Loss of Mtarc1 Protects Against Steatotic Liver Disease in Mice

Abstract

Background & Aims:

Metabolic dysfunction-associated steatotic liver disease (MASLD) spans from simple steatosis to metabolic dysfunction-associated steatohepatitis (MASH) and can progress to cirrhosis or hepatocellular carcinoma. Despite its prevalence, effective therapies are lacking. Recent genome-wide association studies identified a common missense variant (rs2642438) in the Mitochondrial Amidoxime Reducing Component 1 (MTARC1) gene that protects against liver cirrhosis without increasing cardiovascular disease risk. Biochemical and disease risk signatures associated with carriers of this missense variant also aligned with those of a known loss-of-function MTARC1 variant, suggesting mARC1 inhibition as a potential MASLD treatment.

Methods:

To validate mARC1 loss-of-function as protective against MASLD, we generated Mtarc1 knockout (KO) mice and placed them on a choline-deficient, L-amino acid-defined, high-fat diet (CDAHFD). Effects of Mtarc1 KO on obesity and type 2 diabetes were explored using a high-fat diet. Hepatocytes from Mtarc1 KO mice were isolated to explore the molecular mechanisms by which Mtarc1 KO impacts lipid metabolism.

Results:

Mtarc1 KO mice exhibited no vital growth or development defects. With a high-fat diet-induced obesity model, obese Mtarc1 KO mice exhibited reduced liver mass and lower cholesterol levels, with no effect on glucose homeostasis. In a CDAHFD-induced MASLD model, mARC1 deficiency significantly reduced liver steatosis, profibrosis, and inflammation. Untargeted metabolomics profiling further showed hepatic enrichment of phospholipids in Mtarc1 KO mice. Primary hepatocytes isolated from Mtarc1 KO mice exhibited reduced lipid droplet accumulation, decreased fatty acid uptake, and increased lipid secretion.

Conclusions:

These findings support mARC1 inhibition as a promising therapeutic strategy for MASLD/MASH.

Lay Summary

Mtarc1 knockout provides hepatic benefits in obese mice without affecting glucose homeostasis. mARC1 deficiency decreases hepatic steatosis, profibrosis, and inflammation in a diet-induced MASH model. Decreased uptake of circulating fatty acids and enhanced lipid secretion may contribute to the reduced lipid accumulation in Mtarc1 knockout primary mouse hepatocytes.

Introduction

Metabolic dysfunction-associated steatotic liver disease (MASLD) is the leading cause of chronic liver disease globally and is fueled by the epidemics of obesity and type 2 diabetes ^1,2. MASLD includes a variety of conditions caused by an excessive lipid buildup in hepatocytes, ranging from simple steatosis to steatohepatitis (MASH), with some patients progressing to cirrhosis and hepatocellular carcinoma ^3,4. The worldwide prevalence of MASLD has been rising over time, with an estimated 32% of adults affected according to a recent meta-analysis ⁵. MASH is anticipated to become the leading cause of liver transplantation in the US between 2020 and 2025 ⁶ and to increase by 63% between 2015 and 2030 ⁷. Currently, there is only one FDA conditionally approved drug, Resmetirom, available for the treatment of adults with noncirrhotic MASH with moderate to advanced liver fibrosis ⁸.

Multiple genome-wide association studies (GWAS) reported that a common missense variant (rs2642438 G>A, p.A165T) in the MTARC1 (mitochondrial amidoxime-reducing component 1) gene, which encodes the mARC1 enzyme, was associated with a decreased risk of MASLD ^9,10 and liver cirrhosis ¹¹, without raising the risk of cardiovascular outcomes ^11,12. Carriage of this variant is also associated with reductions in liver fat ^11,13, enzymatic markers of liver damage ^11,14, and blood cholesterol ^11,12. Furthermore, several GWAS showed that the phenotypic effects of the two rare loss of function variants (p.R200ter ¹¹ and p.R305ter ¹⁵) closely resemble those of the p.A165T missense variant. Recent studies suggest that the p.A165T variant might cause mARC1 protein instability ^16,17. In aggregate, these findings suggest that inhibition of mARC1 may have therapeutic potential for the treatment of liver disease.

MTARC1 (mouse ortholog Mtarc1) encodes a molybdenum-containing enzyme located in the outer mitochondrial membrane in eukaryotes. mARC1, in combination with NADH cytochrome b5 reductase and cytochrome b5 type B, was initially identified as an enzyme with reductive activity that could activate N-hydroxylated prodrugs or reduce nitrite to produce nitric oxide ^18,19. However, its physiological function in MASLD remains unknown, and it is uncertain whether the rs2642438 variant impacts the N-reductive activity of mARC1 protein. Knockout of mARC2, a mARC1 paralog, was found to protect mice from diet-induced obesity and metabolic disturbances ²⁰, suggesting that mARC2 has an important impact on lipid metabolism. Liver-specific siRNA-mediated Mtarc1 knockdown in mice demonstrated an apparent reversal of diet-induced MASH ^21–23. However, no publicly accessible data exists regarding the comprehensive in vivo assessment of mARC1.

Here, we developed and characterized Mtarc1 whole-body knockout mice. We discovered that Mtarc1 KO promotes a beneficial liver phenotype in obese mice and protects against MASLD. Further, Mtarc1 KO protects against progression to MASH by reducing steatosis and markers of hepatic profibrosis and inflammation, as demonstrated by various diet-induced disease models. We then applied non-targeted metabolomics to obtain a more comprehensive view of the metabolic impact of Mtarc1 knockout and revealed an enrichment in phosphatidylcholine and phosphatidylethanolamine in Mtarc1 KO mouse livers. Lastly, we isolated primary mouse hepatocytes and observed differences in lipid transport and handling involved in decreased lipid accumulation. Collectively, this study presents the initial experimental evidence that the absence of mARC1 exerts a direct protective effect against liver damage, as opposed to merely modulating other risk factors (e.g., obesity).

Methods

Generation of Mtarc1 KO and Mtarc2 KO mice

Whole-body knockout (KO) mice were generated by CRISPR/Cas9-mediated targeting of exon 2 in Mtarc1 or Mtarc2 on the C57BL/6J background (Janelia Gene Targeting & Transgenic Facility). Founders were backcrossed for two generations, and genotyping was performed by PCR (see Supplementary Methods).

Determination of mARC1/2 levels by proteomics

Mouse liver lysates were subjected to targeted LC-MS/MS. Three unique peptides per protein were quantified and normalized to GAPDH (see Supplementary Methods).

In vivo diet-induced disease models

Male Mtarc1 wildtype (WT) and knockout (KO) mice were fed high-fat (HFD, 60% kcal fat) or low-fat (LFD, 10% kcal fat) diets for 16 weeks. Body weight, composition, glucose tolerance, plasma biochemistry, and tissue histology were assessed. To model liver disease, wildtype (WT), heterozygous (Het), and knockout (KO) mice were fed a choline-deficient, amino acid-defined, high-fat diet (CDAHFD) for 2–11 weeks with serial sampling (see Supplementary Methods).

Body composition and metabolic phenotyping

Fat and lean mass were determined by EchoMRI^™ analyzer. Glucose tolerance tests were performed after 6-hour fasting with intraperitoneal glucose injection. Plasma insulin was measured by ELISA (see Supplementary Methods).

Molecular assays

RNA was extracted from liver and adipose tissue for qPCR. Library preparation and sequencing for transcriptomic profiling were conducted at Azenta Life Sciences. Differential expression was analyzed using limma-voom with FDR correction, and pathway enrichment by GSEA (see Supplementary Methods).

Histology and image quantification

Formalin-fixed tissues were paraffin-embedded and stained with H&E or Masson’s Trichrome. Lipid droplets (CellProfiler) and fibrosis (ImageJ) were quantified from digitized slides.

Metabolomics

Liver metabolites from CDAHFD-fed mice were profiled using four liquid chromatography-tandem mass spectrometry (LC-MS) methods (Broad Institute Metabolomics Platform) (see Supplementary Methods).

Primary mouse hepatocyte studies

Hepatocytes were isolated from adult WT and Mtarc1 KO mice by collagenase perfusion. Cells were subjected to free fatty acid treatment, Oil Red O staining, fatty acid uptake assays, and ApoB secretion measurements (see Supplementary Methods).

Statistics

All data represent at least two separate experiments performed in triplicate. Results are presented as mean ± SEM. All statistical analyses were performed with GraphPad Prism 10 software. A P-value of <0.05 was considered statistically significant. For 2-group comparisons, Student’s unpaired 2-tailed t test was used. When comparing more than two groups, a one-way analysis of variance (ANOVA) with Tukey’s post-hoc test was performed. A two-way ANOVA with various multiple comparison tests was performed to compare genotype- and treatment-dependent effects, as indicated in each figure legend. Representative images were selected to represent the means of the quantified data.

Results

Characterization of Mtarc1 KO and Mtarc2 KO mice

To better understand the role of mARC1 in development and liver disease, we generated Mtarc1 whole body knockout (KO) mice by CRISPR/Cas-mediated genome engineering. In brief, C57BL/6J embryos were injected with Cas9 mRNA and two sgRNAs that specifically targeted exon 2 of Mtarc1, inducing an indel/frameshift loss of function. Creation of the indel was confirmed by sequencing of the PCR products using primers that span the targeting site (Supp Fig. 1A). Genotypes were determined by PCR analysis using mouse-tail DNA (Supp Fig. 1C). Homozygous Mtarc1 KO mice were viable and fertile and did not exhibit any noticeable phenotypic abnormalities, indicating that the deletion of the gene does not play a crucial role in development and growth. Specifically, there was no obvious difference in body weights between heterozygotes, homozygotes, and wildtype (WT) mice on a normal diet (Fig. 1A). Loss of function at the mRNA transcript level was validated by quantitative PCR across various tissues (Fig. 1B). In contrast to its human homolog, which was found to be most abundant in adipose tissue (GTEx Analysis Release V8 (dbGaP Accession phs000424.v8.p2)), mouse mARC1 was most highly expressed in the liver, followed by epididymal and inguinal white adipose tissue. Absence of mARC1 at the protein level was validated by targeted proteomics (Fig. 1C).

Figure 1. Characterization of Mtarc1 KO and Mtarc2 KO mice.

A) Body weight of Mtarc1 wildtype (+/+), heterozygous (+/−), and knockout (−/−) mice at 8 weeks. B) Expression of Mtarc1 in liver (left) and other organs (right) by RT-qPCR (n=3). C) Relative abundances of mARC1 and mARC2 protein levels in Mtarc1 mouse liver (n=3) in both sexes. Values were derived from peak areas of unique peptides in a targeted proteomics experiment. D) Body weight of Mtarc2 wildtype (+/+), heterozygous (+/−), and knockout (−/−) mice at 8 weeks. E) Percent survival of Mtarc1 knockout (KO) male (n=10), Mtarc1 KO female (n=10), Mtarc2 KO male (n=5), and Mtarc2 KO female (n=10) mice. Data are presented as mean ± SEM and analyzed by one-way ANOVA followed by Tukey’s post-hoc test (B-D); *P-value < 0.05. eWAT, epididymal white adipose tissue; iWAT, inguinal white adipose tissue; FD, found dead; Mal, malocclusion; HL, hindlimb paralysis (Euthanasia is required for malocclusion and hindlimb paralysis).

All mammalian genomes encode two mARC proteins, mARC1 and mARC2, which share a high degree of sequence similarity. Our proteomics data showed that mARC2 protein level was not affected in Mtarc1 KO mice (Fig. 1C). To compare the physiological functions of mARC1 and mARC2, Mtarc2 KO mice were produced employing the same CRISPR/Cas strategy. The absence of mARC2 expression was verified by gene sequencing (Supp Fig. 1B and 1C) and RT-qPCR (data not shown). It has been reported that Mtarc2 KO mice display elevated glucose levels, reduced total cholesterol levels, lower body weight, and elevated body temperature ²⁰. In line with previous research, our Mtarc2 KO mice exhibited a decrease in body weight for both male and female mice (Fig. 1D). Furthermore, we observed postnatal mortality and episodic hindlimb paralysis in mice lacking mARC2 but not mARC1 (Fig. 1E), consistent with previously reported findings ¹⁷.

Mtarc1 KO modulates the liver phenotype in obese mice, with no impact on glucose homeostasis

In a prior work, the common coding variant (MTARC1 p.A165T) was protective for all-cause cirrhosis (OR 0.91, p=2.3*10⁻¹¹), but was associated with a mild increased risk of developing type 2 diabetes (T2D, OR 1.03, p=0.04) in MTARC1 variant carriers ¹¹. This observation raised the possibility that treatment of cirrhosis might have adverse effects on glucose metabolism. We therefore investigated whether mARC1 deficiency is associated with an increased T2D risk in a diet-induced obesity mouse model. Male Mtarc1 WT or KO mice were fed a high-fat diet (HFD) or low-fat diet (LFD) for 16 weeks. On HFD, Mtarc1 KO mice gained the same amount of weight as WT control mice from baseline, compared to the LFD groups (Fig. 2A). There was no significant difference in body composition between KO and WT mice either on HFD or LFD (Supp Fig. 2A and 2B). After 15 weeks of the diet, there was no difference between KO and WT mice in terms of glucose tolerance test (Fig. 2D), fasting blood glucose or insulin levels (Fig. 2E), suggesting that the absence of mARC1 did not have an impact on glucose homeostasis in vivo. It is noteworthy that liver weights were substantially reduced in Mtarc1 KO mice on HFD, whereas white adipose tissue weights remained unaffected (Fig. 2B and Supp Fig. 2C). Accordingly, there was significantly less lipid droplet accumulation in the Mtarc1 KO liver relative to the WT liver on HFD (Fig. 2C). While not statistically significant, a decline in circulating ALT level (Supp Fig. 2D) as well as in the expression of the inflammation marker gene Ccl2 (Supp Fig. 2E) was observed in liver tissues from Mtarc1 KO mice. Taken together, our findings suggest that mARC1 loss of function did not affect glucose homeostasis, but rather modulated the liver phenotype in obese mice.

Figure 2. Mtarc1 KO provides increased benefits on liver phenotype in obese mice, with no impact on glucose homeostasis.

Mtarc1 wildtype (WT) and knockout (KO) male mice were fed a high-fat diet (HFD) or low-fat diet (LFD) for 16 weeks (n=9-11 per group). A) Body weight growth curve over 16-week diet. B) Liver mass at the end of treatment. C) Representative hematoxylin & eosin staining of mouse liver tissue (left) and quantification of lipid droplet areas (%) on the H&E staining (right). Scale bar: 50 μm. D) Blood glucose level throughout glucose tolerance test (GTT) performed at week 15 on diet (left) and quantification of area under the curve (AUC, right). E) Fasting blood glucose level (left) and fasting plasma insulin level (right) before GTT. Data are presented as mean ± SEM and analyzed by two-way ANOVA followed by Tukey’s post-hoc test; ns, not significant, *P-value < 0.05.

Mtarc1 KO decreases steatosis in a diet-induced MASH model

To assess the specific impact of mARC1 dysfunction in liver tissue, a choline-deficient, L-amino acid-defined, high-fat diet (CDAHFD) was used to develop a mouse model of metabolic dysfunction-associated steatohepatitis (MASH). It has been reported that within 6 weeks, this diet increased liver fat accumulation, inflammatory markers, and circulating levels of aspartate aminotransferase (AST) and alanine aminotransferase (ALT) while maintaining energy intake and body weight. Moreover, this diet caused fibrosis in C57BL/6J and A/J mice at week 6 and week 9, respectively ²⁴. On the other hand, feeding mice on CDAHFD for only 1 week also induced steatohepatitis accompanied by severe oxidative stress and mitochondrial dysfunction, without measurable fibrosis ²⁵. In this study, we administered a CDAHFD for 2 weeks to female mice of various Mtarc1 genotypes. In comparison to WT mice on a chow diet, livers dissected from WT mice on CDAHFD displayed significant color changes from the typical dark red to a brown-white, as well as an enlargement in size, which may indicate lipid accumulation (Fig. 3A). In contrast, Mtarc1 KO mice on CDAHFD exhibited darker and smaller livers than WT mice on CDAHFD. The histological analysis of liver specimens from Mtarc1 KO mice also demonstrated a 39% reduction in the accumulation of lipid droplets compared to WT mice on CDAHFD (Fig. 3A, 14.9% lipid droplets in KO vs. 24.6% in WT, p-value=0.1587). While the results did not reach statistical significance, which likely reflects the inherent variability among mice, the limited number of samples included in the histological analysis, and the lack of a priori power calculation, the observed trends are consistent with our hypothesis. We noticed a decrease in the hepatic expression of the Mtarc1 and Mtarc2 genes in response to CDAHFD treatment. Meanwhile, Mtarc1 KO had no effect on the mRNA level of Mtarc2, suggesting no compensatory role of Mtarc2 (Fig. 3B). In addition, CDAHFD increased the liver-to-body weight ratio in WT mice without affecting the body weight. In contrast, the liver-to-body weight ratio was significantly reduced in Mtarc1 KO mice compared to WT mice on CDAHFD (5.3% vs 6.6%, p-value<0.0001), as was the absolute liver weight (1.156 grams vs 1.398 grams, p-value=0.0051) (Fig. 3C). Furthermore, plasma ALT and triglyceride levels, elevated by CDAHFD, were decreased in Mtarc1 KO mice compared to WT mice on the diet (Fig. 3D; 179.5 mg/dL ALT vs 268.1, p-value=0.0028; 47.4 mg/dL triglycerides vs 60.2, p-value=0.0393). These results suggest that liver damage and the deposition of lipid droplets were induced in mice fed a CDAHFD for 2 weeks, and mARC1 deficiency significantly alleviated these effects. Analysis of mARC1 KO mice on a longer CDAHFD, for either 4 or 11 weeks, further corroborated our findings on a shorter diet exposure (Supp Fig. 3).

Figure 3. Mtarc1 KO decreases steatosis in a diet-induced MASH model.

Mtarc1 wildtype (WT), heterozygous (Het), and knockout (KO) female mice were fed a choline-deficient, L-amino acid-defined, high-fat diet (CDAHFD) or control chow diet for 2 weeks (n=8-15 per group). A) Representative gross liver pictures (top left) and hematoxylin & eosin staining of mouse liver tissue (bottom left), with quantification of lipid droplet area (%) on the H&E staining (right). Scale bar: 100 μm. B) Expression of Mtarc1 and Mtarc2 in the liver by RT-qPCR. C) Percentage of body weight change on 2-week diet (left), liver weight (center), and percentage of liver weight over body weight (right). D) Plasma ALT and triglyceride levels at the end of CDAHFD challenge. Individual datapoints represent each mouse, with mean ± SEM also shown. Data was analyzed by one-way ANOVA followed by Tukey’s post-hoc test; ns, not significant, *P-value < 0.05. H&E, hematoxylin and eosin; ALT, alanine aminotransferase; TRIG, triglyceride.

Mtarc1 KO decreases liver profibrosis and inflammation in a diet-induced MASH model

A previous study demonstrated that liver fibrosis was not induced by 1 week of CDAHFD treatment ²⁵. In our study, we observed a mild increase in hepatic fibrosis by Masson’s trichrome staining in WT mice that were fed CDAHFD for 2 weeks. Functional deletion of mARC1 resulted in a trend toward reduction in fibrosis from histological measurement (p-value>0.05, Fig. 4A). Gene expression analysis in the liver revealed that CDAHFD significantly increased profibrosis and inflammatory markers in WT mice. Furthermore, the expression of profibrosis marker genes (Tgfβ1, αSMA, Col1a1) and inflammation marker genes (Tnf-α, Mcp1) that were induced by CDAHFD was drastically decreased with mARC1 loss of function (Fig. 4B and Fig. 4C, Tgfβ1: 2.4 vs 3.1 fold-change, p-value=0.0008; αSMA: 1.4 vs 2.1 fold-change, p-value=0.0001; Col1a1: 9.4 vs. 18.6, p-value=0.0056; Tnf-α: 10.9 vs 14.2 fold-change, p-value=0.07; Mcp1: 11.4 vs 16.1 fold-change, p-value=0.07). These results demonstrate that C57BL/6J mice developed early MASH without fibrosis after 2 weeks on CDAHFD, and mARC1 deficiency decreased the expression of marker genes for both profibrosis and inflammation.

Figure 4. Mtarc1 KO decreases liver profibrosis and inflammation in a diet-induced MASH model.

Mtarc1 wildtype (WT), heterozygous (Het), and knockout (KO) female mice were fed a choline-deficient, L-amino acid-defined, high-fat diet (CDAHFD) or control chow diet for 2 weeks (n=8-15 per group). A) Representative Masson’s trichrome staining of mouse liver tissue (left) and quantification of fibrotic area (%) on the histological sections (right). Scale bar: 50 μm. Expression of (B) profibrosis marker genes Tgfβ1, αSMA and Col1a1 and (C) inflammatory marker genes Tnf-α and Mcp1 in the liver at the end of CDAHFD challenge. Data are presented as mean ± SEM and analyzed by one-way ANOVA followed by Tukey’s post-hoc test; ns, not significant, *P-value < 0.05.

Mtarc1 KO results in transcriptional changes in metabolic and inflammatory pathways

To determine the global transcriptional impact of mARC1 loss of function, liver tissues from Mtarc1 KO and WT mice fed a CDAHFD or chow diet for 2 weeks were subjected to bulk RNA sequencing. On CDAHFD, Mtarc1 was the only gene identified as significantly differentially expressed in the liver of KO mice compared to WT mice (Supp Fig. 4A). However, gene set enrichment analysis (GSEA) employing the t-statistic ranking revealed that several metabolic pathways (e.g., amino acid metabolic process and tRNA metabolic process) are significantly upregulated in the KO-CDAHFD group relative to the WT-CDAHFD group. Furthermore, several inflammatory pathways were significantly downregulated in KO-CDAHFD mice compared to WT-CDAHFD mice (Supp Fig. 4B). This suggests that mARC1 functions through non-transcriptional mechanisms, exerting its effects at the pathway level rather than by regulating individual genes.

Mtarc1 KO increases hepatic phosphatidylcholine and phosphatidylethanolamine levels

To gain insight into the metabolic effects of Mtarc1 knockout, non-targeted metabolomics based on liquid chromatography-mass spectrometry (LC-MS) was applied to compare the metabolic profiles of liver tissues in Mtarc1 KO and WT mice subjected to a CDAHFD for 2 weeks. A total of 581 distinct known metabolites were identified using 4 different LC-MS methods (C8-pos, C18-neg, HILIC-pos, and HILIC-neg) (Supp Table). Among hundreds of confirmed known metabolites, phosphatidylcholine (PC) and phosphatidylethanolamine (PE) were of particular interest to us, given that alterations in hepatic phospholipid composition have been associated with the development of MASLD in humans ²⁶. Lipidomic analysis revealed that the total hepatic PC level decreased in MASLD patients ²⁷.

It is noteworthy that MTARC1 p.A165T variant carriers were found to have elevated levels of hepatic polyunsaturated PCs in comparison to non-carriers ²⁸. In accordance with this, we observed that not only the total hepatic PCs level (Supp. Fig. 5A), but also a majority of the significant hits in the PC and PE species were enriched in the livers of Mtarc1 KO mice of both sexes after 2 weeks of CDAHFD (Fig. 5A and 5B), whereas the total levels of cholesterol esters, triglycerides, and the different fatty acids remained largely unchanged (Supp. Fig. 5B). In addition, female Mtarc1 KO mice that were fed CDAHFD for either 4 weeks or 11 weeks exhibited comparable hepatic enrichment of PC and PE (Fig. 5C and 5D). However, we did not observe gene expression differences in enzymes involved in PC synthesis from the RNAseq data (Supp. Fig. 6A). Our data indicate that the Mtarc1 KO mouse model replicated the correlation observed in humans between MTARC1 variant carriers and elevated levels of hepatic PC. This suggests that mARC1 may have a significant impact on protection against MASLD by regulating hepatic PCs (and PEs).

Figure 5. Mtarc1 KO increases hepatic phosphatidylcholine and phosphatidylethanolamine levels with CDAHFD challenge.

Mtarc1 wildtype (WT) and knockout (KO) mice were fed a choline-deficient, L-amino acid-defined, high-fat diet (CDAHFD) for indicated weeks (n=9-21 per group). Volcano plots of LPC and PC (top) or LPE and PE (bottom) fold-change in liver of male mice at 2 weeks on CDAHFD (A), or female mice at 2 weeks (B), 4 weeks (C), and 11 weeks (D) on CDAHFD. Red indicates significant enrichment in KO mice, while blue indicates significant enrichment in WT mice. Each data point indicates a different metabolite as detected by C8-pos or HILIC-pos Mass spectrometry method. Data is analyzed by Student’s unpaired t test, corrected for multiple testing by group using a Benjamini-Hochberg FDR adjustment. The gray dotted line represents the cutoff for significance after Benjamini-Hochberg FDR correction. LPC, lysophosphatidylcholine; PC, phosphatidylcholine; LPE, lysophosphatidylethanolamine; PE, phosphatidylethanolamine.

Mtarc1 KO decreases hepatic lipid accumulation via reduced fatty acid uptake and increased fatty acid secretion

To elucidate the molecular mechanism underlying the protection against steatotic liver disease in mice lacking mARC1, primary hepatocytes were isolated from both Mtarc1 KO and WT mice. Consistent with the findings observed in vivo (Fig. 3A), Mtarc1 KO significantly reduced cellular lipid accumulation in the presence of palmitic acid, as compared to WT cells (Fig. 6A), in agreement with previously reported findings in GalNac-siMTARC1-treated primary human hepatocytes ²¹. Four pathways are thought to contribute to the disruptions in hepatic lipid homeostasis: 1) increased flux of fatty acid into the liver; 2) increased de novo lipogenesis (DNL) of hepatic lipids; 3) decreased secretion of triglycerides in the form of very low-density lipoprotein (VLDL) particles; and 4) decreased fatty acid oxidation ²⁹. Strikingly, a substantial 32% reduction in fatty acid uptake was observed in hepatocytes lacking mARC1 (Fig. 6B). Furthermore, evaluation of the conditioned media revealed that Mtarc1 KO hepatocytes significantly increased secretion of ApoB, a factor crucial for the assembly and secretion of VLDL in the liver (Fig. 6C). Our RNAseq data also showed a trend toward decreased Cd36 expression (which encodes a protein crucial for fatty acid transport across the cell membrane) in KO liver tissue compared to WT on CDAHFD (Supp. Fig. 6B), while genes involved in fatty acid oxidation and TG synthesis showed no differences (Supp. Fig. 6C and 6D). These data suggest that Mtarc1 KO reduced hepatic lipid accumulation as a result of decreased fatty acid uptake and increased fatty acid secretion.

Figure 6. Mtarc1 KO decreases hepatic lipid accumulation via reduced fatty acid uptake and increased VLDL secretion.

A) Representative Oil Red O staining of primary Mtarc1 KO and WT mouse hepatocytes treated with either BSA (200 μM) or palmitic acid (PA, 200 μM) for 6 hours (left), with quantification of lipid droplet areas (%) on the ORO staining (right). Scale bar: 50 μm. (n = 3 independent experiments) B) Representative kinetic (real-time) fatty acid uptake in primary Mtarc1 KO and WT mouse hepatocytes (n=3 independent experiments). C) Media ApoB protein level secreted from primary Mtarc1 KO and WT mouse hepatocytes 6 hours after fatty acid treatment (n=4 independent experiments). Data are presented as mean ± SEM and analyzed by two-way ANOVA followed by Sidak’s multiple comparison test (lipid droplet quantification and ApoB secretion) or Bonferroni’s post-hoc test (fatty acid uptake); *P-value < 0.05.

Discussion

Since first being isolated and identified in 2006, the molybdoenzyme mARC1, with its N-reductive capacity, has been explored as an important drug-metabolizing enzyme ¹⁸. Recently, it was demonstrated that mARC1 is a key factor in lipid metabolism and is involved in human liver disease, specifically metabolic dysfunction-associated steatotic liver disease (MASLD) and metabolic dysfunction-associated steatohepatitis (MASH) ^11,12. Nevertheless, the specific molecular mechanism underlying mARC1’s effect on lipid metabolism is still unknown. In this report, we have generated and characterized Mtarc1 knockout mice to determine the role of mARC1 in the development and progression of MASLD. By utilizing a high-fat diet-induced obesity model, we demonstrated that mARC1 loss of function did not affect glucose homeostasis, but rather conferred modulation of liver injury in obese mice. Furthermore, mARC1 deficiency decreased hepatic profibrotic and inflammatory responses and protected against steatosis in a diet-induced MASH model. Mechanistically, we observed that hepatocytes lacking mARC1 accumulated fewer lipids intracellularly due to decreased fatty acid uptake and increased fatty acid secretion. To rule out the effects of other key genes on MASLD in humans, we examined Gpam, Mboat7, Pnpla3, and Tm6sf2 and found no changes in their expression levels with Mtarc1 knockout (Supp. Fig. 6E). These findings demonstrate mARC1 as a key regulator of hepatic lipid homeostasis and establish mARC1 as a promising therapeutic target for MASLD.

Several pharmaceutical companies are currently investigating siRNA-based approaches for liver-specific mARC1 knockdown as a therapeutic intervention for treatment of MASLD (Patents WO2021237097, Alnylam Pharmaceuticals; WO2022036126, Amgen; WO2022183065, Ionis Pharmaceuticals; WO2022248665, Novo Nordisk; WO2023282704, OliX Pharmaceuticals; WO2024192379 Arrowhead Pharmaceuticals, and WO2024222898 Beijing Fuyuan Medicine, etc). Despite the use of distinct siRNA oligos and diet-induced MASH models by each company, the results from the in vivo mice studies are comparable ^21–23. Mtarc1 targeting siRNA duplexes can significantly improve the phenotype in multiple mouse models of MASLD, similar to that of human mARC1 p.A165T variant carriers. Our work described here further confirms the therapeutic benefit of mARC1 deficiency in a mouse model of MASLD.

Interestingly, Smagris et al. showed that mARC2 and not mARC1 is the main MARC family enzyme (regarding N-reductive activity) in mice ¹⁷. Our data indicate that mARC1 modulation in mice still recapitulates genetic findings in humans to a certain extent. The detrimental phenotypes of Mtarc2 KO found in this study and previous reports indicate that therapeutic drug development may require mARC1 selectivity. Further studies are warranted to elucidate if those effects are developmental and if they translate across species. Taken together, our data aligns well with mARC1 knockdown experiments both in vivo and in vitro from several research groups, suggesting a consistent role of mARC1 in lipid metabolism between mice and humans and highlighting the promising therapeutic potential of mARC1 inhibition for liver disease.

The protective effect of MTARC1 rs2642438 on human liver disease was most pronounced among those at the highest risk of MASLD, specifically individuals with obesity or type 2 diabetes mellitus and carriers of PNPLA3 rs738409:G ¹². Each rs2642438 minor allele reduced risk of MASLD development by 21% in individuals with diabetes, and by 27% for those who were both obese and diabetic. Consistent with findings from the human population, it is encouraging to observe that obese mice exhibited greater benefit with mARC1 deficiency, as evidenced by decreased hepatic lipid accumulation, decreased liver mass (Fig. 2), and a trend toward decreased ALT level and Ccl2 inflammatory gene expression (Supp Fig. 2). From a mechanistic standpoint, it has been reported that HFD-fed mice showed elevated levels of mARC2 and a protein assumed to be mARC1 and therefore exhibited increased N-reductive activity within liver tissue. Dietary and cell culture-induced increases in glucose concentration induce modifications in the N-reductive system in mice ³⁰. Consistent with previous research, Mtarc1 expression levels increased in WT mice following HFD, as demonstrated in our study. Additional research is necessary to elucidate the regulatory mechanisms and endogenous substrates that link lipid metabolism and N-reductive activity.

Phosphatidylcholine (PC) and phosphatidylethanolamine (PE) are the most abundant phospholipids in all mammalian cell membranes. It is important to maintain an appropriate hepatic PC/PE ratio, given that a significant proportion of patients with MASLD and MASH have a reduced hepatic PC/PE ratio in comparison to healthy controls ^26,27. It has been reported that MTARC1 rs2642438 carriers have elevated levels of hepatic polyunsaturated PCs compared to non-carriers ²⁸. In the human liver, phosphatidylcholine, together with apolipoprotein ApoB, is required for the assembly and secretion of VLDLs, which are triacylglyceride-rich lipoproteins ³¹. Both rodents ^32,33 and humans ^34,35 developed hepatic steatosis in response to a choline-deficient diet. This was at least partially attributed to decreased VLDL secretion ³³. In our murine model, the accumulation of lipids in the liver was observed as a result of impaired synthesis of PC due to choline deficiency and reduced methionine in CDAHFD. Through non-targeted metabolomics, we provided evidence that mARC1 depletion enriched hepatic PCs and PEs, which may protect from steatosis by increasing VLDL secretion. Furthermore, Mtarc1 KO mice of both sexes that were challenged with CDAHFD at various time points exhibited enrichment of hepatic PCs and PEs, with less lipid accumulation in the liver (Fig. 5 and Supp table). Interestingly, another study in hepatocyte-specific mARC1 knockdown mice using untargeted lipidomics revealed decreased PC levels in plasma, showing an inverse relationship between liver and circulating PC species ³⁶. These observations indicate that hepatic phosphatidylcholines, which are regulated by mARC1, appear to be crucial for maintaining hepatic lipid homeostasis.

Concomitant with the enrichment of hepatic PCs and PEs, we also observed increased secretion of ApoB in the culture medium of Mtarc1 KO primary mouse hepatocytes. However, this finding contradicts previously published data that ApoB secretion was reduced in primary human hepatocytes treated with siMTARC1, as well as in carriers of the A allele of rs2642438 ^12,22. We therefore hypothesize that this discrepancy could be attributed to species differences (such as ApoB isoforms, lipoprotein clearance, cholesteryl ester transfer protein (CETP) absence in mice, hormonal regulation etc), and further investigations should be considered.

A disturbance in hepatic lipid homeostasis not only involves the disruption in the secretion of triglycerides but also includes an alteration in the influx of lipids. Lewis et al. recently reported that mARC1 knockdown had no effect on lipid uptake in primary human hepatocytes, as evidenced by the residual concentration detected in conditioned media ²². Given that the overabundance of free fatty acid (FFA) in an exogenous FFA mixture could obscure minor differences in FFA uptake, we employed a direct and more sensitive method to quantify fatty acid uptake. Prior to the addition of fatty acids, both WT and Mtarc1 KO primary mouse hepatocytes underwent a serum deprivation period for one hour in order to deplete any remaining endogenous fatty acid reserves. In Mtarc1 KO hepatocytes, FFA uptake was significantly reduced, indicating that impaired lipid import also contributes to the observed amelioration of hepatic steatosis.

Ciociola et al. recently reported that MTARC1 downregulation in primary human hepatocytes reduces intracellular lipid content by increasing fatty acid utilization through β-oxidation, as demonstrated by radiolabeled palmitate oxidation assays along with transcriptomic and proteomic analyses ³⁷. In contrast, our data showed no changes in the expression of genes involved in fatty acid oxidation. This discrepancy may stem from the use of bulk RNA-seq on whole mouse liver tissue in our study, which could obscure hepatocyte-specific effects. Additional studies using primary mouse hepatocytes and direct fatty acid utilization assays would help clarify these differences.

Numerous animal models exist for the study of MASLD and have facilitated both mechanistic and therapeutic investigations. The most common preclinical models of MASLD are mice and rats fed various diets including high-fat diet (HFD) and several nutrient-deficient diets. Mice fed an exclusive, long-term feeding of HFD develop liver steatosis and obesity; however, the onset of significant fibrosis may take up to one year ³⁸. Using a choline-deficient, L-amino acid-defined, high-fat diet (CDAHFD), Matsumoto et al. discovered that this diet helped to maintain energy intake and body weight while increasing liver fat accumulation, inflammatory markers, circulating AST and ALT levels, as well as fibrosis within 6 weeks in C57BL/6J mice ²⁴. More recent work from Sugasawa et al. demonstrated that 1 week of CDAHFD induced steatohepatitis with mitochondrial dysfunction and severe oxidative stress without fibrosis in mouse liver; this partially resembles the early stages of MASH in humans ²⁵. It is worth noting that male mice were utilized in both studies, and a 3–5% reduction in body weight from baseline was observed within the first 3 weeks on CDAHFD. In our 2-week dietary model, both sexes were evaluated, and Mtarc1 KO mice challenged with CDAHFD exhibited a significant reduction in liver-to-body weight ratio compared to WT mice. However, male and female mice reacted differently to the CDAHFD intervention. Female WT mice on CDAHFD showed an increased liver mass and a tendency to maintain body weight (Fig. 3). Conversely, the body weight of male mice decreased by 10%, with a variable effect on liver weight (Supp. Fig. 7). Thus, in our study, a MASH phenotype triggered by CDAHFD was more pronounced in female mice. It is recommended that forthcoming research consider this sex discrepancy when employing this dietary model, particularly during short-term feeding interventions.

A limitation of the current study is that the preclinical model induced by the nutrient-deficient diet does not reliably recapitulate obesity, diabetes or other metabolic risk factors necessary to accurately represent the natural progression of MASLD. It is crucial to carefully evaluate these significant variations in metabolic risk factors, as well as the species differences when evaluating translatability. In patients with MASH, liver fibrosis is the main determinant of mortality ^39,40. Published Mendelian Randomization studies showed that reduction of hepatic fat content is beneficial against liver fibrosis ^41,42. In this report, we did not observe significant fibrosis induction with 2 weeks of CDAHFD treatment. Further work is warranted to see whether prolonged treatment on CDAHFD with Mtarc1 KO may effectively prevent the fibrotic liver phenotype. Additionally, while our work has established potential lipid-related pathways by which Mtarc1 KO prevents hepatic fat accumulation, further mechanistic studies are needed to identify how mARC1 directly or indirectly impacts these pathways. It is also noteworthy that interpreting PC levels in the context of a choline deficient diet should be approached with caution.

In summary, we have characterized Mtarc1 in multiple in vivo and in vitro models to investigate the protective mechanisms triggered by mARC1 loss of function in MASLD. In obese mice, Mtarc1 knockout mitigated liver damage without affecting body weight or glucose homeostasis. In addition, mARC1 deficiency significantly reduced biochemical, transcriptional, and histological indicators of liver steatosis, profibrosis, and inflammation in a diet-induced MASH model. Mechanistically, hepatocytes lacking mARC1 exhibited reduced intracellular lipid accumulation due to decreased fatty acid absorption, as well as enhanced fatty acid secretion. This study supports that mARC1 is a promising pharmacological target for innovative therapies for the prevention or treatment of MASLD and MASH.

Abbreviations

MASLD

metabolic dysfunction-associated steatotic liver disease

MASH

metabolic dysfunction-associated steatohepatitis

MTARC1

mitochondrial amidoxime reducing component 1

CDAHFD

choline-deficient, L-amino acid-defined, high-fat diet

GWAS

genome-wide association studies

HFD

high-fat diet

LFD

low-fat diet

IPGTT

intraperitoneal glucose tolerance test

MRI

magnetic resonance imaging

H&E

hematoxylin and eosin

AST

aspartate aminotransferase

ALT

alanine aminotransferase

PC

phosphatidylcholine

PE

phosphatidylethanolamine

DNL

de novo lipogenesis

VLDL

very low-density lipoprotein

FFA

free fatty acid

Data Availability Statement

The RNA sequencing data generated during this study are available in the NCBI Gene Expression Omnibus (GEO) under accession number GSE315678. All data supporting the findings of this study are available from the corresponding author upon reasonable request.