Fig 2: BRD9 defines a unique SWI/SNF subcomplex. a Immunoprecipitation-mass spectrometry of BRD9 showing known SWI/SNF subunits from four RT (SMARCB1-Mutant) cell lines and three non-RT (SMARCB1-WT) cell lines. Source data are provided as a Source Data file. b Immunoprecipitation of BRD9 in RT cell line G401 cells and blotted with indicated SWI/SNF subunits. c Immunoprecipitation-mass spectrometry of BRD9, ARID1A, PBRM1 in RT cell line G401 without or with SMARCB1 re-expression, showing differential IP enrichment of BRD9, BAF, and PBAF complex subunits. d Model of BRD9-SWI/SNF subcomplex in comparison to BAF and PBAF complexes, color for each subunit is consistent with what described in Fig. 2c. e, f Glycerol gradient of SWI/SNF complexes in G401 cells without (top half) or with SMARCB1 addback (bottom half) (e); or in HEK293T parental (top half) or isogenic SMARCB1 KO cells (bottom half) (f). BRD9 complexes are slightly smaller than BAF and PBAF complexes and there is a shift of core subunits (SMARCC1, SMARCA4) from BAF complexes to BRD9 complexes in the absence of SMARCB1

Fig 3: SS18-SSX1 and BRD9 co-bind the synovial sarcoma genome.(A) Genomic tracks showing BRD9 and SS18-SSX1 ChIP-seq signal on the 98 Mb right arm of chromosome eight in HSSYII cells. The chromosome eight ideogram is displayed above the gene tracks with the relevant region highlighted in red. (B) Pie charts representing the distribution of BRD9 and SS18-SSX1 binding sites on the synovial sarcoma genome. (C) Venn diagram overlaps of all identified BRD9 and SS18-SSX1 ChIP-seq peaks in HSSYII cells. (D) Tornado plots showing BRD9, SS18-SSX1, RNAPII and H3K27Ac ChIP-signal ±10 kb of all hg19 gene promoters in HSSYII cells. Promoters are ranked by RNAPII ChIP signal. (E) Tornado plots showing BRD9, SS18-SSX1 and H3K27Ac ChIP-signal ±10 kb of all active enhancers (defined by H327Ac) in HSSYII cells. (F) Tracks showing BRD9 and SS18-SSX1 ChIP-seq occupancy at the indicated genomic loci in HSSYII cells.

Fig 4: Genome-wide binding of BRD9 is highly cell type specific.a Volcano plot from label-free GFP pulldown of GFP-BRD9 HeLa cell nuclear extracts. Bait and its interactors are shown in the upper right corner (SWI/SNF members including BRD9 are shown in red). Statistically, enriched proteins in GFP-BRD9 pulldown were identified by a permutation-based FDR-corrected t-test. Label-free quantification intensity of GFP pull-down relative to control (fold change, x-axis) is plotted against the log2-transformed p-value of t-test (y-axis). BRD9 interacts with proteins previously reported as core member of the BAF-SWI/SNF complex. b K-means clustering analysis of BRD9 ChIP-seq data showing three distinct clusters of BRD9-binding sites with different specificity for U937 and HeLa cells. c Distance of the clusterized BRD9-binding sites to TSS (promoters: =1 kb from TSS, putative enhancers: >1 kb from TSS). d GO analysis of BRD9-binding sites using Genome regions enrichment of annotations tool (GREAT). The most significant GO terms for each of the three clusters are shown

Fig 5: Quantitative live-cell kinetics of VZ185 induced degradation of BRD7 and BRD9. Degradation profile (A), calculation of DC50 values (B), and initial degradation rate (C) across concentration series indicated of VZ185 using continual luminescent reading of CRISPR/Cas9 endogenously tagged HiBiT-BRD7 or HiBiT-BRD9 in HEK293 cells. Error bars are expressed as SEM taken from n = 3 experiments.