Fig. 2: The loss of ASPSCR1::TFE3 expression modulates distribution of super-enhancers (SEs).

a The genomic distribution of ASPSCR1::TFE3 6523 and 3412 bound sites in mouse and human ASPS, respectively, showing predominant binding in distal regions. Common bound sites in two biological replicates were analyzed. b De novo motif enrichment analysis of ASPSCR1::TFE3 binding regions in mouse (left) and human (right) ASPS. The top three motifs identified are shown. Motif enrichment and p-value was calculated using Fischer’s exact test. c Venn diagrams showing the number of common binding regions between ASPSCR1::TFE3 and H3K27ac in mouse and human ASPS. d Composite plots showing a significant reduction in H3K27ac signals in the absence of ASPSCR1::TFE3 in both mouse and human ASPS (left). Heat maps showing ASPSCR1::TFE3, H3K27ac, H3K4me3, and H3K27me3 signals in murine ASPS17 and human ASPS-KY cells. Reduction in H3K27ac signals is observed in both cells (right). e Enhancers are ranked by increasing H3K27ac signals in ASPS17 and ASPS null cells. Using the ROSE algorithm, 527 and 456 enhancers were defined as SEs in ASPS17 and null cells, respectively. f Venn diagram showing overlapping and distinct SEs (left). Enrichment of Gene Ontology biological process for 258 ASPS17-specific SEs, showing inclusion of angiogenesis pathways in red (right). p-value was calculated using a binominal test. g ChIP-seq track at Pdgfb, Vwf, Rab27a, Sytl2, and Myh9 genomic loci, showing the association between ASPSCR1::TFE3 and H3K27ac binding in ASPS17 cells. Significant loss of H3K27ac signals and/or SEs in ASPS null cells are exhibited. The Myh9 genomic locus is shown as an example of SEs unaffected by ASPSCR1::TFE3 loss.