Characterization of Ewing's sarcoma susceptibility loci
Final Report Abstract
Deciphering how somatic driver mutations and germline susceptibility variants cooperate to promote cancer constitutes a considerable challenge. The hallmark of Ewing’s sarcoma (EwS) is a chimeric EWSR1-ETS gene fusion, usually EWSR1-FLI1, which generates a potent oncogenic transcription factor that can bind DNA at GGAA-motifs such as found in GGAA-microsatellites. EwS incidence is highly variable across human populations (e.g. about 8-10 times higher in Europeans than in Africans), which was recently explored by a genome-wide association study (GWAS) that identified susceptibility variants on chromosome 1p36.22 near TARDBP, on chr10q21.3 near the Early Growth Response 2 gene (EGR2) and ADO, as well as on chr15q15.1 near BMF. The aim of the this project was to i) functionally characterize the role of these GWAS candidate genes in EwS, ii) to identify the causal regulatory variants (among hundreds of genetically linked variants) that modulate their expression in EwS, and iii) to decipher the genetic mechanism how these causal regulatory variants control expression of these candidate genes possibly together with EWSR1-FLI1. Here, the funding recipient provides a comprehensive follow-up of the EwS GWAS by integrating targeted next-generation deep-sequencing of germline DNA from 600 individuals with functional characterization of the GWAS candidate genes TARDBP, EGR2, ADO, and BMF in EwS in vitro and in the case of EGR2 also in vivo. Preliminary analyses showed that the siRNA-mediated knockdown of all GWAS candidate genes, except for ADO, inhibits proliferation and cell cycle progression of EwS cells, which is paralleled by induction of apoptosis. Although all GWAS candidate genes exhibit a limited overlap in their transcriptional signatures that they confer to EwS with that conferred by EWSR1-FLI1, the most significant overlap was observed with EGR2. This is in accordance with the fact that only EGR2 appears to be consistently and strongly regulated by EWSR1- FLI1 in EwS cells. In-depth analysis of EGR2 and the chr10 susceptibility locus revealed, that EGR2 is crucial for tumorigenesis of EwS and identified a regulatory variant, which contributes to EWSR1- FLI1-driven EGR2 overexpression in EwS. The funding recipient observed that EGR2 knockdown inhibits proliferation, clonogenicity, and spheroidal growth of EwS cells in vitro, and induces complete regression of EwS xenografts in vivo. Targeted germline deepsequencing of the EGR2 locus in 347 cases and 251 genetically matched controls revealed 267 single nucleotide polymorphisms (SNPs) significantly associated with EwS. At one common A/T SNP (rs79965208), the A-risk-allele is significantly more frequent in EwS than in controls and is in strong linkage disequilibrium with the previously identified EwS- susceptibility SNPs. It connects two adjacent GGAA-repeats by conversion of an interspaced GGAT- into a GGAA-motif. Thereby, the A-allele considerably increases the number of consecutive GGAA-repeats and thus EWSR1-FLI1-dependent enhancer activity of this GGAA-microsatellite that exhibits chromatin characteristics of an active regulatory element. Accordingly, the A-allele is associated with increased overall and allele-specific EGR2 expression in EwS tumors. Strikingly, the frequency of the A-allele is highly significantly different between Europeans (0.57) as compared to Africans (0.25) (P=6.87x10-30), hence suggesting that rs79965208 may contribute to the variable susceptibility to EwS across human populations. Collectively, the data of the funding recipient establish the cooperation between a dominant oncogene and a susceptibility variant that regulates a major driver of EwS tumorigenesis. To the best of the funding recipient’s knowledge this constitutes one of the first reports exemplifying how a regulatory germline variant highly correlated with the reported GWAS signal can inform our understanding of a cancer-specific acquired genetic abnormality and point out a key pathway involved in tumorigenesis. However, as many additional SNPs in the different sub-haploblocks of the chr10 locus are associated with EwS, it remains plausible that other SNPs, in conjunction with rs79965208, could also have a regulatory effect on EGR2 expression through other mechanisms. Furthermore, the precise role of the chr1 and chr15 loci and whether and how they might interact with the chr10 locus need to be determined. These aspects will be examined in subsequent cooperative studies between the funding recipient and the Institut Curie Research Center.
Publications
- Chimeric EWSR1-FLI1 regulates the Ewing sarcoma susceptibility gene EGR2 via a GGAA microsatellite. Nat Genet. 2015 September; 47(9): 1073–1078
Thomas G. P. Grünewald, ... and Olivier Delattre
(See online at https://doi.org/10.1038/ng.3363)
