analyzed and interpreted data

analyzed and interpreted data. Levels of human -fetoprotein (AFP) were monitored in the serum of animals. Immunohistochemical and gene expression analyses were also completed on xenograft tumor samples. BLI signal indicative of tumor growth was seen in 55% of HepG2- and Huh-6-injected animals after a period of four to seven weeks. Increased AFP levels correlated with tumor growth. MRI showed large MI-136 intrahepatic tumors with active neovascularization. HepG2 and Huh-6 xenografts showed expression of -catenin, AFP, and Glypican-3 (GPC3). HepG2 samples displayed a consistent gene expression profile most similar to human HB tumors. Intrahepatic injection of HB cell lines leads to liver tumors in mice with growth patterns and biologic, histologic, and genetic features similar to human HB tumors. This orthotopic xenograft mouse model will enable clinically relevant testing of novel brokers for HB. Introduction Hepatoblastoma (HB) is the most common malignant liver tumor seen in children1. The disease is usually most often diagnosed in patients under five years of age and is usually sporadic but can also be associated HDAC9 with familial adenomatous polyposis, Beckwith-Wiedemann syndrome, or prematurity2. Five-year overall survival (OS) of patients with stage I and II disease is usually above 95%, but patients with stage IV disease have a five-year OS rate of about 40%3. Standard treatment for HB consists of medical procedures and high dose, non-targeted chemotherapy, which leads to multiple damaging and long term side effects, including ototoxicity and cardiotoxicity4C6. Thus, new treatment strategies are needed, especially for high-risk patients. To date, HB research includes studies with hydrodynamic injection of oncogenes for liver specific expression7, as well as subcutaneous and intrasplenic murine xenograft models8C10. Unfortunately, these models do not recapitulate the disease seen in a majority of patients, which is a large primary tumor encompassing one to four segments of the liver3. Mice with tumors generated with hydrodynamic injection develop multifocal nodules within the liver, and the organ is usually eventually entirely replaced by tumor. This may be representative of patients that present with tumor in all four segments of the liver, but this is only a small percentage of patients3. With the subcutaneous and intrasplenic xenograft models, tumors can be quickly generated in genetically identical animals from the human HB cell lines Huh-611, HepT18, and HepG212. In the subcutaneous model, injection of all three cell lines led to growth of tumors, depending on the strain of mice and time elapsed since injection MI-136 of cells8,9. In the intrasplenic model, immunodeficient mice were directly injected with HepG2, Huh-6, or HepT1 cells into the spleen. The Huh-6 and HepT1 tumor cells, but not HepG2 cells, then migrated to the liver, giving rise to intrahepatic tumors9,10. Of note, animals that MI-136 underwent splenectomy just after injection more readily designed intrahepatic tumors10. These tumors were small, multifocal nodules that again do not represent the disease typically seen in children. Notably, there is one published study of injection of HepG2 cells into the portal vein to generate intrahepatic tumors, but the focus in this work is usually use of this model for drug testing for hepatocellular carcinoma (HCC)13. Thus, although these models have contributed to the field, none truly recapitulates the disease. For effective preclinical studies to be performed, a true intrahepatic orthotopic xenograft model that accurately replicates the human disease is essential. We have successfully developed an intrahepatic patient-derived xenograft (PDX) model of HB using patient specimens14. Other groups have also examined subcutaneous and intrahepatic growth of patient-derived liver malignancy tissues as models of HCC, including an interesting study in which tumors composed of sorted human liver malignancy stem cells (hLCSCs) were produced subcutaneously15,16. Since these tissues have limited availability due to the rarity of the disease, we wanted to develop and characterize an intrahepatic, orthotopic xenograft model using commercially available HB cell lines. In addition, cell line derived xenograft models can be better standardized and are not dependent on tissue quality of surgical samples that usually have MI-136 been exposed MI-136 to chemotherapy. In this paper, we describe the development and characterization of such an intrahepatic xenograft HB mouse model. Human HB cells were injected into.