In vivo prime editing for precision cancer mouse models

Project Summary/Abstract Cancer genomic studies have identified a large number of mutations, including point mutations, insertions/deletions (indels), and structural variants in coding and regulatory regions. For many uncharacterized mutations, how they contribute to tumor growth and resistance to therapy remains murky. In vivo functional validation is a pre-requisite for identifying suitable interventions and biomarkers. Rapid and flexible genome editing platforms are therefore needed to build and validate precision mouse models with translational value. Recent advances in CRISPR-based genome editing allow us to directly model cancer mutations in adult mouse tissues, bypassing the need for germline mouse models. Current CRISPR-based approaches however are either inefficient (e.g., homology directed repair) or cannot be used to model all types of mutations (e.g., base editing). A more flexible genome editing platform is needed to speed up the generation of somatic cancer mouse models. We have begun to optimize prime editing as a way to precisely model a wide variety of cancer mutations in mice. Prime editor (PE)—Cas9 nickase fused to reverse transcriptase—utilizes an extended guide RNA (called a pegRNA) that doubles as a template for reverse transcriptase to copy information into the genomic target. We have recently developed an optimized PE that can install cancer mutations in mouse liver. We also engineered a dual-PE approach to precisely introduce large deletions (up to ~10 kb) with short insertions. This work provides a foundation upon which to further optimize the delivery of PE to mice and to expand its application to cancer mouse models. The goal of this project is to develop and optimize new PE tools to speed up the generation of both constitutive and inducible somatic mouse models of liver and lung cancer. Aim 1 will define rules for optimal pegRNA design while validating 100 cancer-associated mutations, establish platforms for somatic cancer modeling in vivo, define pre-clinical features of PE models, and develop delivery vectors for multiplexed prime editing in mouse lung. Aim 2 will develop prime editing tools for inducible cancer models. Dual-PE approaches will be used to make inducible alleles and floxed alleles in mouse liver. We will also generate efficient PE models to recapitulate large genomic deletions in human cancer. This project will result in new mouse models that validate driver mutations in liver and lung cancer and will provide a flexible platform to enhance the translational utility of mouse models of cancer.