Abstract
Hepatocellular carcinoma (HCC) are highly potent to invade the portal venous system and subsequently develop into the portal vein tumor thrombosis (PVTT). PVTT could induce intrahepatic metastasis, which is closely associated with poor prognosis. A comprehensive systematic characterization of long noncoding RNAs (lncRNAs) associated with HCC metastasis has not been reported. Here, we first assayed 60 clinical samples (matched primary tumor, adjacent normal tissue, and PVTT) from 20 HCC patients using total RNA sequencing. We identified and characterized 8,603 novel lncRNAs from 9.6 billion sequenced reads, indicating specific expression of these lncRNAs in our samples. On the other hand, the expression patterns of 3,212 known and novel recurrently deregulated lncRNAs (in >=20% of our patients) were well correlated with clinical data in a TCGA cohort and published liver cancer data. Some lncRNAs (e.g., RP11-166D19.1/MIR100HG) were shown to be useful as putative biomarkers for prognosis and metastasis. Moreover, matched array data from 60 samples showed that copy number variations (CNVs) and alterations in DNA methylation contributed to the observed recurrent deregulation of 716 lncRNAs. Subsequently, using a coding-noncoding co-expression network, we found that many recurrently deregulated lncRNAs were enriched in clusters of genes related to cell adhesion, immune response, and metabolic processes. Candidate lncRNAs related to metastasis, such as HAND2-AS1, were further validated using RNAi-based loss-of-function assays. The results of our integrative analysis provide a valuable resource regarding functional lncRNAs and novel biomarkers associated with HCC tumorigenesis and metastasis.
