47P - Epigenomic analysis of primary breast cancer tumors reveals novel tumor cell vulnerabilities and therapeutic targets

Date 04 May 2017
Event IMPAKT 2017
Session Welcome reception and Poster Walk
Topics Breast Cancer
Presenter Matthew Guenther
Authors M.G. Guenther1, M.W. Chen1, C. Kolodzy1, M. McKeown2, E. Lee2, C. Collins1, D. Orlando3, E. Di Tomaso2, C.C. Fritz4, E.R. Olson5
  • 1Epigenomics And Gene Regulation, Syros Pharmaceuticals, 02139 - Cambridge/US
  • 2Translational Medicine, Syros Pharmaceuticals, 02139 - Cambridge/US
  • 3Computational Biology, Syros Pharmaceuticals, 02139 - Cambridge/US
  • 4Oncology, Syros Pharmaceuticals, 02139 - Cambridge/US
  • 5Discovery, Syros Pharmaceuticals, 02139 - Cambridge/US



To date, a large portion of cancer research has focused on somatic mutations in protein coding regions to identify putative oncogenic drivers. Here, we have investigated the roles of genomic non-coding regions in defining oncogenic cell state drivers and pinpointing novel druggable targets. Abnormally large clusters of cis-acting enhancers, called super-enhancers (SEs), have emerged as regulatory features of oncogenes and other key tumor drivers in cancer cells. Mapping these features through H3K27ac ChIP-seq in primary patient samples and linking them to protein-coding genes provides an inroad to identify novel dependencies and new therapeutic targets in cancer.

We have analyzed 42 primary breast cancer patient samples using H3K27ac ChIP-seq to map enhancers and SEs genome-wide. We find that these SE maps pinpoint known oncogenic drivers and recapitulate established clinical subgroups: Most samples classified as HER2+ contain a SE at the ERBB2 locus, most samples classified as ER+ contain a SE at the ESR1 locus while neither tend to appear in TNBC samples. These findings strengthen the hypothesis that SE analysis can be used to discover breast cancer dependencies de novo, independent of somatic mutations. To validate novel targets that were revealed by SEs in patient samples, we used CRISPR-mediated gene ablation as well as chemical validation in a panel of cell lines that either exhibit or do not exhibit the gene-associated SEs. Using the chemical approach, we identified a SE at the RARA locus that predicts sensitivity to a potent RARα agonist (SY-1425) in a panel of breast cancer cell lines. The sensitivity of these cell lines to SY-1425 is correlated with enhancer size, identifying RARα as an enhancer-correlated vulnerability in breast cancer. We show that this correlation extends to in vivo xenograft models. Using the CRISPR approach, we discovered novel targets that were identified by their association with SEs in primary samples and then validated in a panel of breast cancer cell lines.

Together, these studies indicate that super-enhancer analysis in primary patient samples can be used to define new biomarker-linked breast cancer vulnerabilities for therapeutic intervention.

Clinical trial identification