J Neurosurg80:510-514, 1994 Liposomal transfection of human T-interferon gene into human glioma cells and adoptive immunotherapy using lymphokine-activated killer cells MASAAKI MIZUNO, M.D., JUN YOSHIDA, M.D., TORU TAKAOKA, M.D., AND KENICH[RO SUGITA, M.D. Departments of Neurosurgery, Nagoya University School of Medicine and Chukyo Hospital, Nagoya, Japan ~,, The authors evaluated the effect of liposomal transfection of human T-interferon (HuIFN-3,) gene into human glioma cells and lymphokine-activated killer (LAK) cells, alone and in combination. An HuIFN-y gene inserted in a eukaryotic expression vector was entrapped in liposomes bearing positive surface charges. Liposomal gene transfection induced production of HulFN-'y and its secretion in culture medium of human glioma cell lines (SK-MG-1 and U-251 MG). At 4 days after transfection, the cells produced 10 to 513 U/ml of HulFN-7 in the medium, whereby the major histocompatibility complex (MHC) class I and II antigens, as well as intercellular adhesion molecule-1 (ICAM-1), were induced on the glioma cell surface. The growth-inhibiting effect of transfection-induced HuIFN-3, was much stronger in comparison with control cultures exposed to 500 U/ml of exogenously added HuIFN-7. In addition, 20% to 40% growth inhibition was obtained in the glioma cells when they were treated with LAK cells alone at a 5:1 ratio of effector to target cells. Liposomal transfection of HulFN-y gene into human glioma cells combined with immunotherapy using LAK cells was more effective than either technique alone. The reinforcement of growth inhibition in the case of combined therapy was quenched by anti-icam-1 monoclonal antibody, but not by anti-mhc class I or II monoclonal antibodies. These results suggest that the combined effect of liposomal transfection of HulFN- 7 gene plus LAK cells into human glioma cells is a potentially useful therapy for malignant glioma, and that the mechanisms of the reinforcement of growth inhibition are closely related to the expression of ICAM-1 on the glioma cell surface. KEy WORDS 9 liposomal gene transfection gamma-interferon 9 iymphokine-activated killer cell 9 intercellular adhesion molecule 9 glioma p ATIENTS with malignant glioma have a poor prognosis despite the combined use of surgery, irradiation, chemotherapy, and a variety of immunotherapies. These patients are known to show decreased circulatory immune function and suppression of the tumor-specific immune response. In order to increase host immune function and inhibit tumor growth, many kinds of cytokines, such as interferons, tumor necrosis factors, and interleukins, have been used. In particular, intravenously or intrathecally administered human /3-interferon (HulFN-/3) has been given to patients with malignant glioma since 1980 and its usefulness has been confirmed. 3,1~ Like HulFN-/3, human ~/-interferon (HulFN-7) also has a variety of biological responses in both host and tumor cells. In brain tumors, HulFN- 7 has been reported to inhibit in vitro growth of glioma cell lines, 15 to activate the specific cytotoxicity of lymphocytes in the presence of interleukin-2 (IL-2), 1 and to increase the major histocompatibility complex (MHC) surface antigens and some kinds of cell adhesion molecules. 8,9 The MHC antigens and cell adhesion molecules are essential in the control of immune responses functioning in the recognition of antigens by T lymphocytes, u Alterations of the expression of MHC antigens or cell adhesion molecules may result in altered biological behaviors. On the other hand, lymphokine-activated killer (LAK) cells generated by culturing peripheral blood lymphocytes with IL-2 can lyse a wide variety of tumor cells. Some investigators have reported that IFN- 7 potentiated LAK cell cytotoxicity, 2,v,21 but this remains controversial. 24,25 In the present experiments, we studied the growth-inhibiting effect of in vitro transfection of glioma cells with HulFN-3, gene entrapped in the liposomes and also evaluated the usefulness of combining liposomal transfection and adoptive immunotherapy using LAK cells. 510 J. Neurosurg. / Volume 80 / March, 1994
HulFN-~/and LAK cells in human glioma Cell Lines Materials and Methods Cells of the SK-MG-1 and U-251 MG human glioma cell lines were used in this study. The cells were maintained in Eagle's minimum essential medium supplemented with 10% fetal calf serum, 2 mm nonessential amino acids, 5 mm L-glutamine, and antibiotic agents (streptomycin, 100 ~g,/ml, and penicillin, 100 U/ml). Preparation of Liposome-Entrapped Plasmids Liposome-entrapped plasmids were prepared by an improved procedure s of the reverse-phase evaporation method, 22 as described in our previous papersj 2,1s The liposomes were prepared using the positivelycharged lipid N-(ot-trimethylammonio-acetyl)-didodecyl-D-glutamate chloride (TMAG), dilauroyl phosphatidylcholine (DLPC), and dioleoyl phosphatidylethanolamine (DOPE), and were composed in a molar ratio of 1:2:2 as TMAG:DLPC:DOPE. An HulFN-3~ gene inserted into an SV40-derived expression vector (psvifn-"/) was used for the plasmid. This was constructed from HindlII-BgllI, large fragments of psv dihydrofolate reductase (psvdhfr), HulFN-2/complementary deoxyribonucleic acid (cdna), and deoxyoligonucleotides.* Gene Transfection Two milliliters of SK-MG-1 or U-251 MG cell suspensions was placed in 7.5 104 cells/ml culture medium in each well of Falcon No. 3042 plates and incubated at 37~ for 24 hours in a humidified atmosphere of 5% CO 2 and 95% air. After the liposomes or exogenous HulFN-~/t (500 U/ml, titer of 1 X 106 U/ml and specific activity of 3.0 x 10-7 U/rag of protein) were added to the medium, the solution was incubated at 37~ for 16 hours. Then the medium containing the liposomes was replaced with fresh medium containing no liposomes while the medium containing exogenous HulFN-3, was not replaced, and the incubation was continued for up to 80 hours. The culture medium was collected at 2, 4, 6, and 9 days after mixing with liposomes, and the level of HulFN-'y in the medium was measured by radioimmunoassay. A viable cell count was performed at the same time. Detection of MHC Antigens and Intercellular Adhesion Molecule-1 Expression of MHC antigens was detected by immunohistochemical study. Cells grown on coverslips were fixed in cold acetone. Mouse anti-mhc class I and II (human leukocyte antigen (HLA)-DR) mono- clonal antibodies were applied first for 1 hour, followed by a 30-minute incubation with anti-mouse immunoglobulin (Ig)G monoclonal antibody~ conjugated with horseradish peroxidase. The enzymatic activity was revealed using 0.015 % H202 and 0.4% diaminobenzidine as substrate. Expression of MHC antigens in the glioma cells was studied at 2, 4, 6, and 9 days after adding exogenous HuIFN- 7 or liposome-entrapped psvifn-~/. Expression of intercellular adhesion molecule-1 (ICAM-1) was measured using fluorescence-activated cell sorter analysis. The cultured cells (1 X l0 s) were harvested and resuspended in 25 ~1 of a 1:100 dilution of anti-icam-1 monoclonal antibody in the complete medium containing 10% fetal calf serum. The cells were incubated for 30 minutes on ice, washed with phosphate-buffered saline (PBS), then incubated for 30 minutes on ice with a 1:20 dilution of fluorescein isothiocyanate-labeled goat anti-mouse IgG monoclonal antibody.w Thereafter, the cells were washed with PBS three times and resuspended in 0.5 ml PBS. Fluorescence was quantitated using the EPICS profile software.ii Generation of LAK Cells Peripheral blood lymphocytes were obtained from healthy allogeneic donors. Heparinized peripheral blood was diluted 1:1 with PBS, and lymphocytes were separated by Ficoll-Paque gradient centrtfugation. The cells collected at the gradient interface, designated as peripheral blood lymphocytes, were washed three times with PBS, then resuspended in complete medium consisting of RPMI 1640 containing 10% fetal calf serum, 2 mm L-glutamine, antibiotic agents (streptomycin, 100/xg/ml, and penicillin, 100 U/ml), and IL-2 (10 U/ml). The peripheral blood lymphocytes were activated to generate LAK cells by Day 5 of incubation. Combined Application of HulFN- 3~ Gene Transfection and LAK Cells At first, we applied either exogenous HulFN-~/(500 U/ml) or liposome-entrapped psvifn-y (15 nmol/ml of lipids; 0.6 /xg/ml of DNA) to each well. For the former treatment, glioma cells were incubated for 4 days without medium changes; for the latter, the cells were incubated with liposomes for 16 hours and the medium was then replaced with fresh medium containing no liposomes, after which the incubation was continued for up to 80 hours. The culture medium in both treatments was collected 4 days after adding reagents, and the level of HulFN-',/in the medium was measured. At the same time, the cells were counted and LAK cells were applied at a 5:1 ratio of effector to target cells. After an additional 2 days of incubation, the results of combined application were evaluated. * TMAG obtained from Sogo Pharmaceutical Co., Ltd., Tokyo, Japan; DLPC supplied by Sigma Chemical Co., St. Louis, Missouri; DOPE obtained from Avanti Polar Lipids, Inc., Pelham, Alabama; psvifn-~, plasmid constructed by Toray Industries, Inc., Tokyo, Japan. t Natural HulFN-'y obtained from Ohtsuka Pharmaceutical Co., Ltd., Tokyo, Japan. :~ Mouse anti-mhc and anti-mouse IgG monocional antibodies obtained from Dakopatts, Glostrup, Denmark. w Anti-ICAM-1 monoclonal antibody obtained from British Biotechnology Products, lnc., Oxford, England; goat anti-mouse IgG monoclonal antibody supplied by Medical and Biological Laboratories Co., Ltd., Nagoya, Japan. II EPICS software obtained from Coulter Corp., Hialeah, Florida. J. Neurosurg. / Volume 80 / March, 1994 511
M. Mizuno, et al. TABLE 1 Production of HulFN-3J in cultured glioma cells transfected with psvifn-7* Agent untreated (control) cells empty liposomes Human Glioma Cell Line SK-MG-1 liposome-entrapped psv1fn-'y 41.3 _+ 5.5 U/ml U-251 MG 19.3 -+ 6.1 U/ml * Data are taken from six experiments and expressed as mean -+ standard deviation. HulFN-3, = human "y-interferon; psvifn- 7 = SV40- derived expression vector with HulFN- 7 gene insertion. Growth Inhibition Inhibition of gene transfection and/or LAK cell growth was evaluated by determining the number of viable cells, expressed as the number of trypan blueexcluding cells counted in a hemocytometer. The inhibition (% cytotoxicity) of gene transfection and/or LAK cell growth was determined with the following formula: number of cells in target well % cytotoxicity = 1 number of cells in control well X 100. Results Production of HulFN-7 in Glioma Cells Transfeeted With psvifn- 7 When glioma cells were transfected with psvifn- 7 by means of liposomes and incubated for 4 days, HulFN-7 was detected via radioimmunoassay in the media of both SK-MG-1 and U-251 MG cells. The HulFN-7 was detected at 2 days after transfection and reached a maximum after 4 days. Thereafter, it gradually decreased and was at all by 9 days after transfection. The maximum level of HulFN-7 in the medium was 41.3 5.5 U/ml (mean +_ standard deviation) in SK-MG-1 cells and 19.3 _+ 6.1 U/ml in U-251 MG cells (Table 1). On the other hand, HuIFN-',/ was in the culture medium in either SK- MG-1 or U-251 MG cells when empty liposomes were added to the medium. Expression of MHC Antigens or ICAM-1 on Glioma Cells The MHC class I and II (HLA-DR) antigens were negative at the first passage of culture, but could be induced to express MHC antigens for 2 to 6 days after adding more than 50 U/ml exogenous HuIFN-7. This was confirmed by immunohistochemical staining. Similarly, the antigens could be detected on glioma cell surfaces during the same time periods when liposomeentrapped psvifn-3, was added to the medium, even though the level of HuIFN- 7 in the medium was less than 50 U/ml. Expression of ICAM-1 was detected by fluorescence-activated cell sorter analysis. It was not expressed in either SK-MG-1 or U-251 MG cells when the cells were untreated. However, it was detected on glioma cell surfaces when liposome-entrapped psvifn- 7 was added to the medium and the cells were incubated for 4 days (Fig. 1). Growth Inhibition Empty liposomes did not significantly suppress the growth of either SK-MG-1 or U-251 MG cells; however, liposome-entrapped psvifn- 7 suppressed cell growth remarkably. Compared with the growth-inhibiting effect of exogenous HulFN-7 (500 U/ml), the effect of gene transfection was obviously stronger and was cytocidal (Fig. 2), although the level of HulFN-3, in the culture medium was very low (Table 1). The glioma cells were cultured for 4 days after adding exogenous HulFN-'y or transfection of its gene. Four days after the reagents were added, the culture medium was collected and the number of viable cells counted. Thereafter, LAK cells were applied at an effector-to-target cell ratio of 5:1. After an additional 2 days of incubation, the growth-inhibiting effect was evaluated by determining the number of viable cells. As shown in Fig. 3, glioma cells transfected with psvifn-7 were suppressed by LAK cells more efficiently than those incubated with exogenously added HulFN-% The combined therapy of HulFN-7 gene transfection followed by adoptive immunotherapy using LAK cells resulted in a synergistic growth-inhibiting effect in both SK-MG-1 and U-251 MG cells. The reinforcement of growth inhibition in the case of the combined therapy was partially quenched by anti- ICAM-1 monoclonal antibody (Fig. 4), but not by anti- MHC class I or class II monoclonal antibody (data not shown). Discussion Recently, a number of biological response modifiers have emerged as potentially useful agents in patients with malignant ghoma. In 1980, intravenous or intrathecal administration of IFN's was introduced for the treatment of malignant glioma following both in vitro and in vivo demonstrations of its efficacy. 3,1~ Interferons have a definite direct growth-inhibiting effect on glioma cells and modulate the expression of the cell surface antigens, including MHC antigens and cell adhesion molecules, thus providing a potential regulatory mechanism for local immune reactivity. 16,17,23 Gammainterferon in particular has very strong immunoreactivities. In this study, we investigated the alteration of cell surface antigens and the reinforced effect of cytotoxicity of LAK cells when HulFN-'/gene was transfected into glioma cells, which then continuously produced HulFN-7. As a result, HulFN-',/produced in the cells by its gene transfection induced MHC antigens and ICAM-1 on the glioma cell surface. Antigens of MHC class I and II are essential in the control of immune response. Although LAK cells are cytotoxic to a wide variety of tumor cells (so-called "non-mhc-restricted cytotoxicity"), some reports showed that target susceptibility to natural killer and LAK cytotoxiclty was inversely correlated with the target expression of MHC antigens. 4,6,2~ The relationship of HulFN-',/, MHC antigens, and LAK cells is very complicated and difficult 512 J. Neurosurg. / Volume 80 / March, 1994
HulFN- 7 and LAK cells in human glioma Fl6. 1. Graphs showing expression of intercellular adhesion molecule-1 (ICAM-1) on human glioma cell surfaces. Expression of ICAM-1 was detected at 4 days after transfection by fluorescence-activated cell sorter analysis. AU = arbitrary units. A: Untreated control cells. B: Cells treated with liposome-entrapped SV40- derived expression vector with human ",/-interferon gene insertion (psvifn-30. lo trace. Our data indicated that induction of MHC antigens on the glioma cell surface was not related to the cytotoxicity of LAK ceils, because the reinforcement of LAK cytotoxicity that was induced by HulFN-~," gene transfection into the glioma cells was not neutralized by anti-mhc class I or class II monoclonal antibodies. On the other hand, ICAM-1, which belongs in the immunoglobulin superfamily, seems to be an important agent because the reinforcement of LAK cytotoxicity was partially canceled by anti-icam-1 monoclonal antibody. The adoptive immunotherapy reported by Rosenberg and colleagues 1s,19 was effective in patients with advanced cancer, but several studies have reported that adoptive immunotherapy is not satisfactory because LAK cells may become trapped in the reticular formation and thus cannot reach the target cells. However, in our patients with malignant glioma, it has been possible to administer liposomes and/or LAK cells into the tumor cavity repeatedly through an Ommaya reservoir placed at surgery. The procedure seems to ensure longterm expression of the HulFN-'y gene product in the patient. Moreover, we are studying the reinforcement of LAK cytotoxicity with bifunctional antibodies, which are composed of anti-cd3 monoclonal antibody Ft6. 2. Graphs showing growth inhibition of human glioma cells by exogenous human ~/-interferon (HulFN-y) or transfection of its gone. Open circles = untreated control cells; closed circles = empty liposomes (15 nmol/ml lipids); triangles = exogenous HulFN-y (500 U/ml); and squares = liposome-entrapped SV40-derived expression vector with HulFN- 7 gene insertion (psvifn-~/) (15 nmol/ml lipids, 0.6 /~g/ml DNA). Each symbol represents the mean of six experiments. In both SK-MG-1 and U-251 MG cells, the growthinhibiting effect of HulFN- 7 gene transfection was much stronger than that of 500 U/ml exogenous HulFN-',/, to the extent of being cytocidal. FIc. 3. Graphs showing cytotoxic activity of lymphokineactivated killer (LAK) cells against human gliema cells: A = LAK cells alone; B = exogenous human y-interferon (HulFN-% 500 U/ml) plus LAK cells; C = HulFN-'y gone transfection plus LAK cells. The growth-inhibiting effect was enhanced when LAK cells were applied 4 days after the addition of 500 U/ml exogenous HulFN-'y or transfection of its gene. The effect was synergistic in the case of HulFN- 7 gene transfection. Data are presented as the mean -4- standard deviation of six experiments. J. Neurosurg. / Volume80 / March, 1994 513
M. Mizuno, et al. FIG. 4. Graphs showing the effect of anti-intercellular adhesion molecule-1 (ICAM-1) monoclonal antibody on lymphokine-activated killer (LAK) cell cytotoxicity against human glioma cells: A = LAK cells alone; B = exogenous human ",/-interferon (HuIFN-y, 500 U/ml) plus LAK cells; C = HulFN- 7 gene transfection plus LAK cells. Shaded columns represent no anti-lcam-1 monoclonal antibody; open columns represent the application of anti-icam-1 monoclonal antibody (5/xg/ml). Data are presented as the mean -+ standard deviation of six experiments. Statistical significance: * = p < 0.01; ** = p < 0.001. and anti-g-22 monoclonal antibody 13 that reacts specifically with a surface antigen (G-22) of human glioma. We have confirmed an increase in affinity of gene-transfected glioma cells and LAK cells using the bifunctional antibodies; details will be reported at a later date. These results suggest that combining liposomal transfection of HulFN- 7 gone into human glioma cells with the addition of LAK cells is a potentially useful therapy for malignant glioma. References 1. Bogdahn U, Fleischer B, Rupniak HTR, et al: 7-Interferon in specific T-cell mediated cytotoxicity in human gliomas. Antiviral Res 4:84, 1984 (Abstract) 2. Brenda MJ, Tarnowski D, Davatelis V: Interaction of recombinant interferons with recombinant interleukin-2: differential effects on natural killer cell activity and interleukin-2-activated killer ceils. Int J Cancer 37: 787-793, 1986 3. Cook AW, Carter WA, Nidzgorski F, et ah Human brain tumor-derived cell lines: growth rate reduced by human fibroblast inte.r.feron. Science 219:881-883, 1983 4. Gidlund M, Orn A, Pattengale PK, el al: Natural killer cells kill tumour cells at a given stage of differentiation. Nature 292:848450, 1981 5. Haga N, Yagi K: An improved method for entrapment of plasmids in liposomes. J Clin Biuehem Nutr 7:175-183, 1989 6. Harel-Bellan A, Quillet A, Marchiol C, et al: Natural killer susceptibility of human cells may be regulated by genes in the HLA region on chromosome 6. Proc Natl Acad Sci USA 83:5688-5692, 1986 7. Itoh K, Shiiba K, Shimizu Y, et al: Generation of activated killer (AK) cells by recombinant interleukin 2 (rll 2) in collaboration with intefferon-'y (IFN-7). J lmmunol 134: 3124-3129, 1985 8. Kuppner MC, Hamou MF, de Tribolet N: Activation and adhesion molecule expression on lymphoid infiltrates in human glioblastomas. J Neuroimmunol 29:229-238, 1990 9. Kuppner MC, Van Meir E, Hamou ME et al: Cytokine regu- lation of intercellular adhesion molecule-1 (ICAM-1) expression on human glioblastoma cells. Clin Exp Immnnol 81:142-148, 1990 10. Lundbald D, Lundgren E: Block of a glioma cell line in S by interferon. Int J Cancer 27:74%754, 1981 11. Marlin SD, Springer TA: Purified intercellular adhesion molecule-i (ICAM-I) is a ligand for lymphocyte functionassociated antigen 1 (LFA-1). Cell 51:813-819, 1987 12. Mizuno M, Yoshida J, Sugita K, et ah Growth inhibition of glioma cells of different cell lines by human interferon-/3 roduced in the cells transfected with its gene by means of ~i posomes. J Clin Bioehem Nutr 9:73-77, 1990 13. Mizuno M, Yoshida J, Sugita K, et al: Growth inhibition of glioma cells transfected with the human j3-interferon gene by liposomes coupled with a monoclonal antibody. Cancer Res 50:7826-7829, 1990 14. Nagai M, Arai T: Clinical effect of interferon in malignant brain tumours. Neurosurg Rev 7:55-64, 1984 15. Otsuka S, Handa H, Yamashita J, et al: Single agent therapy of interferon for brain tumours: correlation between natural killer activity and clinical course. Acta Neuroehir 73: 13-23, 1984 16. Piguet V, Carrel S, Diserens AC, et al: Heterogeneity of the induction of HLA-DR expression by human immune interferon on glioma cell lines and their clones. J Natl Cancer last 76:223-228, 1986 17. Pulver M, Carrel S, Mach JP, et al: Cultured human fetal astrocytes can be induced by intefferon-'y to express HLA- DR. J Neuroimmunol 14:123-133, 1987 18. Resenberg SA: Adoptive immunotherapy of cancer: accomplishments and prospects. Cancer Treat Rep 68:233-255, 1984 19. Rosenberg SA, Lotze MT, Muul LM, et al: Observations on the systemic administration of autologous lymphokineactivated killer cells and recombinant interleukin-2 to patients with metastatic cancer. N Engl J Med 313: 1485-1492, 1985 20. Storkus WJ, Howell DN, Salter RD, et al: NK susceptibility varies inversely with target cell class I HLA antigen expression. J Immunoi 138:1657-1659, 1987 21. Svedersky LP, Shepard HM, Spencer SA, et al: Augmentation of human natural cell-mediated cytotoxicity by recombinant human interleukin 2. J Immunal 133:714-718, 1984 22. Szoka F Jr, Papahadjopoulos D: Procedure for preparation of liposomes with large internal aqueous space and high capture by reverse-phase evaporation. Proc Natl Acad Set USA 75:4194-4198, 1978 23. Takiguchi M, Ting JPY, Buessow SC, etal: Response of glioma cells to interferon-gamma: increase in class II RNA, rotein and mixed lymphocyte reaction-stimulating ability. ur J Immunal 15:809-814, 1985 24. Taniguchi K, Petersson M, H6glund P, et al: Interferon 7 induces lung colonization by intravenously inoculated B16 melanoma cells in parallel with enhanced expression of class I major histocompatibility complex antigens. Proc Natl Aead Sci USA 84:3405-3409, 1987 25. Wallach D: Interferon-induced resistance to the killing by NK cells: a preferential effect of IFN-',/. Cell Immune/75: 390-395, 1983 26. Yoshida J, Kate K, WakabayashiT, etal: Antitumor activity of interferon-{3 against malignant glioma in combination with chemotherapeutic agent of nitrosourea (ACNU), in Cantell K, Schellekens H (eds): The Biology of the Interferon System. Boston: Martinus Nijhoff, 1986, pp 399-406 Manuscript received March 16, 1993. Accepted in final form July 28, 1993. Address reprint requests to: Jun Yoshida, M.D., Department of Neurosurgery, Nagoya University School of Medicine, 65 Tsurumai-cho, Showa-ku, Nagoya 466, Japan. 514 J. Neurosurg. / Volume 80 / March, 1994