Summary Recent studies have revealed that ARID1A is frequently mutated across a wide variety of human cancers and also has bona fide tumor suppressor properties. Consequently, identification of vulnerabilities conferred by ARID1A mutation would have major relevance for human cancer. Here, using a broad screening approach, we identify ARID1B, a related but mutually exclusive homolog of ARID1A in the SWI/SNF chromatin remodeling complex, as the number one gene preferentially required for the survival of ARID1A-mutant cancer cell lines. We show that loss of ARID1B in ARID1A-deficient backgrounds destabilizes SWI/SNF and impairs proliferation. Intriguingly, we also find that ARID1A and ARID1B are frequently co-mutated in cancer, but that ARID1A-deficient cancers retain at least one ARID1B allele. These results suggest that loss of ARID1A and ARID1B alleles cooperatively promotes cancer formation but also results in a unique functional dependence. The results further identify ARID1B as a potential therapeutic target for ARID1A-mutant cancers.

SMARCB1 (SNF5/INI1/BAF47), a core subunit of the SWI/SNF (BAF) chromatin remodeling complex1,2, is inactivated in nearly all pediatric rhabdoid tumors3–5. These aggressive cancers are among the most genomically stable6–8, suggesting an epigenetic mechanism by which SMARCB1 loss drives transformation. Here, we show that despite indistinguishable mutational landscapes, human rhabdoid tumors show distinct enhancer H3K27ac signatures, which reveal remnants of differentiation programs. We show that SMARCB1 is required for the integrity of SWI/SNF complexes and that its loss alters enhancer targeting – markedly impairing SWI/SNF binding to typical enhancers, particularly those required for differentiation, while maintaining SWI/SNF binding at super-enhancers. We show that these retained super-enhancers are essential for rhabdoid tumor survival, including some that are shared across all subtypes, such as SPRY1, and other lineage-specific super-enhancers, like SOX2 in brain-derived rhabdoid tumors. Taken together, our findings reveal a novel chromatin-based epigenetic mechanism underlying the tumor suppressive activity of SMARCB1.

Human cancer genome sequencing has recently revealed that genes encoding subunits of SWI/SNF chromatin remodeling complexes are frequently mutated across a wide variety of cancers, and several subunits of the complex have been shown to have bona fide tumor suppressor activity1. However, whether mutations in SWI/SNF subunits result in shared dependencies is unknown. Here we show that EZH2, a catalytic subunit of the Polycomb repressive complex 2 (PRC2), is essential in all tested cancer cell lines and xenografts harboring mutations of the SWI/SNF subunits ARID1A, PBRM1, and SMARCA4, which are several of the most frequently mutated SWI/SNF subunits in human cancer but that co–occurrence of a Ras pathway mutation correlates with abrogation of this dependence. Surprisingly, we demonstrate that SWI/SNF mutant cancer cells are primarily dependent upon a non–catalytic role of EZH2 in stabilization of the PRC2 complex, and only partially dependent on EZH2 histone methyltransferase activity. These results not only reveal a shared dependency of cancers with genetic alterations in SWI/SNF subunits, but also suggest that EZH2 enzymatic inhibitors now in clinical development may not fully suppress the oncogenic activity of EZH2.