Depletion of Mkrn1 suppresses HFHFD-induced MASH

We fed mice a HFHFD to establish a MASH model. When mice were fed a HFHFD for 20 or 30 weeks, the size and weights of mice lacking Mkrn1 (MK1−/−) were significantly lower than those of wild-type (WT) mice under the same average food and fructose water intake conditions (Fig. S1A–I). Similarly, the adipocyte areas of brown adipose tissue (BAT) and white adipose tissue (WAT) were smaller in MK1−/− mice than in WT mice (Fig. S1J–M). Moreover, the liver weights of MK1−/− mice were lower than those of WT mice (Fig. 1A, B). Histological analyses revealed the occurrence of hepatocellular ballooning, inflammation, or Mallory-Denk Bodies with lipid droplet accumulation in the hepatocytes of WT mouse livers. However, these phenomena were absent in MK1−/− mice (Fig. 1C, D). Furthermore, collagen secretion and macrophages were detected in WT mice using Sirius red and f4/80 staining, respectively. As the f4/80 markers are mostly detected in Kupffer cells, the staining primarily reflects Kupffer cell involvement in the inflammatory response [40, 41]. However, MK1−/− mice were protected against fibrosis and inflammation, as evidenced by the lack of positivity for either stain (Fig. 1E, F). These observations were further confirmed via quantitative real-time PCR (qRT-PCR), revealing decreased expression of fibrotic and inflammatory genes in the livers of MK1−/− mice compared to WT mice (Figs. 1G and S2A–D). Accordingly, the serum levels of aspartate aminotransferase (AST), alanine aminotransferase (ALT), triglycerides (TGs), and cholesterol were significantly lower in MK1−/− mice than in WT mice (Fig. 1H, I). Moreover, the stabilization and activation of AMPKα were also enhanced in the livers of MK1−/− mice compared to those in WT mice, leading to the suppression of its downstream target, acetyl-CoA carboxylase (ACC), through increased ACC phosphorylation (Fig. 1J, K) [13, 14, 16]. Quantitative analyses showed that the pAMPK/total AMPK ratio in the liver was similar between WT and MK1−/− mice, indicating that the increased levels of pAMPK were likely due to higher levels of total AMPK stabilized by MKRN1 depletion. These findings suggest that MKRN1 targets total AMPK, instead of pAMPK, for degradation (Fig. S2E, F). Finally, no significant changes in glycogen accumulation in mice fed a HFHFD were observed (Fig. S2G). Overall, Mkrn1 depletion in mice conferred hepatic protection against HFHFD-induced MASH by promoting AMPK activity.

Fig. 1

MASH induced by HFHFD was alleviated in MK1−/− mice. A Representative livers of WT and MK1−/− mice fed NCD or HFHFD (Scale bar = 1 cm). B Liver weights of WT or MK1−/− mice fed NCD or HFHFD (n = 5 mice per group). C H&E staining of the livers (Scale bars; × 40 = 250 µm and × 200 = 25 µm). Black, yellow, and blue arrowheads indicate hepatocellular ballooning, inflammation, and Mallory-Denk bodies, respectively. D Representative images of Oil Red O staining of the livers. (Scale bar = 25 µm). E Representative images of Sirius red staining of the livers. (Scale bar = 25 µm). F Representative images of f4/80 staining of the livers. (Scale bar = 100 µm). G mRNA expression levels of f4/80 in the livers of mice fed HFHFD for 20 (left) or 30 (right) weeks (n = 5 mice per group). H AST and ALT serum levels (n = 5 mice per group). I Plasma concentrations of TG and cholesterol (n = 5 mice per group). J, K Lysates of the liver tissues were analyzed by western blotting. Two-tailed Student’s t-test; *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, n.s. not significant. Mean ± s.d

Protective effect of MASH via Mkrn1 depletion is nullified by simultaneous knockout of Ampkα2

As the results indicated that Mkrn1 depletion promoted the stabilization and activation of AMPK in mouse liver, under metabolic stress, and alleviated MASH, we further investigated the genetic association between AMPK and MKRN1. Considering prior reports indicating that metabolic stress significantly reduces AMPKα2 expression without affecting AMPKα1, we generated double knockout (DKO) mice in which Ampkα2 and Mkrn1 were simultaneously depleted to clearly elucidate the regulatory relationship between MKRN1 and AMPKα2 under metabolic stress (Fig. S3A–C) [42]. When WT, MK1−/−, or DKO mice were fed a HFHFD, MK1−/− mice exhibited lower body weights during metabolic stress, whereas DKO mice had body weights similar to those of the WT mice under the same average food and fructose water intake conditions (Figs. 2A and S3D–F). Similarly, DKO mice exhibited enlarged liver sizes and weights comparable to those of WT mice. In contrast, MK1−/− mice had smaller liver sizes and weights, as well as reduced adipocyte sizes, compared to WT (Figs. 2B, C, and S3G). Consistent with these data, the anti-obesity, anti-fibrotic, and anti-inflammatory effects of MK1−/− mice were completely nullified in the DKO mice in terms of steatosis, fibrosis, and inflammation occurrences, respectively. These patterns were similar to those observed in WT mice (Fig. 2D–J). Serum AST, ALT, TG, and cholesterol levels were decreased in MK1−/− mice and increased in DKO mice, similarly to WT mice (Fig. 2K, L). Moreover, AMPK stabilization and activation were significantly increased in the livers of MK1−/− mice compared to those in the livers of WT or DKO mice (Fig. S3H). Finally, no significant changes were observed in glycogen accumulation (Fig. S3I). These observations provide genetic evidence of a negative correlation between AMPKα2 and MKRN1 and indicate that the suppression of MKRN1 activity could enable AMPKα2 to inhibit MASH.

Fig. 2

Prevention of MASH in MK1−/− mice was abrogated in Mkrn1-Ampkα2 DKO mice. A The body weights of mice administered the HFHFD were measured every four days (n = 3–4 mice per group). B Representative images of the livers of WT, MK1−/−, and DKO mice (Scale bar = 1 cm). C Liver weights. (n = 4–6 mice per group). D Representative images of H&E staining of the livers. (Scale bars; × 40 = 250 µm and × 200 = 25 µm). E Representative images of Oil Red O staining of the livers (Scale bars = 100 µm). F Representative images of Sirius red staining of the livers (Scale bar = 100 µm). G Representative images of f4/80 staining of the livers (Scale bar = 100 µm). H mRNA expression levels of f4/80 in the livers (n = 5–8 mice per group). I mRNA expression levels for fibrosis-related markers (n = 3–4 mice per group). J mRNA expression levels for inflammation-related markers (n = 3–4 mice per group). K AST and ALT serum levels (n = 5 mice per group). L Plasma concentrations of TG and cholesterol (n = 5 mice per group). One-way ANOVA with Dunnett’s multiple comparison test; *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, n.s. not significant

Liver-specific depletion of Mkrn1 suppresses HFHFD-induced MASH

To analyze the acute effects of Mkrn1 depletion on MASH, we used an AAV8 system expressing Mkrn1 shRNA to knockdown Mkrn1 in the liver. Mice fed a HFHFD for 8 or 10 weeks were injected with AAV8 shMkrn1 and subsequently fed a HFHFD for 12 or 20 weeks, respectively (Fig. 3A, B). The infection of the liver by AAV8 was confirmed via GFP detection of the AAV8 marker (Fig. S4A). Decreased levels of Mkrn1 were further confirmed by the detecting both mRNA and protein expression levels in the liver (Figs. 3C and S4B). Upon expression of AAV8 shMKRN1, no notable differences were observed in the weight of the liver or adipose tissues of the HFHFD group compared to those of the control (Fig. S4C–G). No significant differences in glycogen accumulation in the liver or food and fructose water intake were observed between Mkrn1 KD and control mice (Fig. S4H-N). Notably, histochemical and qRT-PCR analyses confirmed that Mkrn1 KD in the liver suppressed steatosis, fibrosis, and inflammation compared to controls in mice fed a HFHFD (Fig. 3D–K). Accordingly, serum AST, ALT, TG, and cholesterol levels were decreased in Mkrn1 KD mice (Fig. 3L, M). Finally, an increase in AMPK stabilization and activation were detected in the liver of Mkrn1-depleted mice (Figs. 3N and S4O). Overall, these observations indicate that the liver-specific depletion of Mkrn1 suppressed MASH induced by HFHFD.

Fig. 3

Ablation of hepatic Mkrn1 improves hepatic steatosis, fibrosis, and inflammation in mice with diet-induced MASH. A B Schematic illustration of the experiment in which mice were fed HFHFD for 20 (A) or 30 (B) weeks and injected with AAV8 (AAV8 control or AAV8 shMK1) at week 8 (A) or 10 (B). C mRNA expression levels of MKRN1 in the livers (n = 8 mice per group). D, E Representative images of H&E staining of the livers of mice fed the HFHFD for 20 (D) or 30 (E) weeks (Scale bars; × 40 = 250 µm and × 200 = 25 µm). F Representative images of Oil Red O staining of the livers of mice fed the HFHFD for 20 (left) or 30 (right) weeks (Scale bars = 100 µm). G H Representative images of Sirius red staining of the livers of mice fed the HFHFD for 20 (G) or 30 (H) weeks (Scale bars; × 100 = 100 µm; and × 200 = 25 µm). I Representative images of f4/80 staining of the livers of mice fed the HFHFD for 20 (left) or 30 (right) weeks. (Scale bar = 100 µm). J mRNA expression levels for fibrosis-related markers (n = 7 mice per group). K mRNA expression levels for inflammation-related markers (n = 7 mice per group). L AST and ALT serum levels (n = 4 mice per group). M Plasma concentrations of TG and cholesterol (n = 4 mice per group). N Lysates of the liver tissues analyzed by western blotting. Two-tailed Student’s t-test; *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, n.s. not significant. Mean ± s.d

Ebastine suppresses lipid accumulation by targeting the MKRN1 C-terminal domain

As liver-specific Mkrn1 KD using AAV8 suppressed diet-induced MASH, MKRN1 was demonstrated to be a potential therapeutic target. Although AAV vectors are promising gene delivery tools for treating multiple diseases, clinical trials have exposed several limitations of AAV gene transfer, including immune-mediated toxicities [43,44,45,46]. Recently, drug repurposing research on antihistamines for various diseases has been ongoing, with several antihistamine drugs confirmed for safety and efficacy through clinical trials [35,36,37]. Therefore, we aimed to identify antihistamines that are effective in regulating lipid accumulation in HepG2 cells. Among them, ebastine effectively inhibited lipid accumulation (Fig. S5A, B) in a manner similar to Mkrn1 KD (Fig. 4A, B). Notably, ebastine binds to MKRN1 with a binding energy of −5.40 kcal/mol, near the active site of the protein containing two specific hydrogen bonds (H-bonds) with arginine 298 (R298) and one H-bond with lysine 360 (K360), as revealed by molecular docking analysis (Figs. 4C, D and S6A). To determine whether the R298 and K360 sites of MKRN1 are critical for the regulation of lipid accumulation by ebastine, we reconstituted HepG2 cells with a mutant MKRN1 (MKRN1 AA) containing double mutations (R298A-K360A). In cells expressing MKRN1 WT, ebastine treatment reduced the increased lipid accumulation, whereas, in cells expressing MKRN1 AA, ebastine failed to reduce lipid accumulation (Fig. 4E, F). Based on these observations, we employed an Octet R8 Bio-layer interferometry (BLI) system to evaluate the binding affinity between the MKRN1 C-terminal fragment and ebastine, using the MKRN1 N-terminal fragment as a control. The binding data analysis revealed that ebastine was able to bind to the C-terminus with a KD of 7.6 μM but not interact with the N-terminus of MKRN1 (Fig. 4G). These results are consistent with the finding that the site where ebastine docks with MKRN1 is located in the C-terminus. Overall, these findings suggest that ebastine effectively inhibits lipid accumulation by targeting the C-terminus of MKRN1, making it a promising candidate for therapeutic intervention in MASH.

Fig. 4

Ebastine suppresses lipid accumulation by targeting the MKRN1 C-terminal domain. A Representative fluorescence microscopy images of HepG2 cells treated with ebastine or siMKRN1 (Scale bar = 10 µm). B Quantitative analysis of the relative fluorescence intensity of OPA-treated cells in A. C Alpha-fold protein structure of protein MKRN1 (left) retrieved from AlphaFold Protein Structure Database, and 3-dimensional structure of ebastine (right) retrieved from DrugBank database. D Visual representation of MKRN1 protein after docking with ebastine. Molecular docking analysis was conducted using Autodock 1.5.7 and Chimera 1.17.3. E Representative fluorescence microscopy images of HepG2 cells transfected with plasmids expressing HA/MKRN1 WT or HA/MKRN1 AA mutant with or without ebastine. (Scale bar = 10 µm). F Quantitative analysis of the relative fluorescence intensity of OPA-treated cells in E. G BLI analysis using recombinant His/MKRN1 1–263 or His/MKRN1 264–482 with ebastine. At least three independent experiments were performed, with a minimum of 300 cells analyzed per group for quantitative analysis. One-way ANOVA with Dunnett’s multiple comparison test; *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.0001, n.s. not significant

Ebastine induces destabilization of MKRN1 and activates AMPK

To investigate the regulatory relationship between ebastine and MKRN1 at the molecular level, HepG2 cells were treated with ebastine, and MKRN1 protein levels were examined. Treatment with ebastine decreased MKRN1 protein levels, which was reversed by the proteasome inhibitor MG132, indicating that ebastine induces destabilization in a proteasome-dependent manner without affecting MKRN1 mRNA levels (Fig. 5A–C). To elucidate the mechanism underlying ebastine-induced MKRN1 degradation, we assessed MKRN1 self-ubiquitination in the presence of ebastine. Ubiquitination analysis revealed that ebastine promoted the self-ubiquitination of MKRN1, whereas the MKRN1 E3 ligase-defective mutant (H307E) was unaffected (Fig. 5D, E). Moreover, using deletion mutants of MKRN1, we observed that ebastine induced the degradation of the MKRN1 C-terminal fragment (264–482), which contains a RING domain with E3 ligase activity (Fig. 5F, G), and directly induced self-ubiquitination of the MKRN1 C-terminus (Fig. 5H). In addition, the MKRN1 AA mutant did not undergo further ubiquitination or degradation by ebastine, corroborating the results of the molecular docking analysis. (Figs. 4C–F and 5I, J).

Fig. 5

Ebastine induces MKRN1 destabilization. A HepG2 cells were treated with 0, 5, or 10 μM ebastine for 16 h or 24 h followed by western blotting. B HepG2 cells were treated with 10 μM ebastine for 24 h in the presence or absence of 20 μM MG132 for 6h followed by western blotting. C The MKRN1 mRNA levels were analyzed via qRT-PCR in the presence or absence of ebastine in HepG2 cells (n = 3 per group). D 293T cells were transfected with plasmids expressing HA/MKRN1 WT, HA/MKRN1 H307E mutant, and HA/Ub with or without ebastine. After 18 h, the transfected cells were treated with MG132 for 6h in the presence or absence of ebastine. Cell lysates were immunoprecipitated using anti-MKRN1 antibodies followed by western blotting using anti-Ub antibodies. E Ubiquitinated endogenous MKRN1 was determined under denaturing conditions using MG132-treated HepG2 cells. Immunoprecipitation of cell lysates with anti-MKRN1 antibodies was followed by western blotting with anti-Ub antibodies. F Schematic illustrations of MKRN1 WT and its mutants. G 293T cells were transfected with plasmids expressing HA/MKRN1 WT and its mutants followed by ebastine treatment. H In vitro ubiquitination analysis of recombinant MKRN1 C-terminus (264–482) in the presence or absence of E1, E2, Ub, ATP, or ebastine. I 293T cells were transfected with plasmids expressing HA/MKRN1 WT or HA/MKRN1 AA mutant followed by ebastine treatment. J 293T cells were transfected with plasmids expressing HA/MKRN1 WT, HA/MKRN1 AA mutant, and HA/Ub with or without ebastine. After 18 h, the transfected cells were treated with MG132 for 6h in the presence or absence of ebastine. Cell lysates were immunoprecipitated using anti-MKRN1 antibodies followed by western blotting using anti-HA antibodies. One-way ANOVA with Dunnett’s multiple comparison test; n.s. not significant

The dissolution of lipid droplets was impeded by the use of an AMPK inhibitor (compound C), regardless of the presence of ebastine or absence of MKRN1, indicating that AMPK is a major factor influencing the reduction in lipid droplets (Fig. 6A, B). Both ebastine administration and MKRN1 suppression in HepG2 cells led to the stabilization and activation of AMPKα, without any detectable impact on its mRNA levels (Figs. 6C and S7A–C). Furthermore, when 293T cells expressing MKRN1 WT were treated with ebastine, AMPKα2 levels increased, accompanied by a decrease in MKRN1 WT expression. However, no changes in the levels of MKRN1 AA mutant and AMPKα2 were noted (Fig. S7D).

Fig. 6

Ebastine prevents MKRN1-mediated ubiquitination and degradation of AMPKα2. A, B HepG2 cells treated with DMSO, ebastine, or siMKRN1 were incubated in the absence (top) or presence (bottom) of compound C with BSA or 1 mM OPA. Representative fluorescence microscopy images of HepG2 cells stained with Hoechst and Nile red (Scale bars = 10 µm) (A). Analysis of the relative fluorescence intensity of OPA-treated cells (B). C HepG2 cells were treated with 0, 5, or 10 μM ebastine for 16 h or 24 h followed by western blotting. D 293T cells were transfected with plasmids expressing FLAG/AMPKα2 and/or HA/MKRN1 with or without ebastine followed by western blotting. E 293T cells were transfected with plasmids expressing MKRN1 WT, FLAG/AMPKα2, and HA/Ub in the presence or absence of ebastine. Cell lysates were immunoprecipitated using anti-FLAG antibodies followed by western blotting using anti-HA antibodies. F HA/Ub-expressing plasmid was transfected into HepG2 cells in the presence or absence of ebastine. Immunoprecipitation of cell lysates with anti-AMPKα2 antibodies was followed by western blotting with anti-HA antibodies. G, H Representative fluorescence microscopy images of WT or MK1−/− MEFs stained with Hoechst and Nile red after treatment with BSA or 1 mM OPA and DMSO or ebastine (G), and quantitative analysis of the relative fluorescence intensity of OPA-treated cells (Scale bars = 10 µm) (H). I WT or MK1−/− MEFs were treated with ebastine followed by western blotting. At least three independent experiments were performed, with a minimum of 100 cells analyzed per group for quantitative analysis. One-way ANOVA with Dunnett’s multiple comparison test; *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, n.s. not significant

We further evaluated the effect of ebastine on the association between MKRN1 and AMPKα expression. Ebastine prevented the MKRN1-mediated degradation of AMPKα2 (Fig. 6D) as well as the MKRN1-mediated and endogenous ubiquitination of AMPKα2 (Fig. 6E, F). To determine whether ebastine could also inhibit MKRN1 function in mouse cells, WT or MK1−/− mouse embryonic fibroblasts (MEFs) were generated, and lipid droplet accumulation was measured. MK1−/− MEFs displayed decreased levels of lipid droplets compared to the WT MEFs. While ebastine was able to decrease lipid droplets in WT MEFs, it exhibited no effect on lipid accumulation in the MK1−/− MEFs (Fig. 6G, H). When WT MEFs were treated with ebastine, AMPKα2 levels were increased, but there was no effect in MK1−/− MEFs (Fig. 6I). Overall, ebastine promoted the degradation of MKRN1, resulting in the stabilization and activation of AMPKα and a subsequent reduction in lipid accumulation.

Ebastine treatment protects against HFHFD-induced MASH

As the results indicated that ebastine could prevent lipid accumulation in mouse cells, possibly by suppressing MKRN1, we subsequently tested its effects in mice. Mice were fed a HFHFD for 16 weeks followed by 4 additional weeks of HFHFD accompanied by daily intraperitoneal injections of ebastine at concentrations of 1 and 5 mg/kg (Fig. 7A). Mice fed a HFHFD for 16 weeks exhibited steatosis with fibrosis (Fig. 7A). However, there were no noticeable changes in body weight between the ebastine injection and control groups, with both groups maintaining a consistent average food and fructose water consumption (Figs. 7B and S8A, B). Notably, the livers of mice subjected to ebastine injection were smaller in size compared to those of the control group (Fig. 7C, D). Furthermore, this reduction in liver size coincided with a decrease in MKRN1 levels and an increase in the accumulation of AMPKα in the liver (Figs. 7E, F and S8C). Histochemistry and qRT-PCR results indicated that the mouse livers injected with ebastine were protected against steatosis, fibrosis, and inflammation (Fig. 7G–M). Accordingly, serum AST, ALT, TG, and cholesterol levels were decreased in mice injected with ebastine (Fig. S8D, E). Overall, these results suggest that ebastine binds to the C-terminus of MKRN1, specifically at the R298 and K360 sites, promoting self-ubiquitination followed by MKRN1 degradation, and leading to the stabilization of AMPKα. Furthermore, ebastine protected the mouse liver from the HFHFD-induced MASH, indicating a possible therapeutic application of this drug for MASH treatment.

Fig. 7

Ebastine treatment protects against diet-induced MASH in vivo. A Schematic illustration of ebastine administration in mice fed a HFHFD. B The body weights of mice administered HFHFD were measured every four days (Mock, n = 8; ebastine 1mg/kg, n = 8; and ebastine 5mg/kg, n = 7). C Liver weights (n = 6 mice per group). D Representative image of the livers after treatment with mock or ebastine (Scale bar = 1 cm). E IHC staining for MKRN1 of the livers (Scale bar = 100 µm). F Lysates of liver tissues analyzed using western blotting. G Representative images of H&E staining of the livers (Scale bars; × 40 = 250 µm and × 200 = 25 µm). H Representative images of Oil Red O staining of the livers (Scale bars = 100 µm). I Representative images of Sirius red staining of the livers (Scale bars = 100 µm). J Representative images of f4/80 staining of the livers (Scale bar = 100 µm). K mRNA expression levels of f4/80 in the livers (n = 4 mice per group). L mRNA expression levels for fibrosis-related markers (n = 4 mice per group). M mRNA expression levels of inflammation-related markers (n = 4 mice per group). One-way ANOVA with Dunnett’s multiple comparison test; *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, n.s. not significant

We further investigated the effect of ebastine in MASH using a choline-deficient, L-amino acid-defined, HFD (CDAHFD) for 6 weeks. Initially, when WT or MK1−/− mice were fed CDAHFD, they maintained normal body and liver weights (Fig. S9A-C). However, MK1−/− mice were protected against steatosis in the liver phenotype and had lower serum AST, ALT, TG, and cholesterol levels than those in WT mice (Fig. S9D-G). pAMPKα levels were also increased in the livers of MK1−/− mice (Fig. S9H). MASH signatures were consequently analyzed in CDAHFD-fed mice, revealing pronounced fibrosis and inflammation in WT mice than in MK1−/− mice, as detected by Sirius red, f4/80, and qRT-PCR analyses (Fig. 9SI–M). To determine whether ebastine could inhibit MASH induced by CDAHFD, we subjected WT and AMPKα2−/− mice to intraperitoneal injections of ebastine at a concentration of 1 mg/kg for 2 weeks (Fig. S9N). Both WT and AMPKα2−/− mice maintained normal body and liver weights regardless of ebastine injection (Fig. S9O-Q). However, histological and mRNA analyses revealed that ebastine injections alleviated steatosis, fibrosis, and inflammation in the livers of WT mice, which are characteristic features of MASH (Fig. S9R, S). Moreover, ebastine injections increased AMPK activity in the livers of WT mice; however, these effects were absent in AMPKα2−/− mice (Fig. S9T). In summary, depletion of Mkrn1 or administration of ebastine effectively protected the liver from HFHFD- and CDAHFD-induced MASH by promoting AMPK activity.