Abstract

   Metabolic dysfunction-associated steatotic liver disease (MASLD),
   previously known as non-alcoholic fatty liver disease, encompasses
   steatosis and metabolic dysfunction-associated steatohepatitis (MASH),
   leading to cirrhosis and hepatocellular carcinoma. Preclinical MASLD
   research is mainly performed in rodents; however, the model that best
   recapitulates human disease is yet to be defined. We conducted a
   wide-ranging retrospective review (metabolic phenotype, liver
   histopathology, transcriptome benchmarked against humans) of murine
   models (mostly male) and ranked them using an unbiased MASLD ‘human
   proximity score’ to define their metabolic relevance and ability to
   induce MASH-fibrosis. Here, we show that Western diets align closely
   with human MASH; high cholesterol content, extended study duration
   and/or genetic manipulation of disease-promoting pathways are required
   to intensify liver damage and accelerate significant (F2+) fibrosis
   development. Choline-deficient models rapidly induce MASH-fibrosis
   while showing relatively poor translatability. Our ranking of commonly
   used MASLD models, based on their proximity to human MASLD, helps with
   the selection of appropriate in vivo models to accelerate preclinical
   research.

   Subject terms: Non-alcoholic fatty liver disease, Metabolic diseases,
   Experimental models of disease, Translational research
     __________________________________________________________________

   The LITMUS consortium provides a resource of rodent MASLD models
   benchmarked against metabolic, histologic and transcriptomic features
   that are relevant for human MASLD. The work is useful for selecting
   relevant rodent models for studying this common disease.

Main

   Non-alcoholic fatty liver disease (NAFLD) is the hepatic manifestation
   of the metabolic syndrome^[138]1. It clusters with obesity, insulin
   resistance, diabetes, dyslipidemia, atherosclerosis and
   cancer^[139]2,[140]3. Given its metabolic context, a multi-society
   consensus statement on fatty liver disease proposed the new
   nomenclature, metabolic dysfunction-associated steatotic liver disease
   (MASLD)^[141]4.

   The MASLD spectrum ranges from simple steatosis to steatohepatitis
   (NASH/MASH) and MASH-fibrosis^[142]5,[143]6. Under the pressure of
   environmental factors (lifestyle, nutrition, microbiome) and genetic
   predisposition (for example, the rs738409 C>G polymorphism in the
   PNPLA3 gene), the disease can progress to MASH, promoted by lipotoxic
   insults driving hepatocyte injury, inflammation and chronic activation
   of wound-healing responses^[144]7,[145]8. Progressive MASH places
   patients at risk of cirrhosis and hepatocellular carcinoma, which may
   result in liver-related mortality and the need for liver
   transplantation^[146]1,[147]9.

   The lack of a standard translationally relevant preclinical model has
   hindered the field’s ability to study the chronic and complex
   pathophysiology of MASH. Furthermore, overinflated preclinical efficacy
   data generated from models in which human pathophysiology is not
   accurately replicated probably contribute to negative clinical trial
   results in the MASH field. Many diets differing in macronutrient
   composition have been tested in a wide range of rodent models aiming to
   feature the whole spectrum of metabolic disease and/or hepatic
   damage^[148]10–[149]12. Disappointingly, within each diet
   macro-category, relatively subtle differences in micronutrient and
   macronutrient composition (for example, amount of cholesterol or
   choline) and study designs are sufficient to introduce variability in
   MASLD disease endpoints^[150]12–[151]16. Several stressors have been
   used to accelerate disease progression and homogenize phenotypes,
   including genetically modified mice and rats (featuring obesity^[152]17
   or hypercholesterolemia^[153]18,[154]19), chemically induced reduction
   of insulin secretory capacity^[155]20 and/or the addition of toxic
   chemicals (for example, carbon tetrachloride, CCl[4] (ref. ^[156]21)),
   and the use of genetically modified mice that spontaneously develop
   progressive steatohepatitis (for example, NEMO, PTEN knockout (KO)
   mice^[157]16,[158]22). Moreover, some rodent models seem to
   recapitulate the metabolic aspects of MASLD, whereas others are better
   at developing its fibro-inflammatory features^[159]12,[160]13,[161]23.
   Given the vast proliferation of models, variability and lack of
   standardization^[162]12–[163]16, a systematic comparison validating
   metabolic features, histology and transcriptomics against human disease
   is warranted but currently lacking.

   To resolve the uncertainty about the most relevant preclinical mouse
   models, we have performed a wide-ranging retrospective analysis of the
   most common murine models used in academia and the pharmaceutical
   industry that were available to our consortium and collaborators,
   benchmarked against human MASLD and ranked according to the following
   characteristics: (1) obesity and/or metabolic syndrome; (2) development
   of steatohepatitis with progressive fibrosis (following hard outcomes
   of clinical trials defined for an amelioration, or at least a
   non-worsening, of fibrosis^[164]24); and (3) similarity of the
   histological features and molecular events to human MASH.

   We developed a bioinformatic pipeline integrating metabolic phenotype
   data, liver histology (with centralized staining and assessment) and
   liver transcriptome benchmarked against human disease to create a MASLD
   ‘human proximity score’ (MHPS; Fig. [165]1). The MHPS was invaluable in
   generating a dual ranking of models based on their metabolic relevance
   and/or ability to induce MASH-fibrosis, highlighting the models that
   more closely resembled the metabolic and/or the fibrotic features of
   the disease. Our approach identified murine MASLD models showing
   (phenotypic and/or histologic and/or transcriptomic) profiles relevant
   to human MASH, which are suitable for most preclinical experimental
   applications.

Fig. 1. Study design.

   [166]Fig. 1
   [167]Open in a new tab

   In this study, we collected retrospective information from 598 animals
   (509 WT/GA mice, 89 WT/GA rats): 336 animals subjected to treatment
   (MASLD-inducing conditions: 315 animals; or CCl[4]: 21 animals) and 262
   animals as controls for MASLD-inducing conditions (247 animals) or
   CCl[4] (15 animals), returning 39 models (that is, study designs) aimed
   at modeling MASLD, and two time points for CCl[4] (positive controls of
   MASLD-independent fibrosis). Details of the study designs (numerosity,
   species, background, genetic manipulation, diet, time point and room
   temperature) are provided in Supplementary Table [168]1. For all the
   studies, phenotypic information (Supplementary Table [169]5),
   centralized histopathology assessment (Supplementary Table [170]6) and
   liver transcriptomics (Supplementary Table [171]4) were available.
   These data were integrated into an unbiased binary score (MHPS) ranking
   the models in terms of their metabolic relevance and ability to induce
   MASH-fibrosis. Created with [172]BioRender (agreement number
   GG26BHMS6Y).

Results

Main phenotypic and histologic attributes characterizing human MASLD in
murine models

   We collated retrospective data and samples from 39 commonly used murine
   genetic or dietary MASLD models (treatment: 315 animals; control diet:
   247 animals; see Fig. [173]2 and Supplementary Tables [174]1 and [175]2
   for more details) available to the consortium and/or collaborators, and
   clustered them according to macro-categories (diet and/or genetic
   background) as follows: (1) genetically modified models of obesity or
   MASLD; (2) high-fat diet (HFD); (3) Western (that is, atherogenic) diet
   (WD; HFD enriched in refined carbohydrates and cholesterol); these
   models were subclustered according to cholesterol concentration (0–2%)
   and/or the use of chemicals (streptozotocin, STZ); (4) American
   lifestyle diet (AMLD) including HFD (HFDAMLD) or WD (WDAMLD)
   supplemented with refined sugars in the drinking water, with or without
   the use of low-dose CCl[4]; and (5) choline-deficient dietary models
   including ‘canonical’ choline-deficient diets (CDD) and
   choline-deficient diets supplemented with fat (CDHFD), cholesterol
   (CDHCD; 1% or 2%) or both (CDHFHCD). (6) As positive controls for
   isolated fibrosis, we used CCl[4]-treated mice (CCl[4] treatment: 21
   animals; control treatment: 15 animals; two time points).

Fig. 2. Phenotypic and histologic characterization of the models.

   [176]Fig. 2
   [177]Open in a new tab

   Phenotypic changes observed in the MASLD models compared to their
   matched controls were profiled as the log[2] fold change (log[2]FC)
   across measures of BW, blood triglycerides (TGs) and cholesterol,
   LW:BW% ratio, and ALT and AST. The red–blue color gradient indicates
   the level of increase–decrease of the measure in the MASLD models
   compared to their controls, while an asterisk indicates a significant
   change at P < 0.05 (two-sided Mann–Whitney U-test). The two panels of
   horizontal bars give an overview of the complete histological profiles,
   in which the total length indicates the activity score (CRN NAS) and
   fibrosis^[178]25. In addition, NAS components (steatosis, ballooning
   and inflammation) are represented by the stacked bar (yellow, green and
   blue, respectively) lengths. All models are grouped according to their
   macro-categories (detailed by the leftmost annotations).

   [179]Source data

   The anticipated outcome of a MASLD murine model exhibiting body weight
   (BW) gain was readily achieved in genetically modified
   (leptin-deficient (ob/ob)) mice, WD and HFD models. Notably, the weight
   gain was less prominent in HFD models, and in WD or WDAMLD supplemented
   with chemicals (CCl[4] or STZ). By contrast, choline-deficient models
   generally reduced BW with the exception of CDHFHCD1% in low-density
   lipoprotein receptor (LDLR) KO mice. Most models did not significantly
   elevate circulating triglyceride levels, except for ZSF1 (diabetic)
   rats on a WD2% diet (RZ-AMLNREP-C2-23W) and a CDHFD mouse model. As
   expected, most HFD models and all diets with increased cholesterol
   content (0.2–2%) (including WD, WDAMLD, CDHCD and CDHFHCD) as well as
   wild-type rats treated with CDAA exhibited hypercholesterolemia.

   Apart from HFDs (and HFDAMLD), most experimental designs resulted in
   notable liver enlargement (that is, the ratio of liver weight to BW
   (LW:BW%)) and elevated aspartate aminotransferase (AST) and alanine
   aminotransferase (ALT) levels, especially at the final time point
   studied. Data on glucose metabolism was limited to a subset of our
   models (Supplementary Table [180]1). However, most HFD and WD diets
   examined have been documented in the literature to decrease insulin
   tolerance and impair glucose metabolism; conversely, choline-deficient
   models are not as well-characterized in this context (see Supplementary
   Table [181]2 for details and references).