The FATZO mouse, a next generation model of type 2 diabetes, develops NAFLD and NASH when fed a Western diet supplemented with fructose

Non-alcoholic fatty liver disease (NAFLD) is an all-encompassing term used to describe the fatty liver environment in the absence of excessive alcohol consumption. It is estimated that 25% of the world’s general population meet the criteria for a diagnosis of NAFLD [1‐3]. Metabolic disturbances are prominent risk factors for the development of NAFLD [3‐10]. NAFLD with concurrent obesity, insulin resistance and hyperlipidemia is associated with an increased risk of progression of NAFLD to NASH. However, whether NAFLD develops prior to or as a result of metabolic dysregulation is unknown and debated [11‐14].

The initial stage of NAFLD is characterized by the accumulation of ectopic fat in hepatocytes (steatosis). Steatosis is generally a benign, asymptomatic condition; however, with concurrent obesity/metabolic disturbances, steatosis can progress to non-alcoholic steatohepatitis (NASH) and in severe cases hepatocellular carcinoma (HCC) and liver failure [15]. Histologically NASH is characterized by hepatocellular ballooning, inflammation and increased liver fibrosis [16‐18]. In the context of insulin resistance, obesity and dyslipidemia, an inflammatory response is initiated which is thought to be causative in the progression of NAFLD to more severe NASH [3, 6, 7, 19]. Unlike benign steatosis, NASH represents a significant health threat that progresses to fibrosis/cirrhosis in 10–28% of patients [12, 20‐23]. Further progression from NASH to fibrosis/cirrhosis is highly predictive of mortality in these patients [24].

The study of human NAFLD and its progression is hampered by the slow (decades) development of disease as well as tools available for staging the disease [3]. While much research is ongoing to identify non-invasive tools for staging and reliable clinical biomarkers are not yet available, biopsy remains the gold standard. Thus attempts have been made to develop rodent models of fatty liver disease to aid in the investigation of the pathophysiological and morphological findings characteristic of NAFLD. As metabolic syndrome is the most prominent risk factor for the development of NAFLD in humans, the ideal animal model should develop NAFLD in the context of metabolic disease. Furthermore, the model should display histological characteristics such as steatosis, interlobular inflammation, hepatocellular ballooning, fibrosis and be susceptible to liver tumors seen in humans [12].

Over the last several years, investigators have taken different approaches to developing mouse models of NAFLD and NASH; methionine-choline deficient (MCD) diet [25], high fat diets with or without fructose in C57BL/6J and ob/ob mice [26‐29] and the STAM model where 4dayold mice are given streptozotocin plus high fat diet [30, 31]. Initial attention has been placed on producing fibrosis as quickly as possible with MCD diet [27]. The mice on the MCD diet are not obese, actually losing significant body weight (30%), and are not insulin resistant or hyperlipidemia during disease progression [25]. The STAM model is characterized by type 1 diabetes induced with streptozotocin, rather than type 2 diabetes (T2D) on a high fat diet and produces fibrosis after 12 weeks on diet and eventually HCC [30, 31]. Streptozotocin, a known carcinogen, could exacerbate the development of HCC in this model. In C57BL/6 and Lep^ob/Lep^ob models using high fat diets either alone or supplemented with fructose, investigators have observed NAFLD and NASH in the presence of insulin [26, 32‐34].

The FATZO mouse was developed by crossing C57BL/6J and AKR/J mice that have a strong propensity to develop obesity when fed a high fat diet. Selective inbreeding of the offspring resulted in animals that have a strong propensity to develop many of the characteristics of “metabolic syndrome” early in life [35, 36]. Thus FATZO mice have a strong genetic pre-disposition towards obesity and develop insulin resistance, dyslipidemia, and obesity when fed a normal chow diet [35, 36].

Excess consumption of fructose has been shown to promote liver steatosis and fibrosis in humans [37‐39] and normal rodents [40‐42]. Therefore, we hypothesized that the pre-disposition for metabolic disruptions in the FATZO mice in conjunction with feeding of the WDF diet with fructose supplementation would generate a more translational model of NAFLD/NASH with characterization that resembles the progression of human disease. Thus, the aims of the present study were to examine 1) if FATZO mice fed WFD develop NAFLD/NASH similar to the pathologic changes in human; and 2) if obeticholic acid (OCA) treatment, one of the most advanced NASH specific drug in clinical trial, is able to improve liver function as well as morphological changes in the liver of FATZO mice fed WDF.

Currently, the global hypothesis for the pathogenesis of NASH is the “multi-hit hypothesis,” with metabolic syndrome playing a major role, due to insulin resistance followed by pro-inflammatory processes. Unlike monogenic leptin deficient ob/ob or db/db mice, the FATZO mice area polygenic model of obesity and type 2 diabetes when fed regular rodent diet, with an intact leptin pathway, thereby making it more translatable to the human disease [35, 36]. The goal of the present investigation was to determine if the FATZO mice, which inherently develop metabolic syndrome and type 2 diabetes, would develop NAFLD and NASH when fed a western diet supplemented with fructose.

In recent reviews [12, 44], the murine models that most closely resembled the human disease, were those that used high fat diets supplemented with fructose. These diets best simulate the high prevalence of high fat food and corn syrup sweetened beverages in the Western diet. Fructose has been shown to enhance the development of NAFLD and NASH without [29], or with fibrosis in C57BL/6 [26, 28, 34, 45], ob/ob [28, 34] and DIAMOND™ [32] mice on high fat diets.

In the present investigation, FATZO mice fed WDF diet developed NAFLD and NASH with progressive steatosis, ballooning, inflammation and mild fibrosis over 20 weeks when compared to the CD. In the plasma, increases in the liver enzymes, ALT/AST, and cholesterol were observed in FATZO mice on WDF diet for as early as 4 weeks and remained significantly higher compared to the values from the mice on CD over 20 weeks. Plasma triglycerides were not elevated in WDF fed animals when compared to CD, which is consistent with reports in the ob/ob NASH models [28, 34]. However, liver triglycerides were elevated 1.4–2.9 folds at 12, 16 and 20 weeks in mice fed WDF compared to CD. On gross necropsy, the livers from WDF fed mice were pale in color and had significantly higher liver over body weight ratios when compared to the corresponding CD fed group. Histologically, the livers from WDF fed mice presented steatosis as early as 4 weeks on diet which progressed to steatohepatitis characterized by balloon degeneration, lobular inflammation and fibrosis. The composite NAS score in the FATZO mice fed WDF was equivalent to “~ 5” at 16 and 20 weeks on diet; indicative of “definitive” NASH [17, 18]. Mild fibrosis was observed as early as 16 weeks on diet in 50% of WDF fed mice and progressed to 100% of animals demonstrating moderate fibrosis scores of 1.5 at 20 weeks.

Recently, Machado et al. reported that the models with Western diet developed a more common and relatively non-progressive subtype of NASH, whereas MCD diet model developed a less common and more rapidly progressive/aggressive NASH subtype [25]. The main differences between the 2 models are significant in regards to which more closely mimics the human condition. Western diet fed animals are obese, insulin resistant and hyperlipidemia whereas the MCD diet fed animals had weight loss, but are not insulin resistant or hyperlipidemia [25]. FATZO mice on WDF benefit with progressive NASH phenotypes, increased body weight and hyperlipidemia. More recently, when 22%fructose and trans-fat are added to the Western diet, ob/ob mice develop steatosis, lobular inflammation and mild fibrosis similar to FATZO mice on WDF [28, 34]. However, compared to FATZO mice on the WDF, the hepatic ballooning was not present in the ob/ob mice on Western diet via the same length of diet induction [28]. More importantly, the use of ob/ob mice with intrinsic leptin deficiency in the metabolic disease should always be cautious, as the involvement of leptin in multiple metabolic signaling pathways should not be overlooked. As a result, the appearance of NASH phenotypes in the models with gene mutations might not fully reflect the actual disease pathogenesis, which might create concerns for the model to be used in testing anti-NASH therapy. Compared to other rodent NASH models where animals with intrinsic gene mutation (eg: ob/ob mice) or chemical induction, FATZO mice have more physiological relevance to human patients with metabolic syndromes, while maintaining critical biochemical and histopathological changes of phenotypes representative of NASH by applying the common risk factor of inducing NASH such as fructose.

The diet induced obesity (DIO) model in C57BL/6J mice also present liver TG accumulation and hepatic steatoses, however, unlike the FATZO mice, C57BL/6J mice are not diabetic, and develop less severe liver histopathologic changes.

Multiple drugs are in the development stage for treatment of NASH. Obeticholic acid (OCA) is a semi-synthetic bile acid that acts on the nuclear farnesoid X receptor (FXR) which is expressed predominantly in the liver, kidney and intestine to regulate bile acid homeostasis, hepatic lipid metabolism and immune function [46, 47]. It was originally developed for the treatment of primary biliary cholangitis [48] and is currently in the most advanced stage being tested for NASH in several clinical trials with evidence of significant alleviation of plasma liver ALT and AST levels and mild improvement in steatosis, hepatic ballooning, lobular inflammation and fibrosis [49]. In pre-clinical rodent studies, OCA has been shown to reduce hepatic lipid accumulation, liver enzyme activities, steatosis and fibrosis, though the models and dosing regimen selected might largely affect the final manifestation of the drug efficacy [50, 51]. FATZO mice which possess aberrant lipid metabolism develop advanced NASH phenotypes upon WD feeding. The development of the phenotypes was largely accelerated by the impact of fructose on hepatic lipid metabolism including lipogenesis and reduced fatty acid oxidation. The treatment of FATZO mice with OCA is likely to reduce such impact by inhibiting lipogenesis and fatty acid synthesis. Indeed, in the present study in FATZO mice on WFD for 8 weeks, OCA treatment for 15 weeks significantly reduced plasma ALT and AST levels almost to baseline values before WDF induction (Fig. 4e and e). This result is consistent with the report from NASH patients treated with OCA [49]. In addition, treatment of OCA for 8 weeks appeared to be more efficacious in reducing liver enzymes in FATZO fed WDF compared to ob/ob mice fed with AMLN diet, where impacts of OCA on plasma liver enzymes in the latter model were minimal [51]. Moreover, OCA treatment improved hepatic ballooning leading to overall reduction in NAS score and increased the numbers of animals with absence of fibrosis in WDF fed FATZO mice (Fig. 5 and 6). The data suggested that the leading anti-NASH treatment OCA can improve NASH phenotypes in FATZO mice fed with WDF similar as seen in human patients. More importantly, some of the known effects of OCA such as reducing liver enzyme ALT in NASH patients which were not shown in ob/ob mice were also evident in FATZO model, suggesting more clinical relevance of the model. Therefore, FATZO mice fed with WDF develop NAFLD/NASH phenotypes in a time frame that might be suitable for testing anti-NASH drug intervention.

In the present investigation, the polygenic FATZO mouse model of obesity and type 2 diabetes with an intact leptin pathway diet developed progressive NAFLD/NASH similar to humans when fed with WDF. The FATZO WDF model of NAFLD/NASH represents a new and improved scientific tool for the advancement of NAFLD/NASH research which is potentially more translatable to human disease than many other models.

Based on the results of this study, other previously published studies and for a better description of model, the common name of this model will be subsequently changed from FATZO/Pco to MS-NASH/PcoJ and as such will be marketed through Jackson Labs.