Identification of prognostic and metastasis-related alternative splicing signatures in hepatocellular carcinoma.

Runzhi Huang,Dianwen Song,Peng Hu,Tong Meng,Aiqing Xie,Hanlin Sun,Penghui Yan,Jie Zhang,Siqiao Wang,Huabin Yin,Zongqiang Huang,Juanwei Zhuang,Rui Kong,Gaili Yan

doi:10.1042/bsr20201001

Abstract

As the most common neoplasm in digestive system, hepatocellular carcinoma (HCC) is one of the most important leading cause of cancer deaths worldwide. Its high-frequency metastasis and relapse rate lead to the poor survival of HCC patients. However, the mechanism of HCC metastasis is still unclear. Alternative splicing events (ASEs) have a great effect in cancer development, progression and metastasis. We downloaded RNA sequencing and seven types of ASEs data of HCC samples, in order to explore the mechanism of ASEs underlying tumorigenesis and metastasis of HCC. The data were taken from the The Cancer Genome Atlas (TCGA) and TCGASpliceSeq databases. Univariate Cox regression analysis was used to determine a total of 3197 overall survival-related ASEs (OS-SEs). And based on five OS-SEs screened by Lasso regression, we constructed a prediction model with the Area Under Curve of 0.765. With a good reliability of the model, the risk score was also proved to be an independent predictor. Among identified 390 candidate SFs, Y-box protein 3 (YBX3) was significantly correlated with OS and metastasis. Among 177 ASEs, ATP-binding cassette subfamily A member 6 (ABCA6)-43162-AT and PLIN5-46808-AT were identified both associated with OS, bone metastasis and co-expressed with SFs. Then we identified primary bile acid biosynthesis as survival-related (KEGG) pathway by Gene Set Variation Analysis (GSVA) and univariate regression analysis, which was correlated with ABCA6-43162-AT and PLIN5-46808-AT. Finally, we proposed that ABCA6-43162-AT and PLIN5-46808-AT may contribute to HCC poor prognosis and metastasis under the regulation of aberrant YBX3 through the pathway of primary bile acid biosynthesis.

Highlights

Hepatocellular carcinoma (HCC) is the most common primary liver tumor and one of the most important lethal malignancies worldwide [1]
AS event (ASE) data of 377 HCC samples were collected from TCGASpliceSeq and we identified a total of 34163 ASEs in 8985 genes among the HCC cases: 12327 ES events in 5343 genes, 8087 alternative terminator (AT) events in 3532 genes, 6352 alternative promoter (AP) events in 2566 genes, 2666 AA events in 1937 genes, 2331 alternative donor (AD) events in 1663 genes, 2263 reserved intron (RI) events in 1561 genes and 137 mutually exclusive exon (ME) events in 135 genes
We totally identified 3197 HCC overall survival-related ASE (OS-SE) by univariate Cox regression analysis and we constructed a prognosis model based on the eight OS-SEs filtered by the Lasso regression

Summary

Introduction

Hepatocellular carcinoma (HCC) is the most common primary liver tumor and one of the most important lethal malignancies worldwide [1]. Optional treatments for liver cancer are very limited and the prognosis is generally unsatisfactory [5,6]. The highly metastatic capability and recurrence rate of HCC result in the low survival rate of HCC patients. HCC has become a severe public health problem worldwide [7]. Many advances have been achieved in the diagnosis and treatment of HCC, the overall prognosis for HCC patients remains poor. Tumor metastasis contributes greatly to the poor prognosis of HCC patients. Current therapeutic strategies for HCC have been demonstrated to promote HCC metastasis instead of repressing it [8].

Methods

Results

Conclusion