Endocrine Abstracts

Metabolic-associated fatty liver disease (MAFLD) is a growing cause of hepatocellular carcinoma (HCC), but the molecular mechanisms associated with the pathological progression from MAFLD to HCC are still to be fully elucidated. The genomic and transcriptomic profile of HCC samples have been widely described; however, the proteomic landscape of MAFLD-derived HCC samples is mostly unknown. Here, we sought to perform the first quantitative proteomic analysis of HCC samples from different aetiologies using a representative, well-characterized cohort of patients. To that end, cytosolic and nuclear proteome of liver tissues from HCC patients (n=42; HCC vs adjacent tissue) and healthy controls (n=5) were determined by SWATH-MS-based proteomics and the results were analyzed using different bioinformatics tools. Non-targeted proteomics revealed the dysregulation of the cytosolic (n=507 proteins) and nuclear (n=925 proteins) tumor proteomes. Enrichment analysis revealed altered cellular functions such as mRNA processing, including mRNA splicing. Indeed, the dysregulation of the splicing machinery was confirmed in two retrospective human cohorts [cohort 1: HCC vs adjacent (n=93); cohort 2: HCC vs adjacent (n=58), cirrhosis (n=39), and healthy livers (n=5)] and in different in silico HCC cohorts. Specifically, from the 94 splicing factors identified in the proteomic analysis, 31 were validated in the retrospective and in silico cohorts, and were associated with key clinical parameters of aggresiveness (tumoral diameter, survival). Interestingly, some of these splicing factors (ej. KHSRP) were associated to metabolic alterations such as diabetes. When comparing between aetologies, we identified cellular pathways (ej. “nucleotide excision repair”, EIF4 and p70S6K signalling) speciffically dysregulated in MAFLD-derived HCC patients, and a battery of more than 20 splicing factors specifically dysregulated in this aetiology [e.g. SNRPD1 and LSM2, two critical components of the small nucleolar ribonucleoproteins (snRNPs) U1/U2/U4/U5 and U6]. Accordingly, culstering analysis defined a HCC subgroup with lower survival and higher recurrence that was associated to the dysregulation of the splicing machinery and included all patients with MAFLD-derived HCC. Finally, in vitro assays (proliferation, colonies/tumorspheres) were performed in liver cancer cell lines (HepG2, Hep3B, SNU-387) after silencing SNRPD1 and LSM2 with specific siRNAs. Remarkably, in vitro silencing of these splicing factors significantly reduced proliferation and dedifferentiation capacity in cancer cells. Our study demonstrates the usefullness of quantitative proteomics for the identification of tumoral subgroups in HCC and the potential of the splicing machinery as a tool for HCC management, specially in less described aetiologies, such as MAFLD.

Fundings: ISCIII (PI20/01301), JdA (PEMP-0036-2020, BIO-0139), FSEEN and CIBERobn.