Deconvoluting the Ewing sarcoma genetic program using ancestry-informed human iPSC modeling

Abstract Ewing sarcoma survival has not improved in decades despite long knowledge of its singular driving somatic mutation, the pernicious EWS-FLI1 fusion oncoprotein. As an aberrant transcription factor EWS-FLI1 alters expression of thousands of genes enacting a complex program toxic to most cell types, so much so that the cell of origin for Ewing sarcoma is still a matter of debate. Ewing sarcoma occurs at starkly higher rates (roughly 10- fold) in children of European ancestry compared to those with primarily African ancestry, while Latino children have roughly ⅔ and Asian children ½ the risk of Ewing sarcoma than do white children. Ancestry is in fact the strongest known risk factor for Ewing sarcoma, but the molecular basis of these differences in risk have been investigated only sparingly. As the mechanisms by which EWS-FLI1 interacts with the genome have become clear, it is feasible with an appropriate model to investigate how genomic ancestry in general and at specific loci modulates EWS-FLI1 activity including its downstream effects on epigenetic and transcriptional programs. To this end we have devised a strategy to introduce EWS-FLI1 in derivatives of induced pluripotent stem cells (iPSC) in order to characterize downstream molecular and functional consequences of EWS-FLI1 expression. Using accessible variant array data from several large stem cell repositories we identified banked iPSC lines derived from individuals with a range of European, African, and Amerind ancestry. We will introduce EWS-FLI1 expression at intermediate stages of development relevant to Ewing sarcomagenesis—namely neural crest cells and mesenchymal stem cells. Measures of functional tolerance and molecular state will be compared to corresponding samples from subjects with solely European ancestry. Globally we will examine whether gene expression and chromatin state exhibits similarity to the Ewing expression signature in proportion to European ancestry percentage. Genome-wide chromatin occupancy of EWS-FLI1 will be profiled and its relationship to local ancestry defined using long-read sequencing. Genes that are differentially influenced by EWS-FLI1 in “permissive” (European ancestry), “moderately permissive” (Amerind ancestry) and “impermissive” (African ancestry) genomes will be considered targets of potential therapeutic value and will undergo validation using CRISPR/Cas9 genome engineering and functional assays. The resulting data will represent the first and only effort, to our knowledge, to take advantage of the known differences in risk for Ewing sarcoma by ancestry to study EWS-FLI1 binding and downstream effects. In addition, we will make available the genome-scale data produced by our study and freely distribute our iPSC models for wider use by Ewing sarcoma researchers.