Time-resolved single-cell transcriptome deconvolution unravels specific and cell cycle independent dynamics of EWS-FLI1-mediated transcriptional regulation
Aynaud M-M*1,4,5, Mirabeau O*1,4,5, Gruel N1,4,5,9, Grossetête-Lalami S1,Gribkova S2,3,4,5, Boeva V2,3,4,5, 5, Surdez D1,4,5 ,Saulnier O1,4,5,8, Durand S1,4,5,8 Kairov U6, Raynal V1,4,5, Tirode F1,4,5,Grünewald TGP7, Vert J-P 2,3,4,5, Barillot E2,3,4,5, Delattre O*1,4,5, Zinovyev A*2,3,4,5
1INSERM U830, 26, rue d’Ulm 75005 Paris; 2INSERM U900, 26, rue d’Ulm 75005 Paris; 3Mines ParisTech, 35, rue Saint Honoré 77305 Fontainebleau, 4PSL Research University, F-75005 Paris, France; 5Institut Curie, 26 rue d'Ulm, F-75005 Paris, France; 6Department of Genomic and Personalized Medicine, Center for Life Sciences, Nazarbayev University, Astana, Kazakhstan; 7Laboratory for Pediatric Sarcoma Biology, Institute of Pathology of the LMU Munich, Munich, Germany; 8Université Paris-Diderot, Sorbonne Paris Cité, INSERM U830, F-75005 Paris, France; 9Departement de recherche translationnelle F-75005 Paris, France; 5Institut Curie, 26 rue d'Ulm, F-75005 Paris, France
Ewing sarcoma is the second most common pediatric bone-associated cancer. It is driven in 85% of cases by a chromosomal translocation, which generates the chimeric transcription factor EWS-FLI1. EWS-FLI1 regulates directly or indirectly a large number of genes (over 1000) which affect a cascade of biological processes leading to abnormal cell proliferation. So far, all studies of the oncogenic effect of EWS-FLI1 on the transcriptome were limited to analyses of cell bulk which obscured the effect of a possible heterogeneity of the cellular response, and hampered the study of specific EWS-FLI1 effects in different cell cycle phases and distinguishing them from its other actions.
To further investigate the oncogenic mechanisms of action, we have taken advantage of a doxycline-regulated, sh-based system that enables to control for EWS-FLI1 expression in Ewing cells. We sequenced the transcriptomes of 742 single cells: at different time points during EWS-FLI1 expression in vitro and in vivo and from patient derived xenograft. We have used a matrix factorization technique, Independent Component Analysis (ICA) to identify independent biological pathways activated in single cells. In parallel, we performed FLI1 chromatin immunoprecipitation sequencing at different time points during EWS-FLI1 expression. Our analysis enabled us to identify two distinct groups of genes, 1) potential EWS-FLI1 direct targets of EWS-FLI1 whose re-expression profile correlated with EWS-FLI1 in time-course experiments, which are specifically expressed in Ewing sarcoma tumors and which contain EWS-FLI1 binding sites in the vicinity of their TSS associated with active epigenetic marks. 2) genes mainly involved in cell cycle regulation and expressed in a highly coordinated manner within single cells.
Our study demonstrates the power of combining single cell transcriptomics, ICA and ChIP sequencing analysis to distinguish the direct and indirect effect of an oncogene, and to gain insights into the cell cycle machinery and how it is coupled to the direct action of an oncogene. In addition it provided us with a strong list of putative direct targets of EWS-FLI1, which are candidate transcriptional entry points for the action of this oncogene in Ewing sarcoma cells.