Pivotal role of the Keap1-Nrf2 pathway in the pathogenesis and prevention of non-alcoholic steatohepatitis induced cirrhosis

Non-alcoholic steatohepatitis (NASH) is a common disease that involves fat accumulation in liver cells and inflammation of the liver. Its incidence is increasing rapidly because it is prevalent in individuals with diabetes and those who are obese; it has been estimated that approximately 2.0% of the adult population in the UK may have NASH. This trend is alarming, and has huge cost implications for the National Health Service, because NASH can give rise to liver fibrosis, cirrhosis and cancer, none of which are treatable. It is thought that NASH arises in certain individuals because their livers do not cope with metabolic changes produced by diet, diabetes or the environment. In each of these cases, it is envisaged that oxidative stress represents a common consequence of the physiological and environmental insults that produce liver disease. The activity of a protein known as Nrf2 is emerging as the most important determinant of susceptibility to NASH for three reasons: (i) it keeps fat production and storage in the liver low; (ii) it helps maintain high levels of anti-inflammatory proteins and (iii) it is a vital component in maintaining high levels of natural anti-oxidant agents in the liver.

We hypothesize that Nrf2 normally functions to minimise the likelihood of NASH arising, and that by increasing its activity using drugs will protect 'at risk' individuals of developing NASH and the disease progressing further. To date, experiments that have implicated Nrf2 in determining susceptibility to NASH or prevention of the disease have used a specialized, non-standard diet (deficient in certain key molecules), which is not relevant to the human disease. Unfortunately, no standard animal model exists that recapitulates the entire spectrum of liver pathology (i.e., from steatosis, to NASH, to fibrosis, to cirrhosis and to hepatoma), which we can use to test the role of Nrf2 in this disease. In a pilot study, we have found that feeding a 'Western' high fat (HF) diet to mice for 24 weeks produced mild NASH, whereas feeding the same diet to mice totally deficient of Nrf2 for the same period produced full-blown NASH with fibrosis and cirrhosis. We now wish to confirm these preliminary findings by studying a larger number of animals on the HF diet, and by examining the time-course of the disease using biomarkers (i.e. monitoring levels of certain disease-associated molecules in the blood).

In order to prove whether Nrf2 is important in the development of NASH and its progression to fibrosis and cirrhosis, we first require at least one reliable animal model that reflects the human disease. To this end, we will generate two models, which we anticipate will develop liver disease at different rates: firstly, we will feed normal mice a 'Western' high fat (HF) diet; secondly, we will feed normal mice a 'Western' high fat + fructose (i.e. high levels of sugar; HFF) diet. A detailed time course of the development of fatty liver, NASH, fibrosis and cirrhosis in mice fed these two diets will be established using clinical chemistry, histology and biochemical measurements of pathological changes associated with development of the disease. Once the rate of progression of the disease has been established in the normal mice placed on the 'Western' HF and HFF diets, we will examine the contribution that Nrf2 makes to protection against NASH by testing whether removal of this protein results in an accelerated appearance of NASH, fibrosis and cirrhosis when these mice are fed the same diets. Lastly, we will examine whether activation of Nrf2, either by genetic means or by treatment with a certain type of drug, inhibits the appearance of NASH, fibrosis, or cirrhosis when the animals are fed the same 'Western' HF and HFF diets.

Part 1. We will perform a longditudinal study of C57BL/6 Nrf2+/+ mice in which the development of disease will be monitored by measuring plasma markers of steatosis, insulin sensitivity, inflammation, oxidative stress and fibrosis. Upon sacrifice of the animals, we will perform various histological examinations (H&E, Van Glieson, reticulin, Oil Red O and TUNEL staining), and measure changes in expression of key hepatic genes associated with lipid metabolism, inflammation, redox status and fibrosis. Furthermore, in the first part of the project we will also test whether administration of a 'Western' high fat/high fructose (HFF) diet increases the rate at which NASH, fibrosis and cirrhosis develop in C57BL/6 Nrf2+/+ mice, and examine whether both the HF and HFF diets provide appropriate models for human liver disease. The rate at which NASH developes and progresses in wild-type mice fed the HFF diet will be evaluated using the same measurements employed for mice on the HF diet, and compared with the latter.

Part 2. We will use the same criteria set out above to test the prediction that loss of Nrf2 results in a rapid acceleration in the rate at which NASH, fibrosis and cirrhosis develop in C57BL/6 Nrf2-/- mice placed on the HF and HFF diets.

Part 3. We will test the hypothesis that either genetic or pharmacological upregulation of Nrf2 provides protection against NASH, fibrosis and cirrhosis in mice placed on the HF or HFF diets: genetic upregulation of Nrf2 will be studied in C57BL/6 Keap1-floxed (i.e., Keap1loxP/loxP) mice in which Keap1, an E3 ubiquitin ligase substrate adaptor protein that represses Nrf2, is substantially downregulated; pharmacological upregulation of Nrf2 will be achieved using the potent chemopreventive agent TBE31, which is a strong inhibitor of Keap1. As one control, we will generate Keap1-floxed::Nrf2-/- and place them on the HF and HFF diets. As a second control, we will treat Nrf2-/- mice placed on the HF and HFF diets with TBE31.

This project seeks to determine whether NF-E2-related factor 2 (Nrf2) normally functions to decrease the likelihood of NASH, fibrosis and cirrhosis arising through exposure to a 'Western' diet, namely one that is high in fat and in sugar. Nrf2 is a transcription factor that is critical to the processes through which cells in the body respond to stress and inflammation. The studies described in this application will demonstrate whether in the absence of Nrf2 transgenic mice show accelerated NASH and cirrhosis, and conversely whether NASH and cirrhosis can be prevented through genetic or pharmacological activation of Nrf2.

It is estimated that 20% of the adult UK population have fatty liver as a consequence of the marked increase in obesity and type II diabetes mellitus. Of these, at least 10% (i.e., 2% of the Nation's adult population) are thought to have NASH, a significant percentage of whom will progress on to develop liver fibrosis and cirrhosis and ultimately end-stage liver disease. More alarmingly, the rising prevalence of obesity and type 2 diabetes in childhood in the UK means that NASH is increasingly seen in young adults. This will substantially shorten life quality, life expectancy, and the contribution that affected individuals make to society. At present, there is no effective treatment for fibrosis and cirrhosis other than liver transplantation, and there are clearly only a finite number of suitable donor livers for this purpose. The potential cost implications to the NHS of a rapidly rising incidence of NASH is truly grave.

In this context, it is clear that alternative strategies to prevent NASH are urgently required and would be of great benefit to society. The preliminary studies performed in our laboratory provide strong support for a role for Nrf2 in the prevention of NASH in an animal model of obesity. The current project aims to build on this exciting data and will directly test the hypothesis that Nrf2 markedly influences susceptibility to NASH, and also that up-regulation of Nrf2 inhibits the onset of NASH and its progression to fibrosis and cirrhosis. Most excitingly, a variety of xenobiotics, including the drug dimethyl fumarate (BG-12) and the food preservative butylated hydroxyanisole (BHA, E320), are known to up-regulate Nrf2. Assuming our hypothesis is correct, a clear opportunity then exists to intervene pharmacologically to activate Nrf2 and stimulate expression of its target genes that collectively inhibit hepatic lipid accumulation, inflammation, oxidative stress and apoptosis.

Within our research collaboration grouping, John Dillon and Rory McCrimmon are best placed to translate the findings from our mouse experiments to the clinic because they run Hepatology and diabetes outpatient clinics in Tayside, and through the Dundee Clinical Research Centre, have access to a new state-of-the-art facility for conducting human trails. We envisage that our initial clinical studies will focus on dimethyl fumarate, which is already being used in phase III trials to treat multiple sclerosis, or the food preservative BHA. This means a drug is, almost immediately, available for use in a human clinical trial to prove the therapeutic concept of up-regulation of Nrf2, leading to rapid translation of our experimental findings into human disease.

In summary, the work described herein may result in outcomes that:
(i) Offer novel pharmacological-based treatments for NASH, which can be fast-tracked into small clinical trials, with the strong possibility that better drugs may be rapidly developed in association with Pharmaceutical companies.
(ii) Enhance the quality of life (and may extend life) for a substantial proportion of the adult (young and old) population.
(iii) Improve the effectiveness of NHS care and treatment and reduce NHS costs associated with liver disease treatment and management.