Buddleoside alleviates nonalcoholic steatohepatitis by targeting the AMPK-TFEB signaling pathway

ABSTRACT

Nonalcoholic steatohepatitis (NASH) is a combination of hepatic steatosis, inflammation, and fibrosis, and it often follows simple hepatic steatosis in nonalcoholic fatty liver disease (NAFLD). However, no pharmacological treatment is currently available for NASH. Given the important role of TFEB (transcription factor EB) in regulating the macroautophagy/autophagy-lysosomal pathway, TFEB is potentially a novel therapeutic target for treatment of NASH, which function can be regulated by AMP-activated protein kinase (AMPK) and MTOR (mechanistic target of rapamycin kinase) complex 1 (MTORC1). Buddleoside (Bud), a natural flavonoid compound, has recently emerged as a promising drug candidate for liver diseases. Here, we shown that Bud treatment alleviated hepatic steatosis, insulin resistance, inflammation, and fibrosis in mice fed a high-fat and high-cholesterol (HFHC) diet. Notably, Bud activated AMPK, inhibited MTORC1, and enhanced TFEB transcriptional activity as well as autophagic flux in vivo and in vitro. Inhibition of AMPK or knockout of hepatic Tfeb abrogated the alleviation effects of Bud on hepatic steatosis, insulin resistance, inflammation, and fibrosis. Mechanistic investigation revealed that Bud bound to the PRKAB1 subunit via Val81, Arg83, and Ser108 residues and activated AMPK, thereby eliciting phosphorylation of RPTOR (regulatory associated protein of MTOR complex 1) and inhibiting the kinase MTORC1, which activated the TFEB-mediated autophagy-lysosomal pathway and further ameliorated HFHC-induced NASH in mice. Altogether, our results indicate that Bud ameliorates NASH by activating hepatic the AMPK-TFEB axis, suggesting that Bud is a potential therapeutic strategy for NASH.

Abbreviations: ACAC, acetyl-CoA carboxylase; ADaM, allosteric drug and metabolite; AICAR, 5-aminoimidazole-4-carboxamide1-β-D-ribofuranoside; AKT, AKT serine/threonine kinase; ALP, autophagy-lysosomal pathway; AMPK, AMP-activated protein kinase; Bud, buddleoside; CAMKK2, calcium/calmodulin dependent protein kinase kinase 2; CC, compound C; CETSA, cellular thermal shift assay; C_max, maximum concentration; CQ, chloroquine; DARTS, drug affinity responsive target stability assay; EIF4EBP1, eukaryotic translation factor 4E binding protein 1; GOT1, glutamic-oxaloacetic transaminase 1; GPT, glutamic–pyruvic transaminase; GSK3B, glycogen synthase kinase 3 beta; GTT, glucose-tolerance test; HFD, high fat diet; HFHC, high-fat and high-cholesterol; HOMA-IR, homeostasis model assessment of insulin resistance; IKBKB, inhibitor of nuclear factor kappa B kinase subunit beta; INSR, insulin receptor; ITT, insulin-tolerance test; LDH, lactate dehydrogenase; STK11, serine/threonine kinase 11; MAP1LC3/LC3, microtubule associated protein 1 light chain 3; MTORC1, MTOR complex 1; NAFLD, non-alcoholic fatty liver disease; NASH, non-alcoholic steatohepatitis; ND, normal diet; NFKB, nuclear factor kappa B; PA, palmitic acid; PSR, picrosirius red; RRAG, Ras related GTP binding; RPTOR, regulatory associated protein of MTOR complex 1; RPS6, ribosomal protein S6; RPS6KB, ribosomal protein S6 kinase B; SMAD2, SMAD family member 2; SMAD3, SMAD family member 3; SQSTM1, sequestosome 1; TFEB, transcription factor EB; tfeb-HKO, hepatocyte-specific tfeb knockout; TSC2, TSC complex subunit 2.

Introduction

Nonalcoholic fatty liver disease (NAFLD) is one of the most frequent causes of chronic liver disease worldwide. In China, the prevalence of NAFLD is estimated to be 29.2% [1], which surpasses the global prevalence of NAFLD [2]. The advanced stage of NAFLD, nonalcoholic steatohepatitis (NASH), is characterized by hepatic steatosis, insulin resistance, inflammation, and different degrees of fibrosis. Furthermore, progressive liver injury in NASH can lead to cirrhosis, thereby increasing the risk of developing hepatocellular carcinoma. Given the complex pathophysiology of NASH, finding a safe and effective NASH drug presents a significant challenge.

In recent years, TFEB (transcription factor EB) has received a lot of attention as a potential target for the treatment of NAFLD and NASH [3,4]. TFEB is considered the master transcriptional regulator of the autophagy-lysosomal pathway (ALP), which upregulates target genes expression through binding to coordinated lysosomal expression and regulation motifs [5–7]. Notably, TFEB transcriptional activity and autophagic flux were found to be impaired in livers of NAFLD and NASH patients [8–10]. Liver-specific knockout of Tfeb aggravates high fat diet (HFD)-induced hepatic steatosis [11], and pharmacological activation of TFEB prevents NAFLD in mice [3,12], suggesting TFEB plays an important role in regulating NASH development and progression.

In nutrient-rich conditions, MTOR (mechanistic target of rapamycin kinase) complex 1 (MTORC1), comprises three core components, MTOR, RPTOR (regulatory associated protein of MTOR complex 1), and MLST8 (MTOR associated protein, LST8 homolog), is recruited to the lysosomal surface via RRAG GTPases, where it is activated and promotes anabolism in the cell [13–15]. TFEB is predominantly phosphorylated by MTORC1 to promote its binding with YWHA/14-3-3 proteins, which keeps TFEB inactive in the cytoplasm [16,17]. Under energy-deprived conditions, AMP-activated protein kinase (AMPK) inhibits MTORC1 activity through direct phosphorylation of TSC2 (TSC complex subunit 2) and RPTOR [18,19], which in turn enhances autophagic flux by inducing the nuclear translocation of TFEB. Of note, hyperactivation of MTORC1 and suppression of AMPK are thought to be major causes of NASH [19,20]. Indeed, persistent activation of AMPK or inhibition of MTORC1 is sufficient to alleviate hepatic steatosis and hepatocellular injury in mice [21,22]. Thus, manipulation of the AMPK-MTORC1-TFEB signaling pathway may offer therapeutic options for NASH.

In recent years, several natural bioactive components derived from plant secondary metabolites have been pronounced as valuable alternatives for alleviating NASH [23,24]. Buddleoside (Bud, also known as Linarin, Figure S1A), a natural flavonoid compound, has been isolated from various plants mainly belonging to the Asteraceae and Lamiaceae, Scrophulariaceae and Valerianaceae families. Bud has a broad spectrum of biological activities, such as anti-inflammation [25], anti-oxidation [26], and neuroprotective effects [27]. Moreover, studies have proven that Bud exerts protective effects on the liver in a variety of models, including a mouse model of D-galactosamine and lipopolysaccharide-induced fulminant hepatic failure [28] and a duck model of viral hepatitis [29]. However, the protective effects of Bud against NASH and the underlying mechanisms of action of these effects remain unknown. In this study, Bud was identified as a promising candidate for treating NASH because it strongly inhibited hepatic steatosis, inflammation, insulin resistance, and fibrosis in mice. Moreover, we showed that Bud bound to Val81, Arg83 and Ser108 of PRKAB1 subunit and elevated AMPK activity, which in turn blocked RPTOR-MTOR interaction and the subsequent TFEB phosphorylation, thereby attenuating NASH in mice. Our finding demonstrates that Bud is a promising therapeutic agent for NASH and provides a therapeutic strategy for NASH by targeting the AMPK-TFEB signaling pathway.

Results

Pharmacokinetic and liver distribution of Bud

A single dose of Bud (30 mg/kg) was given by intraperitoneal injection, and blood and liver samples were collected at various time points (0, 0.25, 0.5, 0.75, 1, 2, 4, 6, 8, and 12 h) for pharmacokinetic studies in plasma and liver. Plasma maximum concentration (C_max) of Bud was determined as 29.3 ng/mL, whereas C_max in the liver was 4,942 ng/g with time at C_max (T_max) value of 0.25 h and 0.75 h in plasma and liver, respectively (Figure S1B, C). Half-life (T_1/2) of Bud in the liver was 1.36 times higher at 5.38 h than that in plasma with 3.96 h (Figure S1B, C). After single-dose intraperitoneal injection, Bud was detected in the liver, fat, muscle, kidney, lung, and heart. The liver exhibited the highest concentration of Bud, followed by the lung, kidney, heart, fat, and muscle (Figure S1D). These data indicate that the liver may be the target of Bud.

Bud alleviates high-fat and high-cholesterol (HFHC)-induced hepatic steatosis, insulin resistance, inflammation, and fibrosis

Mice were fed either a normal diet (ND) or a HFHC diet for 12 weeks. Then, mice were treated with Bud (30 mg/kg/2 days) by intraperitoneal injection while continuing to receive the ND or HFHC diets for additional 4 weeks (Figure 1A). The body weight (Figure 1B), liver weight (Figure 1C), and liver weight:body weight ratio (Figure 1D) were decreased after 4 weeks of Bud treatment in mice fed a HFHC diet. The accumulation of hepatic lipids, as indicated by hepatic triglyceride content (Figure 1E), hematoxylin and eosin (H&E), and Oil red O staining (Figure 1F), was markedly reduced by Bud administration in HFHC-fed mice. Additionally, Bud treatment improved HFHC-induced liver injury, as evidenced by decreased serum activities of GPT (glutamic – pyruvic transaminase) and GOT1 (glutamic-oxaloacetic transaminase 1; Figure 1G). In mice fed a HFHC diet, fasting blood concentrations of glucose (Figure 1H) and insulin (Figure 1I) as well as homeostasis model assessment of insulin resistance (HOMA-IR, Figure 1J) were all decreased by Bud treatment. Tolerance tests for glucose and insulin further verified improved glucose tolerance (Figure 1K) and insulin sensitivity (Figure 1L) in Bud-treated HFHC-fed mice. In accordance with this result, the HFHC-induced impairment of hepatic insulin signaling was attenuated by Bud treatment, as measured by levels of phosphorylated INSR (insulin receptor), AKT (AKT serine/threonine kinase), and GSK3B (glycogen synthase kinase 3 beta; Figure 1M). In the liver of mice fed a HFHC diet, Bud treatment lessened macrophage infiltration (Figure 1N), protein abundance of phosphorylated IKBKB (inhibitor of nuclear factor kappa B kinase subunit beta) and NFKB (nuclear factor kappa B) as well as mRNA abundance of inflammatory genes, including Tnf, Il18, Il1b, Ccl2 and Nos2, but upregulated protein abundance of inhibitor of NFKBIA (NFKB inhibitor alpha) and mRNA abundance of Il10 (Figure 1O, P). Collagen deposition (Figure 1Q), phosphorylation of SMAD2 (SMAD family member 2) and SMAD3 (Figure 1R) as well as mRNA abundance of fibrotic genes, including Col1a1, Col3a1, Acta2, Ccn2, Tgfb and Timp1 (Figure 1S) were much lower in the liver of Bud-treated HFHC mice than those of HFHC mice. Collectively, these data indicate that Bud treatment protects mice against HFHC-induced NASH.

Figure 1.

Bud alleviates lipid accumulation, insulin resistance, inflammation, and fibrosis in the liver of mice fed a HFHC diet. (A) Scheme of the experimental design. Mice were fed either a ND or a HFHC diet for 12 weeks. Then, mice were treated with Bud (30 mg/kg/2 days) by intraperitoneal injection while continuing to receive the ND or HFHC diets for additional 4 weeks. (B-D) Body weight, liver weight, and ratio of liver weight to body weight. (E) Hepatic triglyceride content. (F) Representative images of H&E and Oil-red O staining of liver sections. Scale bar: 50 μm. (G) Serum activities of GPT and GOT1. (H, I) Fasting blood glucose and insulin levels. (J) HOMA-IR. (K, L) GTT and ITT. (M) Representative immunoblotting of INSR, p-INSR (Ser612), AKT, p-AKT (Ser473), GSK3B and p-GSK3B (Ser9) in the livers from different groups and quantification of p-INSR (Ser612):INSR, p-AKT (Ser473):AKT, and p-GSK3B (Ser9):GSK3B. (N) Representative images of ITGAM and ADGRE1 immunofluorescence staining of liver sections. Scale bar: 20 μm. (O) Representative immunoblotting of IKBKB, p-IKBKB (Ser176/180), NFKBIA, NFKB and p-NFKB (Ser536) in the livers from different groups and quantification of p-IKBKB (Ser176/180):IKBKB, NFKBIA:ACTB, and p-NFKB (Ser536):NFKB. (P) Relative mRNA levels of the inflammation related genes. (Q) Representative images of PSR staining of liver sections. Scale bar: 50 μm. (R) Representative immunoblotting of SMAD2, p-SMAD2 (Ser250), SMAD3 and p-SMAD3 (Ser423/425) in the livers from different groups and quantification of p-SMAD2 (Ser250):SMAD2 and p-SMAD3 (Ser423/425):SMAD3. (S) Relative mRNA levels of the fibrosis related genes. n = 6 mice per group. Data were expressed as the mean ± SEM; *p < 0.05, **p < 0.01.

The therapeutic effects of Bud on hepatic steatosis, insulin resistance, and inflammation were further evaluated with HFD-induced NAFLD model (Figure S1E). Similar to the results found in the NASH model, Bud treatment significantly decreased body weight (Figure S1F), liver weight (Figure S1G), and liver weight:body weight ratio (Figure S1H) in mice fed a HFD diet. Furthermore, Bud administration alleviated hepatic steatosis and injury in HFD mice, as indicated by the decreased hepatic lipid accumulation (Figure S1I, J) and activities of GPT and GOT1 (Figure S1K). Bud treatment decreased blood concentrations of glucose and insulin (Figure S1L, M) and HOMA-IR (Figure S1N), and improved glucose intolerance (Figure S1O) and insulin resistance (Figure S1P, Q) upon HFD challenge. Infiltration of inflammatory cells, activation of NFKB signaling pathway, and expression of inflammatory genes were reduced by Bud treatment in HFD mice (Figure S1R-T). Moreover, results from primary mouse hepatocytes revealed that Bud had no effect on cell viability (Figure S2A-D). Lipid accumulation in palmitic acid (PA)-challenged hepatocytes was reduced by Bud treatment in a dose and time-dependent manner (Figure S2E-G). Bud administration alleviated PA-induced insulin signaling impairment and inflammatory response in mouse hepatocytes (Figure S2H, I). Collectively, these data indicate that Bud ameliorates lipid accumulation, insulin resistance, and inflammatory response in the liver of HFD mice and PA-treated hepatocytes.

Bud activates TFEB-mediated ALP in vivo and in vitro

To uncover the key downstream effector (s) that contribute to Bud-induced protection from NASH, we performed a transcriptomic assay on primary mouse hepatocytes in the presence or absence of Bud treatment. Principal component analysis clearly separated the samples into two clusters (Figure 2A). Compared with control group, Bud treatment upregulated 363 genes and downregulated 267 genes (Figure 2B). Gene set enrichment analysis revealed that pathway related to autophagy was enriched (Figure 2C) and markedly upregulated by Bud treatment compared with the control group (Figure 2D,E). Quantitative reverse transcription (qRT)-PCR analysis verified the augment in autophagy regulated genes induce by Bud treatment (Figure 2F). Given the well-established role of TFEB in regulating ALP [30], we wonder whether Bud activates TFEB in hepatocytes. Bud treatment enhanced nuclear localization of TFEB and decreased phosphorylation of TFEB at Ser211 in hepatocytes (Figure 2G,H). By using the lysosomal protease inhibitor chloroquine (CQ), we found that Bud treatment increased the degradation of SQSTM1 (sequestosome 1) and formation of MAP1LC3/LC3 (microtubule associated protein 1 light chain 3)-II in hepatocytes (Figure 2I,J). Enhanced lysosomal function was observed in hepatocytes, as indicated by the elevated LysoTracker staining (Figure 2G). These results underscore the activation role of Bud on TFEB-mediated ALP in hepatocytes.

Figure 2.

Bud activates TFEB mediated ALP in primary mouse hepatocytes. (A-H) Hepatocytes were treated with or without 40 μM Bud for 12 h. (A) PCA of RNA-seq data for primary hepatocytes treated with or without Bud. (B) Volcano plot showing the changes of hepatocytes genes (fold change ≥ 2). (C) Gene set enrichment analysis showing the cellular pathways. (D) GSEA of pathway related to autophagy. (E) Heatmap showing the expression profile of genes related to autophagy in Ctrl and Bud groups. (F) Relative mRNA levels of the autophagy related genes. (G) Representative fluorescence images of TFEB and LysoTracker in hepatocytes. Scale bar: 20 μm. (H) Representative immunoblotting of TFEB and p-TFEB (Ser211), and nuclear and cytosolic TFEB in hepatocytes and quantification of p-TFEB (Ser211):TFEB, TFEB:ACTB, and TFEB nuclear:cytosolic ratio. (I, J) Hepatocytes were pre-treated with or without 50 μM CQ for 4 h and then treated with or without 40 μM Bud for 12 h. (I) Representative immunofluorescence staining of LC3 in hepatocytes. Scale bar: 20 μm. (J) Representative immunoblotting of SQSTM1 and LC3 in hepatocytes and quantification of SQSTM1:ACTB and LC3-II:ACTB. Data were expressed as the mean ± SEM; *p < 0.05, **p < 0.01.

In the liver of mice fed a HFHC or HFD, Bud administration enhanced transcriptional activity of TFEB, as indicated by reduced phosphorylation of TFEB (Figure S3A, B) and increased nuclear localization of TFEB (Figure S3C, D). Furthermore, Bud administration increased the degradation of SQSTM1 and formation of LC3-II in the liver of HFHC or HFD mice (Figure S3E, F). These data indicate that Bud activates TFEB mediated ALP in the livers of NASH and NAFLD mice.

Knockout of hepatic Tfeb blocks the alleviation effects of Bud on NASH

To further explore whether the function of Bud in NASH depends on TFEB, we generated hepatocyte-specific tfeb-knockout (tfeb-HKO) mice (Figure S4A). Knockout of Tfeb showed a complete abolishment of Bud-mediated activation on ALP in the liver of mice fed a HFHC diet (Figure S4B). In tfeb-HKO mice, Bud treatment (Figure 3A) failed to preclude HFHC-increased body weight (Figure 3B), liver weight (Figure 3C), ratio of liver weight:body weight (Figure 3D), hepatic steatosis (Figure 3E,F), and injury (Figure 3G). The ability of Bud to reduce blood concentrations of glucose and insulin, and HOMA-IR was abolished by knockout of hepatic Tfeb (Figure 3H-J). Deletion of hepatic Tfeb eliminated the alleviation effects of Bud on glucose intolerance (Figure 3K), insulin resistance (Figure 3L,M), inflammatory cells infiltration (Figure 3N), and NFKB signaling activation (Figure 3O,P) in mice fed a HFHC diet. The protective effect of Bud against hepatic fibrosis was diminished by knockout of Tfeb in the liver of HFHC mice, as evidenced by increased collagen deposition (Figure 3Q), SMAD2/SMAD3 phosphorylation (Figure 3R), and expression of fibrotic genes (Figure 3S). Additionally, Bud treatment did not alter these parameters in tfeb-HKO mice fed with ND (Figure S4C-T). Of note, enforced expression of TFEB in tfeb-knockout hepatocytes (Figure S4U) restored the protection effects of Bud on lipid accumulation (Figure S4V, W), insulin insensitivity (Figure S4X), and inflammatory response (Figure S4Y). In hepatocytes, knockdown of Atg5 blocked the autophagic flux even in the presence of Bud (Figure S5A, B). The activation effect of Bud on TFEB was not affected by siAtg5 (Figure S5B, C). However, knockdown of Atg5 abrogated the alleviation effects of Bud on lipid accumulation, insulin resistance, and inflammatory response of hepatocytes (Figure S5D-G). These results indicate that TFEB activation and the resultant of ALP serve as downstream effectors of Bud, contributing to protect against HFHC diet-induced NASH.

Figure 3.

Knockout of hepatic Tfeb blocks the alleviation effects of Bud on NASH in mice. (A) Scheme for the experimental strategy. Mice were fed a HFHC diet for 12 weeks. Then, mice were treated with Bud (30 mg/kg/2 days) by intraperitoneal injection while continuing to receive a HFHC diet for additional 4 weeks. (B-D) Body weight, liver weight, and ratio of liver weight to body weight. (E) Hepatic triglyceride content. (F) Representative images of H&E and Oil-red O staining of liver sections. Scale bar: 50 μm. (G) Serum activities of GPT and GOT1. (H, I) Fasting blood glucose and insulin levels. (J) HOMA-IR. (K, L) GTT and ITT. (M) Representative immunoblotting of INSR, p-INSR (Ser612), AKT, p-AKT (Ser473), GSK3B and p-GSK3B (Ser9) in the livers from different groups and quantification of p-INSR (Ser612):INSR, p-AKT (Ser473):AKT, and p-GSK3B (Ser9):GSK3B. (N) Representative images of ITGAM and ADGRE1 immunofluorescence staining of liver sections. Scale bar: 20 μm. (O) Representative immunoblotting of TFEB, IKBKB, p-IKBKB (Ser176/180), NFKBIA, NFKB and p-NFKB (Ser536) in the livers from different groups and quantification of p-IKBKB (Ser176/180):IKBKB, NFKBIA:ACTB, and p-NFKB (Ser536):NFKB. (P) Relative mRNA levels of the inflammation related genes. (Q) Representative images of PSR staining of liver sections. Scale bar: 50 μm. (R) Representative immunoblotting of SMAD2, p-SMAD2 (Ser250), SMAD3 and p-SMAD3 (Ser423/425) in the livers from different groups and quantification of p-SMAD2 (Ser250):SMAD2 and p-SMAD3 (Ser423/425):SMAD3. (S) Relative mRNA levels of the fibrosis related genes. n = 6 mice per group. Data were expressed as the mean ± SEM; *p < 0.05, **p < 0.01.

Bud inhibits MTORC1 by elevating AMPK-mediated phosphorylation of RPTOR

The activity of TFEB is mainly regulated by the kinase MTORC1 and the phosphatase PPP3/calcineurin. MTORC1 phosphorylates TFEB to promote cytosolic retention while PPP3/calcineurin dephosphorylates TFEB and promotes its nuclear translocation [16,31,32]. In hepatocytes transfected with a constitutively active form of the human PPP3/calcineurin catalytic subunit (∆CaN), activity of PPP3/calcineurin was markedly increased, whereas Bud treatment did not alter the activity of PPP3/calcineurin (Figure S6A). Inhibition of PPP3/calcineurin by Ca²⁺ chelator BAPTA-AM or FK506 failed to preclude TFEB nuclear translocation in Bud-treated hepatocytes (Figure S6B), indicating that PPP3/calcineurin is not responsible for Bud-induced TFEB cytoplasm-to-nucleus translocation. The RRAGs, composed of RRAGA or RRAGB in complex with RRAGC or RRAGD, are tethered to the surface of lysosomes via Ragulator, a heteropentamer in which LAMTOR1 wraps around LAMTOR2 to LAMTOR5, to activate MTORC1 [13,33]. Overexpression of RRAGC^S75L, a constitutively active form RRAGC, markedly increased phosphorylation of MTOR and its substrates RPS6KB (ribosomal protein S6 kinase beta), RPS6 (ribosomal protein S6), and EIF4EEBP1 (eukaryotic translation factor 4E binding protein 1; Figure 4A). In the presence of RRAGC^S75L, Bud treatment diminished the phosphorylation of MTOR and its substrates (Figure 4A), reduced the lysosomal localization of MTOR and the interaction of RRAGC with both MTOR and TFEB, but did not affect the interaction of RRAGC with LAMTOR1 (Figure 4B,C), suggesting that Bud inhibits MTORC1 activity via a RRAGs-independent manner. Phosphorylation of RPTOR by AMPK induced the binding of RPTOR to YWHA/14-3-3 and resulted in the suppression of MTORC1 kinase activity [19]. In the present study, Bud treatment increased RPTOR phosphorylation, decreased the interaction between RPTOR and MTOR, and elevated the interaction of RPTOR with PRKAA and YWHA/14-3-3 (Figure 4D). Of note, Bud administration activated AMPK signaling pathway, enhanced phosphorylation of RPTOR, and decreased MTORC1 kinase activity in the livers of NASH and NAFLD mice (Figure 4E,F) as well as in hepatocytes (Figure 4G,H). These data demonstrate that Bud inhibits MTORC1 via activation of AMPK.

Figure 4.

Bud activates AMPK in the liver and mouse primary hepatocytes. (A-C) HEK293T cells were transfected with wild type RRAGC or active RRAGC^S75L and then treated with or without 40 μM Bud for 12 h. (A) Representative immunoblotting of HA-RRAGC, MTOR, p-MTOR (Ser2448), RPS6KB, p-RPS6KB (Thr389), RPS6, p-RPS6 (Ser240/244), EIF4EBP1 and p-EIF4EBP1 (Ser65) in the HEK293T cells and quantification of p-MTOR (Ser2448):MTOR, p-RPS6KB (Thr389):RPS6KB, p-RPS6 (Ser240/244):RPS6, and p-EIF4EBP1 (Ser65):EIF4EBP1. (B) Representative images of MTOR immunofluorescence staining and LysoTracker in HEK293T cells and quantification of colocalization coefficient. Scale bar: 10 μm. (C) Representative co-ip analysis to assay interactions between RRAGC and LAMTOR1, MTOR and TFEB in HEK293T cells. (D) Representative co-ip analysis to assay interactions between RPTOR and MTOR, PRKAA, and YWHA/14-3-3 in hepatocytes. Hepatocytes were treated with or without 40 μM Bud for 12 h. (E-G) Representative immunoblotting of PRKAA, p-PRKAA (Thr172), ACAC, p-ACAC (Ser79), RPTOR, p-RPTOR (Ser792), MTOR, p-MTOR (Ser2448), RPS6KB and p-RPS6KB (Thr389) in the livers or hepatocytes and quantification of p-PRKAA (Thr172):PRKAA, p-ACAC (Ser79):ACAC, p-RPTOR (Ser792):RPTOR, p-MTOR (Ser2448):MTOR, and p-RPS6KB (Thr389):RPS6KB. E: Mice were treated as described in Figure 1. F: Mice were fed a ND or a HFD for 12 weeks. Then, mice were treated with Bud (30 mg/kg/2 days) by intraperitoneal injection while continuing to receive a HFD for additional 4 weeks. G: Hepatocytes were treated with different concentrations of Bud (0, 5, 10, 20, 40 μM) for 12 h or treated with 2 mm Met for 2 h. (H) Representative immunoblotting of MTOR, p-MTOR (Ser2448), RPS6KB, p-RPS6KB (Thr389), RPS6, p-RPS6 (Ser240/244), EIF4EBP1 and p-EIF4EBP1 (Ser65) in the hepatocytes and quantification of p-MTOR (Ser2448):MTOR, p-RPS6KB (Thr389):RPS6KB, p-RPS6 (Ser240/244):RPS6, and p-EIF4EBP1 (Ser65):EIF4EBP1. Hepatocytes were treated with different concentrations of Bud (0, 5, 10, 20, 40 μM) for 12 h or treated with 250 nM torin1 for 3 h. n = 6 mice per group for E and F. Data were expressed as the mean ± SEM; *p < 0.05, **p < 0.01.

AMPK is indispensable for the beneficial effects of Bud on NASH

Next, we examined whether Bud alleviates HFHC diet-induced NASH through activation of AMPK. In the presence or absence of Bud, inhibition of AMPK by compound C (CC) decreased phosphorylation of RPTOR and increased phosphorylation of MTORC1 substrates TFEB and RPS6KB (Figure S6C). Bud-induced TFEB nuclear translocation (Figure S6D), SQSTM1 degradation, and LC3-II formation (Figure S6E) were blocked by CC in mice fed a HFHC diet. In mice treated with Bud (Figure 5A), reductions in body weight (Figure 5B), liver weight (Figure 5C), ratio of liver weight:body weight (Figure 5D), hepatic steatosis (Figure 5E, F) and injury (Figure 5G) were abrogated by CC. In mice treated with CC, Bud had no effects on blood concentrations of glucose and insulin and HOMA-IR (Figure 5H–J). Consistently, CC treatment blocked the beneficial effects of Bud on insulin resistance (Figure 5K–M), hepatic inflammatory response (Figure 5N–P) and fibrosis (Figure 5Q–S) in mice fed a HFHC diet. Similar results were observed in ND mice treated with Bud and CC (Figure S6F-W). Additionally, knockdown of Prkaa1 and Prkaa2 dampened the efficacy of Bud on RPTOR phosphorylation, MTORC1 kinase activity, TFEB transcriptional activity and ALP in hepatocytes (Figure S7A-C). Knockdown of Prkaa1 and Prkaa2 impeded the improvement effects of Bud on lipid accumulation (Figure S7D, E), insulin resistance (Figure S7F), and inflammatory response (Figure S7G) in hepatocytes challenged by PA. Taken together, activation of the AMPK-TFEB signaling pathway is required for the protective effects of Bud against HFHC diet-induced NASH in mice.

Figure 5.

Inhibition of AMPK blocks the protection effects of Bud on NASH in mice. (A) Scheme for the experimental strategy. Mice were fed a HFHC diet for 12 weeks. During the twelfth week, mice were treated with CC (10 mg/kg/2 days) or PBS. Then, mice were treated with CC (10 mg/kg/2 days) or Bud (30 mg/kg/2 days) alone, or co-treated with CC (10 mg/kg/2 days) and Bud (30 mg/kg/2 days) by intraperitoneal injection while continuing to receive a HFHC diet for additional 4 weeks. (B-D) Body weight, liver weight, and ratio of liver weight to body weight. (E) Hepatic triglyceride content. (F) Representative images of H&E and Oil-red O staining of liver sections. Scale bar: 50 μm. (G) Serum activities of GPT and GOT1. (H, I) Fasting blood glucose and insulin levels. (J) HOMA-ir. (K, L) GTT and ITT. (M) Representative immunoblotting of INSR, p-INSR (Ser612), AKT, p-AKT (Ser473), GSK3B and p-GSK3B (Ser9) in the livers from different groups and quantification of p-INSR (Ser612):INSR, p-AKT (Ser473):AKT, and p-GSK3B (Ser9):GSK3B. (N) Representative images of ITGAM and ADGRE1 immunofluorescence staining of liver sections. Scale bar: 20 μm. (O) Representative immunoblotting of IKBKB, p-IKBKB (Ser176/180), NFKBIA, NFKB and p-NFKB (Ser536) in the livers from different groups and quantification of p-IKBKB (Ser176/180):IKBKB, NFKBIA:ACTB, and p-NFKB (Ser536):NFKB. (P) Relative mRNA levels of the inflammation related genes. (Q) Representative images of PSR staining of liver sections. Scale bar: 50 μm. (R) Representative immunoblotting of SMAD2, p-SMAD2 (Ser250), SMAD3 and p-SMAD3 (Ser423/425) in the livers from different groups and quantification of p-SMAD2 (Ser250):SMAD2 and p-SMAD3 (Ser423/425):SMAD3. (S) Relative mRNA levels of the fibrosis related genes. n = 6 mice per group. Data were expressed as the mean ± SEM; *p < 0.05, **p < 0.01.

Bud elevates AMPK activity via targeting to the PRKAB1 subunit

In general, AMPK is activated by sensing increases in AMP:ATP and ADP:ATP ratios, whereas Bud treatment did not affect these ratios (Figure 6A). The kinases STK11 (serine/threonine kinase 11), CAMKK2 (calcium/calmodulin dependent protein kinase kinase 2), MAP3K7 (mitogen-activated protein kinase kinase kinase 7) largely account for the phosphorylation and activation of AMPK [34–36]. Nevertheless, knockdown of Stk11, Camkk2 or Map3k7 (Figure S8A-C) did not alter the activation effect of Bud on AMPK, as evidenced by elevated phosphorylation of PRKAA, ACAC (acetyl-CoA carboxylase) and RPTOR (Figure 6B). The interaction of PRKAA with STK11, CAMKK2, and MAP3K7 was not affect by Bud treatment (Figure 6C). These data indicate that Bud can elevate AMPK activity directly and not as a consequence of stimulation of upstream kinases, or by altering the AMP:ATP and ADP:ATP ratios. Subsequently, we evaluated whether Bud activates AMPK via interacting with AMPK subunits. Molecular dynamics simulation revealed the binding mode with an established structure of AMPK (PDB: 5ISO). The results showed that five residues of PRKAA2 subunit (Val11, Leu18, Gly19, Ile46 and Phe90) and six residues of PRKAB1 subunit (Val81, Arg83, Ser108, His109, Phe112 and Val113) formed interactions with Bud (Figure 6D). Results from cellular thermal shift assay (CETSA) and drug affinity responsive target stability assay (DARTS) revealed that Bud enhanced the thermal stability of PRKAB1 subunit and protected the digestion of PRKAB1 subunit caused by proteases (Figure 6E,F). Surface plasmon resonance analysis (SPR) analysis revealed that recombinant PRKAB1 subunit bound to Bud with high affinity, yielding a dissociation constant (K_D) of 2.23 ± 0.13 μM (Figure 6G). We next generated a series of PRKAB1 subunit mutants and found that mutation of Val81, Arg83 or Ser108 alone or simultaneously disrupted the interaction between Bud and PRKAB1 subunit, as indicated by CETSA and DARTS assays (Figure 6H, I). Notably, Bud-induced interaction between AMPK and RPTOR was decreased by mutation of Val81, Arg83, and Ser108 (Figure 6J). These data indicate that Bud can activate AMPK by directly binding to Val81, Arg83, and Ser108 of PRKAB1 subunit.

Figure 6.

Bud binds to Val81, Arg83, and Ser108 of PRKAB1. (A) The ratios of AMP:ATP and ADP:ATP in the primary mouse hepatocytes. Hepatocytes were treated with different concentrations of Bud (0, 5, 10, 20, 40 μM) for 12 h. (B) Representative immunoblotting of PRKAA, p-PRKAA (Thr172), ACAC, p-ACAC (Ser79), RPTOR and p-RPTOR (Ser792) in the different groups and quantification of p-PRKAA (Thr172):PRKAA, p-ACAC (Ser79):ACAC, and p-RPTOR (Ser792):RPTOR. Hepatocytes transfected with siCtrl, siStk11, siCamkk2 or siMap3k7 were treated with or without 40 μM Bud for 12 h. (C) Representative co-ip analysis to assay interactions between PRKAA and STK11, CAMKK2, MAP3K7 and RPTOR in hepatocytes. Hepatocytes were treated with or without 40 μM Bud for 12 h. (D) Stable three-dimensional structure of AMPK binding with Bud based on molecular dynamics simulation and the detailed presentation of the binding sites. (E, F) CETSA and DARTS were performed to measure the binding ability of Bud to PRKAA2, PRKAB1 and PRKAG1 in HEK293T cells. (G) SPR assay of PRKAB1 interaction with Bud. (H, I) CETSA and DARTS were performed to measure the binding ability of Bud to PRKAB1 mutants. (J) Representative co-ip analysis to assay the interaction between HA-PRKAB1 and RPTOR in HEK293T cells. The cells were transfected with wild type PRKAB1 or mutant PRKAB1 (V81A, R83A, S108A), and then treated with or without 40 μM Bud for 12 h. Data were expressed as the mean ± SEM. *p < 0.05; **p < 0.01.

Val81, Arg83, and Ser108 of PRKAB1 subunit are required for the beneficial effects of Bud

To determine whether Val81, Arg83, and Ser108 are responsible for the protection effects of Bud on PA-induced lipotoxicity, we used small interfering RNA (siRNA) to knockdown of Prkab1 (Figure S8D) and then re-introduced PRKAB1 or its mutant (V81A, R83A, and S108A) in hepatocytes. Mutation of PRKAB1 did not affect the activity of AMPK, as evidenced by the fact that AMPK was activated and inhibited in hepatocytes treated with AICAR and CC, respectively (Figure S8E). In hepatocytes transfected with mutant PRKAB1, Bud treatment did not alter AMPK and MTORC1 activity, TFEB transcriptional activity, and autophagic flux as well as lysosomal function (Figure 7A–C). Moreover, mutation of Val81, Arg83, and Ser108 dampened the improvement effects of Bud on lipid accumulation (Figure 7D, E), insulin resistance (Figure 7F) and inflammatory response (Figure 7G) in PA-treated hepatocytes. Altogether, Bud-alleviated PA-induced lipotoxicity required Val81, Arg83, and Ser108 of PRKAB1 subunit.

Figure 7.

Val81, Arg83, and Ser108 of PRKAB1 is response for the beneficial effects of Bud on pa-induced lipid accumulation, insulin resistance, and inflammatory response in primary mouse hepatocytes. (A-C) Hepatocytes that transfected with siPrkab1 were transfected with wild type PRKAB1 or mutant PRKAB1 (V81A, R83A, S108A). Then, cells were treated with or without 40 μM Bud for 12 h. To block autophagic flux, hepatocytes were treated with 50 μM CQ for 4 h before Bud treatment. (A) Representative immunoblotting of PRKAA, p-PRKAA (Thr172), ACAC, p-ACAC (Ser79), RPTOR, p-RPTOR (Ser792), TFEB, p-TFEB (Ser211), MTOR, p-MTOR (Ser2448), RPS6KB and p-RPS6KB (Thr389) in different groups and quantification of p-PRKAA (Thr172):PRKAA, p-ACAC (Ser79):ACAC, p-RPTOR (Ser792):RPTOR, p-TFEB (Ser211):TFEB, p-MTOR (Ser2448):MTOR and p-RPS6KB (Thr389):RPS6KB. (B) Representative images of TFEB and LysoTracker fluorescence staining in hepatocytes. Scale bar: 20 μm. (C) Representative immunoblotting of SQSTM1 and LC3 in hepatocytes and quantification of SQSTM1:ACTB and LC3-ii:actb. (D-G) Hepatocytes that transfected with siPrkab1 were transfected with wild type PRKAB1 or mutant PRKAB1 (V81A, R83A, S108A). Then, cells were treated with or without 40 μM Bud for 12 h after PA administration (400 μM, 12 h). (D) The triglyceride content in hepatocytes. (E) Representative images of lipid droplet fluorescence staining with BODIPY493/503 in hepatocytes. Scale bar: 20 μm. (F) Representative immunoblotting of INSR, p-INSR (Ser612), AKT, p-AKT (Ser473), GSK3B and p-GSK3B (Ser9) in different groups and quantification of p-INSR (Ser612):INSR, p-AKT (Ser473):AKT and p-GSK3B (Ser9):GSK3B. (G) Relative mRNA levels of the inflammation related genes. Data were expressed as the mean ± SEM. *p < 0.05; **p < 0.01.

Discussion

The AMPK-MTORC1 switching orchestrates multiple interconnected aspects of metabolic homeostasis of the liver. In the present study, we demonstrate that Bud interacts with PRKAB1 subunit, thus activating AMPK and suppressing MTORC1 pathway. This results in activation of TFEB-mediated ALP, which alleviates hepatic steatosis, insulin resistance, inflammatory response, and fibrosis of mice with NASH (Figure 8). Our discovery of Bud as an AMPK activator opens new avenues for pharmacological treatment of NASH.

Figure 8.

Bud alleviates NASH by targeting the AMPK-TFEB signaling pathway in the liver of mice. Bud binds to PRKAB1 subunit and activates AMPK, thereby increasing phosphorylation of RPTOR and inhibiting MTORC1. This results in activation of tfeb-mediated ALP, which alleviates hepatic steatosis, insulin resistance, inflammatory response, and fibrosis of mice with NASH (created with BioRender.com).

The hepatic inflammatory response is regarded as an essential driving force of NAFLD progression, as it promotes sustained fibrogenesis, which ultimately leads to hepatic carcinogenesis [37]. Multiple lines of evidence, including pro-inflammatory factors detection [25], NFKB nuclear localization analysis [38], inflammatory response assay and cellular functional evaluation [39], underscoring a suppression effect of Bud on inflammation. Of note, Bud enriched Flos Chrysanthemi extract attenuated liver inflammation and injury in rats fed a HFHC diet [40]. In line with previous observations, transcription of pro-inflammatory factors was markedly downregulated in the liver of mice treated with Bud, accompanied by decreased phosphorylation of IKBKB and NFKB. More importantly, the mitigation of hepatic inflammatory response along with alleviated hepatic steatosis, fibrosis, and injury as well as enhanced insulin sensitivity in mice treated with Bud agree with previous studies in which Bud treatment exerted hepatoprotective effects in a variety of models [28,29,41]. However, whether Bud has the potential to treat other inflammation-driven diseases, such as rheumatoid arthritis, chronic kidney disease and cardiovascular disease, will be the subject of future studies.

Emerging experimental evidence has indicated that hepatic autophagy in mice with NASH is impaired [8,42], and activation of autophagy ameliorates NASH phenotypes, such as hepatic steatosis, inflammation, fibrosis, and insulin resistance [43–45]. Results from RNA-seq revealed that genes associated with autophagy were transcriptionally upregulated in Bud-treated hepatocytes. Moreover, Bud treatment enhanced hepatic autophagic flux in mice fed with HFD or HFHC. These findings are in line with previous study reporting that Bud treatment increased LC3-II formation in the liver of mice injected with carbon tetrachloride [46]. Of note, Bud increased TFEB transcriptional activity and knockout of Tfeb blocked the activation effect of Bud on ALP as well as the improvement effects of Bud on NASH. It has been reported that overexpression of Tfeb upregulated genes involved in the ALP and alleviated HFD-induced hepatic steatosis and insulin resistance in mice [11]. Thus, manipulation of TFEB-mediated ALP via pharmacological means could be an effective approach to alleviate NASH.

TFEB phosphorylation (Ser211) and cytosolic retention are primarily controlled by MTORC1 [16,17,30], which activity can be inhibited by AMPK via phosphorylating TSC2 and RPTOR [18,19]. It has been reported that depletion of Tsc2 did not affect amino acid deprivation-induced phosphorylation and subcellular localization of TFEB [47], implying that MTORC1-dependent phosphorylation of TFEB is insensitive to TSC2. The fact that Bud increases the phosphorylation of AMPK and RPTOR and the interaction of RPTOR with AMPK and YWHA/14-3-3, and reduces the interaction of RPTOR with MTOR underscores that Bud activates AMPK, and subsequent phosphorylates RPTOR and inhibits MTORC1. Furthermore, in mice fed a HFHC diet, inhibition of AMPK abolished the regulation effects of Bud on MTORC1 and TFEB-mediated ALP, and impeded the mitigation effect of Bud on NASH, demonstrating that AMPK-TFEB axis mediates the therapeutic effects of Bud on NASH. Paquette et al (2021) reported that AMPK regulates TFEB via two separate mechanisms: (1) inhibition of MTORC1 activity that subsequently promotes TFEB nuclear translocation and (2) stimulates transcriptional activation via phosphorylating TFEB at serine residues S466, S467, and S469 [48]. Additionally, activation of AMPK results in the stabilization of CARM1 (coactivator-associated arginine methyltransferase 1) in the nucleus, which in turn functions as a co-activator of TFEB in enhancement of autophagy [49]. Thus, Bud-increased TFEB transcriptional activity may be control by AMPK at multiple levels.

The discovery of A-769662 provided a novel avenue into development of direct AMPK activators by revealing that activation of AMPK with non-nucleotide small molecules is possible [50]. In the present study, Bud does not change cellular AMP:ATP and ADP:ATP ratios. Moreover, knockdown of Stk11, Camkk2, and Map3k7 did not alter the activation effect of Bud on AMPK. By using molecular dynamics simulations and mutant of PRKAB1, we demonstrated that Bud binds to PRKAB1 via three residues Val81, Arg83, and Ser108, which locate at the beta-subunit carbohydrate-binding module (B-CBM). Structural studies of AMPK have identified the allosteric drug and metabolite (ADaM) site, located between the B-CBM and N-lobe on the A-subunit [51,52]. Many pharmacological activators, including A-769662, MK-8722, and salicylate, allosterically activate AMPK by binding to ADaM site and protect against dephosphorylation of PRKAA subunit [50,53,54]. Hence, it could be possible that a similar mechanism exists in the activation effect of Bud on AMPK.

Many natural compounds such as resveratrol and berberine have the characteristics of regulating multiple targets [55,56]. It has been reported that Bud binds to KEAP1 (kelch like ECH associated protein 1) and alleviates oxidative stress in hepatocytes [57]. There were previously reported that Bud treatment suppressed inflammatory response by inhibiting TLR4-MYD88 pathway [38,46,58]. It should be noted that KEAP1 and TLR4 are key facilitators and therapeutic targets in NASH [59,60]. In the present study, inhibition of AMPK or knockout of Tfeb abrogated the protective effects of Bud against the development of hepatic steatosis, insulin resistance, inflammation, and fibrosis in mice fed a HFHC diet. Thus, pharmacological and genetic evidence indicate that activation of AMPK-TFEB axis as a necessary downstream signaling effector for Bud-mediated protection against NASH.

It has been reported that Bud activated autophagy and inhibited inflammatory response in a mouse model of CCl₄-induced acute liver injury [46]. In the liver and lung of diabetic mice, Bud increased phosphorylation of PRKAA via binding directly of ADaM site of AMPK (LysA31 and ArgB83) [41,61]. Furthermore, Bud enriched Flos Chrysanthemi Indici extract (72% of linarin) attenuated liver injury and inflammation in rats fed with a HFHC diet [40]. However, due to the low purity of Bud in this extract (only 72% linarin) [40], the observed biological functions may also come from ingredients other than Bud. The present study showed that Bud elevated AMPK activity via binding to Val81, Arg83 and Ser108 of PRKAB1, which in turn reduced MTORC1-meidated TFEB phosphorylation, thereby activating hepatic ALP and attenuating NASH in mice. Thus, the animal models and experimental objectives were different between our experiment and previous studies, and further studies are needed to gain a more comprehensive understanding of the biological functions of Bud.

Collectively, the present study not only provide mechanistic insights to explain the bioactivity of Bud in alleviating NASH but also open exciting opportunities of developing AMPK-TFEB axis regulators for pharmacological treatment of NASH.

Materials and methods

Antibodies and reagents

Antibodies specific for PRKAA (2532; 1:1000 WB), phospho (p)-PRKAA-Thr172 (2531; 1:1000 WB), AKT (9272; 1:1000 WB), p-AKT-Ser473 (4060; 1:1000 WB), p-GSK3B-Ser9 (5558; 1:1000 WB), TFEB (4240; 1:1000 WB; 1:100 IF), p-TFEB-Ser211 (37681; 1:1000 WB), MTOR (2983; 1:1000 WB; 1:200 IF), p-MTOR-Ser2448 (2971; 1:1000 WB), RPS6KB (9202; 1:1000 WB), p-RPS6KB-Thr389 (9206; 1:1000 WB), EIF4EBP1 (9644; 1:1000 WB), p-EIF4EBP1-Ser65 (9456; 1:1000 WB), RPS6 (2317; 1:1000 WB), p-RPS6-Ser240/244 (5364; 1:1000 WB), RRAGC (3360; 1:1000 WB; 1:200 IF) and NFKBIA (4814; 1:1000 WB) were purchased from Cell Signaling Technology. Antibodies specific for INSR (ab69508; 1:1000 WB), GSK3B (ab69739; 1:1000 WB), LC3 (ab128025; 1:1000 WB; 1:100 IF), Histone-H3 (ab1220; 1:1000 WB), ACTB (ab8226; 1:2000 WB), HRP-conjugated secondary antibodies to mouse (ab205719; 1:5000 dilution) and rabbit (ab205718; 1:5000 dilution) IgGs were purchased from Abcam. Antibodies specific for SMAD2 (A7699; 1:1000 WB) and p-SMAD2-Ser250 (AP1007; 1:1000 WB), SMAD3 (A19115; 1:1000 WB), p-SMAD3-Ser423/425 (AP0727; 1:1000 WB), IKBKB (A19606; 1:1000 WB), p-IKBKB-Ser176/180 (AP0546; 1:1000 WB), NFKB (A19653; 1:1000 WB), ITGAM (A1581; 1:100 IF), SQSTM1 (A19700; 1:1000 WB), p-INSR-Ser612 (AP0553; 1:1000 WB), p-ACAC (AP0298; 1:1000 WB), p-RPTOR (AP09928; 1:1000 WB), LAMTOR1 (A11619; 1:1000 WB), STK11 (A2122; 1:1000 WB), CAMKK2 (A9899; 1:500 WB) and MAP3K7 (A19077; 1:1000 WB) were purchased from ABclonal. Antibody specific for p-NFKB-Ser536 (AF2006; 1:1000 WB) was purchased from Affinity Biosciences. Antibodies specific for TUBB3 (10068–1-AP; 1:5000 WB), HA (66006–2-Ig; 1:10,000 WB), ACAC (21923–1-AP; 1:1000 WB), RPTOR (20984–1-AP; 1:1000 WB), ADGRE1 (27044–1-AP; 1:100 IF), PRKAA2 (18167–1-AP; 1:1000 WB), PRKAB1 (10308–1-AP; 1:1000 WB), PRKAG1 (10290–1-AP; 1:1000 WB) were purchased from Proteintech. Antibody specific for YWHA/14-3-3 (sc -133,232; 1:200 WB) was purchased from Santa Cruz Biotechnology. FITC-conjugated goat anti-rabbit IgG (111-095-003; 1:100 dilution) and cyanine Cy™3-conjugated goat anti-rabbit IgG (111-165-003; 1:100 dilution) were purchased from Jackson ImmunoResearch Laboratories.

Bud (B20860; greater than 98%, HPLC) was purchased from Shanghai Yuanye Bio-Technology Co., Ltd., dissolved in dimethyl sulfoxide (DMSO; Solarbio, D8370). PA (Sigma-Aldrich, P0500) was dissolved in 0.1 M NaOH at 70°C and then complexed with 10% bovine serum albumin (BSA; Sigma-Aldrich, B2064) at 55°C for 10 min to achieve the final palmitate concentration (100 mm). CQ (PHR1258) was purchased from Sigma-Aldrich. CC (HY-13418A) was purchased from MedChemExpress. Leupeptin (S7380), FK506 (S5003), and BAPTA-AM (S7534) were purchased from Selleck. 5-aminoimidazole-4-carboxamide1-β-D-ribofuranoside (AICAR; S1515), metformin (Met; S1741), and torin1 (SC0245) were purchased from Beyotime Institute of Biotechnology.

Animals and treatment

The Institutional Animal Care and Use Committee of Jilin University approved the study protocol (SY202004001, Changchun, China). The animals received humane care according to the Guide for the Care and Use of Laboratory Animals published by the National Academy of Sciences and the National Institutes of Health. All mice were kept in a standard environment with a 12-h dark/light cycle (lights on at 06:30 am). The temperature and humidity were maintained at 23 ± 3°C and 35 ± 5%, respectively. Male 8- to 10-week-old C57BL/6 mice were purchased from the Beijing Vital River Laboratory Animal Technology Co., Ltd. Tfeb^flox/flox mice were generated as described previously [11] and were crossed with Alb (albumin) Cre (The Jackson Laboratory) to generate tfeb-HKO mice. Age matched litter mates from Tfeb^flox/flox and Alb Cre-negative mice were used as control wild-type mice for the knockout mice.

To establish a NASH mouse model, 8-week-old male C57BL/6 mice were fed a HFHC diet (TrophicDiet, TP26304). A NAFLD mouse model was established by feeding the C57BL/6 mice a HFD (60% kcal from fat; Research Diet, D12492). Mice that received a ND (10% kcal from fat; Research Diet, D12450B) served as controls. To evaluate the effects of Bud on NAFLD and NASH, mice were administered with Bud (30 mg/kg, i.p.) every other day for 4 weeks. To inhibit AMPK activity in vivo, mice were treated with CC (10 mg/kg/2 days, i.p.). Mice treated with vehicle (DMSO or PBS [Solarbio, P1020]) were served as control and subjected to the same experimental approaches in parallel with those in Bud or CC group. To block autophagic flux, mice were treated with 40 mg/kg (i.p.) leupeptin for 8 h. The detailed groups and the number of mice included are shown in the figure legends.

Pharmacokinetic study

Pharmacokinetic study of Bud was performed by using male C57BL/6 mice after single-dose intraperitoneal injection at 30 mg/kg. Plasma and tissue samples were collected at 0, 0.25, 0.5, 0.75, 1, 2, 4, 6, 8, and 12 h following the injection. The concentration of the Bud in mice plasma and tissue homogenates was analyzed using liquid chromatography-mass spectrometry (LC-MS; AB SCIEX, Framingham, USA). The pharmacokinetic parameters C_max, T_max, and T_1/2 were analyzed using WinNonlin software.

Metabolic assay and measurement of liver injury

The fasting blood glucose and insulin levels were determined using a glucometer (Abbott Diabetes Care Inc., Alameda, CA, USA) and enzyme-linked immunosorbent assay (Millipore, E-ELEM1382), respectively. The HOMA-IR index using the following mathematical formulation Eq: HOMA-IR = Insulin (μU/mL) × Glucose (mmol/L)/22.5. Glucose-tolerance test (GTT) was carried out on mice that had been fasted overnight for 16 h. After determination of fasted blood glucose level, each mouse received an intraperitoneal injection of 2 g/kg body weight of glucose. Subsequently, blood glucose level was detected from tail vein after 15, 30, 60, and 120 min. Insulin-tolerance test (ITT) were carried out in random-fed mice. After measuring basal blood glucose level, each mouse was treated with 0.75 U/kg body weight of insulin. Blood glucose level was recorded after 15, 30, 60, and 120 min.

The activities of GPT and GOT1 in serum or medium, and the activity of LDH (lactate dehydrogenase) in medium were measured using commercial kits (Jiancheng Bioengineering Institute, C009-3-1, GPT; C010-3-1, GOT1; A020-2-2, LDH) according to the manufacturer’s instruction.

Histological analysis

Liver tissue was fixed in 10% formaldehyde neutral buffer solution, embedded in paraffin, cut into 8-μm sections and stained with H&E (Solarbio, G1120) or picrosirius red (PSR; Solarbio, G1472). For Oil red O staining, liver tissue was frozen in OCT compound (Sakura Finetek Co., 4583), sectioned at an 8-μm thickness at −18°C and fixed with 75% alcohol at room temperature for 15 min. Then, the slides were stained with Oil red O (Sigma-Aldrich, O0625) and counterstained with hematoxylin. The images were collected by inverted microscope (Olympus, Japan).

Fluorescence assay

Liver or cells were fixed with 4% paraformaldehyde. Antigen retrieval was performed using EDTA-Na₂ at 95°C for 5 min. The liver or cells were permeabilized using 0.1% Triton X-100 (Beyotime Institute of Biotechnology, P0096) and then incubated with antibody diluted in PBS containing 5% goat serum (Beyotime Institute of Biotechnology, C0265) overnight at 4°C. The liver or cells were then incubated with secondary antibodies conjugated to FITC (Jackson ImmunoResearch Laboratories, 111-095-003) or Cy3 (Jackson ImmunoResearch Laboratories, 111-165-003). Nuclei were stained with DAPI (Sigma-Aldrich, D9542). For fluorescence staining of lysosomes and lipid droplets, the cells were washed in PBS and then stained with LysoTracker Red DND-99 (50 nmol/L, 1 h; Thermo Fisher Scientific, L7528) and BODIPY493/503 (1 μg/mL, 15 min; Thermo Fisher Scientific, D3922). The stained cells and slides were observed by laser confocal microscopy (Fluoview FV1200, OLYMPUS).

Cell culture and treatment

Primary hepatocytes were isolated from 6- to 8-weeks-old male mice by the collagenase perfusion method. Only hepatocytes with viability > 90%, as judged by Trypan blue exclusion, were used. The isolated mouse primary hepatocytes were seeded into 6-well tissue culture plates at 1 × 10⁶ cells/mL using Dulbecco’s modified Eagle medium (DMEM) medium (HyClone; Thermo Scientific, SH30022.01) containing 10% fetal bovine serum (FBS; Gibco 10,099,141), 100 nM insulin (Beyotime Institute of Biotechnology, P3376), 100 nM dexamethasone (Beyotime Institute of Biotechnology, ST1254) and 1% penicillin-streptomycin (Beyotime Institute of Biotechnology, C0222) for 5 h and then cultured in DMEM medium containing 10% FBS and 1% penicillin-streptomycin at 37°C under 5% CO₂. HEK293T cells were purchased from Cell Bank of the Chinese Academy of Sciences (SCSP-502). The cells were cultured in DMEM medium containing 10% FBS, 1% non-essential amino acids (Solarbio, N1250), and 1% penicillin-streptomycin at 37°C under 5% CO₂. HepG2 cells were purchased from Cell Bank of the Chinese Academy of Sciences (SCSP-510). The cells were maintained in DMEM supplemented with 10% FBS and 1% penicillin-streptomycin at 37°C under 5% CO₂. Mycoplasma contamination was negative for all cells.

To mimic in vivo hepatic steatosis, primary hepatocytes were maintained in medium containing 2% BSA and treated with 400 μM PA for 12 h. To activate the insulin signaling pathway, hepatocytes were treated with 100 nM insulin for 30 min. To study the function of Bud, cells were treated with different concentrations of Bud (0, 1, 5, 10, 20, 40, 80 or 160 μM) for 12 h, 40 μM Bud for different time points (0, 3, 6, 12 or 24 h) or 40 μM Bud for 12 h. To activate AMPK activity, the cells were treated with 1 mm AICAR for 12 h or 2 mm Met for 2 h. To inhibit AMPK activity, the cells were treated with 10 μM CC for 2 h. To inhibit MTORC1 activity, the cells were treated with 250 nM torin1 for 3 h. To block autophagy, cells were treated with 50 μM CQ for 4 h. To reduce cytoplasmic Ca²⁺ level, cells were treated with BAPTA-AM (10 μM for 6 h). To inhibit PPP3/calcineurin activity, cells were treated with FK506 (5 μM for 3 h). The cell treatments and detailed group information are shown in the figure legends.

Plasmids, site-directed mutagenesis, and siRNA

The plasmid encoding full-length TFEB-GFP has previously been described [31] and was used as template to amplify Tfeb cDNA, which was subcloned into pc-DNA3.1-HA (Addgene 128,034; deposited by Oskar Laur) to generate the pcDNA3.1-HA-TFEB. pPK5-HA-GST-RRAGC-S75L (19305) and pRK5-HA-GST-RRAGC-WT (19304) were kind gifts from D. Sabatini (Addgene). The plasmid carrying a constitutively active form of the human PPP3/calcineurin catalytic subunit (∆CaN) was donated by B. Rothermel (The University of Texas Southwestern Medical Center, USA). The plasmid of human pCMV-HA-PRKAA2 (P47894), pCMV-HA-PRKAB1 (P47874) and pCMV-HA-PRKAG1 (P47906) were purchased from Miaoling Biotech. Mutant of the human pCMV-HA-PRKAB1 was constructed by site-directed mutagenesis using QuikChange II kit (Agilent Technologies Inc., 200524) according to the manufacturer’s instructions. The corresponding primers for each mutant are listed in Table S1. All constructs were confirmed by full-length DNA sequencing. The specific siRNA targeting mouse Stk11 (5’-UCUUCUUGAGGAUCUUGACCG-3’), mouse Camkk2 (5’- AGUAUUGUCAUUUUCAUUGUA-3’), mouse Map3k7 (5’-UCAAGAAGACCUGUUUGUCUU-3’), mouse Prkaa1 (5’- GCAUAUGCUGCAGGUAGAUTT-3’), mouse Prkaa2 (5’- UCAAUGAUAAGAUGAUAAGCC-3’), mouse Prkab1 (5’- UCAUAUUCUGAAUCUCAUCCU-3’) and mouse Atg5 (5’-AAAGCAAAUAGUAUGGUUCUG-3’) were chemically synthesized by Sangon Biotech. The nontargeting siRNA were purchased from Sangon Biotech.

Protein extraction, nucleus and cytoplasm separation, and western blotting

Liver tissue and hepatocytes total protein were extracted using a protein extraction kit (Sangon Biotech, C510003). Nucleus and cytoplasm separation were conducted using a Nuclear-Cytosol Extraction Kit (Applygen Technologies, P1200). The protein concentration was estimated by the BCA method (Applygen Technologies, P1511). The samples were separated on 10–15% Tris-glycine gels with a known pre-stained protein ladder (Thermo Fisher Scientific 26,616) and electrophoretically transferred to a polyvinylidene difluoride membrane. The membranes were blocked in 3% BSA (Sigma-Aldrich, B2064) in TBS-T buffer (Beyotime Institute of Biotechnology, ST673) for 4 h at room temperature. The blocked membrane was incubated overnight at 4°C with the primary antibody. The membranes were then incubated with HRP-conjugated anti-rabbit or anti-mouse IgG at room temperature for 45 min. Immunoreactive bands were visualized by enhanced chemiluminescence solution (Millipore, WBKLS0500). All bands were analyzed using Image-Pro Plus 6.0 (Media Cybernetics, Rockville, MD, USA).

Co-immunoprecipitation

The cells were lysed with NP-40 buffer (Beyotime Institute of Biotechnology, P0013F) supplemented with protease and phosphatase inhibitors (Thermo Fisher Scientific 78,429). The supernatants were incubated with anti-HA affinity gel (Sigma-Aldrich, E6779) overnight at 4°C. Immunoprecipitation protein was fractionated by SDS-PAGE and analyzed by western blotting with the indicated primary antibodies. For endogenous proteins, co-IP was performed according to the user guide of the Pierce Crosslink Immunoprecipitation Kit (Thermo Fisher Scientific 26,147). IgG (ABclonal, AC005) was used as control. Input and immunoprecipitated proteins were analyzed by western blotting.

qRT-PCR assay

The total RNA from hepatocytes and liver tissue was extracted using RNAiso Plus (TaKaRa Biotechnology, 9109) according to the manufacturer’s instructions. The RNA concentration and quality were measured using a Nanophotometer N50 Touch (Implen GmbH) and electrophoresis (1% agarose gels). Then, 1 μg of total RNA in each sample was reverse-transcribed to cDNA in a 20-μL reaction using a reverse transcription kit (TaKaRa Biotechnology, RR047A) according to the supplier’s protocol. We evaluated mRNA abundance using qRT-PCR technology with FastStart Universal SYBR Green Master Mix (Roche Diagnostics 04,913,850,001) and a 7500 Real-Time PCR System (Applied Biosystems Inc., Foster City, CA, USA). The relative expression of each target gene was normalized to two reference genes, Actb (actin beta) and Gapdh (glyceraldehyde-3-phosphate dehydrogenase) and calculated using the 2^−∆∆CT method. The primers used in this study are shown in Table S2. The cycles-to-threshold values of Actb and Gapdh were not affected by the experimental treatment, which validated the usefulness of these genes as control genes.

RNA-sequencing

Total RNA was extracted from the hepatocytes using TRIzol® Reagent (Thermo Fisher Scientific, 15596026CN) according to the manufacturer’s instructions. The concentration and purity of RNA was determined using a NanoDrop microspectrophotometer (Thermo Fisher Scientific). Only high-quality RNA sample (OD260:280 = 1.8 ~ 2.2, OD260:230 ≥ 2.0) was used to construct sequencing library. RNA purification, reverse transcription, library construction and sequencing were performed at Shanghai Majorbio Bio-pharm Biotechnology Co., Ltd. All data were analyzed on the online platform of Majorbio Cloud Platform (https://cloud.majorbio.com/page/tools.html).

Cell viability, triglyceride content assessment, and PPP3/calcineurin activity

The cell viability was assessed by CCK-8 kit (Dojindo Molecular Technologies Inc., CK04) according to the manufacturer’s instructions. Shortly, cells were seeded at 5 × 10³ cells/well in 96-well plates incubated at 37°C in 5% CO₂. After treatment, 20 μL of CCK-8 was added to plates. Then the cells were incubated for an extra 4 h. The optical density was measured at 450 nm on a spectrophotometer (Thermo Fisher Scientific).

Liver tissue or hepatocytes were homogenized in 5% Triton X-100 and then heated in a water bath (85°C) for 3 min. After cooling at room temperature, the sample was vortexed and centrifuged at 2,000 × g for 5 min at 4°C. The supernatant was collected and stored at −80°C before the triglyceride assay. The triglyceride content was measured using an enzymatic kit (Applygen Technologies, E1013) following the manufacturer’s instructions.

Activity of PPP3/calcineurin was measured by using a biochemical kit (Jiancheng Bioengineering Institute, A068-1-1) following the manufacturer’s instructions.

CETSA and DARTS

The stabilization of targets in cells by Bud interaction was evaluated by CETSA and DARTS as described previously [62,63]. In brief, cells were harvested and lysed with a freeze-thawed method using liquid nitrogen. The cell lysates were divided into two fractions and were incubated with either DMSO or Bud (40 μM) at 37°C for 1 h. Then two groups of lysates were aliquoted to 8 PCR tubes, respectively, followed by heating at 54, 56, 58, 62, 66, 70, 74 or 78°C for 4 min. After cooling at room temperature for 5 min, the lysates were centrifuged at 20,000 × g for 20 min at 4°C. The supernatants were collected, and equal amount of protein were analyzed by western blotting. To perform DARTS, total protein was extracted using a protein extraction kit (Sangon Biotech, C510003) containing protease and phosphatase inhibitors (Thermo Fisher Scientific 78,429). Then, 10 × TNC buffer (1 M Tris-HCl, 5 mL; 5 M NaCl, 1 mL; 1 M CaCl₂, 1 mL; double-distilled H₂O, 3 mL; pH 7.4) was added to the cell lysate to reach the final concentrations (5 µg/µL) and gently mixed. Bud or the same volume of DMSO was added to cell lysate and incubated for 1 h at 37°C. Pronase E (dissolved in 1× TNC buffer; MedChemExpress, HY-114158) at a 1:1,000 ratio (wt:wt) was added into both the DMSO and Bud groups for 15 min at 37°C. Digestion was stopped by adding 5 × protein loading buffer and the sample was boiled for 5 min. Then, protein was analyzed by western blotting.

Molecular dynamics simulation

The structure of mammalian AMPK complex (PDB code: 5ISO) [64] was used to the initial structure for molecular docking calculation by AutoDock Vina software. The structure of Bud was optimized at the B3LYP/6–31 G* level by using Gaussian 09 program, and molecular mechanical parameters of Bud was determinate based on the antechamber programs and RESP partial atomic charges from the AmberTools20 software. Subsequently, the standard docking procedures for AMPK and Bud were performed using AutoDock vina software. To obtain the stable structure of AMPK and Bud, the structure obtained from molecular docking was used as the initial structure, and the 400-ns standard molecular dynamics simulation was performed for the 3D structure of AMPK with Bud using Amber20 software package. The detailed docking process was referenced from previous studies [65].

SPR

Biacore 3000 instrument (GE Healthcare, Piscataway, NJ, USA). Recombinant PRKAA2, PRKAB1 or PRKAG1 was directly immobilized onto a CMD500M chip (XanTec bioanalytics GmbH, SC CMD500M0517.a) through amine coupling after activation of the chip with 50 mm N-hydroxysuccinimide (GE Healthcare, BR-1000-50) and 200 mm 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide (Sigma-Aldrich, E7750). Bud was injected for 120 s in running buffer (PBS containing 1% DMSO and 0.05% Surfactant P20 [Cytiva Life Sciences Solutions, BR100054]) followed by a 60-s dissociation step. In parallel, a reference channel with no protein immobilized was injected with Bud under the same conditions for background subtraction. Data analysis was performed using BIA evaluation software.

Statistical analysis

Data were expressed as the mean ± standard error of mean (SEM). All analyses were performed using GraphPad Prism 8.0 (Graph Pad Software, San Diego, CA, USA) or Statistical Package for the Social Sciences (SPSS) 19.0 software (IBM, Chicago, IL, USA). Statistical significance was calculated using two-tailed Student’s t-tests for comparisons between two groups and one-way ANOVA followed by post hoc test adjusted using Bonferroni correction for comparisons among more than two groups. *p < 0.05 was considered statistically significant, and **p < 0.01 was considered highly significant.

Funding Statement

This work was supported by the National Key R&D Program of China [Beijing, China; grant no. 2023YFE0116900 and 2023YFD1801400], National Natural Science Foundation of China [Beijing, China; grant nos. 32172927, 32473105, and 32402959], the Jilin Province Young and Middle-aged Outstanding Talent in Science and Technology Innovation and Entrepreneurship [20230508020RC], and China Postdoctoral Science Foundation [GZC20230951 and 2024M751094].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data from this study are available upon reasonable request.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15548627.2025.2466145