Transcriptional profiling analysis and functional prediction of long noncoding RNAs in cancer.

Jiao Yuan,Xin Tong,Youyong Lu,Meiying Zhang,Mingzhou Guo,Jianjun Luo,Dongdong Zhang,Xiaomin Chen,Haiyan Yue,Wei Wu,Runsheng Chen,Tengfei Xiao,Nan Wu,Rui Xing,Xiaowei Chen,Jian Zhao,Lihui Liu

doi:10.18632/oncotarget.6993

Abstract

Long noncoding RNAs (lncRNAs), which are noncoding RNAs (ncRNAs) with length more than 200 nucleotides (nt), have been demonstrated to be involved in various types of cancer. Consequently, it has been frequently discussed that lncRNAs with aberrant expression in cancer serve as potential diagnostic biomarkers and therapeutic targets. However, one major challenge of developing cancer biomarkers is tumor heterogeneity which means that tumor cells show different cellular morphology, metastatic potential as well as gene expression. In this study, a custom designed microarray platform covering both mRNAs and lncRNAs was applied to tumor tissues of gastric, colon, liver and lung. 316 and 157 differentially expressed (DE-) protein coding genes and lncRNAs common to these four types of cancer were identified respectively. Besides, the functional roles of common DE-lncRNAs were inferred based on their expression and genomic position correlation with mRNAs. Moreover, mRNAs and lncRNAs with tissue specificity were also identified, suggesting their particular roles with regard to specific biogenesis and functions of different organs. Based on the large-scale survey of mRNAs and lncRNAs in four types of cancer, this study may offer new biomarkers common or specific for various types of cancer.

Highlights

Cancer has been a major health problem worldwide, with an estimate of more than 4,500 new cases each day in 2014 [1]
Identification of protein coding genes (PCGs) and Long noncoding RNAs (lncRNAs) as candidate common biomarkers for cancer The global expression profile of both PCGs and lncRNAs in four types of cancer tissues and adjacent non-cancerous tissues were examined by a custom microarray platform
In order to validate the alterations of PCG and lncRNA expression obtained from microarray data, we validated a subset of them across a panel of external samples by quantitative realtime PCR (qRT-PCR) (Figure 2A; the validation in gastric cancer was absent due to lack of additional samples)

Summary

Introduction

Cancer has been a major health problem worldwide, with an estimate of more than 4,500 new cases each day in 2014 [1]. The development of high-throughput profiling technology has enabled characterization of cancer cells from perspective of genome, epigenome and transcriptome [6, 7]. Several studies have succeeded in identifying tumor biomarkers for cancer detection, diagnosis or prognosis determination for specific types of cancer, such as estrogen receptor and progesterone receptor in breast cancer [8] and prostate-specific antigen in prostate cancer [9]. High heterogeneity was observed between the transcriptomic landscape of distinct www.impactjournals.com/oncotarget types of cancer [10], cancer cells share characteristics such as dys-regulated cell growth and potential to invade compared to normal cells [11]. Biomarkers might be either specific to a particular type of cancer or general to multiple types of cancer

Methods

Results

Discussion

Conclusion