A druggable genome screen identifies b-catenin transcription targets as desmoid cell vulnerabilities

Background: Desmoid-type fibromatosis (DT) is a rare mesenchymal neoplasm associated with CTNNB1 activating. DT have variable biologic behavior not explained by secondary genomic events, and in aggressive tumors, local progression can cause significant morbidity. We performed a druggable genome screen to better understand beta-catenin function in DT and nominate potential therapies for DT treatment. Experimental design: A lentiviral shRNA screen (LT3GEPIR backbone) was performed in a primary DT cell line overexpressing TERT (DES9525T). Cell proliferation, cell cycle, gene expression and protein accumulation/localization after pharmacologic inhibition or gene silencing were assessed by CyQuant, Guava Cell Cycle reagent, RNA-seq, and immunoblot in whole cells or after subcellular fractionation (NE-PER extraction), respectively. Results: 75 significantly enriched and 20 significantly depleted genes were identified in DT cells. Among the genes that had the most significant effects on proliferation were MDM2 , BRD4 (p<0.001, FC=0.25, 0.31) and XPO1 (encoding exportin-1; p<0.002, FC=0.26) while shRNA targeting TP53 and CDKN1A induced proliferation. Inhibition of exportin-1 with selinexor or shRNA directed at XPO1 inhibited proliferation (by 51%, p<0.05), but failed to affect beta-catenin subcellular localization as in other systems. It did have canonical effects on p53/p21 signaling with increased protein levels observed in treated DT cells. Inhibition of MDM2 with milademetan or shRNA similarly increased p53 and p21 and reduced proliferation (by 60%, p<0.05). In both cases, p53 induction was associated with G 1 cell cycle arrest. BRD4 inhibition via shRNA or treatment with birabresib reduced proliferation (85%, p<0.01) but did not affect p53/p21 signaling. Instead BRD4 inhibition decreased transcription of genes indirectly regulated by beta-catenin including potential oncogenes TNC , TGFBI , and PITX2 . RNA-seq, ChIP-seq and ChIP-PCR showed that BRD4 transcription was directly regulated by beta-catenin binding to the gene. MDM2 transcription was also directly modulated by beta-catenin, but in its case beta-catenin binding to the gene appeared to result in transcriptional repression. Conclusion: BRD4, MDM2 and XPO1 were identified as potential therapeutic targets in DT with gene products modulating p53/p21 or beta-catenin activity. Both BRD4 and MDM2 represent direct transcriptional targets of beta-catenin though negative regulation of MDM2 in the context of DT cell dependency on its protein product suggest secondary genetic events or environmental signaling may be necessary to counteract this potential tumor suppressive effect of beta-catenin.