Multicomponent gene therapy vaccines for lung ... - Semantic Scholar

Gene Therapy (1997) 4, 1361–1370  1997 Stockton Press All rights reserved 0969-7128/97 $12.00

Multicomponent gene therapy vaccines for lung cancer: effective eradication of established murine tumors in vivo with interleukin-7/herpes simplex thymidine kinase-transduced autologous tumor and ex vivo activated dendritic cells S Sharma1, PW Miller1, M Stolina1, L Zhu1, M Huang1, RW Paul2 and SM Dubinett1,3 1

Pulmonary Immunology Laboratory, Division of Pulmonary and Critical Care Medicine, Department of Medicine, 3 Jonsson Comprehensive Cancer Center, UCLA School of Medicine and the West Los Angeles VA Medical Center, Los Angeles, CA; and 2 Targeted Genetics, Seattle, WA, USA

Multiple antitumor modalities may be necessary to overcome lung tumor-mediated immunosuppression and effectively treat non-small cell lung cancer (NSCLC). To evaluate a multimodality gene therapy approach for control of local tumor growth, a weakly immunogenic murine alveolar cell carcinoma, L1C2, was transduced with either the interleukin-7/hygromycin-herpes simplex thymidine kinase (IL-7/HyHSVtk) internal ribosome entry site (IRES) retroviral vector or a vector containing the HyHSVtk, but not the IL-7 gene. Of the many cytokines available for gene transfer, IL-7 was chosen for these studies because it both stimulates CTL responses and down-regulates tumor production of the immunosuppressive peptide TGF-b. Following selection in hygromycin, IL-7 transduction was confirmed by ELISA. Clones produced 1.25 to 10 ng of IL7/ml/106 cells per 24 h. In vitro, genetically modified tumor cells were significantly more sensitive to ganciclovir (GCV)

than unmodified parental tumor cells. The in vivo growth of ex vivo modified L1C2 cells was evaluated. There was a dose–response relationship between the amount of IL-7 secreted in vitro and the growth of genetically modified murine tumor in vivo. Transduced tumor cells regressed in mice following GCV therapy. Although ex vivo gene modification of tumor cells led to complete resolution of the tumor following implantation in vivo, IL-7 and HSVtk gene modified tumor cells were not effective in treating established parental tumors. However when 5 × 105 bone marrow-derived, in vitro activated dendritic cells (DC) were administered in combination with transduced tumor and GCV, 5 day old established tumors were eradicated in 80% of mice. These studies suggest that multicomponent vaccines may facilitate improved host responses by replacing host immune deficits and thus could have a role in adjuvant therapy and local control of NSCLC.

Keywords: gene therapy; interleukin-7; dendritic cells; lung cancer

Introduction Lung cancer is the leading cause of cancer death in the USA.1 Several new treatment strategies are in development because overall 5-year survival is less than 15%.2–5 Lung cancer has been unresponsive to immunotherapy with systemic recombinant cytokine administration or adoptive transfer of activated lymphocytes.6,7 This may be due in part to local immunosuppression induced by lung tumors, but systemic administration of cytokines can only partially reverse lung tumor-mediated immunosuppression.8–10 Transfer of cytokine genes into tumor cells is a strategy for cytokine delivery that can augment tumor immunogenicity.11 However, lung tumor cells may Correspondence: SM Dubinett, Division of Pulmonary and Critical Care Medicine, Department of Medicine, UCLA School of Medicine, West Los Angeles VA, W111Q, 11301 Wilshire Boulevard, Los Angeles, CA 90073, USA Received 24 February 1997; accepted 11 July 1997

function poorly as the sole component of an antitumor vaccine because they have defects in antigen processing and presentation and produce immune inhibitory cytokines.12 An ideal antitumor vaccine would contain a source of immunopotentiating cytokine, activated professional antigen presenting cells (APC) and a source of tumor antigen. In the present study, the system developed to address tumor-mediated immune deficits includes autologous tumor cells transduced ex vivo with a polycistronic retroviral vector containing both IL-7 and HyHSVtk genes. IL-7 was utilized in the current study because it has the capacity to both stimulate CTL activity and to decrease the tumor’s production of the potent immunosuppressive cytokine TGF-b.13,14 In addition, we have found that IL-7 gene transfer in human NSCLC can augment tumor expression of MHC class I, ICAM-1, LFA-3 and p185neu, all of which led to a more immunogenic phenotype for the vaccine.15 Tumor-derived IL-7 also decreases lung tumor proliferation.15 Our studies indicate that IL-7 transduction up-regulates tumor cell

Multicomponent gene therapy vaccines for lung cancer S Sharma et al

1362

expression of FAS and this results in increased tumor cell sensitivity to apoptosis. 16 Thus, IL-7 has several immunepotentiating activities that may be of particular benefit in lung cancer therapy. The third component of the therapeutic system is coadministration of syngeneic bone marrow-derived, ex vivo activated dendritic cells. Dendritic cells, the most potent antigen-presenting cells, maintain a high-level expression of the adhesion and costimulatory molecules critical for T cell dependent immune responses.17 Hence, immunization with dendritic cells in combination with cytokine-transduced tumor cells has the advantage of supplying potent Ag presentation in the presence of cytokine-activated cytotoxic T lymphocytes (CTL). The central importance of functional APC in the immune response against cancer was well defined in recent studies by Huang et al.18 These investigations reveal that even highly immunogenic tumors require host APC for antigen presentation. Thus host APC, rather than tumor cells, present tumor antigen. These findings are consistent with the known defects in human lung cancer: lung cancer cells lack critical portions of the requisite antigen processing and presentation pathways functioning in professional APC.12 The rationale for using ex vivo activated dendritic cells in this setting is further supported by studies indicating that tumor cells produce peptides that interfere with dendritic cell function and maturation.19 The combination of these therapeutic components at the tumor site can lead to heightened specific immune responses by supplementing cytokine production and effective antigen presentation which have been found to be deficient in lung cancer.12 In this study we report that: (1) the IL-7/HyHSVtk IRES vector effectively and coordinately delivers the active cytokine and drug sensitivity genes to a murine alveolar carcinoma cell line; and (2) that gene-modified murine alveolar carcinoma cells used in combination with GCV and dendritic cells effectively eradicate 5-dayold established tumors in vivo while promoting long-term specific antitumor immunity.

Results Functional in vitro evaluation of transduced genes expressed by the bicistronic vector in L1C2 cells Bicistronic vectors have the advantage of coordinate expression of two genes (Figure 1). To determine if both the cytokine and drug sensitivity genes were functional in transduced cells, we assessed the biological activity of tumor-derived IL-7 and GCV sensitivity of tumor cells in vitro. Following retrovirus-mediated IL-7 transduction of L1C2 alveolar carcinoma cells and selection in hygromycin, IL-7 secreted in the medium was determined by murine IL-7-specific ELISA. L1C2-IL-7 clones produced 1.25–10 ng of IL-7/ml/106 cells per 24 h. To determine if tumor-derived IL-7 was biologically active syngeneic splenic lymphocytes were cultured in cell free supernatants from IL-7-transduced tumor cells. Compared to supernatants from HyHSVtk transduced or parental tumor cells, IL-7/HyHSVtk transduced tumor cell supernatants significantly enhanced lymphocyte proliferation (data not shown). As previously demonstrated in human lung tumors, IL-7/HyHSVtk-transduced tumor cells had a decreased

proliferative capacity in vitro (Figure 2a)compared to the HyHSVtk transduced or parental L1C2 cells. The bicistronic vector used in this study contains the HSVtk gene. Cells expressing HSVtk are sensitized to the toxic effects of GCV.20 The GCV sensitivity of the genetically modified cells was assessed by the MTT assay. In comparison to the parental L1C2 cells, both the HyHSVtk and IL-7/HyHSVtk-transduced cells were exquisitely sensitive to the toxic effects of GCV (Figure 2b). Following exposure to GCV (10 mm), HyHSVtk and IL7/HyHSVtk-transduced tumor cells showed a significant decrease in survival compared to parental L1C2 cells (P , 0.05). An additional measure of HSVtk function is the capacity of transduced cells to mediate the bystander effect. The bystander effect refers to the capacity of HSVtk-transduced cells to mediate killing of nontransduced cells which are in physical proximity.21 This may occur via the transfer of apoptotic bodies and/or the passage of phosphorylated GCV products via gap junctions.21–23 Consistent with a functional bystander effect, when mixed with HyHSVtk or IL-7/HyHSVtk transduced cells, parental L1C2 cells were sensitive to the toxic effects of GCV (50 mm) (Figure 2c). Parental L1C2 cells were more sensitive to the toxic effects of GCV as the percentage of transduced cells in the mixing ratio increased. As few as 10% transduced cells in the mixture had the capacity to sensitize parental cells to the toxic effects of GCV.

Dose responsive decrement in tumorigenicity of transduced cells in vivo: IL-7-HSVtk transduction decreases tumorigenicity and increases immunogenicity To determine the in vivo behavior of ex vivo transduced L1C2 tumor cells, transduced cells were inoculated s.c. in BALB/C mice. The low IL-7 (1.25 ng/ml/106 cells per 24 h) secreting L1C2 cells formed tumors and were lethal in all mice. There was a dose–response relationship between the amount of IL-7 secreted and the number of mice rejecting the tumor. Clones producing intermediate amounts of IL-7 (5 ng/ml/106 cells per 24 h) were rejected in 60% of mice, whereas high IL-7 (10 ng/ml/106 cells per 24 h) producing clones were rejected in all mice (Table 1). Subcutaneous inoculations of 5 × 104 parental L1C2 cells were tumorigenic in 100% of mice. Administration of GCV (2 mg i.p. twice daily for 4 days) to mice bearing parental L1C2 did not significantly affect in vivo tumor growth (Figure 3). In contrast, administration of GCV to mice bearing HyHSVtk or IL-7/HyHSVtk-modified tumors resulted in their complete regression. The effect

Figure 1 Diagram of the murine IL-7/Hytk vector. The vector backbone is derived from Moloney murine leukemia virus (MoMLV). The IRES sequence is from the encephalomyocarditis virus (EMCV). The hygromycin phosphotransferase gene (hph) is fused in-frame to the herpes simplex thymidine kinase gene. The fusion gene confers a negative and positive selectable phenotype. The control vector has the same structure but lacks the murine IL-7 and IRES sequence.


1363

Figure 2 (a) IL-7 transduction reduces L1C2 proliferation. The growth patterns of parental, control vector and IL-7-transduced L1C2 cells were determined by the MTT assay. Results are representative of three assays. (*P , 0.05). (b) Transduced L1C2 cells are sensitized to GCV. Parental and transduced L1C2 cells were incubated with 10 mm GCV and survival of the cells monitored by the MTT assay. (P , 0.05). Results are representative of three assays. (c) Transduced tumor cells demonstrate a functional bystander effect in mixing studies in vitro. Parental L1C2 cells, control vector (HyHSVtk) and IL-7 (IL-7/HyHSVtk)-transduced cells were either exposed to GCV alone or mixed in various ratios and sensitivity to GCV determined by the MTT assay after 96 h in culture. Mixing HyHSVtk or IL-7/HyHSVtk transduced cells with parental cells sensitized parental cells to the effects of GCV.

of GCV in mice bearing IL-7/HyHSVtk-transduced tumors was independent of the amount of IL-7 secreted by these tumors; both high and low IL-7 producing clones are rejected following GCV administration. To determine the in vivo efficacy of the bystander effect in this model, in vivo mixing experiments were performed. In the absence of GCV administration parental L1C2 cells mixed with HyHSVtk transduced cells were tumorigenic in all mice. However, when GCV was administered to mice inoculated with HyHSVtk cells mixed with parental L1C2 cells in a ratio of 10:90, 20% of the mice rejected the tumors. The percentage of mice rejecting the parental tumor cells on GCV administration increased as the percentage of HyHSVtk transduced cells increased in the mixture (80% of mice rejected tumors with a mixing ratio of 50:50 and all mice rejected tumors with a mixing ratio of 90:10). In the absence of GCV administration mixing the high IL-7 producing clone with parental cells in ratios of 50:50 and 90:10, lead to 40 and 60% rejection respectively. All mice rejected the tumor with a 50:50 ratio of IL-7/HyHSVtk transduced to parental tumor cell and GCV administration (Table 2). To determine the immunogenicity of the genetically

Table 1 IL-7-transduced tumor cells demonstrate a dose responsive decrement in tumorigenicity in vivo Cells

L1C2 L1C2 L1C2 L1C2 L1C2

No. complete responses out of 10

HSVtk IL-7 1.25 ng/ml/106 cells IL-7 5 ng/ml/10 6 cells IL-7 10 ng/ml/106 cells

0 0 0 6 10

105 Parental, HyHSVtk vector (L1C2-HSVtk) and IL-7/HyHSVtk transduced (L1C2 IL-7/HStk) cells were injected s.c. and tumor growth monitored. There is a dose–response relationship between the amount of IL-7 secreted and tumor rejection in vivo.

modified tumor cells, mice that had rejected the following tumors L1C2-IL-7/HyHSVtk, L1C2-IL-7/HyHSVtk + GCV, L1C2-HyHSVtk + GCV or irradiated parental L1C2 cells were rechallenged with 5 × 104 parental cells. Eighty per cent of mice that initially rejected the L1C2-IL-


1364

lished mice were treated with either GCV or diluent injections (Figure 4). Parental tumors grew in SCID mice and ganciclovir had no effect on the growth of the parental cells. HyHSVtk vector modified cells grew in SCID mice but on GCV administration, the growth of the HyHSVtk vector modified cells was significantly reduced and survival of the mice extended. The L1C2-IL-7/HyHSVtk cells grew more slowly than the parental L1C2 or L1C2HyHSVtk cells in the SCID mice and the survival of the mice bearing the L1C2-IL-7/HyHSVtk tumors was extended compared to control. Growth of L1C2-IL7/HyHSVtk tumors was further reduced in SCID mice on ganciclovir treatment and mice in this group had the longest survival. However, in contrast to immunocompetent BALB/C mice, there were no complete antitumor responses in SCID mice and all mice in each group eventually died due to progressive tumor growth.

Figure 3 Transduced tumor cells are rejected in vivo after administration of GCV. 105 L1C2 parental, control vector transduced and IL-7-transduced L1C2 cells were injected subcutaneously. Ten days following tumor inoculation, GCV was administered i.p. (2 mg per dose, twice daily for 4 days) and tumor volumes monitored three times per week. Inoculation of HyHSVtk-modified or IL-7/HyHSVtk-modified L1C2 cells with GCV administration results in complete regression of the genetically modified tumor.

Table 2 Transduced tumor cells demonstrate a functional bystander effect in mixing studies in vivo Mixing ratio transduced: parent cells 0:100 10:90 50:50 90:10 100:0

No. complete responses out of five CV alone CV + GCV

0 0 0 0 0

0 1 4 5 5

IL-7 alone

0 0 2 3 5

IL-7 + GCV

0 1 5 5 5

Combined gene therapy for the treatment of established tumors in vivo: dendritic cells enable complete eradication of established parental tumors when coadministered with genetically modified tumors To determine if this gene therapy approach was effective in the eradication of established tumor, 5-day established subcutaneous parental tumors were treated with direct intratumoral injection of gene modified L1C2 cells. IL-7, HSVtk/GCV or combination gene therapy with IL-7/ HSVtk/GCV failed to control tumor growth. We postulated that therapeutic failure in this setting could be due to decreased antigen-presenting capacity at the tumor site. To determine if this system of gene therapy could be enhanced by repletion of APC functions, in vitro activated syngeneic bone marrow DC were administered at the tumor site. The addition of activated DC to IL-7/ HSVtk/GCV led to an 80% complete response rate (Table 4). Splenocytes from mice with complete responses secreted significantly more IFNg and GM-CSF than did splenocytes from normal controls and tumor bearers.

Parental L1C2 tumors were rejected in vivo when mixed with control vector or IL-7 (10 ng/ml/106 cells per 24 h) transduced cells with GCV i.p. Parental L1C2 tumor cells were mixed with control vector (HyHSVtk) and IL-7 transduced (IL-7/HyHSVtk) cells in various ratios and 105 cells were inoculated s.c. Seven days following tumor inoculation, mice were treated with GCV (2 mg/ml per dose, twice daily for 4 days) or diluent.

7/HyHSVtk cells were able to reject parental tumor cells on rechallenge while 75% of mice from the L1C2 IL-7 + GCV group rejected parental tumors on rechallenge (Table 3). In contrast only 25% of mice from both the L1C2-HyHSVtk + GCV or the irradiated control groups were able to reject the rechallenge dose of parental L1C2 tumors.

Rejection of transduced tumors in vivo requires intact host immunity To determine the contribution of host immunity in the in vivo rejection of transduced L1C2 tumors SCID mice were inoculated s.c. with parental L1C2 cells, L1C2-HyHSVtk or L1C2-IL-7/HyHSVtk and after tumors were estab-

Figure 4 Rejection of genetically modified tumor cells requires an intact immune system. 105 L1C2 parental, HyHSVtk (control vector) or IL-7/ HyHSVtk-modified L1C2 tumor cells were inoculated s.c. in BALB/C SCID mice. A week following tumor inoculation, mice were treated with GCV (2 mg/ml per dose, twice daily for 4 days) or diluent and tumor volumes and survival were monitored.


Table 3 Immunogenicity is enhanced following rejection of genetically modified tumors Original treatment resulting in rejection of modified tumors

Complete responses following rechallenge (responses/total rechallenged)

L1C2-IL-7 L1C2-IL-7 + GCV L1C2-CV + GCV Irradiated control

8/10 9/12 3/12 3/12

Mice that had rejected genetically modified tumor cells were protected against a rechallenge with L1C2 tumor cells. Mice were rechallenged with 5 × 104 parental tumor cells (10 times the TD/LD50) and tumor growth and survival monitored.

Table 4 Dendritic cells co-administered with IL-7-transduced tumor cells and GCV leads to complete tumor eradication in vivo Groups

% Complete antitumor response

Parent (P) P + DC IL-7 + GCV IL-7 + DC DC + GCV IL-7 + DC + GCV

0 20 0 20 20 80

5 × 104 Parental L1C2 tumor cells were inoculated s.c. in BALB/C mice. Five days following tumor inoculation, IL-7transduced tumor cells and in vitro activated dendritic cells were injected at the site. One day later, GCV was administered intratumorally (2 mg per dose, twice daily for 4 days). IL-7-transduced tumors used in these studies produced 10 ng/ml/106 cells per 24 h of IL-7. (n = 6–12 mice). Mice with complete responses showed no evidence of tumor for at least 8 months following therapy.

Only those therapies which resulted in complete responses were associated with splenocyte production of GM-CSF and IFNg. The greatest production of GM-CSF and IFNg was obtained from splenocytes of mice receiving IL-7/HyHSVtk tumors + DC + GCV (Table 5).

Table 5 Cytokine production by spleen of mice following complete tumor eradication Groups Tumor P + DC IL-2 + GCV IL-7 + DC DC + GCV IL-7 + DC + GCV Control

GM-CSF pg/ml

gIFN pg/ml

0 450 ± 52 0 598 ± 30 227 ± 2 1078 ± 100 0

0 419 ± 15 0 306 ± 8 0 660 ± 7 0

Spleens were isolated from mice cured of their established tumors and 2 × 106/ml splenocytes cultured in CM. After 24 h in culture, IFNg and GM-CSF released from splenocytes were measured by cytokine specific ELISA. P , 0.05 for IL-7 + DC + GCV group versus the other treatment groups and control.

Discussion In this study we have shown the function and utility of a bicistronic vector that co-expresses murine IL-7 and HSVtk. After transduction of a murine alveolar carcinoma cell line, L1C2, both genes were functional in vitro and in vivo. As previously described, the IRES allows for coordinate expression of both genes.24 The ex vivo retroviral transduction of lung tumor cells for vaccine gene therapy in this system has the following advantages. First, the stable transduction of tumor cells with both cytokine and drug sensitivity genes allows the delivery of a precise amount of cytokine for a specific duration of therapy. This system would allow the clinician to deliver gene therapy with similar control of dosing and scheduling as currently employed for standard drug therapy regimens. The level of cytokine production can be controlled either by cloning tumor cells for cytokine production, as in the current study, or by increasing the number of cells delivered in the vaccine. The duration of cytokine delivery can be controlled by the drug sensitivity component of the vector. Thus a prolonged production of cytokine which may be needed to overcome tumor-mediated immunosuppression can be achieved. In the current system the administration of GCV also serves as a safety mechanism, in that IL-7 production from cells will cease following GCV delivery. The HSVtk system includes additional benefits in the multicomponent vaccine; the HSVtkmediated bystander effect amplifies the efficacy of the drug sensitivity system by causing death of nontransduced tumor cells in physical proximity to the transduced vaccine tumor cells. The bystander effect may be extremely potent and in certain instances may be responsible for tumor eradication.25,26 This feature would offer benefit in a clinical setting in which the control of local tumor growth was the goal of the therapeutic intervention (ie obstruction due to endobronchial malignancy). In addition, tumor cell death induced in this manner could lead to the release of tumor antigens, previously unrecognized by the host, thereby augmenting activation of specific host antitumor immunity and response against metastatic disease. Current studies in our laboratory address the potential for this form of therapy to treat metastatic disease. After finding that the IL-7/HyHSVtk vector was functionally active in the transduced murine lung tumor cell line, the tumorigenicity of the genetically modified cells was determined in syngeneic BALB/C mice. Parental nontransduced L1C2 cells inoculated subcutaneously were tumorigenic in all mice (LD50/TD50 = 53 –104 tumor cells). However, following IL-7 transduction of L1C2 cells, there was a dose responsive relationship between the amounts of IL-7 secreted and tumor rejection in vivo. This implies that there is a critical amount of IL-7 required for effective T cell activation at the tumor site for tumor cell rejection. IL-7 has previously been shown to have anti-tumor effects in vivo.27,28 We have previously found that IL-7-transduced fibrosarcoma tumors were heavily infiltrated with host T cells which appeared to be responsible for tumor regression.27 However, as indicated by our current results using SCID mice, IL-7 also has nonT cell-dependent antitumor activities. This could include induction of the release of other cytokines from macrophages such as TNFa as well as a decrease in tumor production of TGFb. 13,14,29 Our current and previous

1365


1366

studies also suggest that IL-7 directly decreases tumor proliferation.15 Transduced tumor cells maintained their GCV sensitivity in vivo as inoculation of HyHSVtk or IL-7/HyHSVtk modified L1C2 cells combined with GCV administration led to complete rejection of the genetically modified tumors. In fact, admixture of IL-7-producing tumor cells with parental nontransduced cells led to tumor reduction even without GCV administration. This could be due to one or more of several mechanisms: (1) a transduced tumor-mediated bystander effect;25 (2) an IL-7-mediated immune response;27,28 and/or (3) an IL-7-mediated antiproliferative response.15 The bystander effect was also operational in vivo and was increased when IL-7 was produced from admixed tumor cells. These findings suggest that high levels of IL-7 secreted at the tumor site by IL7-modified tumors is sufficient for tumor rejection when a fraction of the tumor mass is genetically modified. Long-term systemic immunity was generated in mice that had rejected IL-7/HyHSVtk-modified L1C2 cells or HyHSVtk and IL-7/HyHSVtk + GCV. Mice that had rejected genetically modified tumor cells were able to reject 10 times the TD/LD50 dose. The IL-7 component of the vector contributed predominantly to long-term systemic effects since vaccination with IL-7/HyHSVtk modified cells led to protection in 80% on rechallenge whereas vaccination with HyHSVtk modified cells plus GCV led to protection in only 25% of mice. To determine the importance of a cell-mediated immune component involved in the rejection of genetically modified tumors in immunocompetent BALB/C mice, the tumorigenicity of the genetically modified tumor cells was tested in SCID mice. IL-7/HyHSVtk-modified L1C2 cells grew slowly in SCID mice and prolonged survival was evident. Proliferation of the genetically modified cells in SCID mice was further reduced and survival of mice increased when the mice bearing genetically modified cells were treated with GCV. However, all mice eventually died secondary to tumor growth. This suggests that the complete rejection of IL-7-modified tumor cells requires intact host immunity but non-T cell-mediated pathways for inhibition of tumor growth are operative. In this study we found that effective in vivo therapy of established lung cancer required replacement of each of the immune components known to be deficient at the lung tumor site. The lung tumor environment is known to be a rich source of immune inhibitory cytokines10,30,31 and both lymphocytes and APC are known to be defective. Tumor-infiltrating lymphocyte activities are depressed and lung tumors have been found to have decreased MHC expression and antigen-processing defects.8–10,12,30–32 The use of both IL-7 and HSVtk in a tumor cell vaccine could have the following advantages. First, IL-7 activates the local host response and reverses tumor-mediated immunosuppression. After this activation has begun, administration of GCV causes lysis of transduced tumor cells and release of tumor antigens in the setting of IL-7-induced heightened antitumor immunity. By injecting the vaccine at the tumor site, the added benefit of the bystander effect can augment the local immune response, causing more tumor lysis and the potential of more antigen presentation by activated antigen-presenting cells co-administered at the tumor site. A variety of cytokines have been used in studies of gene therapy with tumor cells.33 It has not yet been

determined which cytokines are optimal for this type of therapy, especially given the possibility that each type of primary tumor may have a different immunoregulatory response to a given cytokine. We have chosen to focus on IL-7 for NSCLC for the following reasons. IL-7, originally described as a growth factor for pre-B cells,34 has been found to stimulate the growth of thymocytes and mature T cells.35 Resting T cells proliferate in direct response to IL-7, as well as through an IL-2-dependent pathway.36 In addition, IL-7 enhances the generation of cytotoxic T cells37 and LAK cells,38 and induces cytokine secretion and tumoricidal activity in human peripheral blood monocytes.29 The co-stimulatory molecule B7 can be induced on T cells by IL-7.39 Because activated T cells can coexpress both B7 and its counter-receptor, CD28, this suggests that IL-7-activated T cells may be capable of autocrine costimulation via the CD28 activation pathway.40 IL-7 synergizes with IL-12 in the induction of T cell proliferation, cytotoxicity, and IFN-g release.41 IL-7 may contribute to this synergy by up-regulating IL-12R expression. In agreement with these findings are recent studies indicating that IL-7 plays a role in cell-mediated immune responses characteristic of type 1 cytokines42 and enhances Ag-specific T cell cytotoxicity.43 In human NSCLC cells, we found that IL-7 transduction slowed the in vitro proliferation of tumor cells and augmented surface molecule expression of MHC class I and ICAM-1.15 This suggests that IL-7 may enhance the immunogenicity of NSCLC cells. In addition, allogeneic peripheral blood lymphocytes co-cultured with IL-7-transduced human tumor cells showed enhanced cytolytic and proliferative capacities compared to co-culture with parental cells.15 IL-7 can potentiate the growth and production of IFN-g from tumor-reactive TIL.44 This is important because IFNg production has been found to be requisite for TIL antitumor efficacy. When used in adoptive transfer, IL-7-generated CTL have been shown to reduce metastatic tumor as effectively as do CTL generated in IL-2.37 In accordance with these findings, IL-7 has been suggested both as an additional cytokine for the in vitro expansion of lymphocytes for adoptive transfer23,24 and as a systemic agent for cancer immunotherapy.38 We and others27,45 have found that IL-7 mediates significant antitumor responses in vivo in murine models.28,46 In addition, we have found that IL-7 down-regulates both macrophage47 and tumor13,14 production of TGF-b, a potent inhibitor of antitumor immunity. Although we focused on IL-7 in this study, other cytokines may also be efficacious when used in this multicomponent system. Because it directly stimulates APC activities and differentiation, we also evaluated GM-CSF as the cytokine component in preliminary studies. We found that IL-7 was significantly more effective than GM-CSF in this system (data not shown). Dendritic cells are characterized by high-level expression of costimulatory molecules and the potent capacity to present antigen and prime naive CTL in vitro and in vivo.17,48,49 Several recent studies have shown the successful use of peptide pulsed dendritic cells to treat established tumors or reduce metastases.50,51 This therapy utilizes antigen-pulsed dendritic cells with either previously identified tumor peptide antigens, acid-elutriated peptides or tumor RNA. 18,50–53 In this study we found that in vitro activated bone marrow-derived DC had the capacity to facilitate the eradication of an established, weakly immunogenic murine alveolar cell carcinoma


when combined with IL-7/HyHSVtk-transduced autologous tumor. Immunization with dendritic cells in the setting of cytokine-transduced tumor cells has the advantage of restoring potent tumor-antigen presentation which cannot be mediated by the tumor alone.18 The presence of all these components at the tumor site can lead to heightened specific immune responses. Accordingly we found that the successful treatment of established tumors with this multicomponent therapy was accompanied by resistance to tumor rechallenge and an enhanced release of splenocyte IFN-g and GM-CSF. The release of these cytokines in response to autologous tumor has been found to be indicative of specific T cellmediated antitumor responses.54,55 Thus this therapy resulted in complete antitumor responses without the use or knowledge of specific tumor antigens. This may be important because, although tumor antigens have been identified in human lung cancer, they are present in a minority of tumors and are recognized only in association with certain specific HLA haplotypes.56,57 In summary, we have developed a model system to evaluate genetic modulation of the autologous tumor to serve as an antigen source and immune enhancer for DC. Gene-modified tumors may enhance the induction of antitumor immunity by augmenting antigen presentation and CTL activation. Thus, the genetic manipulation of the autologous tumor can act to enhance the generation of antitumor immunity and serve as the source of antigen for DC. In contrast to conventional immunization with purified peptide Ag, autologous tumor has the capacity to provide the activated DC access to the entire repertoire of available antigens. This could increase the likelihood of a response and reduce the potential for tumor resistance due to phenotypic modulation. The use of gene therapy or cytokine immunotherapy to control local tumor growth in non-small cell lung cancer has been the focus of recent clinical trials and preclinical studies have shown the efficacy of several gene therapy combinations.30,58–62 Utilization of multicomponent vaccine therapies may allow control of local tumor growth and promote specific systemic antitumor responses in lung cancer.

Materials and methods Mice Pathogen-free female BALB/C mice (H-2d) (8–12 weeks of age) were obtained from NIH and Simenson Laboratories (Gilroy, CA, USA). Mice were maintained in the West LA VA Animal Research Facility. BALB/C SCID mice were obtained from UCLA and maintained in cages housed in laminar flow hoods under pathogen-free conditions. Cell culture Line 1 alveolar cell carcinoma (L1C2) was used to establish the in vivo model. This is a weakly immunogenic, natural killer cell-resistant, and highly malignant tumor with minimal metastatic potential, which arose spontaneously in a female BALB/C mouse.63 The cells were cultured in RPMI (Irvine Scientific, Santa Ana, CA, USA) supplemented with 10% fetal calf serum (Gemini Bioproducts, Calabasas, CA, USA), antibiotics (penicillin and streptomycin) (complete medium (CM) ) and kept at 37°C in a humidified atmosphere containing 5% CO2.

Retroviral construct and transduction of tumor cell lines The IL-7 retroviral vector backbone is derived from Moloney murine leukemia virus (MoMLV) (Figure 1).64,65 Both the murine IL-7 cDNA and the hygromycin-herpes simplex thymidine kinase fusion genes are driven by the retrovirus LTR and there is an internal ribosomal entry site in the vector construct that allows translation of multiple reading frames from a polycistronic message.23 The control vector construct contains the hygromycin gene selectable marker and the HSVtk gene but lacks the murine IL-7 cDNA.15 Transduction L1C2 cell line was maintained in exponential growth phase in 25-cm2 T flasks and transduction was performed using retroviral supernatants at a multiplicity of infection of 4:1 in the presence of protamine sulfate (10 mg/ml, Sigma, St Louis, MO, USA). After 24 h, the medium was replaced with CM and the cells were incubated for an additional 24 h. The cells were then transferred to a 75cm2 flask and grown to confluency. Supernatants were collected from the bulk transduced cells and IL-7 protein was measured by cytokine-specific ELISA. After bulk transduction of L1C2 cells with the IL-7 retroviral vector the cells produced 300–500 pg/ml/106 cells per 24 h. The IL-7/HyHSVtk and HyHSVtk transduced cells were selected in 800 mg/ml of hygromycin for 10 days. After selection in hygromycin the bulk selected IL-7-transduced cells produced 5 ng/ml/106 cells per 24 h. The IL-7-transduced cells were cloned for IL-7 production by limiting dilution and several clones were isolated that produced 1 to 10 ng of IL-7/ml/106 cells per 24 h. Isolation and in vitro propagation of dendritic cells Lymphocyte-depleted mouse bone marrow cells were cultured with murine GM-CSF (2 ng/ml) and IL-4 (20 ng/ml) (R&D, Minneapolis, MN, USA) for 8 days as previously reported.50,52 Medium was replenished every other day. On day 8, nonadherent dendritic cells were isolated by pipetting. The dendritic cells were counted, washed twice in PBS and 5 × 105 cells injected at the tumor site. Consistent with previous studies, dendritic cells characterized by flow cytometry were found to have high-level expression of: B7–1, B7–2, CD11c and MHC II and to be .90% dendritic cells (data not shown). IL-7, IFNg and GM-CSF ELISA IL-7 protein concentrations from transduced supernatants were determined by IL-7-specific ELISA as previously described.15 Briefly, 96-well Costar (Cambridge, MA, USA) plates were coated overnight with 2 mg/ml of anti-IL-7 mAb (Genzyme, Cambridge, MA, USA). The wells of the plate were blocked with 10% fetal bovine serum (Gemini Bioproducts) in PBS for 30 min. The plate was then incubated with the antigen for 1 h and excess antigen was washed off with PBS/Tween. The plate was incubated with 1 mg/ml of mAb to IL-7 (R&D) for 30 min and excess antibody was washed off with PBS/Tween. The plate was incubated with biotinylated anti-goat IgG (Zymed, South San Francisco, CA, USA) for 30 min and excess antibody washed off with PBS/Tween. The plate was incubated with avidin peroxidase and following incubation in OPD substrate the subsequent change in color was read at 490 nm with a Dynatech MR5000 spectrophotometer (Chantilly, VI, USA). The recombinant IL-

1367


1368

7 used in the assay as a standard was obtained from Genzyme. Sensitivity of the IL-7 ELISA was 50 pg/ml. For IFNg and GM-CSF determinations, the above procedure was repeated but with either IFNg or GM-CSF antibody pairs and standards from Pharmingen (San Diego, CA, USA). The sensitivity of the IFNg and GM-CSF ELISA was 15 pg/ml.

Tumor proliferation in vitro The MTT assay was performed to determine the proliferation rate of the parental, HyHSVtk and IL-7/HyHSVtk transduced L1C2 cell lines. The cell lines to be tested were maintained in exponential growth prior to inoculating 96-well plates for growth rate determination by the MTT assay as previously described.66 Briefly, 103 cells were inoculated in a volume of 200 ml in 96-well plates and each day 25 ml of MTT (5 mg/ml) was added to the wells and incubated at 37°C for 4 h. The plates were centrifuged for 10 min at 1000 g and the medium discarded. To solubilize the crystals, 200 ml of DMSO and 25 ml of the glycine buffer (0.1 m Glycine, 0.1 m NaCl, pH 10.5) were added to the wells. The OD was read at 570 nm. The OD, which is proportional to the number of cells, was a linear relationship from 102 to 105 cells. The MTT assay was also performed to assess the sensitivity of IL7/HyHSVtk or HyHSVtk transduced and parental cells to GCV. Briefly, cells were incubated in GCV and the number of cells monitored by the MTT assay. To determine the in vitro bystander effect parental L1C2 cells were mixed in different proportions with HyHSVtk and IL7/HyHSVtk transduced cells and exposed to GCV (50 mm). Lymphocyte proliferation Splenocytes were isolated from normal mice using a standard protocol.67 105 Cells in 100 ml of CM were incubated in 96-well plates with 100 ml of medium from exponentially growing parental, HyHSVtk transduced and IL-7/ HyHSVtk transduced L1C2 cells. Murine recombinant IL7 (10 ng/ml) was used as a control. On day 7, the lymphocytes were pulsed with 1 mCi of tritiated thymidine (Amersham, Arlington Heights, IL, USA; specific activity 62 Ci/mm) for 6 h and the cells were harvested using a Skatron cell harvester (Skatron, Sterling, VA, USA) on to filter mats. The discs were air dried, placed in scintillation vials with 1 ml of scintillation fluor and then counted in a Beckman scintillation counter (Fullerton, CA, USA). Evaluation of tumorigenicity and immunogenicity To determine the tumorigenicity of HyHSVtk and IL-7/ HyHSVtk modified L1C2 cells, 105 exponentially growing genetically modified L1C2 cells were inoculated subcutaneously on the right suprascapular area in BALB/C mice and tumor sizes and survival monitored. Tumor growth was assessed three times a week following tumor implantation. Two bisecting diameters of each tumor were measured with calipers. The volume was calculated using the formula (0.4) (ab2), with a as the larger diameter and b as the smaller diameter. The LD/TD50 of L1C2 parental cells in BALB/C mice was between 5000–10 000 cells. To determine the effect of combined IL-7/HyHSVtk and HyHSVtk genes on tumorigenicity in vivo, 105 exponentially growing L1C2 parent, control vector and IL-7

modified cells were inoculated in the right suprascapular area of BALB/C mice and GCV was administered i.p. (2 mg per dose, twice daily for 4 days). Tumor growth and survival were monitored. Mice that had rejected genetically modified tumor cells were rechallenged 30 days following tumor rejection with 5 × 104 parental L1C2 cells by s.c. inoculation on the left flank. Tumor volumes were monitored three times per week.

In vivo evaluation of the bystander effect To determine the effect of in vivo mixing on the growth of parental L1C2 tumors, exponentially growing parental cells were mixed with HyHSVtk and IL-7/HyHSVtktransduced L1C2 cells and 105 cells were inoculated in the right suprascapular area. Mixing ratios were as follows: 0:100, 10:90, 50:50, 90:10 and 100:0. Mice were treated with or without GCV (2 mg/ml per dose, twice daily for 4 days) and tumor sizes monitored three times per week. Growth of genetically modified tumor cells in SCID mice 105 Exponentially growing L1C2 parental, HyHSVtkmodified or IL-7/HyHSVtk-modified L1C2 tumor cells were inoculated s.c. in SCID mice. When tumors were established on day 10, mice were treated with or without ganciclovir (2 mg/ml per dose, twice daily for 4 days). Tumor size and survival were monitored four times per week following tumor implantation.

Acknowledgements This work was supported by the UCLA Jonsson Comprehensive Cancer Center, American Lung Association, Department of Veterans Affairs, the Helen Neufeld Research Career Development Award (Stop Cancer Award) and National Institutes of Health grant CA71818. Sherven Sharma is a recipient of a Research Award from the American Lung Association of California and was supported by NIH Institutional Training Grant in Pulmonary Medicine HL07014 and a Parker B Francis Fellowship Award of the Francis Families Foundation. Min Huang was supported by the NIH Institutional Training Grant in Tumor Immunology CA09120. Marina Stolina was supported by NIH Institutional Training Grant in Pulmonary Medicine HL07014. Steven Dubinett is a Career Investigator of the American Lung Association.

References 1 Johnson BE, Johnson DH (eds). Lung Cancer. Wiley-Liss: New York, 1995. 2 Grim J et al. Induction of apoptotic cell death in erbB-2 overexpressing tumor cells of diverse histologic subtypes mediated by intracellular localization of an anti-erbB-2 SFV. Cancer Gene Ther 1994; 1: 333–334 (Abstr. V-100). 3 Smith MJ et al. Surfactant protein A-directed toxin gene kills lung cancer cells in vitro. Hum Gene Ther 1994; 5: 29–35. 4 Tang D-C, Carbone DP. Potential application of gene therapy to lung cancer. Semin Oncol 1993; 20: 368–373. 5 Dubinett SM, Kradin RL. Cytokine immunotherapy of nonsmall cell lung cancer. Regional Immunol 1993; 5: 232–242. 6 Kradin RL, Boyle LA Preffer FI et al. Tumor-derived interleukin2-dependent lymphocytes in adoptive immunotherapy of lung cancer. Cancer Immunol Immunother 1987; 24: 76–85. 7 Kradin RL, Kurnick JT, Lazarus DS et al. Tumor-infiltrating lymphocytes and interleukin-2 in the treatment of patients with advanced cancer. Lancet 1989; i: 577–580.


8 Yoshino I, Yano T, Murata M et al. Tumor-reactive T cells accumulate in lung cancer tissues but fail to respond due to tumor cell-derived factor. Cancer Res 1992; 52: 775–781. 9 Anderson TM et al. Effects of systemic recombinant interleukin2 on natural killer and lymphokine activated killer activity of human tumor-infiltrating lymphocytes. Cancer Res 1988; 48: 1180–1183. 10 Huang M et al. Human non-small cell lung cancer cells express a type 2 cytokine pattern. Cancer Res 1995; 55: 3847–3853. 11 Pardoll DM. Paracrine cytokine adjuvants in cancer immunotherapy. Annu Rev Immunol 1995; 13: 399–415. 12 Restifo NP et al. Identification of human cancers deficient in antigen processing. J Exp Med 1993; 177: 265–272. 13 Miller AR et al. Transduction of human melanoma cell lines with the human interleukin-7 gene using retrovirus-mediated gene transfer: comparison of immunological properties with IL-2. Blood 1993; 82: 3686–3694. 14 Dubinett S et al. Down-regulation of murine fibrosarcoma transforming growth factor-b1 expression by interleukin-7. J Natl Cancer Inst 1995; 87: 593–597. 15 Sharma S et al. Interleukin-7 gene transfer in non-small cell lung cancer decreases tumor proliferation, modifies cell surface molecule expression, and enhances antitumor reactivity. Cancer Gene Ther 1996; 3: 302–313. 16 Sharma S et al. IL-7 gene transfer enhances Fas-mediated apoptotic cell death in non-small cell lung cancer. Cancer Gene Ther 1995; 2: 312 (Abstr: P-14). 17 Steinman R. The dendritic cell system and its role in immunogenicity. Annu Rev Immunol 1991; 9: 271–296. 18 Huang AYC et al. Role of bone marrow-derived cells in presenting MHC class I-restricted tumor antigens. Science 1994; 264: 961–965. 19 Gabrilovich DI et al. Production of vascular endothelial growth factor by human tumors inhibits the functional maturation of dendritic cells. Nature Med 1996; 2: 1096–1103. 20 Moolten FL. Drug sensitivity (‘suicide’) genes for selective cancer chemotherapy. Cancer Gene Ther 1994; 1: 279–287. 21 Freeman SM et al. The ‘bystander effect’: tumor regression when a fraction of the tumor mass is genetically modified. Cancer Res 1993; 53: 5274–5283. 22 Bi WL, Parysek LM, Warnick R, Stambrook PJ. In vitro evidence that metabolic cooperation is responsible for the bystander effect observed with HSVtk retroviral gene therapy. Hum Gene Ther 1993; 4: 725–731. 23 Elshami AA et al. Gap junctions play a role in the ‘bystander effect’ of the herpes simplex virus thymidine kinase/ganciclovir system in vitro. Gene Therapy 1996; 3: 85–92. 24 Parks G, Duke G, Palmenberg A. Encephalomyocarditis virus 3C protease: efficient cell-free expression from clones which link viral 5′ noncoding sequences to the P3 region. J Virol 1986; 60: 376–384. 25 Rizk N et al. Evaluation of the bystander effect in adenovirusmediated p53 gene therapy. Cancer Gene Ther 1996; 3: S36. 26 Marini FC, Nelson JA, Lapeyre J-N. Assessment of bystander effect potency produced by intratumoral implantation of HSVtkexpressing cells using surrogate marker secretion to monitor tumor growth kinetics. Gene Therapy 1995; 2: 655–659. 27 Hock H, Dorsch M, Diamantstein T, Blankenstein T. Interleukin7 induces CD4 T cell-dependent tumor rejection. J Exp Med 1991; 174: 1291–1298. 28 McBride WH et al. Genetic modification of a murine fibrosarcoma to produce IL-7 stimulates host cell infiltration and tumor immunity. Cancer Res 1992; 52: 3931–3937. 29 Alderson MR, Tough TW, Ziegler SF, Grabstein KH. Interleukin7 induces cytokine secretion and tumoricidal activity by human peripheral blood monocytes. J Exp Med 1991; 173: 923–930. 30 Dubinett S. The immune response of lung tumors and the effects of cytokine administration on pulmonary immune cells. In: Kradin R, Robinson B (eds). Immunopathology of Lung Disease. Butterworth-Heinemann: Boston, 1996, pp 469–490.

31 Smith DR et al. Production of interleukin-10 by human bronchogenic carcinoma. Am J Pathol 1994; 145: 18–25. 32 Anderson TM, Ibayashi Y, Holmes EC, Golub SH. Modification of natural killer activity of lymphocytes infiltrating human lung cancers. Cancer Immunol Immunother 1987; 25: 65–68. 33 Tepper R, Mule J. Experimental and clinical studies of cytokine gene-modified tumor cells. Hum Gene Ther 1994; 5: 153–164. 34 Namen A et al. Stimulation of B-cell progenitors by cloned murine interleukin-7. Nature 1988; 333: 571–573. 35 Welch PA et al. Human IL-7: a novel T cell growth factor. J Immunol 1989; 143: 3562–3567. 36 Chazen GD et al. Interleukin-7 is a T cell growth factor. Proc Natl Acad Sci USA 1989; 86: 5923–5927. 37 Jicha DL, Mule JJ, Rosenberg SA. Interleukin-7 generates antitumor cytotoxic T lymphocytes against murine sarcomas with efficacy in cellular adoptive immunotherapy. J Exp Med 1991; 174: 1511–1515. 38 Stotter H et al. IL-7 induces human lymphokine-activated killer cell activity and is regulated by IL-4. J Immunol 1991; 146: 150–155. 39 Yssel H, Schneider P, Lanier L. Interleukin-7 specifically induces the B7/BB1 antigen on human cord blood and peripheral blood T cells and T cell clones. Int Immunol 1993; 5: 753–759. 40 Azuma M et al. Functional expression of B7/BB1 on activated T lymphocytes. J Exp Med 1993; 177: 845–850. 41 Mehrotra PT, Grant AJ, Siegel JP. Synergistic effects of IL-7 and IL-12 on human T cell activation. J Immunol 1995; 154: 5093–5102. 42 Sieling PA et al. IL-7 in the cell-mediated immune response to a human pathogen. J Immunol 1995; 154: 2775–2783. 43 Kos FJ, Mullbacher A. IL-2-independent activity of IL-7 in the generation of secondary antigen-specific cytotoxic T cell responses in vitro. J Immunol 1993; 150: 387–393. 44 Sica D et al. Interleukin-7 enhances the proliferation and effector function of tumor-infiltrating lymphocytes from renal-cell carcinoma. Int J Cancer 1993; 53: 941–947. 45 Komschlies KL et al. Administration of recombinant human IL7 to mice alters the composition of B-lineage cells and T cell subsets, enhances T cell function and induces regression of established metastases. J Immunol 1994; 152: 5776–5784. 46 Kelley DM, McBride WH, Dubinett SM. In vivo antitumor response to interleukin-7. Clin Res 1992; 40: 63A. 47 Dubinett SM et al. Down-regulation of macrophage transforming growth factor-b messenger RNA expression by interleukin-7. J Immunol 1993; 151: 6670–6680. 48 Arthur J et al. A comparison of gene transfer methods in human dendritic cells. Cancer Gene Ther 1997; 4: 17–25. 49 Inaba K et al. Generation of large numbers of dendritic cells from mouse bone marrow cultures supplemented with granulocyte/macrophages colony-stimulated factor. J Exp Med 1992; 176: 1693–1702. 50 Mayordomo JI et al. Bone marrow-derived dendritic cells pulsed with synthetic tumor peptides elicit protective and therapeutic antitumor immunity. Nature Med 1995; 1: 1297–1302. 51 Zitvogel L et al. Therapy of murine tumors with tumor peptidepulsed dendritic cells: dependence on T cells, B7 costimulation, and T helper cell 1-associated cytokines. J Exp Med 1996; 183: 87–97. 52 Celluzzi CM et al. Peptide-pulsed dendritic cells induce antigenspecific, CTL-mediated protective tumor immunity. J Exp Med 1996; 183: 283–287. 53 Boczkowski D, Nair S, Snyder D, Gilboa E. Dendritic cells pulsed with RNA are potent antigen-presenting cells in vitro and in vivo. J Exp Med 1996; 184: 465–472. 54 Schwartzentruber D, Topalian S, Mancini M, Rosenberg S. Specific release of granulocyte–macrophage colony-stimulating factor, tumor necrosis factor-alpha, and IFN-gamma by human tumor-infiltrating lymphocytes after autologous tumor stimulation. J Immunol 1991; 146: 3674–3681. 55 Schwartzentruber D, Stetler-Stevenson M, Rosenberg S, Topalian S. Tumor-infiltrating lymphocytes derived from select B cell lymphomas secrete granulocyte–macrophage colony-stimulating factor and tumor necrosis factor-alpha in response to autologous tumor stimulation. Blood 1993; 82: 1204–1211.

1369


1370

56 Yoshino I et al. HER2/neu-derived peptides are shared antigens among human non-small cell lung cancer and ovarian cancer. Cancer Res 1994; 54: 3387–3390. 57 Slingluff CLJ et al. Cytotoxic T lymphocyte response to autologous human squamous cell cancer of the lung: reconstitution with peptides extracted from HLA-Aw68. Cancer Res 1994; 54: 2731–2737. 58 Roth J et al. Retrovirus-mediated wild-type p53 gene transfer to tumors of patients with lung cancer. Nature Med 1996; 2: 985– 991. 59 Wakimoto H et al. Intensified antitumor immunity by a cancer vaccine that produces granulocyte–macrophage colony-stimulating factor plus interleukin 4. Cancer Res 1996; 56: 1828–1833. 60 Dilloo D et al. Combined chemokine and cytokine gene transfer enhances antitumor immunity. Nature Med 1996; 2: 1090–1095. 61 Santodonato L et al. Cure of mice with established metastatic Friend leukemia cell tumors by a combined therapy with tumor cells expressing both interferon-a1 and herpes simplex thymidine kinase followed by ganciclovir. Hum Gene Ther 1996; 7: 1–10. 62 Chen S-H et al. Combination suicide and cytokine gene therapy

63

64

65

66

67

for hepatic metastases of colon carcinoma: sustained antitumor immunity prolongs animal survival. Cancer Res 1996; 56: 3758– 3762. Yuhas J, Toya R, Wagner E. Specific and nonspecific stimulation of resistance to the growth and metastasis of the line 1 lung carcinoma. Cancer Res 1975; 35: 242–244. Lupton SD, Brunton LL, Kalberg VA, Overell RW. Dominant positive and negative selection using a hygromycin phosphotransferase-thymidine kinase fusion gene. Mol Cell Biol 1991; 11: 3374–3378. Miller AD, Buttimore C. Redesign of retrovirus packaging cell lines to avoid recombination leading to helper virus production. Mol Cell Biol 1986; 6: 2895–2902. Bunnak J, Takigami M, Ito H, Shinozawa T. Gamma irradiation effects on cultured cells: investigated by the MTT method. J Rad Res 1994; 35: 205–212. Dubinett SM, Kurnick JT, Kradin RL. Adoptive immunotherapy of murine pulmonary metastases with interleukin 2 and interferon-gamma. Am J Resp Cell Mol Biol 1989; 1: 361–369.