Project Summary Melanoma continues to be a devastating cancer. Early stage melanomas represent the majority of melanomas that are diagnosed and managed in the US. It is becoming clearer that metastatic dissemination and seeding occurs very early during tumor progression while the tumor is still localized to the skin without signs of other organ involvement. To effectively battle with this disease we must understand the molecular underpinnings of biologically early tumors, in particular host defense mechanisms against the tumor cells. Basic research in early melanomas has not made significant progress over the past decades mostly due to lack of murine systems that model early disease. Genetically engineered murine models of melanoma represent many shortcomings, and at best they mimic late aggressive tumors (murine genome) within the mice immune system. In this proposal, we will mimic early malignant states in novel in vivo humanized murine models that we have developed, in which highly immune deficient mice are reconstituted with human CD34+ hematopoietic stem cells and challenged with HLA-A allele-matched human melanoma cells. By ultraviolet (UVB/A) irradiation, we will induce additional mutations as in human skin, and study the progression of tumors as they grow in time and space. Our preliminary studies indicate that a transcriptional switch occurs within the tumor cells, it accompanies unique host immune responses, and this crosstalk dictates the fate of the tumor towards progression or elimination. Stimulator of interferon – STING – signaling gets activated once melanoma cells acquire an aggressive phenotype suggesting a role during tumor progression. Here, we propose to establish paradigm-shifts in the concept of non-aggressive melanomas transitioning into an aggressive phenotype that will directly impact clinical practice. We will investigate: 1) changes within the tumor (mutations, pathways) that co-occur within the immune microenvironment (cell lineages, signals) during early disease progression using humanized mouse models (Aim1), 2) pro- or anti-tumor responses in the presence or absence of STING activation in early disease leveraging the humanized mouse and 3D skin models (Aim 1), 3) reversal of phenotypes and testing causality by modulating immune cell subpopulations (Aim 2), and 4) development and validation of a prognostic assay applied to early melanomas that predicts survival (Aim 3). Successful completion of this project promises to bring new model systems to the melanoma (and cancer) field that enable studying human tumor and human immune system interactions. If offers a major leap in early stage melanoma research, and deepens our understanding by revealing new tumor cell intrinsic or extrinsic mechanisms of disease progression.