In vitro and in vivo correlation of the effect of granulocyte-macrophage colony-stimulating factor gene transfer on the tumorigenicity and immunogenicity of B16 melanoma

Transduction of murine B16 melanoma cells with a GM-CSF gene, the B16-MG tumor line, showed reduced tumorigenicity. In vitro studies demonstrated no remarkable difference between the parent and transduced tumor lines in their ability to induce secondary response to generate the anti-tumor killer cells (immunogenicity), or in their susceptibility to the killing by anti-tumor killer cells (immunosensitivity). Both CD4+ and CD8+ cells were required for the generation of the effectors. Nevertheless the effectors were determined to be Thy 1.2+, CD8-, and NK1.1-. At least two antigenic specificities could be defined in the cytolytic reactions. One was a broadly cross-reactive antigen shared by a variety of tumor cells, and the other apparently a tumor-specific antigen which was only present in B16 tumors. Cold target inhibition experiment confirmed these specificities. In the in vivo tumor transplantation study, the B16-MG cell line was not only more immunogenic but also was more immunosensitive than the parent line. More than 50% of the mice which were immunized with B16-MG remained tumor free after challenge with the parent tumor B16, indicating that GM-CSF gene transfer makes an effective tumor vaccine. The in vivo protective effect was specific for B16 tumor, thus only the tumor-specific antigen could function as transplantation antigen. Both CD4+ and CD8+ cells were required for providing the in vivo protection. Both the B16 and B16-MG tumor bearing hosts could generate anti-tumor killer cells, hence the development of progressive growth of B16 tumor was not due to the lack of anti-tumor immune response. It appears that the overall effect of in vivo tumor immunity is determined by a complex network of interactions among different compartments of host immune cells and different immuno-regulatory molecules derived from the host and from the tumor.