Chromatin remodellers as therapeutic targets

Abstract

Large-scale cancer genome sequencing studies have revealed that chromatin regulators are frequently mutated in cancer. In particular, more than 20% of cancers harbour mutations in genes that encode subunits of SWI/SNF (BAF) chromatin remodelling complexes. Additional links of SWI/SNF complexes to disease have emerged with the findings that some oncogenes drive transformation by co-opting SWI/SNF function and that germline mutations in select SWI/SNF subunits are the basis of several neurodevelopmental disorders. Other chromatin remodellers, including members of the ISWI, CHD and INO80/SWR complexes, have also been linked to cancer and developmental disorders. Consequently, therapeutic manipulation of SWI/SNF and other remodelling complexes has become of great interest, and drugs that target SWI/SNF subunits have entered clinical trials. Genome-wide perturbation screens in cancer cell lines with SWI/SNF mutations have identified additional synthetic lethal targets and led to further compounds in clinical trials, including one that has progressed to FDA approval. Here, we review the progress in understanding the structure and function of SWI/SNF and other chromatin remodelling complexes, mechanisms by which SWI/SNF mutations cause cancer and neurological diseases, vulnerabilities that arise because of these mutations and efforts to target SWI/SNF complexes and synthetic lethal targets for therapeutic benefit.

Figure 1:. Structure and function of SWI/SNF complexes

A, SWI/SNF complexes localize to enhancers and promoters and hydrolyze ATP to move or eject nucleosomes, generating accessibility for transcription factor (TF) binding. B, The three SWI/SNF subfamilies, canonical BAF (CBAF), polybromo-associated BAF (PBAF), and non-canonical BAF (ncBAF/GBAF), are composed of shared and subfamily-specific subunits. C, Illustration of CBAF, PBAF, and ncBAF/GBAF complexes bound to nucleosomes with subfamily-specific subunits in blue, red, and green, respectively. SMARCA2/4 include catalytic ATPase domains that facilitate remodeling. CBAF and PBAF bilaterally engage nucleosomes through SMARCA2/4 and SMARCB1, but it is not known if ncBAF can bilaterally engage nucleosomes without SMARCB1 as cryo-EM structures of ncBAF have not been reported. SWI/SNF subunits include reader domains, including bromodomains, PHD finger domains, and chromodomains, that can recognize post-translational modifications on histone tails. Flexible features of SWI/SNF complexes and histone tails have not been resolved in existing structural models. SWI/SNF: SWItch/Sucrose Non-Fermentable; ATP: adenosine triphosphate; ADP: adenosine triphosphate; TF: transcription factor; BAF: BRG-/BRM-associated factor; SMARC: SWI/SNF related, Matrix associated, Actin dependent Regulator of Chromatin; Cryo-EM: cryogenic electron microscopy; PHD: plant homeodomain

Figure 2:. SWI/SNF mutations in cancer

A, Frequency of mutations in SWI/SNF subunits across all cancers in the TCGA PanCancer Atlas (n=10967). Shared SWI/SNF subunits are shown in gray, and CBAF-, PBAF-, and ncBAF-specific subunits are shown in blue, red, and green, respectively. B, Frequency of mutations in SWI/SNF subunits in select cancers SWI/SNF: SWItch/Sucrose Non-Fermentable; TCGA: The Cancer Genome Atlas program; SCCOHT: small cell carcinoma of the ovary hypercalcemic type

Figure 3:

Goldilocks phenomenon: Targeting SWI/SNF defects via rescue or further impairment Cancer-associated mutations in SWI/SNF subunits typically impair (but do not outright eliminate) chromatin remodeling and transcriptional activation, leading to a functional profile that is “just right” for tumorigenesis in select cell populations. Studies have revealed that cancer cells with SWI/SNF mutations are sensitive to both rescue or further impairment of chromatin regulation. Chromatin state can be rescued by enhancing the activity of residual complexes or inhibiting antagonistic chromatin regulators, such as the Polycomb Repressive Complex, and often leads to differentiation^,,. Chromatin state can be impaired further by targeting residual SWI/SNF complexes, and more frequently leads to cell cycle arrest or cell death^,. SWI/SNF: SWItch/Sucrose Non-Fermentable

Figure 4:

Strategies to inhibit SWI/SNF complexes A, Small molecules that bind to bromodomains of BRD9 (I-BRD9) and SMARCA2/4 (PFI-3) were developed in an attempt to block chromatin localization of SWI/SNF complexes. These inhibitors were ineffective, presumably because SWI/SNF complexes contain multiple reader domains that remain functional in the presence of the compound. B, Small molecules that inhibit the ATPase activity of SMARCA2/4 (BRM014, FHD-286, FHT-1015) catalytically inactivate SWI/SNF complexes. C, PROTACs that use ligands recognizing bromodomains or ATPase domains degrade SMARCA2, SMARCA4, PBRM1, and BRD9 with varying efficacy to inactivate SWI/SNF complexes. D, Small molecules that block interactions between transcription factors and SWI/SNF complexes have been proposed in cancers driven by oncogenic transcription factors. *Denotes drugs that are currently in clinical trials. SWI/SNF: SWItch/Sucrose Non-Fermentable; BRD9: bromodomain-containing protein 9; SMARC: SWI/SNF related, Matrix associated, Actin dependent Regulator of Chromatin; PROTAC: proteolysis targeting chimera; TF: transcription factor