UNIVERSITY OF PITTSBURGH THE
The liver performs over 3,000 functions critical to maintaining the organism, ranging from production and secretion of proteins such as albumin, clotting factors and antiproteases, to metabolism and excretion of exogenous compounds such as drugs or toxins as well as endogenous compounds like hormones, bilirubin and bile acids and cholesterol. Liver-based metabolic disease can result from mutations in genes in these critical pathways. Transgenic or knock-out mouse models have been created for many of these diseases, but they do not always faithfully reproduce the human disease. For example mutations in the bile salt export pump (BSEP) results in severe cholestasis and fibrosis in human patients requiring whole organ transplants, yet the knockout mouse is nearly without a phenotype. While many patients with alpha-1-antitrypsin (A1AT) deficiency develop liver fibrosis/cirrhosis many times resulting in liver transplantation, the transgenic mouse model carrying mutant human genes shows an extremely mild phenotype. The mouse model for a deficiency of ornithine transcarbamylase (OTC) activity, the rate limiting step in ammonia metabolism, is a fair model, however the animals tolerate a diet containing normal amounts of protein, while severely affected human patients require severe protein restriction to prevent lethal hyperammonemia. Mice (and rats) are particularly poor models for studying liver fibrosis and cirrhosis, common features of many human liver diseases. Humanized mice may offer a platform to both the study and treatment of hepatic fibrosis and cirrhosis. These are only a few examples of mouse models that do not faithfully recreate the human disease. We propose, the hypothesis, that the best models for human metabolic liver disease are those created from the affected human hepatocytes. Thus, we propose to "humanize" the liver of FRG mice by transplantation of affected human hepatocytes to create authentic models of human metabolic disease. These mice are immunodeficient and also deficient in the tyrosine catabolic enzyme, fumarylacetoacetate hydrolase (Fah -/-) and develop irreversible liver failure if left untreated. However, if Fah-proficient cells are transplanted, they readily and rapidly repopulate the native liver with donor cells, even if the donor cells are of human origin (a). To create these models, the liver of FRG mice will be "humanized" with hepatocytes derived from patients with metabolic disease. In addition, iPSC technology will be utilized to reprogram the liver cells from metabolic disease patients and following hepatic differentiation, additional mice will be humanized with iPS-derived hepatocytes. These humanized mouse models can then be compared to the authentic diseased liver with respect to changes in clinical chemistry (of the patient or animal), the histopathology of the liver and gene and protein expression profiling of liver tissue. In this manner, we will be able to determine if the humanized models developed with these procedures faithfully reproduce the phenotype observed in the patient. In addition to the direct effects on metabolic liver disease, success with these models will facilitate the use of humanized mouse models to investigate other liver based diseases such as Wilson's and Alpha-1-antitrypsin deficiency and may even lead to better-humanized models for primary, acute liver failure and viral, alcoholic and autoimmune hepatitis and even the hepatic stage of malaria, especially if a human immune system were reconstituted in addition to the liver.