The infiltration of T cells into the tumor tissue is in many cases insufficient to prevent tumor progression. In addition, tumor-infiltrating lymphocytes (TILs) in the tumor are chronically exposed to high levels of tumor antigen, which can lead to their elimination. Therefore, it is unclear how well T cells with therapeutically effective T cell receptors (TCRs) are represented in progressively growing tumors. For the successful treatment of cancer by adoptive transfer of TCR-engineered T cells (TCR- Ts), it is crucial to use TCRs of sufficiently high affinity. To determine whether the quality of tumor antigen-specific TCRs from TILs is generally inferior to that of T cells generated from antigen-negative donors, I used TCRs recognizing either of two well- characterized MHC-I-presented neoantigens. First, I analyzed a panel of TCRs directed against a neoantigen from the mutated gene cyclin-dependent kinase 4 (CDK4). These TCRs were derived from TILs, healthy donors, or from mice with human TCR gene loci. On the other hand, I isolated TCRs specific for a neoantigen derived from the mutated gene p68 (mp68) from tumor-bearing and tumor-free, immunized mice. To induce the expansion of mp68-specific T cell clones in tumor-bearing mice, I developed a cancer model in which antigen expression could be induced 3 weeks after tumor cell transplantation, thus preventing priming of mp68-specific T cells during transplantation-associated inflammation. I transferred all TCRs into donor T cells to evaluate their ability to secrete cytokines after co-culture with target cells as well as their potential to kill tumor cells in vitro. To assess the therapeutic efficacy of the TCRs, I performed adoptive T cell transfer on tumor-bearing mice. I found that the in vitro experiments that best predicted in vivo tumor control were long-term cytotoxicity assays. Furthermore, I observed that both, TIL- and immunization-derived TCRs were of variable quality which underscores the importance of testing TCRs experimentally before therapy. Most importantly, I could show that the majority of TIL-derived TCRs were therapeutically effective, and that their quality was not inferior to that of healthy donor-derived TCRs. Therefore, my results support the use of TIL-derived TCRs for adoptive transfer of TCR-Ts directed against neoantigens.
