Evolution of fibrosis, antifibrotic therapies and mechanisms of iron homeostasis in the Mdr2-/- mouse model of hepatic fibrogenesis

Abstract

Liver fibrosis is the excessive accumulation of fibrous scar tissue throughout the liver in response to chronic injury. Its most advanced form, cirrhosis, is among the top ten causes of death in the Western world and is an increasing burden on our healthcare system. For many human liver diseases, particularly cholestasis, there is no effective therapy available leaving transplantation as the only option for many patients. The increasing shortage of donor organs highlights the need for antifibrotic therapies which can inhibit or reverse hepatic fibrosis. Unfortunately, while many agents have shown excellent antifibrotic activity in animal models, there has been little translation to clinical efficacy. Regardless of disease aetiology, many of the same pathways contribute to fibrogenesis including activation of hepatic stellate cells and portal fibroblasts, the ductular reaction, inflammation and matrix remodelling. An important stimulus for the hepatic fibrogenic process is excess iron, as is seen in genetic haemochromatosis. Iron catalyzes the formation of free radicals leading to peroxidative damage to cell membranes and stimulation of inflammatory and fibrogenic responses. Alterations in iron homeostasis have also been described in advanced liver injury independent of genetic mutations associated with hereditary haemochromatosis. Hepatic iron overload is common in advanced fibrosis and cirrhosis associated with hepatocellular injury and has been postulated to be due to a variety of factors including decreased hepcidin expression associated with declining hepatic synthetic function. Hepatic iron deposition is however rarely seen in biliary cirrhosis despite similar degrees of hepatic dysfunction. The multidrug-resistance 2 (Mdr2-/-) null mouse is a recently described model of biliary liver injury and fibrosis. It is the genetic equivalent of progressive familial intrahepatic cholestasis type 3 (PFIC3) and histologically resembles primary sclerosing cholangitis (PSC). These mice lack the Mdr2 canalicular flippase and consequently secrete phospholipid-free bile which leads to portal inflammation and significant fibrosis. The resemblance to human disease and the consistency and reproducibility of this model make it ideal for studies of the mechanisms of hepatic fibrosis and for testing potential antifibrotic agents. This thesis characterizes the evolution of fibrosis in Mdr2-/- mice up to 16 weeks of age and uses this model to test the efficacy of potential antifibrotic agents and to investigate iron homeostasis in cholestasis. The evolution of hepatic fibrosis in Mdr2-/- mice was studied from 3 to 16 weeks of age. Hepatic hydroxyproline increased with age in Mdr2-/- mice until it reached a plateau from 12 to 16 weeks. In contrast, Metavir fibrosis stage, the ductular reaction, portal inflammation and fibrogenic gene expression reached maximal levels at 5 to 8 weeks of age and then decreased significantly at 12 and 16 weeks. Mdr2-/- mice were administered mTOR inhibitors (rapamycin or everolimus) or renin-angiotensin system inhibitors (captopril or irbesartan) to test the antifibrotic activity of these agents in this model in both early and advanced fibrosis. Despite significant antifibrotic activity in other animal models, treatment did not result in a decrease in hepatic hydroxyproline content or fibrosis in Mdr2-/- mice. High dose rapamycin or combination therapy with rapamycin and irbesartan in 3 week old Mdr2-/- mice likewise did not reduce hepatic hydroxyproline content, fibrosis stage or fibrogenic gene expression. Three week old Mdr2-/- mice were also administered a sub-chelating dose of deferasirox. Long term therapy with deferasirox has been shown to reduce hepatic fibrosis in patients with transfusional iron overload independent of its iron chelating properties however no antifibrotic activity was observed in Mdr2-/- mice. While the reasons for the lack of efficacy of these agents in the Mdr2-/- mouse model remain unexplained, this study highlights the importance of testing potential agents in a variety of animal models (particularly those which resemble human pathology), the need to use physiologically relevant doses and the importance of testing at various stages of injury. Mdr2-/- mice had reduced hepatic iron stores when compared to age-matched controls despite lower hepatic hepcidin expression, increased duodenal iron absorption and serum iron levels and increased hepatic transferrin receptor 1 expression. To understand the mechanisms of this aberrant phenotype, Mdr2-/- mice were challenged with either 1% carbonyl iron or an iron deficient diet. Mdr2-/- mice fed a 1% carbonyl iron diet were resistant to hepatic iron accumulation suggesting impaired hepatocyte iron uptake in this model. This novel observation may explain the relative paucity of iron in cholestatic injury compared to hepatocellular disease. The clinical implications of impaired hepatocyte iron uptake in cholestasis are not clarified but these patients may be susceptible to iron deficiency due to reduced mobilizable iron stores. The findings of this thesis have extended our understanding of the pathogenesis of cholestatic liver injury and may have important implications for the clinical management of patients with biliary disease.