Abstract 3492: Subtype-specific expression of small non-coding RNAs in breast cancer

Abstract

Abstract We have performed a small RNA-sequencing (RNA-Seq) experiment to find candidate small non-coding RNAs (ncRNAs) in breast cancer (BC) for the identification of signatures that aid in the prediction of therapy responsiveness, prognosis or patient outcome. We used 195 invasive BC tumor and 20 benign breast tissue samples from the Kuopio Breast Cancer Project (KBCP) study material. KBCP material is collected from a genetically homogeneous population of Eastern Finland, and includes fresh-frozen and FFPE breast tissue samples, EDTA blood samples and serum, as well as comprehensive background/lifestyle information, clinical, treatment and follow-up data extending over 20 years. This well-defined material enables the identification of factors associating with BC risk, different stages of disease progression and outcome. Total RNA was extracted from the fresh-frozen BC and benign breast tissue using the Ambion mirVana miRNA Isolation Kit. Small RNA-Seq was performed using the Illumina TruSeq Small RNA Library Prep kit and the Illumina MiSeq next-generation sequencing instrument. Analyses were concentrated in PIWI-interacting RNAs (piRNAs) and other non-miRNA small ncRNAs (sncRNAs). Bioinformatic analysis commenced with quality control and preprocessing steps including read quality assessment (FastQC), adapter trimming (TRIMMOMATIC), and removal of e.g. ribosomal RNA reads (Bowtie). Preprocessed reads were aligned (Tophat) to human reference transcriptomes (piRNABank and GENCODE non-miRNA sncRNAs), followed by data conversions e.g. for visualization (samtools, IGVtools). Gene-wise read counts were collected using summarizeOverlaps R package, and statistically differentially expressed (DE) genes were identified using DESeq2 R package. We first compared the small RNA expression in the benign breast tissue with invasive BC and observed in total 635 sncRNAs and 94 piRNAs with differential expression (PAdj<0.05). Of these 520 sncRNAs and 73 piRNAs were upregulated, and 115 sncRNAs and 21 piRNAs downregulated in the invasive tumors compared to benign breast tissue. When comparing the luminal BC subtype with triple-negative BC (TNBC) we observed 204 DE (PAdj<0.05) sncRNAs and 32 piRNAs. Of these, 103 sncRNAs and 18 piRNAs were upregulated, and 101 sncRNAs and 14 piRNAs downregulated in TNBC compared to luminal tumors. Differential expression was also observed between ER (estrogen receptor) negative and ER positive invasive tumors. Of the 242 DE sncRNAs and 32 piRNAs, 105 sncRNAs and 12 piRNAs were upregulated, and 137 sncRNAs and 20 piRNAs downregulated in ER positive BC compared to ER negative. Preliminary results of hierarchical clustering and principal component analysis of normalized log2 piRNA and sncRNA expression levels revealed the potential for finding differences between BC subtypes. The subsequent statistical analysis enabled us to identify individual piRNAs and other sncRNAs as candidate druggable biomarkers for e.g. TNBC patients. Citation Format: Jaana M. Hartikainen, Sami Heikkinen, Maria Tengström, Veli-Matti Kosma, Arto Mannermaa. Subtype-specific expression of small non-coding RNAs in breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3492. doi:10.1158/1538-7445.AM2017-3492