Webinar Q&A — The Diet Induced NASH B6: A Translational NASH Model for Drug Discovery


Webinar Q&A — The Diet Induced NASH B6: A Translational NASH Model for Drug Discovery Dr. Janell Richardson recently presented a webinar on the Diet Induced NASH B6, a translational model of NASH for drug discovery. She provided an excellent introduction to the global health problem of nonalcoholic steatohepatitis (NASH), gave an overview of different types of preclinical NASH models and focused in on the Diet Induced NASH B6 as a relevant model of NASH in the context of a metabolic disease phenotype.

Due to time constraints, many of the questions submitted during the webinar went unanswered. We present a full Q&A here.
  • Q: How does the AMLN or modified AMLN diet compare to a choline deficient diet?

Dr. Janell Richardson (JR): A choline deficient diet is used with a fibrotic endpoint. This can be incorporated into an obesogenic diet. The classic choline deficient diet is the methionine choline deficient diet (MCD). The benefit to MCD or other choline deficient diets is that the penetrance of fibrosis is very clear compared to an AMLN-based diet. The negative is understanding the translatability. Having a nutritional deficit as the mechanism behind overall pathophysiology is going to impact the translatability to the human condition where the etiology is very different. The benefit to a choline deficient diet is that it can give a clear yes or no answer to whether a therapeutic can modulate fibrogenesis prior to undertaking additional studies. It can also be used for proof of concept testing of imagining techniques and/or biomarkers to identify the direct or indirect presence of fibrosis, as well as determine augmentation of the fibrotic pathway.
  • Q: Could other types of rodent models be used with the modified AMLN diet?

JR: Yes and no. It depends heavily on the rodent strain. Rodents in general have variation in response to obesogenic diets, including development of steatosis and NASH. The reason for this is that not all rodent strains are susceptible to diet induced obesity. For MCD diets, the pathology does not rely on an obesogenic phenotype but it can still demonstrate strain differences in degree of susceptibility as can chemical/toxin induction of fibrosis. For strains which are susceptible to diet induced obesity, the degree of susceptibility varies. For example, B6 are very susceptible and have the most favorable metabolic parameters as compared to BALB/c or nude mice. Nude mice will only show an obesogenic-based phenotype under thermoneutral conditions. So there's quite a bit of variation, even between substrains of the same inbred strain (e.g. data presented on differences between B6N and B6J). For example, a super immunodeficient animal such as the CIEA NOG mouse® is not susceptible to weight gain on a high fat diet. They can show increased adiposity, but their overall body weights would change little and they would not be expected to develop an insulin resistance phenotype, which is another key aspect which underlies the natural history of the clinical aspects of NASH.
  • Q: The Diet Induced NASH B6 can lose weight during transportation. Do we know what drives the weight loss? Do they eat less, increase caloric expenditure, etc?

JR: Weight loss in mice can be from multiple aspects, for example decreased meal frequency or decreased meal size — both impact food intake resulting in weight loss. So even with respect to decreased food intake, it can be further differentiated and impacted in parallel by yet further variables. We don't have a granular understanding of weight loss related to transit in this model. Overall, it's important to understand that stress on any rodent model for which weight gain is a key property, is going to be problematic. Rodents, being a non-emetic species, typically deal with stress by significant reduction in body weight loss typically either through increased energy expenditure via locomotion or reduction in food intake. We are not able to track either of these variables while the mice are in transit.

Overall, the net effect is that transit stresses the animal, and this stress results in larger weight loss in obese animals. The severity varies from animal to animal but if chronic high stress is present, it can mitigate any body weight gains for these obese animals (bring them to near control body weight levels). For the most part, control animals lose some weight but it rectifies rapidly, whereas obese animals have more room to drop weight and rectification is much slower.
  • Q: Liver biopsy is of course the gold standard for humans. Can liver biopsy be performed multiple times on a single animal?

JR: Yes, it can be performed multiple times in the same animal based on the size of the tissue excised, the skill of the surgeon, approval of IACUC committee, the ability to quickly clot any resulting bleeding, and the overall time required under anesthetic. Of course, any time an animal model, such as the NASH B6, is placed under such an invasive procedure, weight loss should be expected. Again, the resulting weight loss, recovery time and how successive biopsies will impact the overall etiology of NASH would need to be deciphered. Chae and colleagues published their findings this year on using multiple liver biopsies to track NAFLD in preclinical rodent models (WD, HFD, and MCD). Additionally, there was a recent publication/video from MedImmune on the protocol of survivable 1x liver biopsy in AMLN diet-induced B6 and efficacy results of liraglutide.
  • Q: Are there alternatives to an invasive liver biopsy for characterizing fibrosis in individual animals?

JR: Yes. There are some alternatives. This is a big issue, both in the preclinical field and in the clinic. Patients are often reluctant to undergo liver biopsies, which are very invasive. There is a real movement to define biomarkers which could identify and understand with better granularity the NASH stage for a particular patient. In humans, liver biopsy is the gold standard for diagnosis of NASH because it's the best we have at this point. It gives a one time glimpse at the liver; a biopsy is typically 1/50,000 of the total liver volume (in a mouse it tends to be a little bigger). Imaging techniques are another contender for replacing the liver biopsy. In humans and preclinical models, there has been recent work on characterizing NASH via elastography (a technique to measure the elasticity or stiffness of the liver). A liver with fibrosis is more rigid due to collagen deposition and overall parenchymal rigidity; fibrosis leads to decreased elasticity of the tissue. There are a few different forms of elastography. The one that I've seen some of the most NASH work done in right now are transient elastography, shear wave elastography and overall high frequency ultrasound and photoacoustics. Each one of those techniques has its pros and cons, but because these techniques are still new, it's important to link them back to histological confirmation of disease stage.
  • Q: What is the breakdown of fibrosis scores achieved in the Diet Induced NASH B6?

JR: Unfortunately, we do not yet know what the % of fibrosis penetrance is in our model, the scores achieved, the mean, the frequency, nor the spread. We are actively collecting this data and hope to provide added clarity on this question within the upcoming months. That being said, we do not anticipate anything beyond a fibrosis grade of 2 in this particular model.
  • Q: Are there scoring differences between humans and rodents with respect to NAS? How are they different?

JR: Yes, Liang and colleagues from Leiden University in 2014 published a study asking this question: could one use the NAFLD Activity Scoring (NAS) system outlined by the NASH Clinical Research Network in preclinical rodent models. They compared human NAFLD liver pathology with that of several NAFLD mouse models (e.g. Leiden mouse, high fat diet B6, KKAY, methionine choline diet B6, etc.) stained (HE and PSR) and looked at key features of human NAS for presence or absence in mice. They summarized their findings (found in slide 26 of the webinar) and found that the overall histological cross-sections of the mouse NAFLD were similar to NAFLD patients. However, not all key features of the human disease were recapitulated in these mouse models. Key aspects such as steatosis, inflammation, and fibrosis were maintained between the two. In contrast, hepatocellular injury features, apoptosis, pigmented macrophages, and megamitochondria were absent in mouse models vs. humans. Hepatocyte ballooning was occasionally found, however, it was not nearly as prominent as what is seen in the clinic. Based on overlapping features the authors selected a Rodent NAS system that identifies the degree of:

  1. Steatosis (macrovesicular, microvesicular, hypertrophy, % area affected) and
  2. Inflammation (average # inflammatory cell aggregate counts within 5 different microscope fields).
So in summary rodents share many similarities to components that are used in human NAS system but not all. A separate system that accounts for these differences, as well as, the overall decreased frequency of severe NAFLD occurence in preclinical models when compared to the clinic.
  • Q: Progression to hepatocellular carcinoma (HCC) is a feature of NASH and is also seen in other models such as the American Lifestyle diet (high fat and high fructose) at 12 months — is that a feature of the Diet Induced NASH B6 model?

JR: It's not clear yet whether HCC develops in the Diet Induced NASH B6. The American Lifestyle-Induced Obesity Syndrome (ALIOS) diet provided a key background for which Amylin Pharmaceuticals then modified to develop the AMLN diet. The diet we use is a modified version of this AMLN diet. We have not yet evaluated our model at prolonged diet administration beyond 26 weeks to evaluate whether HCC develops in this model. In general, this might not be likely. In humans, the progression from the NAFLD spectrum and NASH to HCC is well established, but in preclinical rodent models this appears to be a limitation and often present in only rare occurrences. We don't yet have data to say whether that's a limitation in our model.
  • Is ballooning degeneration observed in the Diet Induced NASH B6?

JR: We haven't yet looked at ballooning in this animal model, but I would not expect ballooning degeneration to be present in this model. One of the few models which displays that phenotype is the ob/ob on an AMLN-based diet. However, ob/ob animals are very susceptible to stress and can be a technically challenging model to work with.
  • Do Diet Induced NASH B6 mice display glucose intolerance at later time points such as 34 weeks on diet?

Fasting blood glucose levels did not significantly change from controls at any of the time points measured when exposed to modified AMLN dietJR: Fasting blood glucose levels (as indicated by the graph) did not significantly change from controls at any of the time points measured when exposed to modified AMLN diet, including week 34. This is to be expected, as glucose intolerance (high blood glucose levels) have not been shown in the original AMLN diet B6 model.
  • Q: Some users of AMLN-based NASH models report a high occurrence of dermatitis. Is dermatitis observed in the Taconic Biosciences' model, and since dermatitis has an inflammatory component, how might it affect the model?

JR: Dermatitis could absolutely affect the model. Stress, whatever the cause, can impact this animal model and can impact the overall phenotype and phenotypic penetrance and overall pathogenesis of the model. Understanding and trying to mitigate stress factors is important for good study design, but it's nearly impossible to mitigate them all. For example, ones that are easier to mitigate are location on an individually ventilated cage (IVC) rack. Animals closer to a rack blower could show an increase in environmental stress and could vary in terms of phenotype from animals in cages which sit lower on the rack. Aggressor animals in cages, extreme barbering, bedding type...there are a multitude of factors which experimenters must plan for.
  • Q: Have B6NTac x B6J F1 or F2 progeny been phenotyped for NASH and genetic linkage analysis?

JR: I'm not aware of any studies using B6NTac x B6J F1 or F2 hybrids for NASH based studies nor what the expected phenotype or genetic linkage would be.
  • Q: Can the NASH phenotype in the Diet Induced NASH B6 be reversed by moving the mice back to a chow diet? If so, at what stages is the disease phenotype reversible versus irreversible?

JR: We do not currently have any data on what timepoint/pathogenesis our NASH model becomes irreversible. Traditionally, end stage fibrosis or cirrhosis has been looked at as an irreversible condition. However, in humans, certainly rectification from NASH can and does occur "and the possibility that cirrhosis can be reversed has been shown in patients with viral hepatitis and other forms of chronic liver disease when the underlying cause of liver injury is eliminated. Data in patients with NASH cirrhosis before bariatric surgery also suggests that cirrhosis resolution can occur in NASH, but in what proportion of patients remains to be established." — Liver forum clinical trial endpoints 2019.
  • Q: Can Taconic perform studies using NASH mice?

JR: Taconic Biosciences can provide limited auxiliary services with the Diet Induced NASH B6 mice. Taconic is also able to generate other NASH models, for example by feeding different specified diets to B6 or other strains. Contact us for more details on available services.
  • Q: There is not yet an FDA-approved drug for NASH, which may be due in part to a lack of translational models for this disease. Could use of rodent models in conjunction with large animal models such as pig and dog in preclinical trials lead to more predictive clinical trial outcomes?

JR: It's certainly a good idea to consider. There are hard limitations to rodent models of metabolic syndrome indications, even with genetic or dietary modifications, that are due to the nature of rodents. These are non-emetic species, and that is important as it relates to their overall behavior with respect to food. A non-emetic animal will be naturally guarded when it comes to any type of food source. Presenting brand new food to rodents can induce hesitation in feeding. In higher order mammals such as dogs or pigs, which are emetic, do not show this type of avoidance or guarded behavior. They also show mechanisms which are much more in line with humans. For example, both dogs and swine can naturally suffer from obesity. Companion animals are nearly as obese as their owners! To understand and identify therapeutic outcomes of NASH, a multifactorial disease, it may be beneficial to add in larger animal models which better align with human NASH pathways, particularly in later stages of drug discovery.