Comprehensive analysis of full-length transcripts reveals novel splicing abnormalities and oncogenic transcripts in liver cancer

概要

Hepatocellular carcinoma (HCC) is the third-leading cause of death worldwide and the seventh most common form of cancer [1]. Common etiological factors in liver carcinogenesis
include infection by hepatitis B virus (HBV) or hepatitis C virus (HCV), but other factors,
such as alcohol intake, metabolic diseases, and exposure to specific carcinogens also play significant roles [2]. These factors cause liver inflammation, leading to cirrhosis and result in
malignant transformations in hepatocytes [2]. To elucidate the molecular mechanisms underlying liver carcinogenesis, genetic and transcriptional aberrations have been investigated [3–
7]. These previous studies identified numerous somatic mutations and differentially expressed
genes (DEGs), and have led to the discovery of liver cancer-associated pathways, such as apoptosis, Wnt signaling, chromatin remodeling, and lengthening telomeres [3–7]. In HBVinfected liver cancers, HBV integrations and HBV-human fusion transcripts have been identified [8]. Fusion genes have been investigated, but recurrent fusions were rare [9]. Although
these studies have expanded our knowledge of HCCs, our understanding of liver carcinogenesis is far from complete [10,11]. To obtain deeper insight into the mechanism responsible for
HCCs, applications of novel technologies and analysis of new aspects of genetic and transcriptional aberrations are needed.
We aimed to analyze full-length transcripts for HCCs in the present study. The majority of
previous transcriptome studies have used microarrays or short-reads sequencing technologies,
and therefore lacked information on full-length transcripts that may assist in the detection of
splicing variants expressed from each gene. Protein sequences are specified by the mRNA
sequences of transcripts; thus, the direct observation of transcripts should provide essential
information about carcinogenesis. Indeed, several recent studies have suggested that transcript-specific functions contribute to carcinogenesis in breast and ovarian cancers [12–14]. In
HCCs, a recent study analyzed short-reads RNA-seq data and detected oncogenic splicing
changes in the AFMID gene [15]. These splicing changes were estimated to occur at the early
stage of HCC carcinogenesis and were associated with patient survival [15]. Overall, previous
studies strongly suggest that splicing variants have important roles that are not recognized
without an analysis of full-length transcripts.
One promising approach to observe transcripts is RNA-seq using long-reads sequencers
[16], but their application in cancer research is still limited [17–21] for reasons such as high
error rates in the long-reads sequencing [22] and the complexity or heterogeneity of the cancer
genome. ...

この論文で使われている画像