Dietary Mouse Models of NASH: Comparing the AMLN Diets and the MS-NASH Mouse
Developing Preclinical NASH Dietary Mouse Models
The main challenge in preclinical development of NASH drugs is the current suitability of small animal models. Ideal disease presentation would feature multiple characteristics of human NASH, including a preamble of metabolic symptoms and NAFLD. Disease would be induced through feeding the harmful components of a Western diet.
Early NASH model development efforts focused on offering animals an excessive caloric density in the form of fat, along with high amounts of refined sugar, especially fructose. These diets recapitulated NAFLD-like disease in mice, but involved a multimonth regimen and achieved only mild to moderate degrees of NASH-like fibrosis. Parallel efforts introduced inflammatory chemical insults and/or micronutrient limitations to exacerbate fibrosis, but these manipulations compromise translatability in the context of pure metabolic disease.
Animal Selection for NASH Dietary Mouse Models
The C57Bl/6 (“B6”) lineage of mice is a workhorse of early stage preclinical research, including for dietary models of obesity and metabolic disease. Popular sublineages include the original Jackson Laboratory mouse (B6J) and the NIH sublineage (B6N). These differ through a variety of genetic variants that affect weight gain, insulin resistance, sensitivity to oxidative stress, and more.
Historically, the B6J mouse was the most widely used strain. The Taconic Biosciences strain of the B6N mouse (“B6NTac”) is currently gaining popularity for models of obesity, metabolic syndrome, NAFLD, and NASH. This sub-sublineage demonstrates considerably greater fatty weight gain, insulin secretion, and insulin resistance to the B6J, and even outperforms the B6N ancestral lineage in these regards.
Seeking a highly translational model of NASH in a wild type B6 mouse has been a holy grail quest for the metabolic research community since the emergence of the obesity epidemic. Breakthroughs in nutrition science and animal husbandry have advanced preclinical NASH modeling significantly in recent years, providing attractive commercial options including induction through AMLN and GAN diets, as well as next generation polygenic models.
The AMLN Diet
In 2012, a novel dietary model for obese and fibrotic mouse NASH was introduced by researchers at Amylin Pharmaceuticals. This involved ad libitum feeding of Research Diets Inc. chow formulation D09100301, comprising:
- 40% kcal fat (primarily Primex hydrogenated vegetable oil shortening, i.e. trans-fat)
- 20% kcal fructose
- 2% w/w cholesterol
This formulation has come to be commonly known as the AMLN (Amylin Liver NASH) diet.
B6J mice on the AMLN diet experience rapid, robust weight gain, approximately three times as much as counterparts on a standard chow diet in 12 weeks (10% fat, no fructose or cholesterol). This weight gain is skewed toward accumulation of fat mass.
AMLN mice exhibit hepatomegaly, with livers reaching about 8% of total body weight compared to 4% in standard chow controls. Circulating liver injury indices (amino acid transaminases) are markedly elevated within 12 weeks. The steatosis of the AMLN diet is attributable to the trans-fat composition of D09100301, as mice on an equicaloric regimen substituted with saturated lard fat exhibited much milder symptoms across the spectrum of disease characteristics.
Histopathological NASH is poorly pronounced and variable in under 16 weeks on an AMLN diet, and frequent and appreciably mild to moderate after 30 weeks. Hepatic features include:
- Macro- and microvesicular lipid accumulation
- Inflammatory infiltrates
- Augmented collagen deposition
- Fibrosis, including periportal and perisinusoidal in location.
Aside from the time required to achieve overt NASH (26-30 weeks), heterogeneity is also a confounding issue. B6J mice exhibit ranging disease severity across parameters of the clinical NAFLD activity score (NAS). This includes presence of hepatocyte ballooning only in subsets of animals. This variability is significant enough to encourage staging individual animals into severity cohorts. This challenging feat is best achieved through invasive biopsy, an expensive and technically challenging option not available to most researchers.
Interrogation of substrain differences between 30 week AMLN-fed B6J and B6N mice has uncovered some phenotypic differences, including in weight gain (B6J < B6N), hepatic triglyceride content (B6J > B6N), and serum biomarkers of liver injury (B6J < B6N). Clinical scoring of NAFLD and fibrosis are roughly equal in magnitude and variability, so no quick fix is available through selection of the NIH substrain.
The GAN Diet
A recent FDA ban on trans-fats as food additives prompted the development of a modified AMLN diet with a suitably deleterious surrogate fat to Primex. This emerged through Research Diets Inc. chow formulation D09100310, comprising:
- 40% kcal fat (primarily palm oil)
- 20% kcal fructose
- 2% w/w cholesterol.
This diet is commonly known as the GAN (Gubra Amylin NASH) diet.
B6J mice fed GAN experience even greater weight gain than on AMLN. Otherwise, a head-to-head comparison of the B6J mouse on 28 week GAN and AMLN diets resulted in equally marked steatosis, with comparable NAS scoring and degrees of fibrosis. Heterogeneity of NAS scoring and fibrotic accumulation, however, remain an equally confounding factor, with individual animals exhibiting marked degrees of staging. Therefore, palm oil supplementation fully compensated for trans-fat in the GAN diet, but did not resolve limitations of the model.
Within Taconic Biosciences, efforts are currently underway to characterize the B6NTac mouse on GAN. Early returns are promising, confirming that key stages of liver disease occur at time points that mirror historical characterization in B6J and B6N mice. These include steatosis at 12 weeks, inflammation at 20 weeks, and fibrosis at 26 weeks.
It remains to be seen if heterogeneity issues persist in this mouse lineage. However, if these are resolved by use of this strain, the organization’s goal is to provide off-the-shelf animals conditioned to the buyer’s ideal starting point for liver disease projects, enabling rapid initiation of projects by circumventing the lengthy dietary induction period.
The MS-NASH Mouse
A new polygenic mouse strain is emerging as a champion of metabolic and liver disease research. The MS-NASH mouse, available through Crown Bioscience and the Jackson Laboratory, was generated through initial crossing of the B6J and the AKR/J strains. Selective inbreeding over 30 generations generated a strain that is spontaneously obese, insulin resistant, and hyperglycemic, and can be triggered to these metabolic phenotypes more rapidly and more severely on a high-fat diet (60% kcal fat).
The MS-NASH mouse rapidly progresses through metabolic dysregulation, obesity, and steatosis, toward NASH in as little as 20 weeks on a Western diet + fructose, including:
- 41% kcal fat (primarily butter)
- 29% kcal sucrose
- 0.21% cholesterol w/w
- 5% fructose w/vol ad libitum in drinking water
Liver injury markers are elevated within 8 weeks on diet, with clinically appreciable scoring of NAFLD parameters possible in as short as 16 weeks, and fibrosis in 20 weeks.
With its shorter induction time on a translationally relevant diet, its polygenic basis for metabolic dysregulation, and its broad commercial availability, the MS-NASH mouse is poised to herald the next big breakthrough in preclinical NASH research.
AMLN-GAN Diet vs MS-NASH Mouse: Head to Head
|Mouse strain||C57Bl/6 (J, N, NTac)||MS-NASH|
|Polygenic predisposition to obesity||No||Yes|
|Diet||AMLN: Research Diets D09100301
GAN: Research Diets D09100310
|Research Diets D12079B with fructose in drinking water|
AMLN: primarily Primex trans-fat
GAN: primarily palm oil
|Fructose||20% kcal||29% kcal in chow (sucrose)
5% w/v in drinking water
|Cholesterol||2% w/w||0.21% w/w|
|Age at diet start||6-10 weeks||6-10 weeks|
|Circulating liver injury markers||12 weeks||8 weeks|
|Steatosis||12 weeks||12 weeks|
|NAFLD||20 weeks||16 weeks|
|Fibrosis||26 weeks||20 weeks|
|Inter-animal variability||AMLN: yes
GAN: studies ongoing
The table above shows a direct comparison of the MS-NASH mouse and NASH models induced by the AMLN and GAN diets. The polygenic MS-NASH mouse shows lower inter-animal variability compared with monogenic models induced by the AMLN diet (with GAN diet studies ongoing). Disease features including evidence of NAFLD, fibrosis, and circulating injury markers are also all apparent in a shortened timeframe for the MS-NASH mouse versus AMLN/GAN diet induced models.