EWS-FLI1 fluctuations in Ewing sarcoma

In patients with solid cancers, the development of metastases is frequently associated with therapy resistance and, in consequence, bad prognosis. Therefore, there is a burning need for new therapies specifically preventing metastatic spread. This applies also to Ewing sarcoma, a malignant bone cancer affecting predominantly adolescents. Metastasis is based upon marked cellular plasticity, which allows tumor cells to reshape their cytoskeleton thus enabling invasion into neighboring tissue, entrance into the circulation, homing to distant organs and proliferation at the metastatic site. Ewing sarcoma growth is driven by an oncogenic fusion protein, which is expressed as consequence of rearrangement between the EWS and FLI1 genes. The resulting fusion protein binds DNA and alters the regulation of a large number of genes by various distinct mechanisms. Previous results suggested that fluctuations in EWS-FLI1 expression may promote Ewing sarcoma plasticity. In addition, in primary Ewing sarcomas, a small number of tumor cells was described, which experimentally is associated with low EWS-FLI1 levels and increased metastatic potential. We seek to test this hypothesis and uncover factors driving EWS-FLI1 fluctuations. We will identify EWS-FLI1 modulatory signaling pathways and their upstream environmental stimuli, and study their influence on tumor plasticity and metastasis. On the other hand, we will interrogate the influence of distinct EWS-FLI1 concentrations on downstream gene regulatory mechanisms, and study ensuing consequences for tumor cell growth and metastasis. To that end, we will apply a new Ewing sarcoma model expressing EWS-FLI1 fused to a fluorescent protein domain and tagged with a so-called degron, which allows rapid, tunable partial or complete pharmacological EWS-FLI1 degradation. This way, we will be able to decrease EWS-FLI1 levels gradually and determine limiting concentrations for proliferation, invasion and migration of tumor cells, as well as characterize distinct gene sets with different EWS- FLI1 affinity. Automated monitoring of declining fluorescence of the fusion protein will allow us to identify in high throughput format molecular factors, gene activities and novel drugs modulating EWS-FLI1 protein concentrations and consequently tumor cell invasion and migration. Thus, this project will for the first time investigate causes and consequences of oncogene fluctuations in a cancer using Ewing sarcoma as a paradigm, and define new therapeutic avenues to prevent metastasis of solid tumors.

Ewing sarcoma (ES), the second most common malignant bone tumor in children and young adults, is driven by the EWS::FLI1 (EF) fusion oncoprotein. Accumulating evidence suggests that the expression level of EF determines the functional state of tumor cells. To dissect this relationship, we defined transcriptional programs and cell states associated with distinct EF thresholds and explored pathways, genes and compounds that modulate EF protein abundance. To this end, we generated ES cell line models expressing endogenous EF fused to a fluorophor and a ligand-inducible degron, which allowed us to monitor EF fluctuations with high sensitivity and establish a finely tuned gradient of different EF levels. We found that even a subtle EF reduction resulted in significantly increased tumor cell migration and invasion, while simultaneously abrogating their ability to grow and form anchorage-independent colonies. Nascent transcriptomic analyses revealed different gene sets with distinct sensitivities to acute EF depletion of varying amplitudes, allowing us to link these response patterns to EF dependent and EF co-dependent gene regulatory mechanisms. Strikingly, complete rescue of EF levels after one to four weeks of EF modulation, mimicking oncoprotein fluctuations, resulted in emergence of a proliferative hybrid phenotype with increased metastatic potential. This cell state was characterized by the persistent upregulation of a specific set of genes, several of which correlate with poor patient prognosis. To identify modulators of EF levels, we performed a genome-wide gene knockout screen and a high-throughput compound screen of 2500 compounds in two ES model cell lines. We identified histone deacetylase inhibitors and inhibitors of the PI3K/AKT pathway to downregulate EF, while inhibitors of mitotic spindle assembly and G2/M cell cycle transition increased EF protein levels. The gene knockout screen further confirmed an EF modulatory role of PI3K/AKT signaling, on which several receptor tyrosine kinase pathways converge, including IGF-1 receptor signaling. Notably, using a mouse model, we demonstrated that embryonic EF expressing mesenchymal bone precursors can be fully transformed upon transient exposure to high IGF-1 levels, mimicking the pubertal hormonal environment during which Ewing sarcoma incidence is highest. Integrative genomic and epigenomic analyses uncovered a tumor promoting mechanism wherein IGF-1 signaling activates a set of EF-primed genes through the YAP/TEAD transcriptional complex. Consistently, we found that pharmacological co-inhibition of IGF-1R and YAP/TEAD synergistically kiled ES cells in 3-dimensional spheroid culture. While our findings may identify a potential novel avenue to ES treatment, they also underscore a critical caveat where directly targeting the ES driver oncoprotein EF may unintentionally and paradoxically promote disease progression if it is not completely eradicated.