Non-coding RNAs in Various Stages of Liver Disease Leading to Hepatocellular Carcinoma: Differential Expression of miRNAs, piRNAs, lncRNAs, circRNAs, and sno/mt-RNAs

Srinivas V Koduru,Ashley N Leberfinger,Dino J Ravnic,Niraj J Gusani,Arun J Sanyal,Milind Mahajan,Yuka I Kawasawa

doi:10.1038/s41598-018-26360-1

Abstract

Hepatocellular carcinoma (HCC) was the fifth leading cause of cancer death in men and eighth leading cause of death in women in the United States in 2017. In our study, we sought to identify sncRNAs in various stages of development of HCC. We obtained publicly available small RNA-seq data derived from patients with cirrhosis (n = 14), low-grade dysplastic nodules (LGDN, n = 9), high grade dysplastic nodules (HGDN, n = 6), early hepatocellular carcinoma (eHCC, n = 6), and advanced hepatocellular carcinoma (HCC, n = 20), along with healthy liver tissue samples (n = 9). All samples were analyzed for various types of non-coding RNAs using PartekFlow software. We remapped small RNA-seq to miRBase to obtain differential expressions of miRNAs and found 87 in cirrhosis, 106 in LGDN, 59 in HGDN, 80 in eHCC, and 133 in HCC. Pathway analysis of miRNAs obtained from diseased samples compared to normal samples showed signaling pathways in the microRNA dependent EMT, CD44, and others. Additionally, we analyzed the data sets for piRNAs, lncRNAs, circRNAs, and sno/mt-RNAs. We validated the in silico data using human HCC samples with NanoString miRNA global expression. Our results suggest that publically available data is a valuable resource for sncRNA identification in HCC progression (FDR set to <0.05 for all samples) and that a data mining approach is useful for biomarker development.

Highlights

Hepatocellular carcinoma (HCC) was the fifth leading cause of cancer death in men and eighth leading cause of death in women in the United States in 2017
The present study focused on in-depth analysis of small RNA sequencing data obtained from cirrhosis, low-grade dysplastic nodules (LGDN), high-grade dysplastic nodules (HGDN), early stage hepatocellular carcinoma, and advanced stage hepatocellular carcinoma (HCC)
Our analysis found 87 miRNAs differentially expressed in cirrhosis compared to normal liver tissue (15 upregulated and 72 downregulated; Fig. 1A, Table 1), 106 miRNAs in LGDN (46 upregulated and 60 downregulated; Fig. 1B, Table 2), 59 miRNAs in HGDN (18 upregulated and 41 downregulated; Fig. 1C, Table 3), 80 miRNAs in early stage hepatocellular carcinoma (eHCC) (12 upregulated and 68 downregulated; Fig. 1D, Table 4) and 133 miRNAs in HCC (64 upregulated and 69 downregulated; Fig. 1E, Table 5)

Summary

Introduction

Hepatocellular carcinoma (HCC) was the fifth leading cause of cancer death in men and eighth leading cause of death in women in the United States in 2017. We obtained publicly available small RNA-seq data derived from patients with cirrhosis (n = 14), low-grade dysplastic nodules (LGDN, n = 9), high grade dysplastic nodules (HGDN, n = 6), early hepatocellular carcinoma (eHCC, n = 6), and advanced hepatocellular carcinoma (HCC, n = 20), along with healthy liver tissue samples (n = 9). We remapped small RNA-seq to miRBase to obtain differential expressions of miRNAs and found 87 in cirrhosis, 106 in LGDN, 59 in HGDN, 80 in eHCC, and 133 in HCC. We analyzed the data sets for piRNAs, lncRNAs, circRNAs, and sno/mt-RNAs. We validated the in silico data using human HCC samples with NanoString miRNA global expression. Death rates are increasing by 3% every year, with an estimated 28,920 deaths in 20171 These statistics suggest more research needs to be done towards understanding the biology, diagnosis, and prevention. Poorly differentiated HCC has a more aggressive course and poorer outcomes than well differentiated HCC

Objectives

Methods

Results

Conclusion