Metabolic dysfunction-associated steatotic liver disease (MASLD) is a prevalent liver disorder that encompasses conditions ranging from simple steatosis to hepatocellular carcinoma and affects a substantial portion of the global population (Hutchison et al., 2023). Approximately 32.4% of the world’s population has non-alcoholic fatty liver disease, and its prevalence is rapidly increasing (Tang et al., 2022). MASLD is closely associated with metabolic abnormalities, such as obesity and diabetes (Hutchison et al., 2023). Given the pivotal role of insulin resistance in the pathogenesis of MASLD, enhancing insulin sensitivity is an attractive therapeutic strategy (Armstrong et al., 2016; Hwang and Yang, 2021). Recently, Resmetirom, the first drug approved by the FDA for treating noncirrhotic nonalcoholic steatohepatitis (now renamed metabolic dysfunction-associated steatohepatitis), offers a significant breakthrough (Harrison et al., 2024). However, despite this advancement, treatments for MASLD remain limited.

AMP-activated protein kinase (AMPK) is an important regulator of glucose and lipid metabolism. Reduced AMPK activity is often associated with obesity and insulin resistance. However, AMPK activation counteracts obesity and enhances insulin sensitivity (Pollard et al., 2019; Lee et al., 2023). In particular, AMPK reduces adiposity by decreasing fat cell numbers and increasing glucose uptake in skeletal muscles (Cokorinos et al., 2017). Within the liver, AMPK inhibits glucose production by downregulating the expression of gluconeogenic genes, such as phosphoenolpyruvate carboxykinase (PEPCK) and glucose-6-phosphatase (G6Pase) (Cool et al., 2006). Furthermore, hepatic AMPK activity prevents the development of steatosis (Woods et al., 2017). AMPK also phosphorylates liver X receptor α (LXRα), leading to the inhibition of sterol regulatory element-binding protein 1 (SREBP1c), a key enzyme in lipid synthesis (Hwahng et al., 2009). Moreover, AMPK promotes fatty acid oxidation in the liver (Boudaba et al., 2018). Various studies have shown that pharmacological AMPK activators, such as oltipraz, have beneficial effects on insulin resistance and hepatic steatosis (Bae et al., 2007; Hwahng et al., 2009).

AMPK serves as a cellular energy sensor and becomes activated AMPK in response to an increase in the AMP:ATP ratio (Hardie, 2011). Additionally, several upstream regulators of AMPK have been identified, including liver kinase B1 (LKB1) (Shaw et al., 2004), NAD-dependent deacetylase sirtuin-1 (SIRT1) (Price et al., 2012), calcium/calmodulin-dependent protein kinase kinase (CaMKK) (Abbott et al., 2009), and transforming growth factor beta-activated kinase 1 (TAK1) (Xie et al., 2006). LKB1 plays a pivotal role in AMPK regulation, particularly in response to energy stress, as it phosphorylates threonine (Thr172) of AMPKα (Shaw et al., 2004). The loss of LKB1 function has been shown to inhibit AMPK phosphorylation. Moreover, the phytochemical, resveratrol, is known to activate SIRT1, and its beneficial effects have been extensively studied (He et al., 2023).

In East Asia, Atractylodes lancea is an important traditional herbal medicine that is used to treat a wide range of diseases, particularly related to digestive disorders (Kimura and Sumiyoshi, 2012). Atractylodes lancea is recognized for its anti-cancer, anti-anoxic, and anti-inflammatory properties (Jiao et al., 2014; Jun et al., 2018; Hossen et al., 2021). Moreover, it demonstrated anti-obesity effects in a previous study (Jiao et al., 2014). Atractylodin, a major constituent derived from the rhizomes of Atractylodes lancea, exhibits potent lipase inhibitory activity (Jiao et al., 2014). A previous study showed that atractylodin ameliorated acute liver failure induced by lipopolysaccharides and d-galactosamine, emphasizing its inhibitory role in inflammation and oxidative stress (Lyu et al., 2019). In this study, we investigated the effects of Atractylodes lancea and its major component, atractylodin, on fatty liver and insulin resistance and their underlying mechanisms.

Materials

Atractylodin (Cat. HY-N0238-250MG) was procured from MCE (Monmouth Junction, NJ, USA). To obtain the Atractylodes lancea ethanol extract, Atractylodes dry root material was soaked in 80% (v/v) ethanol at approximately ten times its volume at room temperature. The mixture was agitated for 24 h, and the extraction process was repeated two to three times. The extract was then vacuum-filtered through a Whatman filter paper (No. 2) to remove any impurities. The filtered extract was subjected to ultrafiltration and vacuum concentration using a rotary evaporator and subsequently dried to obtain the solid fraction of the Atractylodes extract.

The normal diet (ND; containing 24.5% protein, 13.1% fat, and 62.4% carbohydrates; Cat. 5053) and HFD (containing 20% protein, 60% fat, and 20% carbohydrates; Cat. D12492) were purchased from Lab Diet (St. Louis, MO, USA) and Research Diets (New Brunswick, NJ, USA), respectively. Polyethylene glycol 400 (Cat. 25322-68-3) was purchased from Duksan (Ansan, Korea). Hematoxylin and eosin (H&E; Cat. 3801698) reagents were purchased from Leica (Buffalo Grove, IL, USA). FUJI DRI-CHEM slides for GPT/ALT-P III (Cat. 3250), GOT/AST III (Cat. 3150), TG-P III (Cat. 1650), and TCHO-P III (Cat. 1450) assessments were obtained from FUJIFILM (Tokyo, Japan). The TRIzol reagent (15596018) was sourced from Invitrogen (Carlsbad, CA, USA). AccuPrep Universal RNA Extraction Kit (K-3140) was obtained from Bioneer (Daejeon, Korea). RQ1 RNase-free DNase (M6101) was purchased from Promega (Madison, WI, USA). The High-Capacity cDNA Reverse Transcription Kit (4368814) was obtained from Thermo Fisher Scientific (Foster City, CA, USA). The ExcelTaq^TM 2X Fast Q-PCR Master Mix (SYBR, TQ1210) was purchased from SMOBIO (Hsinchu, Taiwan).

The glucometer (Accu-Chek Instant) and appropriate test strips were purchased from Roche (Mannheim, Germany). The Ultra-Sensitive Mouse Insulin ELISA kit (Cat. 90080) was purchased from Crystal Chem (Downers Grove, IL, USA). D-glucose (Cat. GL3071) was purchased from GeorgiaChem (Suwanee, GA, USA). Insulin was purchased from Novo Nordisk (Bagsværd, Denmark). The saline solution was obtained from Huons, Inc. (Seongnam, Korea). 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; Cat. 475989-1G) was obtained from Millipore (San Diego, CA, USA). The syringe filters (Cat. 17845-ACK) were sourced from Sartorius (Goettingen, Germany). DMEM/Low Glucose with L-Glutamine and Sodium Pyruvate (Cat. SH30021.01) and MEM/EBSS (Cat. SH30024.01) were purchased from HyClone (Logan, UT, USA). 0.5% Trypsin-EDTA (10X) was obtained from Gibco (New York, NY, USA).

Fetal bovine serum (Cat. 35-015-CV) was obtained from Corning (Glendale, AZ, USA). Penicillin/streptomycin (Cat. SV30010) was purchased from HyClone (Logan, UT, USA). T0901317 (Cat. 575310) was purchased from Millipore (Burlington, MA, USA). The phosphatase inhibitor cocktail (Cat. P5726) was purchased from Sigma-Aldrich (St. Louis, MO, USA). Furthermore, the protease inhibitor cocktail (535140) was purchased from Calbiochem (Darmstadt, Germany). The ECL reagent (RPN2106) was purchased from GE Healthcare (Buckinghamshire, UK). The p-AMPK (T172) (Cat. 2535S), AMPKα (Cat. 2532S), p-ACC (S9) (Cat. 3661S), and acetyl-CoA carboxylase (ACC) (Cat. 3662S) antibodies were obtained from Cell Signaling Technology (Danvers, MA, USA). The beta-actin antibody (Cat. A5441) was purchased from Sigma-Aldrich.

In vivo treatment

Seven-week-old male C57BL/6N mice were purchased from Orient Bio Corporation (Seongnam, Korea) and KOATECH Corporation (Pyeongtaek, Korea). The mice were housed in the Kangwon National University Animal Laboratory Center under controlled environmental conditions at 23 ± 2°C with a 12-h light/12-h dark cycle. After allowing the mice to acclimatize for 1 week, they were randomly allocated to different experimental groups. The mice were provided with either ND or HFD ad libitum for 12 weeks. After 8 weeks of dietary treatment, mice in the experimental groups were orally administered Atractylodes lancea at doses of 30, 60, or 120 mg/kg, or a vehicle, five times per week for 4 weeks. The groups were designated as follows: Group 1, ND+Vehicle (n=8); Group 2, ND+Atractylodes lancea 120 mg/kg (n=8); Group 3, HFD+Vehicle (n=7); Group 4, HFD+Atractylodes lancea 30 mg/kg (n=10); Group 5, HFD+Atractylodes lancea 60 mg/kg (n=7); and Group 6, HFD+Atractylodes lancea 120 mg/kg (n=10).

In a separate experiment, the mice were orally administered atractylodin at doses of 5 or 10 mg/kg, or a vehicle, 5 days per week, after 8 weeks of HFD feeding. The groups in this experiment were designated as follows: Group 1, ND+Vehicle (n=8); Group 2, ND+Atractylodin 10 mg/kg (n=8); Group 3, HFD+Vehicle (n=8); Group 4, HFD+Atractylodin 5 mg/kg (n=8), and Group 5, HFD+Atractylodin 10 mg/kg (n=9).

After 4 weeks of drug administration, the mice were sacrificed. Atractylodes lancea and atractylodin were dissolved in a 40% polyethylene glycol 400 solution, respectively before administration. The experimental protocol was approved by the Institutional Animal Care and Use Committee of the Kangwon National University (KW-211021-4 and KW-220302-1).

Measurement of serum profiles

Blood was collected from the retro-orbital plexus of the mice. Following a 30-min incubation at room temperature, the blood samples were centrifuged at 5,000 rpm for 15 min at 20°C. The resulting supernatant was carefully transferred to fresh tubes to obtain the serum (Hong et al., 2022). Serum triglyceride, total cholesterol, alanine aminotransferase (ALT) and aspartate transaminase (AST) levels were assessed using a Chemistry analyzer (FUJIFILM, DRI-CHEM NX). Serum insulin levels were measured using an Ultra-Sensitive Mouse Insulin ELISA kit. To achieve this, 95 μL of the sample diluent was added to antibody-coated microplates. Then, 5 μL of the samples (or working mouse insulin standard) were added to each well. The microplate was incubated at 4°C for 2 h, washed five times with a washing buffer, and 100 μL of an anti-insulin enzyme conjugate was added to each well. The mixture was allowed to react at room temperature for 30 min. After washing seven times with the washing buffer, 100 μL of the enzyme-substrate solution was added. Then, the mixture was incubated for 40 min at room temperature with protection from light. The reaction was terminated by adding an additional 100 μL of an enzyme reaction stop solution (TMB) per well. Following the absorbance measurement at 450 nm and 630 nm using SpectraMax i3 (Molecular Devices, Sunnyvale, CA, USA), the insulin levels were calculated using a standard curve.

H&E staining

Mouse tissues were fixed in 10% formalin for 2-3 days and subsequently processed into paraffin blocks using a tissue processor (TP1020, Leica). Tissue sections with a thickness of 4 μm were obtained using a microtome (RM2235, Leica) and mounted on glass slides for tissue staining. After deparaffinization, we performed H&E staining as previously reported (Yang et al., 2019): The tissue sections were incubated with Hemalast for 30 s and then immersed in hematoxylin solution for 10 min. After rinsing, the sections were treated with a differentiator solution for 45 s, followed by a 1-min rinse in distilled water. Subsequently, the samples were incubated with a bluing reagent for 1 min and rewashed. The sections were then immersed in 80% ethanol for 1 min and placed in an eosin solution for 2 min. After three dips in 95% ethanol, the samples were transferred to 100% ethanol for 1 min and soaked in fresh 100% ethanol for an additional 2 min. Following two 2-min treatments with xylene, the samples were mounted. Image acquisition was conducted using the ImageXpress Pico Automated Cell Imaging System at the Core-Facility for Innovative Cancer Drug Discovery (CFICDD) at Kangwon National University.

Oil red O staining

Frozen sections were obtained at –80°C and allowed to equilibrate at room temperature for 10-15 min. The frozen sections were washed twice in phosphate-buffered saline (PBS) for 5 min each and briefly immersed in 60% isopropanol/H₂O for 1 min. After complete drying, the sections were exposed to an Oil Red O solution for 15 min at room temperature. Following staining, the sections were rinsed in 60% isopropanol. They were then washed in PBS for 5 min, followed by a 1-min counterstaining with hematoxylin. After a final wash in H₂O, the sections were mounted with glycerol. The Oil Red O-positive area was measured using ImageJ software (Yang et al., 2021).

cDNA preparation and quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR)

The qRT-PCR procedure involved the homogenization of mouse liver tissues or HepG2 cells in 1 mL of TRIzol reagent. As previously reported, upon adding 200 μL of chloroform, the mixture was incubated at room temperature for 3 min and then subjected to centrifugation (Kim et al., 2022). The clear supernatant obtained after centrifugation was transferred to a new tube, and an equal volume of 70% ethanol was added for total RNA extraction using the AccuPrep Universal RNA extraction kit. The extracted RNA samples were incubated at 65°C for 5 min. The RNA concentration and purity were assessed using Quick Drop (Molecular Devices). Genomic DNA was eliminated using RQ1 RNase-free DNase, and cDNA was synthesized using a High-Capacity cDNA Reverse Transcription Kit. qRT-PCR was conducted using SYBR on a CFX96 Real-Time system (Bio-Rad, Hercules, CA, USA). The primer sequences used for qRT-PCR are listed in Table 1.

Table 1 Primer sequences for qRT-PCR

Primer	Forward	Reverse
18s	AGTCCCTGCCCTTTGTACACA	CGATCCGAGGGCCTCACTA
mSrebf1	AACGTCACTTCCAGCTAGAC	CCACTAAGGTGCCTACAGAGC
mDgat2	GAAGATGTCTTGGAGGGCTG	CGCAGCGAAAACAAGAATAA
hSREBF1	CGACATCGAAGACATGCTTCAG	GGAAGGCTTCAAGAGAGGAGC
hFASN	GACATCGTCCATTCGTTTGTG	CGGATCACCTTCTTGAGCTCC
hACACA	GCTGCTCGGATCACTAGTGAA	TTCTGCTATCAGTCTGTCCAG

Glucose tolerance test and insulin tolerance test

The glucose tolerance test (GTT) was carried out 2 weeks after the administration of Atractylodes lancea or atractylodin to the mice. Prior to the experiment, the mice underwent a 16-h fast. After fasting, the body weights and fasting glucose levels were measured. Subsequently, the mice were orally administered 2 g/kg of glucose. Blood glucose levels were assessed at 30, 60, and 120 min after glucose administration using a blood glucose meter.

The insulin tolerance test (ITT) was conducted 2 weeks after atractylodin administration. Following a 6-h fasting period, the body weights and fasting blood glucose levels were measured. Then, 0.75 U/kg of insulin was administered intraperitoneally and blood glucose levels were assessed at 30, 60, 90, and 120 min after insulin injection.

RNA-sequencing

Immediately after dissecting liver tissues from the mice, the tissues were preserved in an RNAlater (Thermo Fisher Scientific) solution at 4°C for 1 day for RNA stabilization and storage. Subsequently, the RNAlater solution was replaced, and the tissues were stored at –80°C until needed. RNA-sequencing was performed by ebiogen (Seoul, Korea). The quality of the total RNA extracted from the mouse liver tissues was assessed using an Agilent 2100 Bioanalyzer (Agilent Technologies, Amstelveen, The Netherlands). Library construction was performed using the QuantSeq 3’ mRNA-Seq Library Preparation Kit (FWD). Subsequently, bulk RNA-sequencing was performed using high-throughput single-end 75 bp sequencing on a NextSeq 500 platform (Illumina, San Diego, CA, USA). Differentially expressed gene analysis was conducted using Exdega v.4.0.3 (ebiogen). Principal component analysis (PCA) was performed using PCA and ExDEGA GraphicPlus. Functional annotation analysis and clustering heatmap analysis were performed using the DAVID database (https://david.ncifcrf.gov/home.jsp) and the MeV program, respectively. Pathway analysis was performed using the Kyoto Encyclopedia of Genes and Genomes (KEGG) mapper (https://www.genome.jp/kegg/mapper/). The RNA-seq data generated in this study have been deposited in the Gene Expression Omnibus (GEO) under the accession number GSE234666.

Western blotting assay

The liver tissues were homogenized (or the cells were lysed) using a cell lysis buffer (containing 10 mM Tris-HCl pH, 7.1, 100 mM NaCl, 1 mM EGTA, 10% glycerol, and 0.5% Triton X-100) with the addition of a phosphatase inhibitor cocktail and a protease inhibitor cocktail. After incubation on ice for 1 h, centrifugation was performed to obtain whole-cell lysates. For the nuclear fraction, NE-PER^TM Nuclear and Cytoplasmic Extraction Reagents were used according to the manufacturer’s guidelines. The protein concentration was quantified using the Bradford assay. Subsequently, 20-35 μg of the protein was loaded onto an SDS-PAGE gel to separate proteins by their molecular weights. The separated proteins were transferred to a nitrocellulose membrane and blocked with 5% skim milk in Tris-buffered saline containing Tween 20 (TBST) for 1 h at room temperature. The membrane was washed three times in phosphate-buffered saline with Tween 20 (PBST) for 5 min and incubated with the primary antibody at 4°C overnight. On the following day, the membrane was incubated with the secondary antibody for 1 h at room temperature. Protein bands were detected using an ECL reagent (Shin et al., 2023).

In vitro treatment

The cells were initially seeded in a 6-well cell culture plate and serum-deprived overnight. Subsequently, the cells were treated with either atractylodin or Atractylodes lancea for the specified durations and concentrations. In a separate experiment, HepG2 cells were treated with atractylodin 1 h prior to exposure to the LXR agonist, T0901317 (T090), and then maintained in the presence of 3 μM of T090 for 12 h.

Cell viability assay

HepG2 cells were cultured in DMEM/Low Glucose with L-Glutamine and Sodium Pyruvate supplemented with 10% FBS and 1% penicillin and streptomycin. Subculturing of the HepG2 cells was performed every 2-3 days. The HepG2 cells were seeded in 96-well cell culture plates. Once the cells reached approximately 60% confluency, the serum was withdrawn overnight. The cells were then subjected to treatment with either Atractylodes lancea (31.25, 62.5, 125, 250, or 500 μg/ml) or atractylodin (5, 10, 20, 40, or 80 μM) for a duration of 24 h. Following this treatment period, the cells were exposed to MTT at a final concentration of 0.5 mg/ml for 4 h. Upon removal of the culture medium, 150 μL of DMSO was introduced. The absorbance was then measured at 540 nm using a SpectraMax i3 spectrophotometer (Molecular Devices).

In silico molecular dynamics study

Ligand docking studies were performed using AutodockTools 1.5.7 (https://autodock.scripps.edu/). The docking model utilized the AMPK structure with ligands from the PDB entry 6E4U. Docking was conducted with a docking box size sufficient to encompass the binding site.

Statistical analysis

Statistical analysis of the experimental results was conducted using GraphPad Prism version 8.0.1 software (GraphPad Software Inc., Boston, MA, USA). The data are presented as the mean ± SEM. Differences between the experimental groups were statistically analyzed using one-way ANOVA and Tukey’s or Dunnett’s tests. The level of statistical significance was set at p<0.05.

^{Atractylodes lancea} treatment mitigated HFD-induced obesity and fatty liver

To assess the effect of Atractylodes lancea on hepatic steatosis and insulin resistance induced by HFD, the mice were maintained on HFD for 12 weeks, with Atractylodes lancea (30, 60, and 120 mg/kg) administered orally during the final 4 weeks of HFD feeding (Fig. 1A). The weight gain of the vehicle-treated, HFD-fed mice (HFD+Veh) was significantly higher than that of the vehicle-treated, ND-fed mice (ND+Veh), indicating successful development of the in vivo model.

Figure 1. HFD-induced obesity and fatty liver is attenuated by Atractylodes lancea treatment. (A) Schematic of the experimental design. Mice were maintained on a normal diet (ND) or high-fat diet (HFD) for 12 weeks, and 30 mg/kg, 60 mg/kg, or 120 mg/kg of Atractylodes lancea (AL) was administered orally during the final 4 weeks of HFD feeding (n=7-10, each group). (B) Mouse body weight at the end of the study. (C) White adipose tissue (WAT) and liver weight. (D) Serum triglyceride (TG) and total cholesterol levels. (E) Serum alanine aminotransferase (ALT) levels. (F) Representative images of Hematoxylin and eosin (H&E) and Oil Red O staining in the liver. Scale bar (H&E): 500 μm, scale bar (ORO): 60 μm. (G) RT-qPCR analyses of Srebf1 and Dgat2 mRNA levels from the mouse liver. (H) Immunoblot analysis showing the expression levels of fatty acid synthase (FAS) protein in the liver tissues from HFD-fed mice treated with Atractylodes lancea compared to vehicle (Veh)-treated controls. Data are presented as mean ± SEM (n=7-10). Statistical analysis was performed using one-way ANOVA with Tukey’s post-hoc test. **p<0.01, ***p<0.001: compared with ND+Veh; ^#p<0.05, ^##p<0.01, ^###p<0.001: compared with HFD+Veh.

Moreover, the administration of 30 mg/kg, 60 mg/kg, or 120 mg/kg of Atractylodes lancea to the treatment groups resulted in significantly reduced body weight when compared to the HFD+Veh (Fig. 1B). Importantly, the food intake remained unchanged after Atractylodes lancea treatment (data not shown). Atractylodes lancea treatment also decreased the weights of the white adipose tissue and livers in the HFD-fed mice (Fig. 1C). Serum triglyceride and total cholesterol levels were reduced by 60 or 120 mg/kg Atractylodes lancea treatment in the HFD-fed mice (Fig. 1D). These results suggest that Atractylodes lancea possesses anti-obesity properties.

Furthermore, the level of serum ALT, an indicator of liver damage, was lower in the HFD-fed mice treated with 30, 60, or 120 mg/kg of Atractylodes lancea compared to the HFD+Veh mice (Fig. 1E). Hepatic damage and lipid accumulation induced by the HFD were alleviated by Atractylodes lancea treatment, as observed following H&E and Oil Red O staining (Fig. 1F). Additionally, RT-qPCR analysis demonstrated a reduction in the expression of sterol regulatory element-binding transcription factor 1 (Srebf1), which encodes SREBP1c, a central regulator of lipogenesis, in the Atractylodes lancea-treated groups of the HFD-fed mice (Fig. 1G, left). Furthermore, Atractylodes lancea reduced the expression levels of diacylglycerol acyltransferase 2 (Dgat2) mRNA (Fig. 1G, right), which catalyzes the final step in triglyceride synthesis. In addition, Atractylodes lancea significantly decreased the protein expression levels of fatty acid synthase (FAS), a key enzyme in de novo lipogenesis (Fig. 1H). These findings indicate that Atractylodes lancea effectively reduced obesity and hepatic steatosis.

Atractylodes lancea treatment attenuated HFD-induced glucose intolerance

Next, we assessed the effects of Atractylodes lancea on insulin resistance. The HFD-induced fasting serum glucose levels were reduced following treatment with 30, 60, and 120 mg/kg of Atractylodes lancea (Fig. 2A). The fasting serum insulin levels also decreased in the HFD-fed mice treated with Atractylodes lancea (Fig. 2B). The HOMA-IR value, a representative clinical indicator of insulin resistance, was significantly reduced following Atractylodes lancea treatment in the HFD-fed mice (Fig. 2C). Next, we performed a GTT 2 weeks after Atractylodes lancea administration. In the HFD-fed mice, Atractylodes lancea treatment significantly lowered the glucose concentrations and the area under the curve (AUC) of the GTT, demonstrating improved insulin sensitivity (Fig. 2D, 2E). These results suggest that Atractylodes lancea treatment effectively ameliorated insulin resistance in the HFD-fed mice.

Figure 2. Atractylodes lancea alleviates glucose intolerance. (A) Fasting serum glucose levels (mg/dL) after a 16 h fast at the end of the study. (B) Fasting serum insulin levels (ng/mL). (C) Homeostatic model assessment of insulin resistance (HOMA-IR) calculated using the formula: [Fasting insulin (μU/mL)×Fasting glucose (mg/dL)]÷405. (D) Glucose tolerance test (GTT) was performed 2 weeks after the administration of Atractylodes lancea. (E) The area under the curve (AUC) of GTT. Data are presented as mean ± SEM (n=7-10). Statistical analysis was performed using one-way ANOVA with Tukey’s post-hoc test. (A-C, E) **p<0.01, ***p<0.001: compared with ND+Veh; ^#p<0.05, ^##p<0.01, ^###p<0.001: compared with HFD+Veh. (D) *p<0.05, **p<0.01: ND+Veh vs. HFD+Veh; ^##p<0.01: HFD+Veh vs. HFD+AL 30 mg/kg; ⁺p<0.05: HFD+Veh vs HFD+AL 60 mg/kg; ^$p<0.05: HFD+Veh vs HFD+AL 120 mg/kg.

Atractylodin, the bioactive component of Atractylodes lancea, inhibited HFD-induced liver damage and fat accumulation

Atractylodin is an important bioactive component of Atractylodes lancea (Kulma et al., 2021). The chemical structure of atractylodin is illustrated in Fig. 3A. In the in vivo experiment, HFD (60% fat) was administered for 12 weeks, and 5 or 10 mg/kg of atractylodin was orally administered from the eighth week (Fig. 3B). Atractylodin treatment attenuated HFD-induced obesity (Fig. 3C) without affecting food intake (data not shown). The weights of the white adipose tissues were decreased following atractylodin treatment in the HFD-fed mice (Fig. 3D). Moreover, atractylodin treatment reduced serum triglyceride and total cholesterol levels in the HFD-fed mice (Fig. 3E), accompanied by decreased ALT and AST levels (Fig. 3F). H&E and Oil Red O staining revealed a reduction in fat droplets in the livers of the atractylodin-treated mice (Fig. 3G, 3H). The HFD-induced liver weights also decreased in the atractylodin-treated groups (Fig. 3I). Similar to the observations in the in vivo study involving Atractylodes lancea, we observed the anti-obesity and anti-steatotic effects of atractylodin.

Figure 3. Atractylodin ameliorates obesity and lipogenesis in HFD-fed mice. (A) Molecular structure of atractylodin. (B) Schematic of the experimental procedure to evaluate the role of atractylodin (AT) in HFD-induced MASLD in mice (n=8-9, each group). (C) Mouse body weight. (D) White adipose tissue (WAT) weight of mice. (E) Serum triglyceride and total cholesterol levels. (F) Serum ALT and AST levels. (G) Representative H&E and Oil red O (ORO) staining images. Scale bar (H&E): 100 μm, scale bar (ORO): 60 μm. (H) The quantification of Oil red O staining (% of tissue area). (I) Liver weight. Data are presented as mean ± SEM (n=8-9). Statistical analysis was performed using one-way ANOVA with Tukey’s post-hoc test. *p<0.05, **p<0.01: compared with ND+Veh; ^#p<0.05, ^##p<0.01: compared with HFD+Veh.

Atractylodin suppressed glucose intolerance

Next, we determined the effect of atractylodin on insulin resistance. The fasting serum glucose and insulin levels were reduced in the HFD-fed, atractylodin-treated mice compared to the HFD-fed, vehicle-treated mice (Fig. 4A, 4B). The HOMA-IR value was notably decreased in the HFD-fed mice treated with 10 mg/kg atractylodin compared to the vehicle-treated group (Fig. 4C). In the HFD-fed mice, the glucose levels and AUC of the GTT were significantly decreased in the groups administered 5 mg/kg and 10 mg/kg of atractylodin compared to the vehicle-treated group (Fig. 4D). Additionally, the ITT revealed a substantial improvement in the insulin sensitivity in the HFD-fed mice treated with atractylodin compared to that in the HFD-fed, vehicle-treated mice (Fig. 4E). These data suggest that atractylodin efficiently reduced insulin resistance. Next, we conducted an RNA-sequencing analysis using liver samples from HFD-fed mice treated with either the vehicle or 10 mg/kg of atractylodin. PCA revealed significant differences in gene expression patterns between the two groups (Fig. 4F). RNA-seq analysis of genes involved in the glucose metabolism revealed that atractylodin treatment increased the expression of glycolysis-related genes including Gclc, Angptl4, and Pdk4 (Fig. 4G). Conversely, expression levels of Gale, known for its role in hepatic gluconeogenesis regulation (Zhu et al., 2017), were decreased by atractylodin treatment (Fig. 4G). These results suggest that atractylodin may modulate glycolytic activity and gluconeogenesis, potentially influencing glucose metabolism and related signaling pathways.

Figure 4. Mice treated with atractylodin overcome glucose intolerance. (A) Fasting serum glucose levels after sacrificing mice that were fasted for more than 16 h. (B) Fasting serum insulin levels. (C) HOMA-IR index. (D) Glucose tolerance test (GTT). GTT curve (left) and area under the curve (AUC, right). (E) Insulin tolerance test (ITT). ITT curve (left) and % of area under the curve (AUC, right). Data are presented as mean ± SEM (n=8-9). Statistical analysis was performed using one-way ANOVA with Tukey’s post-hoc test. **p<0.01: compared with ND+Veh; ^#p<0.05, ^##p<0.01: compared with HFD+Veh. (F) Principal component analysis (PCA) comprising gene expression patterns in the livers of HFD-fed mice treated with vehicle (HFD+Veh) or 10 mg/kg atractylodin (HFD+AT10) (n=3). (G) Heatmap of gene expression changes in the glucose metabolism upon atractylodin treatment.

Atractylodin regulated lipid metabolic processes

To identify the differential gene expression patterns between vehicle-treated and atractylodin-treated groups and understand the role of atractylodin in biological processes, we further analyzed RNA-seq data. The volcano plot displayed the distribution of the differentially expressed genes between the vehicle-treated and atractylodin-treated groups. Notably, atractylodin administration resulted in the downregulation of lipogenic genes, including Fasn, Srebf1, Scd2, and Mogat1. In addition, the expression of Lipin1, an inducible amplifier of the hepatic PGC-1α/PPARα regulatory pathway, was up-regulated (Fig. 5A) (Finck et al., 2006).

Figure 5. The lipid metabolic process was dysregulated by atractylodin. (A) Volcano plot depicting differential expression of genes in the atractylodin-treated group compared to the vehicle-treated group in HFD-fed mice. Genes with significant upregulation are shown in red, downregulated genes in blue, and genes with no significant difference in grey (│fold change│>1.5, p-value <0.05). (B) Top 10 Gene Ontology (GO) biological process (BP) analysis of major signaling pathways. Using the DAVID bioinformatics database. (C) Analysis of significantly enriched GO terms using the GO BP Direct database. (D) Heatmap illustrating the results of RNA-sequencing analysis for the lipid metabolic process (n=3). (E) Reduced hepatic mRNA expression of key enzyme involved in de novo lipogenesis, Srebf1 and Dgat2, in atractylodin-treated HFD-fed mice compared to vehicle-treated HFD-fed mice (n=8-9, each group). Data are presented as mean ± SEM. Statistical analysis was performed using one-way ANOVA with Tukey’s post-hoc test. **p<0.01: compared with ND+Veh; ^##p<0.01: compared with HFD+Veh. (F) Representative Western blotting for fatty acid synthase (FAS).

Gene Ontology (GO) biological process analysis of the major signaling pathways was conducted using the DAVID bioinformatics database (Fig. 5B). Among the TOP10 significantly altered GO terms, it was evident that lipid metabolism, lipid biosynthesis, fatty acid metabolism, and fatty acid biosynthesis were significantly modulated in the livers of mice treated with atractylodin compared to those treated with the vehicle. We then analyzed the significantly enriched GO terms using the GO BP Direct database (Fig. 5C). The lipid metabolic processes exhibited the highest numbers of up-regulated and down-regulated genes. The heatmap revealed that atractylodin treatment led to a decrease in the expression of numerous genes associated with lipid synthesis, whereas genes regulated by PPARα, such as Slc27a1, Ces1, Cyp1a2, Cpt1b, Lipin2, Pck1, and Angptl4, were up-regulated (Fig. 5D). The hepatic mRNA expression of Srebf1 and Dgat2, crucial enzymes in de novo lipogenesis, was markedly downregulated in the HFD-fed mice treated with atractylodin compared to that in HFD-fed mice treated with the vehicle (Fig. 5E). Furthermore, the protein expression of FAS was significantly reduced following treatment with 5 or 10 mg/kg of atractylodin in the HFD-fed mice (Fig. 5F). Our findings indicated that atractylodin significantly influenced lipid metabolic processes, leading to the downregulation of key enzymes involved in de novo lipogenesis, suggesting its role in the regulation of hepatic lipid metabolism.

Atractylodes lancea and atractylodin promoted AMPK phosphorylation

AMPK is a crucial regulator of lipid and glucose metabolism (Cool et al., 2006; Woods et al., 2017). To elucidate the potential binding mode of atractylodin with AMPK, we conducted a docking study using Autodock4, which yielded a predicted binding energy of –5.41 kcal/mol (Fig. 6A). The α subunit domain contains two binding domains. Notably, Arg-72 appeared to play a significant role in mediating key interactions between atractylodin and AMPK.

Figure 6. Atractylodes lancea and atractylodin activate the AMPK signaling pathway. (A) Molecular docking model of atractylodin binding to AMPK. The α subunit is represented in purple, the γ subunit in beige, and the components in yellow. (B, C) Cell viability after Atractylodes lancea (left) or atractylodin treatment (right). HepG2 cells were treated with Atractylodes lancea or atractylodin for 24 h at the indicated doses. The MTT assay was performed, and absorbance at 540 nm OD was measured (n=5). (D, E) Protein levels of phospho-AMPK and phospho-ACC. HepG2 cell were treated with (D) Atractylodes lancea (n=8) or (E) atractylodin (n=3) at the indicated concentrations for 3 h. Data are presented as mean ± SEM. Statistical analysis was performed using one-way ANOVA with Dunnett’s post-hoc test. *p<0.05, ***p<0.001: compared with Control; N.S., not significant.

To investigate whether Atractylodes lancea and atractylodin could activate AMPK, we conducted in vitro assays using HepG2 cells. The MTT assay demonstrated that both Atractylodes lancea and atractylodin exhibited no cytotoxic effects at concentrations of up to 500 μg/ml and 80 μM, respectively (Fig. 6B, 6C). Furthermore, treatment with Atractylodes lancea or atractylodin led to a dose-dependent increase in AMPK phosphorylation. Notably, Atractylodes lancea and atractylodin treatment also resulted in increased phosphorylation of ACC, a known substrate for AMPK (Galic et al., 2018) (Fig. 6D, 6E). These results indicate that Atractylodes lancea and atractylodin could activate the AMPK signaling pathway.

Atractylodin inhibited LXRα-mediated lipogenic gene expression through AMPK activation

LXRα is a key transcription factor that regulates SREBP1c, a protein with a major role in hepatic de novo lipogenesis (Hwahng et al., 2009). We used an LXRα agonist, T0901317 (T090), to gain deeper insights into atractylodin’s regulatory role in lipogenic gene expression. Treatment with T090 increased the mRNA expression of lipogenic genes, including SREBF1, FASN, and ACACA. Moreover, atractylodin treatment effectively reduced the elevated expression of the lipogenic genes induced by T090 (Fig. 7A-7C).

Figure 7. Atractylodin decreases lipogenic gene expression by activating AMPK. (A-C) qRT-PCR analyses of mRNA levels for lipogenic genes (SREBF1, FASN, and ACACA). HepG2 cells were treated with atractylodin for 1 h followed by T090 treatment for 12 h (n=4). (D) qRT-PCR analysis of SREBF1 mRNA levels. HepG2 cells were treated with 5 μM Compound C (C.C.) for 30 min, followed by 20 μM atractylodin for 1 h, and then 3 μM T090 treatment for 12 h (n=5). (E) Representative western blots for precursor and mature SREBP-1c expression. The precursor form of SREBP-1c was assessed using whole cell lysates (WCL) (upper panel), while the mature form was determined from nuclear fractions (NF) (lower panel). Data are presented as mean ± SEM. Statistical analysis was performed using one-way ANOVA with Tukey’s post-hoc test. **p<0.01: compared with Control (Con); ^#p<0.05, ^##p<0.01: compared with T090 treatment, N.S., not significant.

We further explored whether the inhibition of SREBP1c by atractylodin was mediated via AMPK activation. Notably, the ability of atractylodin to suppress T090-induced SREBF1 mRNA as well as precursor and mature SREBP1c protein expression was reversed by compound C, an AMPK inhibitor (Fig. 7D, 7E). These findings strongly suggest that atractylodin downregulated the lipogenic genes by activating the AMPK pathway.

The global prevalence of MASLD is rapidly increasing due to factors such as HFD-associated obesity, sugar consumption, and reduced physical activity. It is more prevalent in Western Pacific region, Southeast Asia and America, and its risk factors include obesity and elevated triglyceride levels (Tang et al., 2022). In a phase 3 clinical trial, obeticholic acid, a representative drug candidate for nonalcoholic steatohepatitis, failed to demonstrate nonalcoholic steatohepatitis resolution with no worsening of fibrosis (Younossi et al., 2019). Anti-diabetic drugs, such as dapagliflozin and semaglutide, are currently in phase 3 clinical trials (He et al., 2022; Loomba et al., 2023). This study suggests that atractylodin can potentially treat insulin resistance and MASLD.

Atractylodin is often present in Atractylodes, including A. lancea, A. chinensis, A. japonica, and A. koreana (Zhang et al., 2021). A previous study that analyzed the components of Atractylodes chinensis, a species similar to Atractylodes lancea, found that the ethyl acetate fraction exhibited the most significant reduction in nitric oxide production compared to other fractions (Ishii et al., 2020). When the ethyl acetate extract was fractionated by silica gel chromatography, four principal constituents in the rhizome of Atractylodes chinensis were separated: Atractylodin, (+)-Hinesol, β-Eudesmol, and (−)-α-Bisabolol. Among them, atractylodin exhibited high potency in suppressing nitric oxide production, with an IC₅₀ value of 8.25 μM. It inhibited the iNOS gene expression and reduced the levels of proinflammatory cytokines induced by IL-1β in hepatocytes (Ishii et al., 2020). The major components of Atractylodes lancea include atractylodin, atractylenolide (I, II, and III), atractylone, hinesol, β-eudesmol, stigmasterol, and β-sitosterol. Daily dosages of 250 mg/kg and 500 mg/kg of the ethanol extract of Atractylodes lancea rhizome has previously demonstrated anti-obesity effects. Moreover, the ethanol extract of Atractylodes lancea can inhibit lipase formation. A previous study showed that among the seven compounds found in the extract, atractylodin displayed the highest lipase inhibitory activity (Jiao et al., 2014). Jiao et al. (2014) showed that atractylodin comprises approximately 2.7% of the Atractylodes lancea extract. In this study, we obtained 5.36 mg of atractylodin from an 80% ethanol extract of 1 g of Atractylodes lancea (data not shown) and investigated the pharmacological effect of Atractylodes lancea and atractylodin.

A previous study highlighted the inhibitory effect of Atractylodes chinensis aqueous extract on obesity, primarily due to its impact on SIRT1/AMPK (Park et al., 2021). SIRT1 acts as an upstream regulator of the AMPK pathway (Price et al., 2012). Moreover, SIRT1 and AMPK are known to play essential roles in regulating various aspects of metabolism, including energy expenditure, fat oxidation, and overall metabolic balance. Diabetes and obesity often lead to a reduction in SIRT1 activity and inhibition of AMPK activity, which, in turn, impair anti-oxidative responses and mitochondrial biogenesis through the inhibition of peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC1α) (Feige et al., 2008). Another study showed that Atractylodes chinensis administered to mice on an HFD effectively suppressed obesity and reduced the weights of white adipose tissues by promoting the activation of SIRT1/AMPK and increasing the expression of genes related to fatty acid oxidation in the white adipose tissues (Park et al., 2021). In addition, atractylodin has been found to increase SIRT1 expression in human umbilical vein endothelial cells (Fujitsuka et al., 2016). The activation of SIRT1 by atractylodin may be a contributing factor to its anti-obesity effect.

Our findings indicated that atractylodin treatment significantly reduced HFD-induced obesity, fatty liver, and insulin resistance, similar to the effects of Atractylodes lancea treatment. These results suggest that atractylodin can effectively treat MASLD. Previous studies demonstrated that 250 mg/kg of an aqueous extract of Atractylodes lancea and 10 mg/kg atractylodin significantly increased the gastric emptying rate (Nakai et al., 2002; Bai et al., 2017). Another study showed that atractylodin promoted the NRF2-mediated ferroptosis (Ye et al., 2023). This study utilized relatively high concentration (20 mg/kg, 40 mg/kg, and 80 mg/kg) of atractylodin compared to the 5 mg/kg and 10 mg/kg doses used in our study. Ye et al. (2023) showed that the doses of 40 mg/kg and 80 mg/kg of atracytlodin attenuate HFD-induced oxidative stress, whereas the dose of 20 mg/kg failed to do so. Our study demonstrated that much lower concentration of atractylodin (5 mg/kg and 10 mg/kg) can improve glucose sensitivity and reduce fatty liver. In addition, we investigated systematic gene regulation by atractylodin and identified AMPK as its key pharmacological target of atractylodin through RNA-seq and molecular docking analysis.

In the present study, we found that atractylodin activated the AMPK signaling pathway (Fig. 6). AMPK is a serine/threonine protein kinase that acts as a critical energy sensor for cellular metabolism (Hwahng et al., 2009). Owing to obesity, the serum levels of free fatty acids may decrease the AMPK activity, which can subsequently lead to the development of insulin resistance, as supported by previous studies. Additionally, a reduction in the AMPK activity is associated with an accumulation of triglycerides in hepatocytes (Yu et al., 2020). AMPK can directly or indirectly inhibit SREBP1c, a key regulator of lipid synthesis from glucose in the liver. SREBP1c serves as a substrate for AMPK, which exerts its inhibitory effect by phosphorylating SREBP1c at Ser372 (Li et al., 2011). Furthermore, AMPK is involved in the transcriptional suppression of SREBP1c by inducing the inhibitory phosphorylation of LXRα (Hwahng et al., 2009). The net result of these actions is a reduction in hepatic lipogenesis and the accumulation of lipids in the liver. Notably, liver-specific AMPK activation in mice can alleviate HFD-induced obesity and fatty liver. Consequently, AMPK has emerged as a promising therapeutic target for fatty liver treatment. Based on our experimental data, we demonstrated that atractylodin activated AMPK (Fig. 6). Furthermore, our results revealed that atractylodin can inhibit LXRα-dependent SREBP1c (Fig. 7). These findings strongly indicate the potential of atractylodin as a therapeutic agent for MASLD treatment.

Numerous genetic and pharmacological studies have consistently emphasized the pivotal role of AMPK in maintaining glucose homeostasis. In the liver, AMPK exerts its influence by inhibiting the expression of gluconeogenic genes, such as PEPCK and G6Pase (Cool et al., 2006; Inoue and Yamauchi, 2006). Furthermore, AMPK plays a role in reducing the expression of the carbohydrate response element-binding protein (ChREBP), which is associated with lipogenesis and glucose production. AMPK can directly phosphorylate ChREBP, thereby reducing its DNA-binding capacity (Sato et al., 2016). Metformin, a widely used drug for type 2 diabetes treatment, is known for its glucose-lowering effects, which are mediated by AMPK activation. The therapeutic benefits of metformin require the presence of LKB1 (Shaw et al., 2005). A recent study demonstrated that the combination of Atractylodes lancea and Magnolia officinalis effectively improved systemic insulin resistance in rats fed a high-fructose diet (Yang et al., 2023). Building upon this foundation, our study revealed that atractylodin treatment improved systemic glucose homeostasis and enhanced insulin sensitivity (Fig. 4). Unlike metformin, our data suggested that atractylodin did not activate LKB1 (data not shown). The beneficial effects of atractylodin on glucose regulation and insulin sensitivity are primarily attributed to its ability to directly activate the AMPK signaling pathway. Furthermore, our RNA-seq analysis revealed that several key genes in the glucose metabolism were significantly altered by the administration of atractylodin (Fig. 4G).

In summary, the results of this study demonstrate that both Atractylodes lancea and its major constituent, atractylodin, can activate the AMPK signaling pathway. These compounds have shown promising effects in alleviating HFD-induced obesity, fatty liver, and insulin resistance (Fig. 8). Importantly, since Atractylodes lancea has a history of traditional use in East Asia, its safety profile is well-established. These findings not only suggest the utilization of Atractylodes lancea and atractylodin as potential MASLD therapeutics but also shed light on their underlying mechanisms.

Figure 8. A schematic illustration showed the therapeutic potential of Atractylodes lancea and its principal constituent, atractylodin, in addressing metabolic dysfunction-associated steatotic liver disease (MASLD). Atractylodes lancea and atractylodin exert significant benefits by activating the AMP-activated protein kinase (AMPK) signaling pathway. These interventions lead to improvements in high-fat-diet-induced obesity, fatty liver, and insulin resistance.

This work was supported by the National Research Foundation of Korea (NRF) funded by the Korea government (MSIT) (No. 2022R1G1A1003601 to S.B.Y. and RS-2023-00210489, RS-2023-00301850 to Y.M.Y.). This research was supported by Korean Fund for Regenerative Medicine (KFRM) funded by Ministry of Science and ICT, and Ministry of Health & Welfare (RS-2022-00070363 to Y.M.Y.). This research was supported by Korea Basic Science Institute (National Research Facilities and Equipment Center) grant funded by the Ministry of Education (No. 2022R1A6C101A739).

The authors have no conflicts of interest to declare.

G.Y.S., S-B.Y., and Y.M.Y. conceived the project and designed the research. G.Y.S., S.M.K., and Y.M.Y. performed the studies, analyzed the data and wrote the manuscript. S.B. performed the molecular docking analysis. S-B.Y. and Y.M.Y. provided administrative support and obtained funding. Y.M.Y. supervised overall data.