Systematic multi-omics cell line profiling uncovers principles of Ewing sarcoma fusion oncogene-mediated gene regulation

Abstract

Cell lines have been essential for major discoveries in cancer including Ewing sarcoma (EwS). EwS is a highly aggressive pediatric bone or soft-tissue cancer characterized by oncogenic EWSR1-ETS fusion transcription factors converting polymorphic GGAA-microsatellites (mSats) into neo-enhancers. However, further detailed mechanistic evaluation of gene regulation in EwS have been hindered by the limited number of well-characterized cell line models. Here, we present the Ewing Sarcoma Cell Line Atlas (ESCLA) comprising 18 EwS cell lines with inducible EWSR1-ETS knockdown that were profiled by whole-genome-sequencing, DNA methylation arrays, gene expression and splicing arrays, mass spectrometry-based proteomics, and ChIP-seq for EWSR1-ETS and histone marks. Systematic analysis of these multi-dimensional data illuminated hundreds of new potential EWSR1-ETS target genes, the nature of EWSR1-ETS-preferred GGAA-mSats, and potential indirect modes of EWSR1-ETS-mediated gene regulation. Moreover, we identified putative co-regulatory transcription factors and heterogeneously regulated EWSR1-ETS target genes that may have implications for the clinical heterogeneity of EwS. Collectively, our freely available ESCLA constitutes an extremely rich resource for EwS research and highlights the power of leveraging multidimensional and comprehensive datasets to unravel principles of heterogeneous gene regulation by dominant fusion oncogenes.

The Transcription Factor FEZF1, a Direct Target of EWSR1-FLI1 in Ewing Sarcoma Cells, Regulates the Expression of Neural-Specific Genes

“…This indicates that FEZF1 is highly expressed in Ewing sarcoma cells and that FEZF1 expression was specific for Ewing sarcoma, at least when compared to other bone sarcomas. To strengthen these results, we analyzed the datasets available from a study published during the preparation of this manuscript and in which the authors carried out a multi-omics approach to characterize an extensive list of Ewing sarcoma cell lines [ 30 ]. According to this study, FEZF1 is expressed at high levels in all Ewing sarcoma cell lines studied, including cells expressing EWSR1-FLI1 and EWSR1-ERG fusion proteins.…”

“…Finally, we asked how many genes on this list had GGAA microsatellites bound by EWSR1-FLI1 and, in this way, determined whether these genes were directly regulated by EWSR1-FLI1 or by FEZF1. For that purpose, we searched the list of 67 genes for EWSR1-FLI1 peaks at a maximum of 10 kb upstream and downstream TSS using an in-house R script and the list of EWSR1-FLI1 ChIP-seq binding sites coincident with the GGAA microsatellites described in Orth et al [ 30 ]. We found that no FEZF1/EWSR1-FLI1 regulated genes had a peak of 10 Kb upstream TSS and only two genes (FEZF1 and PHOSPHO1) had a peak 10 Kb downstream TSS.…”

“…We have shown that FEZF1 and FEZF1-AS1 mRNA levels correlate strongly in Ewing sarcoma cells and that EWSR1-FLI1 knock-down drastically reduces the levels of FEZF1 and FEZF1-AS1 mRNAs, suggesting a coordinated regulation of both genes, at least in Ewing sarcoma cells. In the case of Ewing sarcoma, such coordinated regulation probably involves the binding of EWSR1-FLI1 to the long GGAA-microsatellite located in the exon 1 of FEZF1 , as demonstrated from publicly available ChIP-seq datasets [ 3 , 30 ]. Since we are interested in analyzing the role of FEZF1 in Ewing sarcoma cells, we put special attention into designing a specific shRNA strategy to specifically knock-down FEZF1 mRNA without affecting the expression of FEZF1-AS1.…”

“…ChIP-seq data publicly available were retrieved from Gene Expression Omnibus (GEO accession number GSE61944 and GSE176400) [ 3 , 30 ] and displayed in the UCSC browser (genome.ucsc.edu).…”

“…Gene location (chromosome number, start and end of the gene and strand) was extracted from BioMart using R routines. GGAA microsatellites bound by EWSR1-FLI1 were extracted from two published datasets of EWSR1-FLI1 ChIPseq studies (GEO accession number GSE61944 and GSE176400) [ 3 , 30 ]. Output was stored in tables.…”