Search

Alternative splicing has been shown to participate in tumor progression, including hepatocellular carcinoma. The poor prognosis of patients with HCC calls for molecular classification and biomarker identification to facilitate precision medicine. We performed ssGSEA analysis to quantify the pathway activity of RNA splicing in three HCC cohorts. Kaplan–Meier and Cox methods were used for survival analysis. GO and GSEA were performed to analyze pathway enrichment. We confirmed that RNA splicing is significantly correlated with prognosis, and identified an alternative splicing-associated protein LUC7L3 as a potential HCC prognostic biomarker. Further bioinformatics analysis revealed that high LUC7L3 expression indicated a more progressive HCC subtype and worse clinical features. Cell proliferation-related pathways were enriched in HCC patients with high LUC7L3 expression. Consistently, we proved that LUC7L3 knockdown could significantly inhibit cell proliferation and suppress the activation of associated signaling pathways in vitro. In this research, the relevance between RNA splicing and HCC patient prognosis was outlined. Our newly identified biomarker LUC7L3 could provide stratification for patient survival and recurrence risk, facilitating early medical intervention before recurrence or disease progression.

Abstract Background Intrahepatic cholangiocarcinoma (iCCA) is a highly heterogeneous and lethal hepatobiliary tumor with few therapeutic strategies. The metabolic reprogramming of tumor cells plays an essential role in the development of tumors, while the metabolic molecular classification of iCCA is largely unknown. Here, we performed an integrated multiomics analysis and metabolic classification to depict differences in metabolic characteristics of iCCA patients, hoping to provide a novel perspective to understand and treat iCCA. Methods We performed integrated multiomics analysis in 116 iCCA samples, including whole‐exome sequencing, bulk RNA‐sequencing and proteome analysis. Based on the non‐negative matrix factorization method and the protein abundance of metabolic genes in human genome‐scale metabolic models, the metabolic subtype of iCCA was determined. Survival and prognostic gene analyses were used to compare overall survival (OS) differences between metabolic subtypes. Cell proliferation analysis, 5‐ethynyl‐2'‐deoxyuridine (EdU) assay, colony formation assay, RNA‐sequencing and Western blotting were performed to investigate the molecular mechanisms of diacylglycerol kinase α (DGKA) in iCCA cells. Results Three metabolic subtypes (S1‐S3) with subtype‐specific biomarkers of iCCA were identified. These metabolic subtypes presented with distinct prognoses, metabolic features, immune microenvironments, and genetic alterations. The S2 subtype with the worst survival showed the activation of some special metabolic processes, immune‐suppressed microenvironment and Kirsten rat sarcoma viral oncogene homolog (KRAS)/AT‐rich interactive domain 1A (ARID1A) mutations. Among the S2 subtype‐specific upregulated proteins, DGKA was further identified as a potential drug target for iCCA, which promoted cell proliferation by enhancing phosphatidic acid (PA) metabolism and activating mitogen‐activated protein kinase (MAPK) signaling. Conclusion Via multiomics analyses, we identified three metabolic subtypes of iCCA, revealing that the S2 subtype exhibited the poorest survival outcomes. We further identified DGKA as a potential target for the S2 subtype.