In order to facilitate analysis of the complex immune response to cancer cells, a well-defined tumor model system has been established in our group. In this model the mouse renal cancer Renca, transfected with the viral protein hemagglutinin (Renca-HA) as a surrogate tumor antigen, is used. The transfer of the viral HA gene provides a tumor that is much more immunogenic, due to response by CD 8 cytotoxic T cells of the mouse to the viral protein. In addition transgenic mice have been generated where the T cell receptor (TCR) transgene specifically recognizes a single dominant peptide of the viral HA protein (TCR-HA), thus allowing for this antigen-specific T cell population to be accurately followed and monitored in vivo. These mice have been crossed onto strains which lack some of the cytotoxic effector molecules of T cells such as perforin and FasL important for killing tumor cells in vitro. Adoptive transfer of activated CD 8 T cells from these various novel mouse strains (developed in our group) to wild type mice bearing the Renca-HA tumors shows that T cell production of the effector protein FasL is crucial for rejection of Renca-HA cells in vivo, whereas the cytotoxic pathway using perforin is dispensable. The reason for the dominance of the FasL pathway may be that it is more efficiently triggered under conditions where levels of tumor antigen (HA) are very low resulting in a weak signal to the antitumor T cells. This may more accurately model the normal situation in vivo, where responses to tumor antigens are probably much weaker than response to viral antigens. Therefore our conclusions from this mouse model of renal cancer are that the FasL killing pathway may be more important than was previously thought for tumor destruction in vivo. The role of this cytotoxic FasL pathway for killing human renal cancer cells is under further investigation. This tumor model system has also proven useful for a detailed analysis of how drugs affect a CD8 T cell response, at either the generation of the response or the effector phase of the response could be used together with immunotherapy. This is due to the practical consideration that the development of the specific immune response to a single peptide from HA can be easily followed in vivo using the appropriate reagents in this model system. More recently we have used a Nanog-promoter-driven green fluoresence construct to mark 4T1.2 breast carcinoma cells with a more cancer """"""""stem cell"""""""" phenotype. Preliminary data suggest that the GFP+ cells have many characteristics of cancer stem cells. We have also put the model antigen HA into these cells to allow for an analysis of how cancer stem cells may respond to immune-mediated cytotoxic effects.
