Epithelioid sarcoma (ES) is one of the deadliest subtypes of soft-tissue sarcoma. Despite its benign presentation it often metastasizes to the lymph nodes and lungs and is essentially always recalcitrant to standard chemotherapy. There are two types of ES, classical (distal) and proximal, that are actually distinguished more by their histology rather than their anatomic location. It has a similar behavior to angiosarcoma in that it can start at the distal extremities and progress towards the body’s trunk. However, histologically and molecularly it relates more to malignant rhabdoid tumor and synovial sarcoma. These three sarcomas are from mesenchymal origins but display some propensity for epithelial characteristics such as cytokeratin. They also share a common molecular mechanism for genetic loss of SMARCB1/Snf5 or in the case of synovial sarcoma at the protein level. Regardless of these commonalities, one distinguishing characteristic that sets epithelioid sarcoma apart from malignant rhabdoid tumors and synovial sarcomas is its genetic complexity, which likely contributes to its aggressiveness and recalcitrance. There is a great need to study epithelioid sarcoma and its vulnerabilities and establish a preclinical model for testing potential therapeutic regimens.
One strategy for generating a preclinical model of ES is to temporally and spatially knockout SNF5 in mice with a localized injection of TATCre. Because SNF5 loss is associated with two other sarcomas, to potentiate ES we will combine Snf5 loss with disruption of CDKN2A. CDKN2A gene results in the expression of p16-INK4, which positively regulates p53 and negatively regulates CDK4 kinase. CDKN2A has been recently shown to be deleted in 6 out of 16 epithelioid sarcomas and in malignant rhabdoid tumor cell lines. Thus combinatorial deletion of Snf5 and CDKN2A should generate ES in the proper context. Because of the genetic complexity ES, the likelihood that other genes can potentiate tumorigenesis is high. To explore all possibilities for generating an efficient model that recapitulates the human histological and molecular phenotype, we propose to execute a forward genetic screen using PiggyBac transposon-mediated mutagenesis. With this genome-wide approach we will identify genes that are sufficient for generating the ideal model of epithelioid sarcoma that can be rigorously tested with therapeutic agents. For aims 1 and 2, we will test if Snf5a and CDKN2A loss accelerate tumorigenesis while maintaining epithelioid sarcoma histology and molecular expression. In aim 3, we propose using a forward genetic screen to identify other genetic alterations that can facilitate sarcomagenesis. Mice with the combined alleles of Snf5fl/fl and PiggyBac will receive an injection of TATCre and then administration of tamoxifen. Mice will be allowed to grow sizable primary tumors, which will be sequenced by Splinkerette PCR followed by Ion Torrent sequencing and characterized histologically.