SMARCB1-Dependencies in Malignant Rhabdoid Tumours: A strategy for Pre-Clinical Therapeutic Target Identification in the Absence of Actionable Mutations

ATRTs have low mutation rates and few classically-actionable variants to direct molecularly targeted therapies; loss of SMARCB1 is the sole recurrent mutational event in > 90% of ATRTs. We have designed and implemented a genome-scale strategy for target identification and prioritization in tumors devoid of significant actionable mutations. Re-expression of SMARCB1 causes ATRT cells to cease proliferation and differentiate; we therefore hypothesized that identifying and counteracting critical downstream SMARCB1-dependent events represents a primary route to therapeutic intervention.

We identify such events using an integrated genome-wide approach encompassing genome-scale CRISPR/Cas9 functional screens (GeCKO screening; 122,411 sgRNAs) alongside expression/DNA methylation profiling of primary ATRTs and ATRT cells following SMARCB1 re-expression and/or treatment with demethylating agents. Cross-referencing these analyses, we use a rational selection algorithm, to identify critical tumorigenic genes/pathways and proceed to validate their ability to be targeted as therapeutic targets.

Our strategy identifies, ranks and prioritizes multiple SMARCB1-dependent pathways/genes functionally essential to ATRT, and characteristic of the primary tumour; including those previously described (Rb/CDK4/6, SHH, MYC), those less well evidenced (mTOR, TGF-β, HGF, Stat3/Jak) and novel SMARCB1-dependent synthetic lethalities (NuA4 complex, PIM1). We also describe the repressive genome-wide effect of SMARCB1 mutation on the transcriptome, through altered SWI/SNF binding, associated histone marks and localized hypermethylation and the therapeutic possibilities implied therein.

Data and web-based interactive analysis and visualization tools will be made publically available to help guide and benchmark future pre-clinical testing in ATRT. This study pinpoints SMARCB1-dependent therapeutic susceptibilities in MRTs with the capacity to inform future treatment strategies.