Non-coding single nucleotide variants affecting estrogen receptor binding and activity

Amir Bahreini, Kevin Levine, Lucas Santana-Santos, Panayiotis V. Benos, Peilu Wang, Courtney Andersen, Steffi Oesterreich*, Adrian V. Lee

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

5 Scopus citations


Background: Estrogen receptor (ER) activity is critical for the development and progression of the majority of breast cancers. It is known that ER is differentially bound to DNA leading to transcriptomic and phenotypic changes in different breast cancer models. We investigated whether single nucleotide variants (SNVs) in ER binding sites (regSNVs) contribute to ER action through changes in the ER cistrome, thereby affecting disease progression. Here we developed a computational pipeline to identify SNVs in ER binding sites using chromatin immunoprecipitation sequencing (ChIP-seq) data from ER+ breast cancer models. Methods: ER ChIP-seq data were downloaded from the Gene Expression Omnibus (GEO). GATK pipeline was used to identify SNVs and the MACS algorithm was employed to call DNA-binding sites. Determination of the potential effect of a given SNV in a binding site was inferred using reimplementation of the is-rSNP algorithm. The Cancer Genome Atlas (TCGA) data were integrated to correlate the regSNVs and gene expression in breast tumors. ChIP and luciferase assays were used to assess the allele-specific binding. Results: Analysis of ER ChIP-seq data from MCF7 cells identified an intronic SNV in the IGF1R gene, rs62022087, predicted to increase ER binding. Functional studies confirmed that ER binds preferentially to rs62022087 versus the wild-type allele. By integrating 43 ER ChIP-seq datasets, multi-omics, and clinical data, we identified 17 regSNVs associated with altered expression of adjacent genes in ER+ disease. Of these, the top candidate was in the promoter of the GSTM1 gene and was associated with higher expression of GSTM1 in breast tumors. Survival analysis of patients with ER+ tumors revealed that higher expression of GSTM1, responsible for detoxifying carcinogens, was correlated with better outcome. Conclusions: In conclusion, we have developed a computational approach that is capable of identifying putative regSNVs in ER ChIP-binding sites. These non-coding variants could potentially regulate target genes and may contribute to clinical prognosis in breast cancer.

Original languageEnglish (US)
Article number128
JournalGenome Medicine
Issue number1
StatePublished - Dec 13 2016


  • Breast cancer
  • DNA binding
  • Estrogen receptor
  • IGF1R
  • Non-coding SNVs

ASJC Scopus subject areas

  • Molecular Medicine
  • Molecular Biology
  • Genetics
  • Genetics(clinical)


Dive into the research topics of 'Non-coding single nucleotide variants affecting estrogen receptor binding and activity'. Together they form a unique fingerprint.

Cite this