 7:275-282. Clark, P. R. and Hersh, E. M. (1999) Curr. Opin. Mol. Therapeut. 1:158-176. Ebert, O. et al. (1997) Gene Ther. 4:296-302. Feigner, P. L. et al. (1997) Hum. Gene Ther. 8:511-512. Haensler, J., and Szoka, F. C. Jr. (1993) Bioconj. Chem. 4:372-379. Harrison, G. S., Long, C. J., Curiel, T. J., Maxwell, F., and Maxwell, L. H. (1992) Hum. Gene Ther. 3:461-469. Hege, K. M., and Roberts, M. R. (1996) Curr. Opin. Biotech. 7:629-634. Heslop, H. E. et al. (1996) Nature Med. 2:551-555. Hwu, P. et al. (1993) J. Immunol. 150:4104-4115. Jonak, Z. L., et al. (1993) AIDS Res. Hum. Retroviruses 9:23-32. Konopka, K., Düzgünes, N., Rossi, J., and Lee, N. S. (1998) J. Drug Targeting 5:247-259. Konopka, K., Harrison, G. S., Felgner, P. L., and Düzgünes, N. (1997) Biochim. Biophys. Acta 1356:185197. Konopka, K., Pretzer, E., Felgner, P. L., and Düzgünes, N. (1996) Biochim. Biophys. Acta 1312:186196. Lasic, D. D., and Templeton, N. S. (1996) Adv. Drug Deliv. Rev. 20:221-266. Lisziewicz, J., Sun, D., Lisziewicz, A., and Gallo, R. C. (1995) Gene Ther. 2:218-222. Lusso, P. et al. (1995) J. Virol. 69:3712-3720. Mann, D. L. et al. (1989) AIDS Res. Hum. Retroviruses 5:253-255. Parente, R. A., Nir, S., and Szoka, F. C. Jr. (1988) J. Biol. Chem. 263:4724-4730. Parente, R. A., Nir, S., and Szoka, F. C. Jr. (1990) Biochemistry 29:8720-8728. Philip, R. et al. (1994) Mol. Biol. Cell. 14:2411-2418. Plank, C., Obserhauser, B., Mechtler, K., Koch, K., and Wagner, E. (1994) J. Biol. Chem. 269:1291812924. Saeki, Y., Wataya-Kaneda, M., Tanaka, K., and Kaneda, Y. (1998) Gene Ther. 5:1031-1037. Simões, S., Slepushkin, V., Pretzer, E., Gaspar, R., Pedroso de Lima, M. C., and Düzgünes, N. (1997) FASEB J. 11:A1090. Simões, S., Slepushkin, V., Gaspar, R., Pedroso de Lima, M. C., and Düzgünes, N. (1998) Gene Ther. 5:955-964. Simões, S. et al. (1999a) J. Leukocyte Biol. 65:270-279. Simões, S., Pires, P., Düzgünes, N. and Pedroso de Lima, M. C. (1999b) Curr. Opin. Mol. Therapeut. 1:147-157. Simões, S., Slepushkin, V., Caspar, R., Pedroso de Lima, M. C. and Düzgünes, N. (1999c) Gene Ther. (in press). Singhal, A. and Huang, L. (1994) In: Gene Therapeutics: Methods and Applications of Direct Gener Transfer (J. A. Wolf, ed.) Birkhäuser, Boston, pp. 118-142. Subbarao, N. K., Parente, R. A., Szoka, F. C., Nadasdi, L., and Pongracz, K. (1987) Biochemistry 26:2964-2972. Tran, A.-C., Zhang, D., Byrn, R., and Roberts, M. R. (1994) J. Immunol. 155:1000-1009. Woffendin, C. et al. (1994) Proc. Natl. Acad. Sci. USA 91:11581-11585. Yu, M., Poeschla, E., and Wong-Staal, F. (1994) Gene Ther. 1:13-26. Zhou, C., Bahner, I. C., Larson, G. P., Zaia, J. A., Rossi, J. J., and Kohn, D. B. (1994) Gene 149: 33-39.