Chromatin interactome mapping at 139 independent breast cancer risk signals

Abstract

Genome-wide association studies have identified 196 high confidence independent signals associated with breast cancer susceptibility. Variants within these signals frequently fall in distal regulatory DNA elements that control gene expression. We designed a Capture Hi-C array to enrich for chromatin interactions between the credible causal variants and target genes in six human mammary epithelial and breast cancer cell lines. We show that interacting regions are enriched for open chromatin, histone marks for active enhancers and transcription factors relevant to breast biology. We exploit this comprehensive resource to identify candidate target genes at 139 independent breast cancer risk signals, and explore the functional mechanism underlying altered risk at the 12q24 risk region. Our results demonstrate the power of combining genetics, computational genomics and molecular studies to rationalize the identification of key variants and candidate target genes at breast cancer GWAS signals.The majority of CCVs mapped to non-protein-coding regions of the genome and are enriched at regulatory DNA elements such as enhancers, silencers and insulators 2,5 . It is established that many regulatory elements are located long distances from their target gene promoters, and that regulation of transcription involves direct physical interactions brought about by chromatin looping 6 . Importantly, individual enhancers often loop to and regulate multiple genes, including protein-coding and noncoding RNA genes. Adding to the complexity, enhancers do not necessarily act on the closest promoter but can bypass neighbouring genes to regulate genes located more distally. There is also considerable evidence that most enhancer-promoter interactions occur in cis and within chromatin structures called topologically associating domains (TADs) 7 . TADs are typically several hundred kilobases to a few megabases in size and are relatively stable between cell types and in response to extracellular signals 8,9 .Various chromatin conformation capture (3C)-based methods have been developed to map chromatin contacts at a genome-wide level. The basic principle of 3C involves chromatin fragmentation of formaldehyde-fixed nuclei (usually by restriction digestion), followed by ligation of linked DNA fragments, then detection and quantification of ligation products 10 . One of these methods, Hi-C, is an unbiased but relatively low-resolution approach, that quantifies interactions between all possible DNA fragment pairs in the genome 11 . Hi-C has been used extensively to analyze the three-dimensional organization of genomes, including compartmentalization of chromatin and the position of TADs 12,13 . To increase Hi-C resolution, several groups have developed sequence capture to enrich for chromosomal interactions involving targeted regions of interest [14][15][16][17] .There are several capture methodologies, but typically RNA or DNA oligonucleotide baits are directed to the ends of targeted DNA fragments to enrich for ligation events prior to next gener...

Non-coding RNAs underlie genetic predisposition to breast cancer

“…Therefore, we speculated that CCVs may also act distally through regulatory elements that interact with mencRNA promoters. Using Capture Hi-C data from breast cells [10], we identified 770 mencRNA promoters (defined as ± 500 bp from the transcription start site) that looped to a region containing a CCV (Additional file 2: Tables S11 and S14). For example, at 16q12.2, CCV rs11642015 is an eVariant (p < 5 × 10 −4 ) for the mencRNA XLOC-093918, with the risk haplotype associated with increased XLOC-093918 levels ( Fig.…”

“…We have recently shown that CCVs are enriched in genomic features associated with regulatory activity in a range of breast-derived cell lines including sites of open chromatin, chromatin marks associated with promoter and enhancer activity (H3K4Me3, H3K4Me1, and H3K27Ac), and transcription factor binding sites [9,10]. Capture Hi-C analyses have identified annotated gene promoters that frequently interact with these CCV-containing elements providing a list of candidate risk genes for these signals [10]. However, at 20 signals, we found no evidence of CCVs falling in regions marked by regulatory activity suggesting that some signals alter risk through alternative mechanisms [10].…”

“…Capture Hi-C analyses have identified annotated gene promoters that frequently interact with these CCV-containing elements providing a list of candidate risk genes for these signals [10]. However, at 20 signals, we found no evidence of CCVs falling in regions marked by regulatory activity suggesting that some signals alter risk through alternative mechanisms [10]. Using RNA CaptureSeq, we annotated multiexonic non-coding RNAs (mencRNAs) transcribed from genome intervals surrounding breast cancer risk signals in a range of mammary-derived tissue and cell lines.…”

High-throughput allelic expression imbalance analyses identify candidate breast cancer risk genes

“…At six AEI-identified genes (CASP8, GATAD2A, HSPA4, KLHDC7A, LGR6 and ZC3H11A), breast cancer CCVs are located in the promoter regions, suggesting the CCVs may confer allelic imbalance through altered promoter activity. Moreover, for six genes (BARX2, GATAD2A, HSPA4, LGR6, NTN4 and ZC3H11A), our recent capture Hi-C data has indicated that chromatin looping occurs between a region containing the breast cancer CCVs and the promoter of the corresponding gene 11 .…”

Genome-wide association study identifies 32 novel breast cancer susceptibility loci from overall and subtype-specific analyses

“…Twenty-three target genes in 14 regions were predicted with high confidence (designated "Level 1"), of which 22 target genes in 13 regions were predicted to be distally regulated. These targets include four genes predicted as INQUISIT targets in previous studies 12,13 POLR3C, RNF115, SOX4 and TBX3, a known somatic breast cancer driver gene 14 , and genes implicated by transcriptome-wide association studies (LINC00886 15 and YBEY 16 ).…”