Oncoviruses can integrate into the host genome and cause tumorigenesis. In particular, hepatitis B virus (HBV) infection accounts for more than 50% of hepatocellular carcinoma (HCC) worldwide. We revealed the global geographical disparity of HBV integration that the landscape of HBV integration between HCC tumor and non-tumorous liver varied in regional cohorts, suggesting the different degrees of clonal enrichment. Most HBV integrations were positionally enriched at telomeres and centromeres (T&C) and they highlighted the novel co-involvement of HBV integration, which likely introduces genomic instability in HCC development. This was confirmed by phospho-H2AX staining. We constructed a large meta-cohort of multiple ethnicities to refine the landscape of HBV integration. This enables the gene set/family level exploration. As TERT is the most frequently integrated gene, we further investigated the underlying mechanistic modulation of TERT transcription activation and revealed the concurrent influence by the orientation and relative distance of HBV integration. Additionally, clonal disparity of HBV integration was observed among patients and the higher level of clonal disparity score can indicate poor patients’ prognostication. Taken together, our study uncovered the different levels of clonal enrichment of HBV integration, mechanistic insights, and prognostic biomarker signature, to strengthen our understanding in HBV-associated hepatocarcinogenesis. Export
Background & Aims: Research on metabolic reprogramming in hepatocellular carcinoma (HCC) has increased; however, studies on the metabolism of glycerolipids or interactions between different pathways remain scarce. Enzymes of glycerol phosphate dehydrogenase (GPD) family, which regulate glycerol-3-phosphate shuttle, link the metabolic processes of glycolysis and glycerolipids. Therefore, we aimed to understand the role and regulation of GPDs in HCC. Approach & Results: We performed transcriptomic analysis on clinical HCC samples from in-house and public cohorts and detected upregulation of glycerol-3-phosphate dehydrogenase 1 like (GPD1L) among GPD family genes in HCC. Further analysis showed that high GPD1L expression was associated with more frequent venous invasion and shorter overall survival. Consistent with these clinical findings, GPD1L knockdown suppressed invasiveness of HCC cells, reduced colony- and sphere-forming abilities, and inhibited stemness gene expression in vitro, while also inhibiting tumor growth and metastasis in vivo. On the other hand, we used mass-spectrometry-based metabolomics to confirm that GPD1L facilitated the biogenesis of the glycerolipid precursor glycerol-3-phosphate (G3P) from dihydroxyacetone phosphate (DHAP). Further untargeted lipidomic analysis revealed that GPD1L supported triacylglycerol synthesis. Additionally, our study identified E74 like E-Twenty-Six transcription factor 1 (ELF1) as a direct activator of GPD1L transcription, binding to the GPD1L promoter to boost its transcription while reducing GPD1L expression when ELF1 levels were lowered. Conclusions: GPD1L is overexpressed in human HCCs and associated with worse clinical outcomes. Aberrant GPD1L expression, driven by ELF1, facilitates conversion of DHAP to G3P conversion to support triacylglycerol synthesis in HCC, promoting tumor growth and metastasis. Export
Mitochondria generate energy to support cells. They are important organelles that engage in key biological pathways. The dysfunction of mitochondria can be linked to hepatocarcinogenesis, which has been actively explored in recent years. To investigate the mitochondrial dysfunction caused by genetic variations, target-panel sequencing is a flexible and promising strategy. However, the copy number of mitochondria generally exceeds nuclear DNA, which raises a concern that uneven target enrichment of mitochondrial DNA (mtDNA) and nuclear DNA (ncDNA) in target-panel sequencing would lead to an undesirably biased representation of them. To resolve this issue, we evaluated the optimal pooling of mtDNA probes and ncDNA probes by a series of dilutions of mtDNA probes in both genomic DNA (gDNA) and cell-free DNA (cfDNA) samples. The evaluation was based on read count, average sequencing depth and coverage of targeted regions. We determined that an mtDNA:ncDNA probe ratio of around 1:10 would offer a good balance of sequencing performance and cost effectiveness. Moreover, we estimated the median physiological mtDNA:ncDNA copy ratio as 38.1 and 2.9 in cfDNA and gDNA samples of non-liver cancer subjects, respectively, whereas they were 20.0 and 2.1 in the liver cancer patients. Taken together, this study revealed the appropriate pooling strategy of mtDNA probes and ncDNA probes in target-panel sequencing and suggested the normal range of physiological variation of the mtDNA:ncDNA copy ratio in non-liver cancer individuals. This can serve as a useful reference for future target-panel sequencing investigations of the mitochondrial genome in liver cancer.