Melanoma is one of the deadliest forms of cancer and is poorly responsive to standard chemotherapeutics. Much of our understanding of tumor cell interactions with the microenvironment are either inferred from end-point assays and fixed tumor sections, or have only been visualized with high resolution in in vitro co-culture models. Cancer metastasis is a dynamic process; however, a major limitation to understanding cancer progression is the lack of genetically tractable in vivo model systems that are amenable to high-resolution imaging. We will overcome this obstacle by visualizing and manipulating tumor cells and their microenvironment directly in a human-in-zebrafish xenotransplant model. We will take advantage of the ease of expressing reporters in tumor cells in culture combined with imaging tumor cell and stromal cell interactions in zebrafish to understand heterotypic cell interactions and signaling pathways between the tumor cells and microenvironment that regulate malignant melanoma progression. We will then validate our key findings in mouse models with fixed imaging and/or end-point analyses. We have determined that zebrafish larvae can be injected with human melanoma cells and these xenotransplants exhibit a 14% rate of metastasis. Live cell imaging reveals that tumor cells at the periphery of the tumor mass respond to physical cell contact with macrophages by forming actin-rich protrusions. Depleting larvae of host macrophages resulted in reduced melanoma metastasis. Separately, we have observed that melanoma cells exhibit "angiotropism": expansion along the abluminal surfaces of blood vessels. From these results, we hypothesize that physical contact with macrophages induces actin dynamics in tumor cells to regulate intravasation in vivo. With the long-term goal of identifying potential markers for predicting metastatic risk as well as targets for therapies that block the early steps of melanoma metastasis, this proposal focuses on interactions between tumor associated macrophages and melanoma cells, and seeks to understand how this interaction impacts invadopodium formation during intravasation (Aim I) and angiotropism (Aim II), processes that are difficult to study in other in vivo systems. As melanoma cells adhere to the vasculature prior to penetrating matrix barriers for intravasation, we will further determine whether focal adhesion assembly and invadopodium formation are coupled events (Aim III) for efficient intravasation of melanoma cells. These experiments will reveal the mechanisms by which melanoma cells switch from adhering to matrix to degrading matrix during intravasation, a critical step in metastasis.