Abstract 4375: A molecular carcinogenic approach to modeling hepatocellular carcinoma

The development of accurate disease models is critically important in the ongoing search for better therapies and ultimately cures. Neoplasia manifests through the aberrant development of altered cellular phenotypes. Such phenotypes are characterized by the particular subset of genes expressed. The pathological regulation of gene expression is a function of both cell-innate and tissue-homeostatic responses including inflammation and regeneration. However, little is understood about the actual cellular and biochemical interrelationship between these mechanisms in guiding initiation, promotion, and progression of cancer. The experimental rationale we use in addressing this problem facilitates the interrogation and manipulation of microenvironmental signals (inflammation) and collaborating neoplastic-precursor cell oncogenic lesions resulting in hepatic tumorigenesis. By technical necessity, this approach employs the delivery of disease-specific oncogenes into relevant somatic tissues (hepatocytes). The gene delivery technique used in theses studies, Sleeping Beauty transposition, is highly versatile and efficient, shortening the “gene cocktail to tumor” testing cycle to levels not possible through germ-line transgenesis or any other current in vivo modality. We use this system to first identify collaborating oncogenic- genes/signaling pathways able to initiate hepatic tumorigenesis in vivo. We have found activated AKT, MET, and beta-catenin (CAT) are initiating oncogenes that efficiently induce hepatic tumorigenesis only when co-delivered but not as single agents. Additionally we have been able to establish transplantable hepatocellular carcinoma lines derived by serial transplantation from the AKT/CAT model in vivo. We are currently characterizing and comparing these models to Human hepatocellular carcinoma using QPCR, RNA microarray and immunohistochemical analysis. We will test these individual initiating oncogenes for collaboration with inflammatory microenvironments (hepatotoxins) in the induction of tumorigenesis in order to recapitulate inflammation induced cancer. We are testing these defined tumor initiating models in immunocompromised and transgenic mice to identify the role of host-derived microenvironmental interactions. We are also testing a highly sensitive reporter, Gaussia luciferase, along with the oncoproteins to prospectively monitor tumor growth that otherwise would be undetectable. We are also optimizing highly sensitive PCR based assays