Interleukin (IL) -4 gene therapy has been used clinically against cancer although how IL-4 modulates immune responses in vivo is incompletely understood. We investigated antitumor mechanisms in IL-4 gene therapy in a murine model. A poorly immunogenic murine colorectal cancer cell line, MC38, was transfected to overexpress IL-4 (MC38-IL4, producing 359.8 ± 108.2 ng of IL-4/106 cells/48 h). When 1×105 MC38-IL4 cells were inoculated in immunocompetent mice, tumorigenicity was much reduced from that of parental cells. When tumor-free mice initially injected with MC38-IL4 cells were reinjected contralaterally on day 35 with 3×105 wild-type MC38 cells (MC38-WT), 15 of 18 mice rejected the subsequent challenge. As a model of therapy against established tumors, 1×105 MC38-IL4 cells were inoculated contralaterally 7 days after 1×105 MC38-WT cells had been injected, significantly reducing growth of wild-type tumors (p=0.0301). Injection of MC38-IL4 cells in leukocyte-depleted mice confirmed that neither T cells nor natural killer (NK) cells were involved in IL-4-related primary antitumor effects. Immunohistologic analysis showed numerous granulocytes infiltrating wild-type tumors of mice treated with MC38-IL4, which suggests granulocytes involvement in the antitumor effect. Inoculation of MC38-WT cells in leukocyte-depleted mice initially injected with MC38-IL4 cells suggested that both CD4+ and CD8+ T cells contributed to the antitumor effects. To investigate induction of tumor-specific immune responses, we stimulated splenocytes of MC38-immune mice with MC38-IL4 cells twice weekly in vitro, resulting in MC38-specific lysis 57.5±7.2% of MC38 cells at an effector to target ratio of 20. Thus, IL-4 could elicit and maintain tumor-specific cytotoxic T lymphocytes (CTL). When we treated mice with established wild-type tumors with MC38-IL4 cells in combination with interferon (IFN)-α-overexpressing MC38 cells (MC38-IFNα), the mean area of wild-type tumors was significantly less than in controls (p=0.0092). In vitro IFN-γ production by splenocytes from mice injected with both MC38-IL4 and MC38-IFNα was greatly enhanced in comparison with splenocytes from mice injected with MC38-IL4 alone, while IL-10 production was not increased. Thus, early antitumor effects of IL-4 gene therapy are independent of both T and NK cells, but probably involve granulocytes. Subsequently, IL-4 induces long-lasting, tumor-specific immune responses mediated by T lymphocytes in vivo. IL-4 appears to promote a Th1-type antitumor immune response, which is enhanced in cooperation with IFN-α. IL-4 gene therapy in combination with Th1-type cytokine gene therapy should be evaluated in clinical trials.

[Proc Amer Assoc Cancer Res, Volume 45, 2004]