Modeling Liver Cancer Molecular and Histological Features by CRISPR Editing of Porcine Hepatocytes

/Aims: Hepatocellular carcinoma (HCC) is an aggressive disease with limited response to available therapies. Progress in therapeutic development has been hampered by the lack of preclinical models that recapitulate human HCC and support evaluation of multimodal therapies. Here, we employ CRISPR editing to identify alterations that drive hepatocyte transformation and to establish genetically defined HCC models.

Hepatocytes were isolated from pigs (n=7) and subjected to combinations of 2-4 gene alterations, including CRISPR-mediated knockout of tumor suppressor genes (TP53, PTEN, CDKN2A, AXIN1), with or without c-Myc overexpression. Resulting cells were injected subcutaneously into immunodeficient mice to assess their tumorigenicity and histological features. To establish a large-animal model, autologous transformed hepatocytes were implanted into four intrahepatic sites (n=5 pigs). Transcriptomic profiling of cells, xenografts, and porcine tumors was conducted by RNA sequencing, using FDR<0.05 for DEG analysis.

Eight combinations of gene alterations yielded cells expandable in culture, in contrast to unmodified primary hepatocytes. Cells harboring at least three gene alterations (n=5 lines) formed xenograft tumors (n=22 mice), whereas cells with only two gene knockouts did not (n=5 mice). Xenografts with sgTP53; c-Myc^OE; sgPTEN and/or sgCDKN2A recapitulated histological features of human liver cancer, including clear cell, trabecular, and pseudoglandular HCC and cholangiocarcinoma. Transcriptomic analysis of transformed cells and xenograft tumors revealed activation of cell-cycle pathways and similarity to human HCC profiles. Autologous implantation of transformed hepatocytes generated intrahepatic tumors at 7/8 injection sites (87.5%) in two pigs, providing proof-of-concept for an immune-competent HCC model in wild-type pigs.

This study delineates cooperative oncogenic alterations that drive hepatocyte transformation ex vivo and provides cell and animal models that recapitulate key histological and molecular features of human HCC.

HCC has poor response rates to available treatments, underscoring the urgent need for translational preclinical models. Through ex vivo CRISPR editing of clinically relevant genes in porcine hepatocytes, we established cell and animal models that recapitulate key molecular and histological features of human HCC. This platform provides a foundation for future mechanistic studies, genotype-phenotype mapping, and preclinical testing of precision therapies. Moreover, the porcine model's size and physiological similarity to humans make it a promising model for evaluating locoregional therapies and device-based therapies in an immune-competent setting.