While considerable evidence demonstrates that CD8+ TILs reactive to mutated or non-mutated tumor antigens play an important role in anti-tumor immunity, expanding evidence indicates that bona fide tumor-reactive T cells comprise only a minor fraction of all TILs in many human tumors, indicating that most CD8+ TILs have undefined specificity. Based on this long-standing question, we examined the hypothesis that a substantial fraction of TILs may represent CD8+ memory-phenotype T cells (CD8-MP cells), a unique population of cells of unknown antigen specificity that comprise 5-10% of CD8+ T cells in unprimed mice, exhibit common hallmarks of prior antigen experience, and have the capacity to rapidly expand and produce IFN-? during an immune response. In preliminary work, we found that CD8-MP cells make substantial contributions to the immune infiltrate of oncogene-driven prostate tumors, and express high densities of the PD-1 inhibitory receptor. Given these unique insights, we embarked on parallel studies aimed at further elucidating the fundamental biology of CD8-MP cells, using a clonal approach to define the developmental trajectories of these cells. Our new data reveal that the differentiation of many CD8-MP clones is triggered by recognition of self-ligands in the thymus via a reproducible, orchestrated process, challenging current thought suggesting that CD8-MP cells differentiate in the periphery in response to homeostatic signals.
In Aim 1, we will define the function of CD8-MP cells in anti-tumor immunity, testing the hypothesis that self-ligand recognition drives the early entry of CD8-MP cells into developing tumors, and that CD8-MP cells enhance anti-tumor immunity by catalyzing broader immune cell infiltration. In addition, we will identify novel markers that can be used to identify intratumoral CD8-MP cells, thereby enabling the broader study of CD8-MP cells in murine cancer models and human cancer patients.
In Aim 2, we will elucidate the molecular and cellular mechanisms that direct CD8-MP differentiation, testing the hypothesis that CD8-MP differentiation is a two-step process triggered by TCR-dependent recognition of self-ligands presented by classical dendritic cells in the thymus. We will also utilize a unique T cell antigen discovery assay to identify natural self-peptides recognized by CD8-MP cells, thereby opening new areas of inquiry that were previously inaccessible. Ultimately, defining the function of intratumoral CD8-MP cells and the blueprints of CD8-MP differentiation in mice is expected to open new avenues for the study and manipulation of these cells in humans.
Whereas some tumor-infiltrating T cells have the ability to directly recognize and kill cancer cells, the majority of infiltrating T cells are not reactive to tumor-expressed antigens, implying that these T cells may have unique functions that are not fully appreciated. Here, we demonstrate that 'memory-phenotype' CD8+ T cells reactive to non-mutated endogenous antigens infiltrate mouse prostate tumors and express high levels of the inhibitory receptor PD-1, suggesting a critical role for these cells in the tumor context. This work will define the functional role of memory-phenotype CD8+ T cells in modulating tumor development and anti-tumor immunity, and define the molecular and cellular blueprints of how these cells are generated in the body, thereby revealing new ways to identify and manipulate these cells in human cancer patients.
