MASH Animal Models Compared: AMLN Diet vs. GAN Diet

The MASH Model Turning Point: Why AMLN Died and GAN Rose

In preclinical drug discovery for MASH (Metabolic dysfunction-Associated Steatohepatitis), the GAN (Gubra-Amylin NASH) diet model is currently one of the most widely adopted and reliable dietary models.

However, prior to 2018, its predecessor—the AMLN (Amylin Liver NASH) diet—was the industry standard. Why did this "forced migration" from AMLN to GAN occur? And how did this shift impact the efficacy and readouts of the MASH evaluation system? This article provides a highly detailed comparison between the two models to help researchers and sponsors interpret historical data and design future trials.

1. Behind the Sudden "Supply Disruption"

Around 2018, a massive disruption hit the NASH research community globally: The beloved and highly-validated AMLN diet (Research Diets D09100301) was slated to be discontinued. The cause did not originate within the laboratory animal industry, but rather from a profound regulatory decision by the U.S. Food and Drug Administration (FDA).

FDA's Elimination of Partially Hydrogenated Oils (PHOs)

In 2015, the FDA released its final determination that Partially Hydrogenated Oils (PHOs) were "not Generally Recognized as Safe" (not GRAS), effectively banning their use in human food supplies starting June 2018 to mitigate cardiovascular disease.

The primary driver of severe hepatotoxicity and obesity in the AMLN diet was Primex shortening, a PHO rich in artificial trans fats. As PHOs withdrew from the human food market entirely, sourcing this critical ingredient became unsustainable, sparking an urgent race to develop an alternative MASH model.

2. The GAN Diet: Evolution, Not Just Substitution

In response to this regulatory shock wave, Gubra and Research Diets Inc. co-developed the GAN (Gubra-Amylin NASH) diet (D09100310). By substituting trans fats with Palm Oil (saturated fats), they successfully constructed a model with MASH-inducing potency equal to, or even exceeding, the classical AMLN diet.

Nutritional Composition: Trans Fat vs. Saturated Fat

Nutrient	AMLN Diet (D09100301)	GAN Diet (D09100310)
Protein (kcal%)	20%	20%
Carbohydrate (kcal%)	40%	40%
Fat (kcal%)	40%	40%
Total kcal/gm	4.49	4.60

Main Fat Source	Primex II Shortening	Palm Oil
Trans Fat	Present (18.5% kcal)	None (0%)
Saturated Fat	Low	Rich (Palmitic acid etc.)
Cholesterol	2.0% (wt/wt)	2.0% (wt/wt)
Fructose	22.0% (kCal)	22.0% (kCal)

Notably, the GAN diet completely replaces the trans fat source while intentionally maintaining the high-cholesterol (2%) and high-fructose burdens necessary to drive severe hepatocyte ballooning and inflammation.

3. Phenotypic Comparison: AMLN vs. GAN

How did changing the fat source alter the mouse's pathology? The GAN diet is far from a "downgraded" AMLN. In fact, it drives a remarkably robust and arguably more "modern" MASH phenotype.

① Weight Gain and Insulin Resistance (Metabolic Severity)

The GAN Advantage: Interestingly, numerous studies across C57BL/6J and ob/ob backgrounds report that GAN-fed mice exhibit faster, more pronounced weight gain (obesity) and significantly more severe insulin resistance compared to historical AMLN cohorts.
The Cause: This is largely attributed to the highly obesogenic and pro-inflammatory nature of the saturated fatty acids (particularly palmitic acid) abundant in palm oil.

② Speed of Fibrosis Progression to F2/F3

Initial Onset: Both diets trigger rapid hepatic steatosis and inflammatory cell infiltration within the first few weeks of feeding.
Bridging Fibrosis (≧F2): The classical AMLN diet notoriously required long feeding periods (often 24 to 30 weeks) to achieve solid F2/F3 bridging fibrosis. Conversely, the GAN diet model reliably progresses to significant F2-F3 stage fibrosis within approximately 16 to 20 weeks, exhibiting distinct collagen meshworks when evaluated via Sirius Red quantification. The fibrogenic drive in the GAN model appears slightly more accelerated and consistent.

③ Progression to Hepatocellular Carcinoma (HCC / Tumor Risk)

Assessing the risk during long-term chronic feeding (>40 to 50 weeks):

AMLN Diet: Feeding for over 50 weeks induced liver tumors in a subset of animals.
GAN Diet: Long-term feeding (starting around 36 weeks) of the GAN diet also demonstrates a high penetrance of progression to Hepatocellular Carcinoma (HCC). For researchers looking for a "MASH-driven HCC model," the extended GAN diet is highly validated and exceptionally useful.

4. Our Perspective: A Blessing in Disguise

In modern human diets, the intake of artificial trans fats has plummeted globally due to strict regulations. Meanwhile, the overconsumption of cheap saturated fats (like palm oil) and high-fructose corn syrup remains the overwhelming driver of the global obesity and MASH epidemic.

The "forced transition" mandated by the FDA inadvertently catalyzed an evolution: moving from a "trans-fat dependent model" to a "saturated-fat overload model" that far more accurately mirrors the dietary habits and metabolic profiles of modern MASH patients. Gaining profound insulin resistance alongside more consistent bridging fibrosis was a massive win for the drug discovery industry.

Tips When Outsourcing to CROs

Today, virtually all global preclinical CROs have retired the AMLN model and standardized on the GAN diet protocol (or similar heavily saturated diets like the CDAHFD). If your team is designing a new study based on historical compound data generated in the AMLN era, be mindful that GAN mice will likely be heavier and more insulin-resistant at baseline. Adjusting your vehicle controls and the dosing regimens of metabolic "backbone" drugs (like GLP-1 receptor agonists) accordingly is highly recommended for translational success.

References

FDA. Final Determination Regarding Partially Hydrogenated Oils. 2015.
Tølbøl KS, et al. World J Gastroenterol. 2018. (GAN model validation)
Boland ML, et al. PLoS One. 2019. (Comparison of NASH diets)
Research Diets, Inc. OpenSource Diets data: (D09100301, D09100310)