Mouse Model of Metabolic Dysfunction-Associated Steatotic Liver Disease with Fibrosis

Summary

Mouse model of metabolic dysfunction-associated steatotic liver disease (MASLD) with metabolic dysfunction, hepatic gene expression changes, and liver histopathological alterations that resemble human MASLD, including fibrosis that progresses to advanced fibrosis stage 3. This model can be used in studies of MASLD pathophysiology and in pre-clinical studies of new therapies.

Abstract

Liver fibrosis is the most important predictor of adverse outcomes in metabolic dysfunction-associated steatotic liver disease (MASLD). In pre-clinical studies of MASLD, mice are the most commonly used disease model. However, in those models, liver fibrosis is difficult to induce. This protocol describes a mouse model of MASLD in which hyperphagic Ay mice are fed a diet that is high in fat and fructose, which resembles the average diet in the US. The mice develop hepatic steatosis, injury, inflammation, and fibrosis. Fibrosis progresses from pericellular, stage 1 fibrosis after 16 weeks, to bridging, stage 3 fibrosis after 12 months. These mice also develop obesity, hypertriglyceridemia, glucose intolerance, and hyperinsulinemia. This disease model replicates the liver histopathology, the hepatic gene expression alterations, and the metabolic dysfunction of the human disease. This protocol includes two methods to quantify fibrosis: histological staining of collagen by picro-sirius red, and quantification of the liver content of hydroxyproline. These methods can be used for the study of MASLD pathophysiology and for preclinical studies of potential therapies.

Introduction

Metabolic dysfunction-associated steatotic liver disease (MASLD) is the most common liver disease, with a prevalence of 31% to 40% in the US and 42% worldwide^1,2,3,4. Patients with MASLD can develop metabolic dysfunction-associated steatohepatitis (MASH), previously named non-alcoholic steatohepatitis (NASH), cirrhosis, and hepatocellular carcinoma^5,6. MASLD increases the risk of hepatocellular carcinoma, and following the recent increase in prevalence of MASLD, it has become the most common liver disease in patients who develop hepatocellular carcinoma^7,8. In MASLD, fibrosis is the primary predictor of progression to cirrhosis and of adverse clinical outcomes, including liver-related morbidity (hepatic decompensation, liver failure, or liver cancer) and liver-related mortality^6,9,10,11. Accordingly, the Food and Drug Administration encourages the development of therapeutics focusing on MASH/NASH with fibrosis, as well as the use of appropriate animal models¹².

In preclinical studies of drugs and in mechanistic studies of MASLD, mouse models are most commonly used¹³. Thousands of mouse models of MASLD have been described, suggesting a lack of consensus on an optimal model^13,14. Establishment of optimal models faces several challenges. Although it is relatively easy to induce obesity and steatosis in mice by feeding them diets high in fat, liver fibrosis is more difficult to induce^13,14,15,16. In MASLD models driven by obesogenic diets, fibrosis develops slowly¹⁷ Although this slow progression may better mimic the human disease¹⁸, it increases the duration and cost of animal studies. Diets with high cholesterol content (e.g., 2%) promote the development of fibrosis; however, they also decrease the endogenous synthesis of cholesterol, which is the opposite of the increase described in the human disease^15,16. Some MASLD models use mutant or genetically modified mice; however, they may have caveats: ob/ob mice develop obesity, steatosis, and insulin resistance, but they lack leptin, which is a mediator of fibrosis; PTEN null mice develop steatosis but do not develop insulin resistance^15,17. Therefore, mouse models of MASLD often require long study durations, diets with high cholesterol content, or genetic manipulations to develop significant fibrosis¹⁴. To address some of these problems, we have developed a mouse model in which the animals develop MASLD with fibrosis that progresses from pericellular/perisinusoidal (stage 1) to bridging fibrosis (stage 3)¹⁹.

The goal of this article is to describe this model of MASLD with liver fibrosis and two methods to quantify fibrosis: collagen histological staining by picro-sirius red and quantification of liver hydroxyproline content.

The rationale for developing this model was that the most useful animal models are those that best replicate the human disease, both in drivers of the disease and in alterations in histopathology and gene expression. Therefore, we used a diet that resembles the average diet in the US, including moderately high amounts of fat and fructose^19,20. We fed this diet to mice carrying the Agouti yellow (A^y) mutation, which have been extensively used as models of obesity²¹. These mice ubiquitously express the agouti protein, an antagonist of the melanocortin receptors, including MC4R, causing hyperphagia and replicating the increased food intake that is common in human obesity^21,22,23. This model recapitulates the alterations that define MASH, including steatosis, hepatocellular injury, inflammation, fibrosis, and metabolic dysfunction.

The main advantages of this model are the characteristics that make it similar to the human disease. The model uses a Western-type diet, which has been designed to resemble the typical US diet and has been shown to be the most effective diet in replicating the phenotype of human MASLD^14,20. The mice develop liver histopathological alterations similar to those described in the livers of humans with MASH^19,24. The mice develop metabolic dysfunction, which is an essential criterion for the diagnosis of MASLD⁵. The livers of mice with MASH show changes in gene expression that recapitulate those in humans with MASH/NASH¹⁹. A recent study compared 39 mouse models of MASLD for their similarity to the human disease. It included MC4R knock-out mice, which share the same mechanism of hyperphagia and received the same diet as the model that we describe¹⁴. That model ranked fifth (out of 39) for its similarity to human MASLD as assessed by phenotype, liver histopathology, and transcriptomics¹⁴.

Additional advantages of this model are: Ay mice are commercially available from the Jackson Laboratory, and are available in the C57BL/6J strain, which is commonly used in metabolic studies^21,25; mutant mice are easily identifiable by their hair coat color; the model can be combined with models of gain- or loss-of-function to study the role of specific genes in the development of MASLD; and finally, the model does not use nutrient deficiencies (choline or methionine deficiency), unusual food components (cholic acid, ethionine), an unusually high content of specific nutrients (trans fats), or hepatotoxins (carbon tetrachloride), which are not common causes of MASLD and may not replicate the human disease^13,14,17.

This method is appropriate for the investigation of mechanisms that contribute to the development and progression of MASLD as well as for the evaluation of interventions for its treatment or prevention. This model is particularly useful because the animals not only develop both steatohepatitis and metabolic dysfunction but also develop fibrosis that progresses to stage 3. The model may not be appropriate for the study of mechanisms that affect feeding behavior (e.g., GLP1R agonists), since these mice have hyperphagia.

Protocol

This protocol follows the animal care guidelines of Brooklyn College; IACUC protocol number 318.

1. Mouse model of MASLD with fibrosis

1. Breed mice to produce Ay offspring. Set up breeding pairs with Ay mice (B6.Cg-Ay/J). Record the date of birth of the pups.
2. Ensure that litters have similar numbers of pups to promote adequate and even feeding and growth. Aim to have five to six pups per litter, as it is the average litter size for C57BL/6J mice. Extra pups, preferentially those of the sex not needed, can be euthanized or fostered.
   NOTE: Using Ay males and wild-type females (C57BL/6J Jax 0000664) may improve the efficiency of the breeding since wild-type C57BL/6J female mice are easy to obtain and may be better breeders than Ay female mice.
3. Set up cages for experiments as described below.
   1. Identify the Ay mice by their yellow coat color. Assign a unique identification to each mouse and tag or mark the mice.
      NOTE: The agouti yellow (Ay) allele of the agouti gene confers a yellow coat color. The allele is semidominant in heterozygotes. Homozygosity causes embryonic lethality. Therefore, mice with yellow coat color are heterozygous for the Ay gene.
   2. Wean mice when they are 21 days old. Group together mice that will be housed in the same cage during the study. House three or four Ay mice per cage, according to expected body weight and institutional guidelines.
   3. Feed standard chow and water. House mice at standard temperature (20-23 °C) and relative humidity (40% to 60%) with a 12 h light, 12 h dark cycle.
4. Feed the Ay mice the diet high in fat and fructose (High Fat and Fructose Diet, HFFD) starting at 8 weeks of age as described below.
   1. Provide the Ay mice with the Western diet, which is high in fat and sucrose (Adjusted calories Western Diet Inotiv TD.88137) ad libitum.
   2. Prepare a solution containing fructose (23 g/L) and glucose (19 g/L) in drinking water. Sterilize the solution by filtration through a 0.2 µm filter and store it in sterile bottles. Once opened, keep the solution refrigerated.
   3. Provide the mice with the solution containing fructose and glucose for drinking, in sterile drinking bottles. Replace the solution and bottles at least weekly, checking for growth of microorganisms.
   4. Use cage bedding with no nutritional value (e.g., wood chips).
5. Feed control mice a diet low in fat and fructose as described below.
   NOTE: The type of control mice to use depends on the study design: If Ay-HFFD mice are used to evaluate the effect of drugs, the control mice can be Ay-HFFD mice treated with a placebo. If the model is used to study the effect of a genetic modification on the development of fibrosis, Ay-HFFD mice without the genetic modification can be used as controls. If controls without MASH are desired, follow the steps described here.
   1. Use wild-type mice that are littermates of the Ay mice. They can be identified by their black coat color.
   2. Feed the mice a control diet (Low Fat Control Diet Inotiv TD.08485; referred to as Low Fat and Fructose Diet, LFFD) starting at the same age. Provide the mice with regular drinking water.
6. Monitor the development of obesity by weighing the mice weekly by placing the awake animals in a tared container on a laboratory balance and recording their weight.
7. Ensure that the mice have sufficient food and drinking solution at all times. Monitor and record food and drinking solution consumption to estimate caloric intake, if desired.
8. Collect tissues as described below.
   1. Perform tissue collection after 16 weeks on the HFFD diet. Euthanize the mouse by carbon dioxide inhalation (or other methods approved by the institution's IACUC).
      NOTE: After 16 weeks, Ay mice fed the HFFD diet usually have stage 1 fibrosis; prolonging the time will lead to greater fibrosis, progressing to stage 3 after 9 to 12 months.
   2. Weigh the mouse and record the weight¹⁹.
   3. Collect blood if needed. To do this, draw blood from the inferior vena cava with a 1 mL syringe and 27G needle after opening the abdomen and before dissecting the liver (or use any other method approved by the IACUC).
   4. Dissect the liver by opening the abdomen with a midline incision, holding the liver by its hilum with forceps, and cutting the ligaments that bind the liver to the diaphragm and retroperitoneum with scissors. Weigh the liver and record the weight.
   5. Rinse liver in ice-cold phosphate-buffered saline solution (PBS). Photograph the liver and  include a ruler or other object as a size reference in the image. Place the liver in a Petri dish that is placed on ice.
   6. Dissect samples (approximately 100 to 150 mg per sample) of several lobes, including the median lobe, the left lobe, and the right lobe. Ensure samples have a thickness of no more than 4 mm to facilitate fixation. Place them in fixative solution (formalin 10% or formaldehyde 4%) for 24 h and process them for histopathological analysis (see below).
   7. Dissect additional samples of several lobes. Place them in tubes (1.5 or 2 mL), freeze them by submerging them in liquid nitrogen, and store them at -80 °C. These samples will be used to measure hydroxyproline content.
   8. Collect samples of other organs if desired (adipose tissue, muscle, etc.)

2. Collagen histochemical staining and morphometric quantification

NOTE: The steps that require the use of hazardous and/or volatile chemicals should be done in a fume hood. All slides should be processed in parallel using staining dishes and slide holders.

1. Prepare liver sections as described below.
   1. Fix liver tissue by keeping it in formalin (or formaldehyde) for 24 h at room temperature. Use a volume of fixative 20-fold larger than the tissue volume. Next, transfer tissues to 70% ethanol. The protocol can be paused at this point.
      NOTE: Tissues can be placed in histology cassettes that are then placed in a container with fixative or, alternatively, in individual containers or tubes.
   2. Embed tissues in paraffin. This is usually done by a histopathology core facility or service.
   3. Obtain sections with a thickness of 0.5 µm mounted on slides (non-charged slides are adequate). Allow sections to dry by exposing them to air at room temperature. A histopathology core facility or service usually does this. The protocol can be paused at this point.
2. Deparaffinize and rehydrate sections as described below.
   1. Deparaffinize the sections by submerging them in two changes of deparaffinizing/clearing solution (xylene or citrisolv), for 5-15 min in each.
   2. Rehydrate by successively submerging slides in 100% ethanol, 95% ethanol, 70% ethanol, and water, for 5 min in each.
3. Stain sections with picro-sirius red as described below.
   1. Submerge sections in a solution of Sirius red (direct red 80) 1 g/L in picric acid 1.3% for 1 h. Do not stain much longer, as direct red may stain structures other than collagen (e.g., nuclei). Add fast green (1 g/L) to the staining solution if counterstaining is desired²⁶. Alternatively, stain nuclei with Weigert's hematoxylin²⁷.
   2. Return picro-sirius red solution to a closed, dark glass or plastic container for storage. The solution can be reused multiple times (for at least 20 years²⁷).
4. Submerge samples in ethanol (100%) for 5 min to dehydrate. Repeat with a second change of ethanol.
5. Submerge in clearing solution (xylene or citrisolv) until they are individually removed for mounting.
6. Remove one slide at a time from the clearing solution and remove excess liquid by blotting on paper towels. Add mounting medium (approximately 50 to 100 µL). Cover the section with a coverslip and gently press the coverslip down to remove air bubbles. Let air dry.
   NOTE: The protocol can be paused at this point.
7. Observe sections using a light microscope with brightfield illumination and objectives 4x to 20x - collagen appears red over a light yellow-orange background. Image areas of interest with a color camera, according to the camera instructions.
8. Perform imaging with polarized light microscopy as described below.
   1. Engage and adjust the polarizer and analyzer filters orthogonally to ensure that collagen is visible and that the background signal is minimal. Collagen is visible as white, yellow, or orange on a dark background. In sections with little or no fibrosis, collagen in blood vessel walls will be visible. Keep settings, including illumination intensity, the same for all samples to be compared.
   2. Image using a monochrome camera or monochrome mode in a color camera. Use a 4x magnification objective. Take multiple images per section. Avoid areas near the hilum if they have large blood vessels. Use the same settings for all images to be compared.
9. Perform morphometric quantification using Photoshop (alternative a) as described below.
   1. Quantify the white area in the images, which corresponds to collagen. In Photoshop, click Select > Color range, then select Highlights, and adjust the parameters Fuzziness and Range to select pixels that are white, corresponding to collagen. Next, select Window > Histogram > Expanded View and record the number of pixels selected, shown next to the parameter Pixels. Use the same parameters for all images to be compared.
   2. Quantify the total area of the image. In Photoshop, click Select > All, then record the total number of pixels as previously done.
   3. Calculate the percentage of the image area corresponding to collagen by dividing the number of white pixels by the number of total pixels. Values are usually expressed as area percentages.
10. Perform morphometric quantification using ImageJ (alternative b) as described below.
    1. Select white areas in the images that correspond to collagen. Open the Threshold tool by clicking Image > Adjust > Threshold (or press Shift + T). Set the threshold range to ensure only collagen-stained regions (white areas) are highlighted in red (e.g., minimum threshold 64; maximum threshold: 255) and click Apply. This will binarize the image, converting the collagen region to black and the background to white.
    2. Measure the Total Image Area by clicking Analyze > Set Measurements. Check Area and Display Label. Click OK. Click Analyze > Measure (or press M on the keyboard). The Results window will display the total image area.
    3. To measure the collagen area and percentage, first, enable Limit to Threshold by clicking Analyze > Set Measurements. Check Area, Area Fraction (ImageJ will calculate collagen % automatically), Limit to Threshold, and Display Label. Click OK. Then, click Analyze > Measure (or press M). The Results window will show: Collagen Area (absolute pixel count) and Area Fraction (collagen% of the total image).
    4. Save the results by selecting Save As or copy them to the clipboard and paste them in a spreadsheet.

3. Liver hydroxyproline quantification

NOTE: The steps that require the use of strong bases and acids should be conducted with appropriate personal protective equipment.

1. Weigh liver samples prepared in step 1.8.7 and record weights. Place tissue aliquots in a tube appropriate for homogenization. Add deionized water using a water-to-tissue ratio of 900 µL of water per 100 mg of tissue.
2. Homogenize by bead beating or other method according to the manufacturer's instructions until the tissue is completely homogenized, as assessed by visual observation. The protocol can be paused at this point, and homogenates can be stored at -20 °C.
3. Perform protein precipitation as described below. The following steps are done at room temperature, except when a different temperature is specified.
   1. Transfer approximately 500 µL to 1000 µL of homogenate to 1.5 mL centrifuge tubes. Transfer the same volume for all samples.
   2. Add trichloroacetic acid for a final concentration of 12% (e.g., add 120 µL of trichloroacetic acid to 880 µL of homogenate)²⁸. Vortex briefly to mix.
   3. Incubate samples in an ice/water bath for 30 min. Centrifuge at 6,000 x g at 4 °C for 10 min. A precipitate containing the protein forms. Remove supernatant by aspiration. Keep tubes on ice.
   4. Add 1 mL of ice-cold ethanol (100%) to wash the pellet. Resuspend the pellet by sonication (or other method) and verify visually that the material is resuspended. Centrifuge at 6,000 x g at 4 °C for 10 min and remove supernatant by aspiration. Use sonication settings according to the equipment instructions (e.g., using a Branson SFX 150 sonifier with a 3.2 mm diameter microtip, an amplitude of 70% for 10 s is effective for most samples). Repeat washing 2x.
   5. Let the pellet air-dry for 10 min, or longer if necessary.
4. Perform protein hydrolysis as described below.
   1. Add 800 µL of 6 N hydrochloric acid to the tubes containing the protein precipitate. Resuspend the pellet by sonication (or other method), using the same settings as in step 3.3.4.
   2. Transfer protein suspension to glass tubes or vials with screw caps; close tightly. Incubate tubes in oven at 110 °C for 22 ± 2 h²⁹.
   3. Centrifuge tubes at 16,000 x g for 10 min; particulate will form a pellet. This usually requires transferring the hydrolysate from the glass vials to centrifuge tubes before centrifugation.
   4. Transfer the supernatant, which contains hydroxyproline, to 1.5 mL tubes. The protocol can be paused at this point. Hydrolysates can be stored at 4 °C.
5. Prepare a standard curve for hydroxyproline concentration. For liver samples with fibrosis stage 1 to 3, using the amounts of tissue indicated in this protocol, standard concentrations up to 125 µg/mL are adequate (e.g., serial dilutions of 125, 62.5, 31.3, 15.6, 7.81, 3.91, 1.95, and 0 µg/mL). Prepare the standards by diluting hydroxyproline in 6 N hydrochloric acid.
6. Transfer 40 µL of each sample and standard to 1.5 mL labeled centrifuge tubes. Add 10 µL of 10 N sodium hydroxide to each and vortex briefly to mix.
7. Perform oxidation of hydroxyproline as described below.
   1. Prepare acetate-citrate buffer containing acetic acid 1.2% v/v, citric acid 46 g/L, sodium acetate trihydrate 120 g/L, and sodium hydroxide 34 g/L in deionized water, and adjust buffer pH to 6.5.
   2. Prepare a solution of 12.7 g/L of chloramine T in water:n-propanol:acetate-citrate buffer (1:1:8 v:v:v). First, dissolve chloramine T in water by mixing, then add n-propanol, mix, and then add acetate-citrate buffer and mix.
   3. Add 450 µL of chloramine solution to tubes with samples and standards. Mix by vortexing. Incubate at room temperature for 25 min³⁰.
8. Prepare Ehrlich's reagent with 1 M 4-(Dimethylamino) benzaldehyde in perchloric acid:propanol 1:2 (v/v). Add 500 µL of Ehrlich's reagent to samples and standards. Mix by vortexing.
9. Incubate at 65 °C for 20 min in a heating block, water bath, or incubator. When the incubation ends, let samples and standards cool to room temperature.
10. Measure the absorbance of samples and standards at a wavelength of 550 nm using a spectrophotometer.
11. Plot absorbance versus standard amounts and use a linear regression to make a standard curve. Use the parameters from the standard curve to calculate the concentration of hydroxyproline in the samples.
12. Calculate the amount of hydroxyproline in the liver samples based on the total volume of the hydrolysate (step 3.4) and normalize by the mass of tissue corresponding to the volume of homogenate used for the protein precipitation (step 3.3.1). Express the results as ng of hydroxyproline per mg of tissue. Alternatively, the concentration of protein in homogenates can be measured and the results expressed as mass of hydroxyproline per mass of protein.

Results

Ay mice fed the HFFD gain more weight than wild-type mice fed the LFFD and develop obesity (Figure 1A). They gain weight more rapidly during the first 4 weeks of HFFD feeding, gaining approximately 3-4 g/week. After 8 weeks, they gain weight at a rate similar to control mice (approximately 0.6 g/week)¹⁹. After 16 weeks, the Ay mice fed the HFFD weigh approximately 45 g to 50 g (average of three separate experiments) and weigh approximately 40% to 80% more than control...

Discussion

This protocol describes a mouse model of MASLD with liver fibrosis that replicates the disease in humans^5,6. The mice develop liver steatosis, as shown by mixed micro- and macro-vesicular steatosis in histology and elevated hepatic triacylglycerol content¹⁹. The mice also develop obesity, hypertriglyceridemia, glucose intolerance, and have been reported to have hypertension, which, in combination with liver steatosis, correspond to the dia...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
4-(Dimethylamino)benzaldehyde	Sigma-Aldrich	D2004
Acetic acid	Various	NA
Adjusted calories diet Western diet	Inotiv	TD.88137	Diet containg fat (42 %kcal) and sucrose (34 %w/w). Other suppliers provide diets with the same composition
B6.Cg-Ay/J	The Jackson Laboratory	000021	Strain has the Agouti lethal yellow mutation and a C57BL6/J background. Currently, The Jackson Laboratory has the strain cryproserved.
C57BL/6J	The Jackson Laboratory	000664
Centrifuge	Various	NA	With rotor for 1.5 mL tubes, refrigerated, able to reach 16,000 xg.
Centrifuge	Various	NA
Chloramine-T hydrate	Sigma-Aldrich	857319
Citric acid	Various	NA
Direct Red 80	Sigma	365548-5G
Ethanol	Decon Laboratories	2701
Formalin solution, neutral buffered	Sigma	HT501128
Fructose	Sigma-Aldrich	F0127
Glucose	Sigma-Aldrich	G8270
Histology processing equipment	NA	NA	Embedding and sectioning can be done in the investigator lab or external facility
Homogenizer	Various	NA	Bead beaters are efficient to process multiple samples. Other homogenization methods can be used
Hydrochloric acid	Various	NA
Hydroxyproline	Sigma	H5534
Low Fat Control Diet	Inotiv	TD.08485	Control diet, low in fat (13 %kcal) and sucrose (12 %w/w)
Microscope equiped for polarized light microscopy	Various	NA
N-propanol	Various	NA
Perchloric acid	Various	NA
Permount Mounting Medium	Fisher Scientific	SP15-100
Picric Acid, Saturated Aqueous Solution, Spectrum	Fisher	18612375
Sodium acetate trihydrate	Various	NA
Sodium hydroxyde	Various	NA
Sonicator	Various	NA	Sonicator is useful to resuspend pellets in step 2.5, but not absolutely necessary
Trichloroacetic acid	Various	NA
VWR Micro Cover Glasses, Rectangular	VWR	48393-081
VWR Premium Superfrost Plus Microscope Slides | VWR	VWR	48311-703
Water, deionized	Various	NA	Deionized water from water purification system or purchased from a supplier.
Xylenes	Fisher Scientific	X3F	Citrisolv Clearing Agent (Fisher 22-143-975) is sold a an environmentally-friendly alternative