Supplementary Figure 11: Recurrent somatic duplications at the IGF2 locus associated with IGF2 overexpression.

Related to main Figure 5. Recurrent somatic duplications at the IGF2 locus associating with IGF2 overexpression encompass a contact domain boundary and a non-cognate annotated super enhancer in the adjacent contact domain, but do not encompass the known IGF2 cognate enhancer (light blue). H3K27ac peaks show the presence of a non-cognate enhancer in the contact domain adjacent to the IGF2 locus. H2K27me3 marks for WT samples and samples harboring the duplication show absence of repressive chromatin at the adjacent non-cognate enhancer. 4C-Seq experiments using IGF2 as the viewpoint demonstrate a marked physical interaction between the IGF2 locus and the non-cognate enhancer in samples with the recurrent tandem duplication, but not in samples lacking the tandem duplication (WT). By comparison, no physical interaction is seen between IGF2 and its known cognate enhancer, neither in tandem duplication nor in WT samples. 4C-Seq experiments using the non-cognate enhancer as viewpoint verify the marked physical interaction with IGF2 in tandem duplication carriers and not in WT samples. Collectively, these results demonstrate that hijacking of a non-cognate super-enhancer, mediated by a contact domain spanning recurrent SCNA, drive overexpression of the IGF2 locus (see also Figure 5). CTCF marks for WT samples and samples harboring the duplication are consistent with the presence of TAD boundaries normally separating IGF2 and the non-cognate (super) enhancer. (b) IGF2 expression fold change versus copy number ratio (tumor/normal).