PROJECT SUMMARY/ABSTRACT Tumor heterogeneity, in cell behavior, cell microenvironment and gene expression, has become a hallmark of tumor progression, and provides an important challenge in understanding cancer. For example, the adoption of an invasive phenotype by some tumor cells is a key step as carcinomas advance from the benign tumor state. Migration of tumor cells from the epithelial layers of the epidermis to the dermis provides tumor cell access to the vasculature, offering a direct path for distal metastasis. Understanding how and why some cells are capable of invading surrounding tissues is key to understanding tumor invasion, but current in vivo models of invasion do not readily allow specific invasive cells to be recovered and studied to elucidate drivers of tumor invasion and progression. Our goal is to create a new pipeline for following tumorigenesis, permitting direct studies of tumor cell diversification and invasion. This experimental pipeline is designed to offer capabilities to: (i) track tumorigenesis longitudinally over time from the first origins of a tumor to key events such as invasion; (ii) label and selectively purify tumor cell subsets identified in these time-course studies; (iii) characterize individual tumor cells in isolated populations in unbiased way. This approach would allow identification of the cellular events and developmental trajectories that accompany tumor growth and invasiveness. Molecular correlates characterized in the key sub-fractions of the tumors can be directly related to the event history of the cell sub-fractions, and the importance of these cellular and molecular events can be tested through perturbation, ablation and function blocking experiments. Such a functional and longitudinal experimental pipeline is key to creating more targeted and informative drug screening systems. We have developed a set of transgenic zebrafish lines and imaging tools that allow repeated, non-invasive visualization of endogenous tumor development in adult animals. This permits tumors to be tracked from their origins, over long periods, after oncogene expression in skin epithelial cells. Our preliminary results show that heterogeneity is present from the earliest stages of tumor formation, as only a small fraction of cells induced to express oncogenes go on to form invasive tumors; such results provide the motivation for and the foundation for our analyses of the emergence and elaboration of tumor heterogeneities. Here, we propose to employ imaging and labeling of selected cell populations, in combination with single-cell RNA-sequencing analyses, to study the cell behaviors and differentiation routes key to invasive tumor development. We will establish an experimental pipeline by investigating: the cell-types specific to invasive tumors (Aim 1); the key differentiation pathways leading to an invasive phenotype (Aim 2); and leveraging these insights by functional assays (Aim 3). This pipeline will offer new insigh...