Circular RNAs and their associations with breast cancer subtypes

// Asha A. Nair 1 , Nifang Niu 2 , Xiaojia Tang 1 , Kevin J. Thompson 1 , Liewei Wang 3 , Jean-Pierre Kocher 1 , Subbaya Subramanian 4 , Krishna R. Kalari 1 1 Department of Health Sciences Research, Mayo Clinic, Rochester, MN, USA 2 Division of Genomic and Molecular Pathology, University of Chicago, Chicago, IL, USA 3 Department of Pharmacology, Mayo Clinic, Rochester, MN, USA 4 Division of Basic and Translational Research, University of Minnesota, Minneapolis, MN, USA Correspondence to: Krishna R. Kalari, email: Kalari.Krishna@mayo.edu Subbaya Subramanian, email: subree@umn.edu Keywords: circular RNA, circ-seq, breast cancer, molecular subtypes, proliferation Received: February 23, 2016 Accepted: October 29, 2016 Published: November 05, 2016 ABSTRACT Circular RNAs (circRNAs) are highly stable forms of non-coding RNAs with diverse biological functions. They are implicated in modulation of gene expression thus affecting various cellular and disease processes. Based on existing bioinformatics approaches, we developed a comprehensive workflow called Circ-Seq to identify and report expressed circRNAs. Circ-Seq also provides informative genomic annotation along circRNA fused junctions thus allowing prioritization of circRNA candidates. We applied Circ-Seq first to RNA-sequence data from breast cancer cell lines and validated one of the large circRNAs identified. Circ-Seq was then applied to a larger cohort of breast cancer samples ( n = 885) provided by The Cancer Genome Atlas (TCGA), including tumors and normal-adjacent tissue samples. Notably, circRNA results reveal that normal-adjacent tissues in estrogen receptor positive (ER+) subtype have relatively higher numbers of circRNAs than tumor samples in TCGA. Similar phenomenon of high circRNA numbers were observed in normal breast-mammary tissues from the Genotype-Tissue Expression (GTEx) project. Finally, we observed that number of circRNAs in normal-adjacent samples of ER+ subtype is inversely correlated to the risk-of-relapse proliferation (ROR-P) score for proliferating genes, suggesting that circRNA frequency may be a marker for cell proliferation in breast cancer. The Circ-Seq workflow will function for both single and multi-threaded compute environments. We believe that Circ-Seq will be a valuable tool to identify circRNAs useful in the diagnosis and treatment of other cancers and complex diseases.