Project Summary Transposable elements are a major source of human genetic variation and also contribute to human diseases by causing somatic mutations. The activity of transposition in human cancers is well established. Both expression of the RNA binding protein ORF1p encoded by the L1 element and insertions of L1 and other transposons are commonplace in cancer cells and cancer genomes. However, whether and how retrotransposition contributes to cancer genome evolution and promotes tumorigenesis remain unclear. Here, we hypothesize that retrotransposition may promote cancer evolution by causing long-range genomic rearrangements in addition to insertions. This hypothesis is formulated based on both genomic observations in primary cancer genomes and molecular and genomic data from our in vitro experiments. To test this hypothesis, we will combine experimental approaches in in vitro models (Aim 1) with computational analyses (Aim 2) of retrotransposition and retrotransposition-associated rearrangements in primary cancers and precancer cells (Aim 2). Specifically, we propose that non-canonical resolution of LINE-1 insertions can cause DNA breaks and long-range DNA rearrangements, and a subset of these rearrangements further destabilizes genomes by generating acentric or dicentric chromosomes that undergo waves of rearrangements. To test this, we first assess DNA damage and its dependence on the endonuclease (EN) and reverse transcriptase (RT) activities of L1 ORF2p (Aim 1a). We then determine both immediate and downstream genomic consequences of ORF2p-induced DNA damage using state-of-the-art Look-Seq approaches that combine cell biology and whole-genome sequencing (Aim 1b and 1c). Complementing these in vitro analyses of retrotransposition, we propose to analyze bulk cancer genomes (Aim 2b) and single pre-cancer cell genomes (Aim 2c) to definitively assess the prevalence of insertions and long-range rearrangements that bear the hallmarks of retrotransposition. We focus on esophageal cancers with an extremely high prevalence of retrotransposition (in >90% of cancers) and the precancer condition known as Barrett's Esophagus (BE), where LINE-1 retrotransposition has also been observed previously. The proposed work includes development of new bioinformatic packages for joint rearrangement and transposition detection in shotgun cancer sequencing data that will benefit researchers studying genetic variation, genome instability, and somatic genome evolution (Aim 2a). Together, the proposed studies will thoroughly characterize the relationships between LINE-1 retrotransposition and genome instability writ large and the roles of retrotransposition in cancer evolution. This knowledge may lead to new avenues of cancer prevention and therapeutics.