Metastatic disease remains the defining feature of advanced malignancy, and is responsible for the vast majority of cancer deaths. Metastasis can be conceptualized as an evolutionary landscape, composed of key elements of Darwinian evolution: heritable (epi)genotypes, geographic dispersal, and novel microenvironmental selection pressures. I hypothesize that altering these evolutionary landscapes would provide a strikingly new method for treating cancer, in which the cancer cells can co-exist with the host over long periods of time. To achieve this requires a deep mechanistic understanding of the ways in which tumors generate novel genotypes, and how natural selection in the microenvironment amplifies these mutations. My work utilizes the zebrafish, a small vertebrate organism that has only recently come to light as an important cancer model. The zebrafish offers several unique capacities for studying metastasis: high-throughput transgenesis, unbiased genetic screens, and single cell imaging in the optically transparent casper adult fish. For the period of this proposal, I plan to address three primary questions in metastatic melanoma: 1) can we identify the incipient genetic changes that allow for metastatic progression, whether they arise in the primary tumor or after dissemination, 2) is adaptive mutation required for metastasis, and can this be modified?, and 3) can we identify host microenvironments that disfavor metastatic progression? To do this, I will build upon a zebrafish model of melanoma in which the BRAFV600E allele is expressed in melanocyte progenitors, in the context of p53 loss of function (the BRAFV600E;p53-/- model). By combining the "brainbow" fate mapping system with the BRAFV600E;p53-/- fish, I will use exome sequencing to identify genomic lesions associated with metastasis based on lineage, space and time. These candidate changes can be functionalized using a metastasis assay I have developed in the transparent casper strain. I hypothesize that selection stressors during metastasis lead to a state of adaptive mutation, in which the error rate of DNA replication is temporarily increased to find an evolutionary solution to that stress. To test this, I will generate zebrafish with mutation reportes, and then use this system to probe whether adaptive mutation mediated by error-prone Y family DNA polymerases promotes metastatic progression. Finally, since the host microenvironment provides the ultimate selection pressure, I will perform an unbiased genetic screen to identify novel stromal regulators of metastatic progression. Together, these studies provide a comprehensive framework that considers both tumor cell-intrinsic and microenvironmental dynamics of metastatic disease. The long-term goal of my laboratory is to utilize this information to identify therapies which can convert disseminated disease into a stable state, unable to further progress, leading to large improvements in long-term survival in patients with established metastases.