Comprehensive genomic analysis of hepatoblastoma reveals its genetic heterogeneity and provides novel therapeutic targets

Abstract

Hepatoblastoma (HBL) is the most common pediatric liver cancer, and the prognosis of high-risk cases with clinical poor-prognostic factors remains poor. However, the molecular basis of high-risk HBL is not fully understood. To address this issue, we subjected 59 HBL samples to multiomics analyses, including DNA methylome analysis and RNA sequencing. On the basis of DNA methylation patterns, HBL was classified into three clusters (F, E1, and E2), which exhibited remarkable correlation with clinical, histological, and genetic features. Cluster F was largely composed of cases with fetal histology and good outcome, whereas clusters E1 and E2 largely correspond to cases with embryonal or embryonal/fetal combined histology and poor prognosis. Clusters E1 and E2 had similar genetic profiles and were characterized by the hypermethylation of HNF4A/CEBPA-binding regions followed by fetal liver-like expression patterns, frequent copy-number gains, upregulation of the cell-cycle pathway, and overexpression of NQO1 and ODC1. In particular, ODC1 is the gene encoding the rate-limiting enzyme of the polyamine biosynthesis pathway and associated with a high turnover of tumor cells. In fact, when we performed ODC1 inhibition experiments in vitro, we found that the growth of ODC1-expressing HBL cells was significantly suppressed. From these results, we concluded that ODC1 is a key gene of the aggressive phenotype and a potential therapeutic target of high-risk HBL. The DNA-methylation-based classification of HBL disclosed the molecular basis of HBL, suggesting rational therapeutic strategies for high-risk HBL.