"? DESCRIPTION (provided by applicant): Gastrointestinal (GI) malignancies such as hepatocellular carcinoma, cholangiocarcinoma, and colorectal carcinoma are refractory to most current therapies and are among the largest cancer killers in the western world. While genomic technologies have cataloged many cancer-specific mutations, defining those changes that drive disease progression and represent potential therapeutic targets remains a significant unmet need. Genetically engineered mouse models (GEMMs) offer an ideal setting to interrogate the genetics and biology of cancer initiation and maintenance, and serve as powerful preclinical models to test novel cancer therapies. Nevertheless, the generation and analysis of new GEMMs using conventional methods is simply too slow and costly to evaluate even moderate numbers of candidate genes in vivo, and thus, genetic strategies to validate therapeutic targets in established tumors remain exceedingly challenging. Consequently, GEMM models have been underutilized in drug development efforts. Here we propose a conceptually new approach that enables rapid production of tailored, genetically defined animals in a fraction of the time of conventional methods. Our proposal is based on the derivation and use of disease-specific, multi-allelic embryonic stem cells (GEMM- ESCs) that can be customized to investigate any gene in a defined, disease-prone genetic background, without any breeding. Each model harbors a homing cassette that allows rapid targeting of tetracycline responsive short hairpin RNAs (shRNAs) constructs, allowing the production of cancer prone mice in which any gene can be suppressed in established tumors and normal tissue simply by the addition of doxycycline. This new modeling paradigm is built on a solid foundation of previous innovation in our laboratory that has led to improved RNAi and genome editing tools and will enable rapid evaluation of the potential efficacies and toxicities associated with target inhibition in vivo. Successful completion of the proposed work will dramatically increase the speed and scale at which genetic-associations and therapeutic targets can be investigated GI cancers and provide a paradigm that may transform the way we design and use mice in oncology research. As such, we believe our application meets all three major goals of the FOA, addressing one or more of the technical and experimental parameters that ensure effective translational use of mouse models as well as addressing unmet translational requirements. Our new models also will be ideally suited to contribute to, and benefit from, the Oncology Models Forum NCIP Hub platform. " |