Introduction

Ectopic hepatic lipid deposition is a major point in the development of insulin resistance, nonalcoholic fatty liver (NAFLD), hyperlipidemia, atherosclerosis and other related complications of obesity1. The mechanisms of ectopic hepatic lipid accumulation are controversial, but the interaction of increased lipolysis, lipid transport or hepatic de novo lipogenesis (DNL) is certainly involved2,3. One key transcription factor in DNL is sterol regulatory element binding protein 1c (SREBP-1c), which controls hepatic DNL primarily via regulation of the expression of genes involved in DNL, lipid homeostasis and glucose metabolism4.

There is evidence that lipid deposition is associated with alterations in SREBP-1c expression. Research has confirmed that there is nearly a 5-fold increase in SREBP-1c mRNA in cases of NAFLD compared with healthy controls5. Yang et al also found similar connections between SREBP-1c and cases consistent with the diagnostic standard of fatty liver6. SREBP-1c levels are elevated in the fatty livers of obese (ob/ob) mice with insulin resistance and hyperinsulinemia caused by leptin deficiency7,8. Knebel et al generated mice with liver-specific over-expression of mature human SREBP-1c under control of the albumin promoter and a liver-specific enhancer (alb-SREBP-1c). The alb-SREBP-1c mice developed hepatic lipid accumulation featuring a fatty liver by the age of 24 weeks under normocaloric nutrition9. Mice with germ line deletion of SREBP-1c exhibit diminished fatty acid synthesis in the liver10. There are three forms of SREBP in mammals: SREBP-1a, -1c and -2. These peptides undergo similar proteolytic activation and share some target genes, SREBP-1a and -1c primarily stimulate fatty acid synthesis, whereas SREBP-2 acts primarily on cholesterol biosynthetic genes and the LDL receptor (LDLR) gene11. SREBP-1c is synthesized as a precursor that is attached to the endoplasmic reticulum. The mature protein translocates to the nucleus following cleavage and activates transcription by binding to a sterol regulatory element (SRE). Sterols inhibit the cleavage of the precursor, and the mature nuclear form is rapidly catabolized, thereby reducing transcription12.

SREBP-1c is a target of itself, and it can be induced by the appearance of nuclear SREBP (n-SREBP) in the nucleus13. Studies have also shown that insulin strongly increases SREBP-1c processing to liberate the nuclear form, and it increases the transcription of the SREBP-1c gene, leading to increases in SREBP-1c mRNA and the precursor protein11. Insulin induces SREBP-1c mRNA by as much as 40-fold within 6 h under optimal conditions with freshly isolated rat hepatocytes in cell culture14. The increase can be blocked by wortmannin, an inhibitor of phosphatidylinositol 3-kinase (PI3K)15, which is an early enzyme in the insulin-signaling cascade. Low concentrations of rapamycin can also block the increase, which indicates that enhanced transcription occurs partially through PI3K/AKT (RAC-alpha serine/threonine-protein kinase)/mTORC1 (mammalian target of rapamycin complex 1)16. In addition, insulin stimulation of SREBP-1c processing in hepatocytes requires p70 S6-kinase (p70S6K), which is activated by mTORC117. The insulin pathway bifurcates at AKT. One branch regulates SREBP through the activation of mTORC1; the other branch prevents n-SREBP degradation through glycogen synthase kinase 3 (GSK3). GSK3 phosphorylates n-SREBP, which leads to the ubiquitination and proteasomal degradation of n-SREBP18,19.

mTORC1 is active under nutrient-rich conditions and inactive under nutrient-poor conditions, whereas AMPK (AMP-activated protein kinase) is activated in the inverse pattern20. An AMPK is well known as a master energy sensor that maintains whole-body energy homeostasis energy homeostasis21. Activated AMPK in the liver is involved in the regulation of fatty acid oxidation through the phosphorylation of acetyl-CoA carboxylase (ACC), which subsequently leads to a reduction of malonyl-CoA. Malonyl-CoA is an allosteric inhibitor of carnitine palmitoyl transferase (CPT), the key master of the carnitine-dependent transport across the mitochondrial inner membrane and the rate-limiting step of fatty acid β-oxidation22. AMPK activation also leads to a concomitant inhibition of fatty acid synthesis and the activation of fatty acid oxidation, partially through its inhibition of the rapamycin-sensitive mTORC1 pathway, which controls SREBP-1c transcription. AMPK phosphorylation of the tuberous sclerosis complex 2 (TSC2) tumor suppressor contributes to mTORC1 suppression. Dana M Gwinn reported that AMPK directly phosphorylates the mTOR binding partner regulatory-associated protein of mTOR (Raptor) at two well-conserved serine residues, and this phosphorylation induces 14-3-3 binding to Raptor. The phosphorylation of Raptor by AMPK is required for the inhibition of mTORC1 and cell cycle arrest induced by energy stress23.

In this study, we used a cell-based luciferase reporter driven by an SRE-containing promoter to identify the compounds that regulate SREBP transcriptional activity, and we hypothesized that inhibition of the SREBP-1c pathway, using a novel small molecule, would lower the risk of obesity-induced comorbidities related to hepatic lipid accumulation.

Materials and methods

Cell culture

Rat hepatocytes were isolated using Selgen's two-step perfusion method24 and maintained in DMEM. HepG2 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% FBS (Invitrogen, Carlsbad, CA, USA). Culture and differentiation of 3T3-L1 cells were conducted as described previously25.

Dual-luciferase assay

HepG2 cells were co-transfected with plasmid pGL2 (Promega, Madison, WI, USA) luciferase reporter plasmids containing a human LDLR promoter (−355/+3) as the wild-type sterol regulatory element (SRE) or the mutant reporter with disrupted SRE (Addgene, Cambridge, MA, USA) together with the Renilla luciferase reporter plasmid pRL-SV40 (Promega, Madison, WI, USA). Briefly, HepG2 cells were depleted of sterols by incubation in DMEM containing 5% lipoprotein-deficient serum (LPDS, Hyclone, Logan, UT, USA) and 10 μmol/L compactin (Gene Operation, Ann Arbor, MI, USA) for 16 h. The cells were treated with control DMSO and ZJ001 at the indicated concentrations for 6 h. Then the cells were lysed with reporter lysis buffer (Sigma, St Louis, MO, USA), and luciferase activity was measured according to the manufacturer's protocol.

Lipid synthesis in HepG2 and primary rat hepatocytes

Primary rat hepatocytes were isolated, cultured as described above, and treated with ZJ001 at the indicated concentrations for 20 h. Cells were pulsed with [2-14C] acetate for 4 h. The media was removed, cells were washed three times with cold PBS, and the incorporation of [2-14C] acetate was measured as previously described26.

Measurement of respiration in isolated mitochondria and intact cells

Respiration measurements in intact cells and isolated mitochondria were conducted at 37 °C in a Clark-type oxygen electrode (Strathkelvin Instruments, Lanarkshire, UK). Cells or mitochondria were transferred to the electrode chamber. After a steady rate of oxygen consumption had been attained, ZJ001 was added dose-dependently, and its effect on oxygen consumption was recorded. Mitochondria were isolated from rat liver according to previously reported methods27. For HepG2 cells, rat hepatocytes and 3T3-L1 adipocytes, respiration measurements were conducted using DMEM; for mitochondria, a standard respiration medium was used 28.

Sulforhodamine B (SRB) cytotoxicity assay

HepG2 cells were treated with or without the indicated doses of ZJ001 for 24 h, followed by SRB assay, as previously described29.

Oil Red staining

3T3-L1 cells were treated with either DMSO or ZJ001 for 7 d during adipogenesis. Next, the cells were fixed in 10% formalin for 1 h at room temperature and washed with 60% isopropanol. The cells were stained with Oil Red for 10 min at room temperature, followed by washing with distilled water. Images for each dish were captured using a microscope (Olympus Corporation, Tokyo, Japan), and absorbance was measured using a spectrophotometer (Molecular Device, Sunnyvale, CA, USA) at 500 nm.

Animal experiments

Animal experiments were approved by the Animal Ethics Committee of the Shanghai Institute of Materia Medica. Male C57BL/6J mice (6 weeks old) were purchased from Shanghai SLAC Laboratory Animal Co Ltd. They were housed in a temperature-controlled room (22±2 °C) with a light/dark cycle of 12 h. After 8 weeks of a high-fat diet (20% protein, 20% carbohydrate and 60% fat, Research Diets, New Brunswick, NJ, USA) to induce obesity, the mice were randomized into various treatment groups based on their body weight and random blood glucose levels. The treatment groups for the 7-week chronic study using intragastric gavage (ig) were as follows: vehicle (0.5% microcrystalline cellulose) and ZJ001 (15 mg·kg−1·d−1). Body weight and food intake were recorded daily. After 4 weeks of treatment, and after 6 h of starvation, the intraperitoneal glucose tolerance test was performed via intraperitoneal injection of D-glucose (1.5 g/kg). After 7 weeks of treatment, the mice were sacrificed 4 h after a final dose was administered. The tissues were collected by freeze-clamping for further analysis.

Blood glucose was determined using a free-style blood glucose monitoring system (Abbott Laboratories, Chicago, IL, USA). Plasma insulin and leptin levels were measured using ELISA kits (Linco Research, Saint Charles, MO, USA). Plasma triacylglycerol (TG), cholesterol (TC), HDL-cholesterol (HDL-c), LDL-cholesterol (LDL-c), lactate and non-esterified fatty acids (NEFA) were assayed using kits from Shanghai Fudan-Zhangjiang (Shanghai, China) and Wako Diagnostics (Richmond, VA, USA). Hepatic and muscular TG contents were measured using Folch's method30.

Quantitative RT-PCR

Total RNA was prepared from cultured cells or mouse tissues using Trizol reagents (Invitrogen, Carlsbad, CA, USA). Equal amounts of RNA from individual samples were subjected to quantitative real-time PCR using SYBR green with the Stratagene Mx30005PTM Q-PCR Systems (Agilent Technologies, Santa Clara, CA, USA) according to the manufacturer's protocol. Primer sequences are listed in Supplemental Table S1. All reactions were performed in triplicate. The relative amounts of mRNAs were normalized to GAPDH using the comparative CT method.

Immunoblotting

Total proteins from tissues or cells were prepared in RIPA buffer (50 mmol/L Tris-HCl, pH 8.0, 150 mmol/L NaCl, 1% NP-40, 1 mmol/L Na3VO4, 1 mmol/L DTT, 1 mmol/L EDTA, 1 mmol/L EGTA) containing complete protease inhibitors (Roche, Basel, BS, Switzerland). Protein (20 μg per sample) was electrophoresed through SDS-PAGE after boiling for 5 min in SDS loading buffer. The following primary antibodies were used for immunoblotting: rabbit polyclonal anti-SREBP-1c IgG-H160 and rabbit polyclonal anti-p70S6K H-9 were purchased from Santa Cruz (Santa Cruz, CA, USA); rabbit polyclonal anti-phospho-SREBP-1c (Ser372), rabbit polyclonal anti-AMPKα, rabbit polyclonal anti-phospho-AMPKα (Thr172), rabbit polyclonal anti-ACC, rabbit polyclonal anti-phospho-ACC (Ser79), rabbit polyclonal anti-mTOR, rabbit polyclonal anti-phospho-mTOR (Ser2448), rabbit polyclonal anti-phospho-p70S6K (Ser411), rabbit polyclonal anti-S6, rabbit polyclonal anti-phospho-S6 (Ser235/236), rabbit polyclonal anti-AKT, rabbit polyclonal anti-phospho-AKT (Ser473), rabbit polyclonal anti-Raptor, rabbit polyclonal anti-phospho-Raptor (Ser792), rabbit polyclonal GSK3α/β, rabbit polyclonal anti-phospho-GSK3α/β (Ser21/9), rabbit polyclonal-Tuberin/TSC2, and rabbit polyclonal anti-phospho-Tuberin/TSC2 (Thr1462) were purchased from Cell Signaling Technology (Danvers, MA, USA).

Adenine nucleotide extraction and measurement

Adenine nucleotides were extracted from cells using perchloric acid and measured using HPLC, as previously described31. Before the experiment, the culture medium was changed to serum-free medium for 3 h and DMSO or ZJ001 was added for 1 h. Adenine nucleotides were extracted from cells using perchloric acid and measured using HPLC.

Statistical analysis

The results represent the means±SEM. All the in vitro experiments were conducted at least three times. Differences between two groups were examined using a two-tailed unpaired Student's t-test. Differences between multiple groups were compared using one-way ANOVA, followed by an LSD comparison. P<0.05 was considered statistically significant.

Results

ZJ001 down-regulated the transcription activity of the SRE-containing promoter and inhibited DNL in HepG2 cells

The nuclear form of SREBP binds to SRE and directly activates the expression of lipogenic genes32. To identify small-molecule compounds that inhibit transcriptional activity of SREBP, we transfected a luciferase reporter driven by an SRE-containing promoter into HepG2 cells. These cells were then incubated with compounds, after which luciferase activity was measured. As a positive control, 25-hydroxycholesterol (25-HC) dramatically inhibited the reporter activity. Notably, the transcriptional activation of the wild-type LDLR promoter (−355/+3) was markedly inhibited by a synthesized compound with novel structure, ZJ001 [2-(3-benzoylthioureido)-4,5,6,7-tetrahydrobenzo[b]thiophene-3-carboxylic acid] (Figure 1A). Mutation of the SRE motif (C/A) in the same promoter diminished basal transcription and prevented the further decrease caused by ZJ001, and a similar phenomenon was observed with 25-HC (Figure 1B).

Figure 1
figure 1

ZJ001 inhibited the biosynthesis of cholesterol and fatty acid by decreasing SREBP target genes in HepG2 cells. (A) Chemical structure of ZJ001. (B) ZJ001 down-regulated the activity of an SRE-containing promoter. (C–E) ZJ001 inhibits de novo lipid synthesis. Measurements of fatty acid (C) and cholesterol (D) synthesis by ZJ001. (E) Cell survival results by multi-dose of ZJ001 treatment. (F–H) Gene expression changes in HepG2 cells with ZJ001 treatment. The results are shown as the mean±SEM (n=3). bP<0.05, cP<0.01 compared with the control group.

PowerPoint slide

Full size image

Because ZJ001 was able to inhibit the transcriptional activity of SREBP, we examined the potential of ZJ001 to inhibit DNL by measuring fatty acid and sterol synthesis in HepG2 cells at the indicated non-cytotoxic concentrations (Figure 1C–1E). Regulation of SREBP transcription activity occurs at two levels: transcriptional regulation of SREBP mRNA itself and processing regulation of pre-SREBP protein. We first investigated the mRNA expression of SREBPs and their target genes to elucidate the mechanism that underlies the hepatic lipid content-lowering effect of ZJ001. The results indicated that ZJ001 decreased the mRNA expression of SREBP-2 and SREBP-1c, but not SREBP-1a, in dose-dependent manners. The expression of their target genes, such as fatty acid synthase (FAS), HMG-CoA reductase (HMGCR) and LDLR, respectively, was also inhibited (Figure 1F–1H). These results suggested that ZJ001 down-regulated the mRNA of SREBPs and their target genes, which may cause the inhibition of DNL in hepatic cells.

ZJ001 ameliorated metabolic disorders in diet-induced obese mice

C57BL/6J mice fed fat-rich diets become obese and insulin-resistant and develop moderate hyperglycemia that mimics high-calorie-diet-induced obesity (DIO). To test the chronic effects of ZJ001 on obesity and other comorbidities in vivo, we treated the DIO mice with ZJ001 at a dose of 15 mg·kg−1·d−1 or vehicle by oral gavage for 7 weeks. The distribution of ZJ001 in tissues and plasma 2 h after administration of 15 mg/kg demonstrated that ZJ001 had a favorable tissue distribution in the liver of C57BL/6J mice (6.9 μmol/L, Figure S1), which suggests that 15 mg·kg−1·d−1 may be active in vivo. The two groups of mice showed similar food intake during the treatment. However, obese mice treated with ZJ001 had gained markedly less weight (49%) than those in the vehicle-treated obese group at the end of the 7th week (Figure 2A–2C).

Figure 2
figure 2

ZJ001 ameliorated metabolic status in high-fat-diet-fed mice. Vehicle and ZJ001 (15 mg·kg−1·d−1) was administered to mice via oral gavage twice daily. The mice were sacrificed after 7 weeks of treatment and subjected to various analyses. (A and B) Food intake and body weight per mouse during the 7-week experiment. (C) The increase in body weight per mouse after 7-week treatment. (D) Fasting blood glucose level on the 4th week of treatment after 6-h starvation. (E) Effect of ZJ001 on glucose tolerance in mice fed a high-fat diet, as determined by the glucose tolerance test. (F) Quantification of the area under the curve (AUC) from the glucose tolerance test. Error bars represent mean±SEM (n=6–8). bP<0.05, cP<0.01 compared with the HF-Veh group.

PowerPoint slide

Full size image

Atherosclerosis is characterized by elevated plasma cholesterol and a high LDL-c/HDL-c ratio33. Our model successfully reproduced this tendency, as indicated by the result that the plasma cholesterol level and LDL-c/HDL-c ratio in obese mice were more than twice those in mice fed normal chow. Plasma parameter analyses showed that ZJ001 significantly decreased plasma cholesterol (TC) by 16% and LDL-c/HDL-c ratio by 24%, indicating that ZJ001 may potentially reduce the risk of atherosclerosis. Meanwhile, the liver weight of the ZJ001 treatment group was decreased, and the TC and TG contents in the liver of ZJ001-treated mice were lower than those in vehicle-treated mice, suggesting that ZJ001 may have improvement on fatty liver (Table 1).

Table 1 Chronic effects of ZJ001 on metabolic variables in DIO mice.
Full size table

ZJ001 treatment caused a significant reduction in perirenal, subcutaneous and epididymal fat pad weight, but plasma NEFA was not altered after ZJ001 treatment (Table 1). These results suggest that fat pad loss in white adipose tissue does not occur through triglyceride lipolysis. Furthermore, ZJ001 had no direct effect on adipogenesis in 3T3-L1 cells (Figure S2A), and O2 consumption was not affected by ZJ001 in 3T3-L1 adipocytes (Figure S2B). These data may indicate that ZJ001 ameliorated metabolic disorders in DIO mice through an improvement in ectopic hepatic lipid deposition and that the effects on fat pads may be a feedback effect of chronic treatment.

Furthermore, we investigated whether ZJ001 improved glucose metabolism in vivo. The intraperitoneal glucose tolerance test was performed at the 4th week of treatment. The results showed that ZJ001 treatment significantly improved glucose tolerance capacity as evidenced by the significantly lower blood glucose levels at 15 min and 60 min post-glucose injection and the area under the curve of 0–120 min, but it had no effect on fasting blood glucose level (Figure 2D–2F).

ZJ001 regulated the expression of metabolic genes in vivo

Consistent with previous reports34,35,36, we observed that the mRNA levels of SREBP-1c and SREBP-2, and their target genes, were increased in the liver of long-term high-fat-diet-fed mice. The beneficial effects of ZJ001 on metabolic parameters were reflected by its effects on gene expression changes in the liver of mice. SREBP-2 primarily controls cholesterol homeostasis by activating genes required for cholesterol synthesis and uptake. ZJ001 treatment reduced hepatic SREBP-2 mRNA levels. Consequently, the mRNA levels of genes involved in cholesterol biosynthesis, including HMGCR, HMG-CoA synthase (HMGCS), and squalene synthase (SS), were reduced in the liver of ZJ001-treated mice (Figure 3A). SREBP-1c preferentially regulates the lipogenic process by activating genes involved in fatty acid and triglyceride synthesis. ZJ001 significantly decreased the mRNA of SREBP-1c and its target genes, including FAS and ATP-citrate synthase (ACS) (Figure 3B). SREBP-1a is expressed constitutively at low levels in the liver and most other tissues of adult animals. SREBP-1a is a potent universal activator of all SRE-responsive genes. Similarly, ZJ001 treatment did not alter the mRNA of SREBP-1a (Figure 3C), which was consistent with the in vitro effect in HepG2 cells. Moreover, the mRNA levels of genes involved in fatty acid oxidation, such as peroxisome proliferator-activated receptors α (PPARα) and long-chain acyl-CoA dehydrogenase (LCAD) were slightly decreased, suggesting that ZJ001 may reduce fatty acid synthesis and attenuate lipid oxidation (Figure 3D). However, ZJ001 treatment increased the mRNA of uncoupling protein 2 (UCP2) (Figure 3D). ZJ001 significantly decreased genes in carbohydrate metabolism, such as glucose-6-phosphatase (G6Pase) and phosphoenolpyruvate carboxykinase (PEPCK) mRNA, without altering the expression of pyruvate dehydrogenase kinase isoform 4 (PDK4) (Figure 3E).

Figure 3
figure 3

Changes in mRNA levels in liver of the mice treated by ZJ001. The mice used here are the same as those used in Figure 2. The high-fat-diet-fed mice received vehicle and 15 mg·kg−1·d−1 of ZJ001 were used in analyzing gene expression. For each mouse, equal amounts of total RNA from the liver tissues were subjected to Q-PCR quantification as described in Experimental Procedures. Mouse GAPDH was used as the control. Values represent the amount of mRNA relative to mice fed a high-fat diet in each experiment, which is arbitrarily defined as 1. (A–C) The mRNA level of genes involved in lipid synthesis. (D) The mRNA level of genes involved in fatty acid oxidation. (E) The mRNA level of genes involved in carbohydrate metabolism. The results are shown as the mean±SEM (n=6–8). bP<0.05, cP<0.01 compared with the HF-Veh group.

PowerPoint slide

Full size image

ZJ001 decreased SREBP mRNA and inhibited DNL in hepatocytes.

To further investigate whether the beneficial effects caused by ZJ001 in vivo were related to the SREBP pathway, we utilized primary hepatocytes to dissect the underlying mechanism. Hepatic lipid synthetic processes are controlled reciprocally by insulin and glucagon. Insulin stimulates lipid synthesis in hepatocytes chiefly by increasing the mRNA and the processed nuclear form of SREBP-1c37. Hence, we used rat primary hepatocytes to determine whether ZJ001 suppressed the insulin-mediated SREBP pathway. Here, the hepatocytes were treated with ZJ001 at varied concentrations for 6 h, and 30 nmol/L of insulin was added for 30 min before completion of the experiment. ZJ001 down-regulated insulin-stimulated SREBP-2 and SREBP-1c mRNA levels in cultured cells, which is consistent with the effects in the liver of ZJ001-treated mice (Figure 4A, 4B). ZJ001 treatment significantly reduced the DNL of cholesterol and fatty acid in a concentration-dependent manner, which is consistent with the changes in gene expression (Figure 4C, 4D). Taken together, our data suggest that ZJ001 decreased SREBP mRNAs and their target gene expression, thereby down-regulating cholesterol and fatty acid biosynthesis and consequently decreasing cellular lipid levels.

Figure 4
figure 4

ZJ001 inhibited the biosynthesis of cholesterol and fatty acid by decreasing SREBPs target genes in primary hepatocytes. (A and B) Expression of genes involved in lipid synthesis in primary hepatocytes. Primary hepatocytes were starved for 16 h and incubated in DMEM supplemented with increasing concentrations of ZJ001 for 6 h, followed by the absence or presence of 30 nmol/L insulin for 30 min before completion of the experiment. The cells were lysed in Trizol reagents, and RNA samples were extracted according to the manufacturer's protocol. The expression of multiple genes was analyzed using reverse transcription followed by quantitative-PCR (Q-PCR). (C and D) ZJ001 inhibited de novo lipid synthesis in insulin-treated hepatocyte. Fatty acid (C) and cholesterol (D) were extracted, and lipid synthesis was measured. The results are shown as the mean±SEM (n=3), and all experiments were repeated three times. bP<0.05, cP<0.01 compared with the control group.

PowerPoint slide

Full size image

ZJ001 suppressed mTORC1 activity.

ZJ001 treatment reduced the protein levels of the cytosolic precursor form of pre-SREBP-1c and nuclear n-SREBP-1c, in accordance with the decrease in mRNA of SREBP-1c (Figure 5A). However, ZJ001 did not alter the phosphorylation levels of SREBP-1c (Figure S3). Porstmann et al38 showed that the enhancing effect of insulin on SREBP-1c mRNA is abolished by RNAi-mediated knockdown of mTOR, a main component of mTORC1. mTORC1 also includes Raptor, which acts as a scaffold to recruit downstream substrates, such as eukaryotic initiating factor 4E binding protein 1 (4E-BP1) and p70S6K, that contribute to mTORC1-dependent regulation of protein translation39. mTOR is phosphorylated at Ser2448 via the PI3K/AKT signal pathway and auto-phosphorylated at Ser248140,41. Given that ZJ001 decreased SREBP mRNA and in insulin-treatment hepatocytes, we used rat primary hepatocytes to further investigate whether ZJ001 modulated insulin-mediated PI3KAKT/mTORC1 pathway in hepatocytes. We observed a decrease in phosphorylated levels of mTOR after ZJ001 treatment at the cellular level (Figure 5B). Accordingly, ZJ001 treatment also blocked the activity of p70S6K, one of the major downstream targets of mTORC1, as evidenced by the decline of S6 phosphorylation levels (Figure 5B). Studies have shown that insulin activates mTORC1 through AKT-mediated phosphorylation of TSC2, thereby inhibiting TSC1/2 function42. TSC1/2 suppress Rheb (Ras Homolog Enriched in Brain), which phosphorylates and activates mTOR43,44. However, insulin induced phosphorylation of AKT and ZJ001 did not block TSC2, which suggests that alternative proteins upstream of mTORC1 are involved. The other branch of AKT, GSK3, was also not altered (Figure 5B).

Figure 5A
figure 5

ZJ001 suppressed mTORC1 activity. Primary hepatocytes from rat were prepared and plated on d 0. On d 1, cells were treated with ZJ001 for 6 h, the cells were then treated with 30 nmol/L insulin for 30 min, harvested, and lysed. Each sample was subjected to immunoblot analyses. (A) The protein level of n-SREBP-1c (N) and pre-SREBP-1c (P), the optical density were calculated using Image J software (n=3). (B) The phosphorylation and protein levels of mTOR, p70S6K, S6, AKT, TSC2, and GSK3α/β. The ratio of the phosphorylation level to the protein level was determined (n=3). bP<0.05, cP<0.01 compared with the control group.

PowerPoint slide

Full size image

AMPK responds to energy stress by suppressing cell growth and biosynthetic processes partially through its inhibition of the rapamycin-sensitive mTORC1 pathway20. Hence, we investigated the effect of ZJ001 on the AMPK pathway. As expected, the phosphorylation of AMPK and its direct downstream protein, ACC, were significantly increased in hepatocytes after ZJ001 treatment. However, ZJ001 did not alter TSC2. We observed increased Raptor phosphorylation, which is required for the inhibition of mTORC1. These results may indicate that the activation of AMPK and subsequent phosphorylation of Raptor by ZJ001 potentially inhibited mTORC1 (Figure 5B, 5C).

Figure 5C
figure 6

(C) The phosphorylation and protein levels of AMPK, ACC, and Raptor. The ratio of the phosphorylation level to the protein level was determined (n=3). bP<0.05, cP<0.01 compared with the control group.

PowerPoint slide

Full size image

ZJ001 suppressed lipid synthesis partially via the AMPK pathway

ZJ001 also increased the phosphorylation level of AMPK, ACC and Raptor in non-insulin-treated hepatocytes (Figure 6A). A commonly used competitive AMPK inhibitor, Compound C, was applied to verify the role of AMPK activation in the biological activity of ZJ001. Compound C partially blocked the ZJ001 effect on cholesterol and triglyceride synthesis, suggesting that ZJ001 inhibition of DNL is partially dependent on activation of the AMPK pathway. ZJ001 may exert its lipid-lowing effect through an alternative pathway (Figure 6B). These results suggest that ZJ001 regulates lipid synthesis partially through AMPK/mTORC1 pathways, which was also exemplified by the elevated ratio of phosphorylation levels to the protein levels of ACC and AMPK and decrease in the phosphorylation levels of mTOR in vivo (Figure 6C).

Figure 6A
figure 7

ZJ001 suppressed lipid synthesis partially via the AMPK pathway. (A) Similar ZJ001 treatment as Figure 5 without insulin stimulation. Each sample was subjected to immunoblot analyses of the phosphorylation level and total protein levels of AMPK, ACC and Raptor. (B) Hepatocytes were pre-incubated for 30 min in the absence or presence of Compound C (5 μmol/L) and treated with 40 μmol/L ZJ001 for 20 h, followed by lipid synthesis measurement. (C) mTOR, AMPK, and ACC phosphorylation in the liver of mice fed a high-fat diet. The ratio of the phosphorylation to total protein levels of AMPK, ACC, and mTOR were calculated based on the intensity detected using the Image J software. The results are shown as the mean±SEM. bP<0.05, cP<0.01 compared with the control group.

PowerPoint slide

Full size image

We investigated the direct activation of ZJ001 on recombinant AMPK heterotrimers, but no obvious activation was observed (data not shown). AMPK is a highly conserved sensor of cellular energy status that is activated when the AMP/ATP ratio is increased45. Because the AMP level was too low to be measured accurately by our HPLC, we investigated the ADP/ATP ratio in rat primary hepatocytes. The results showed that ZJ001 dose-dependently increased the ADP/ATP ratio (Figure 6D). Mitochondria are a subcellular organelle in which fatty acids and glucose are used for the production of ATP through the oxidative phosphorylation system46. ZJ001 significantly increased the O2 consumption rate in human HepG2 hepatic cells (Figure 6E). This block was rather mild in rat primary hepatocytes and isolated mitochondria from rat liver (Figure 6F, 6G), which suggests that ZJ001 may stimulate the cellular energy sensor AMPK via a mild uncoupling of the mitochondria electron respiration chain to decrease intracellular ATP production.

Figure 6D
figure 8

(D) Effect of ZJ001 on the ADP/ATP ratio after incubation for 1 h in rat primary hepatocytes (n=3). The effect of ZJ001 on O2 consumption in HepG2 (E), primary hepatocytes (F) and the isolated mitochondria from rat liver (G). The results are shown as the mean±SEM. bP<0.05, cP<0.01 compared with the control group.

PowerPoint slide

Full size image

Discussion

Currently, most drugs for the treatment of lipid-related metabolic disorders target rate-limiting enzymes involved in lipogenesis, such as FAS, which is a key enzyme in fatty acid synthesis; DGAT (acyl CoA: diacylgycerol acyltransferase), which catalyzes the final step, is the only key and rate-limiting enzyme in triglyceride synthesis47. However, the total or partial inhibition of rate-limiting enzymes may result in two adverse issues: the accumulation of substrate may cause lysosomal storage disease48 and the reduction of product will cause compensatory action. SREBPs directly activate the expression of more than 30 genes dedicated to the synthesis and uptake of cholesterol, fatty acids, triglycerides, and phospholipids11. The mechanisms of inactivating rate-limit enzymes that are involved in lipogenesis are different from those causing inhibition of SREBPs, which cause a decline in the overall efficiency of lipogenesis. Statins are the most widely prescribed drugs for the treatment of hypercholesterolemia, and these drugs inhibit HMG-CoA reductase, which catalyzes the rate-limiting step in cholesterol biosynthesis. Decreased cholesterol biosynthesis increases LDL-receptor levels, which results in a positive cycle for lowered cholesterol levels in serum. However, statins cannot decrease fatty acid biosynthesis because they block the cholesterol biosynthesis that stimulates SREBP activity36. The above deficiency of the statin drugs might be compensated for if candidates that specifically modulate the SREBP pathway are identified.

Here, we used a cell-based reporter assay to screen for compounds that regulate SRE-containing promoters, and we found a small molecule, ZJ001. As expected, ZJ001 treatment decreased triglyceride and cholesterol synthesis in vitro and in vivo, with down-regulated expression of SREBP and crucial genes involved in lipogenesis. Chronic ZJ001 treatment improved lipid metabolism in dietary obese mice. These findings of ZJ001 proved the concept that inhibition of the SREBP pathway may be a potential strategy to treat metabolic diseases.

Regulation of SREBP occurs at two levels: transcriptional regulation of SREBP mRNA and posttranscriptional regulation of precursor SREBP protein. Posttranscriptional regulation involves the sterol-mediated suppression of SREBP cleavage49,50. The transcriptional regulation of the SREBP mRNA is more complex. SREBP-1c and SREBP-2 are subject to distinct forms of transcriptional regulation, whereas SREBP-1a appears to be constitutively expressed at low levels in liver and most other tissues of adult animals51. One mechanism of regulation that is shared by SREBP-1c and SREBP-2 involves a feed-forward regulation mediated by SREs that are present in the enhancer/promoters of each gene. This feed-forward loop processes mature n-SREBP, which activates the transcription of their own genes52,53.

AMPK negatively regulates hepatic SREBP-1c activity partially through its inhibition of mTORC1. There is another way that AMPK modulates SREBP-1c activity. Yu Li et al showed that AMPK directly phosphorylates and inhibits SREBP activity to attenuate hepatic steatosis and atherosclerosis in diet-induced insulin-resistant mice54. ZJ001 increased AMPK phosphorylation and its direct downstream proteins, ACC and Raptor, in hepatocytes after ZJ001 treatment. However, ZJ001 did not alter the phosphorylation levels of SREBP-1c (Figure S3), indicating that the inhibition of lipid synthesis by ZJ001 probably occurs through mTORC1. ACC is a biotin-containing enzyme that catalyzes the carboxylation of acetyl-CoA to malonyl-CoA, which is the rate-limiting step in fatty acid synthesis55. SREBP-1c activates the transcription level of ACC gene expression10. The total ACC protein level was significantly decreased in liver after chronic ZJ001 treatment (Figure 6C). Taken together, ZJ001 reduced fatty acid synthesis at least partially through regulation of the AMPK/mTORC1/SREBP-1c axis.

As one of the major downstream target of mTORC1, p70S6K, which phosphorylates the ribosomal protein S6, was also blocked by ZJ001. A previous study showed that p70S6K is also required for insulin stimulation of SREBP processing in rat hepatocytes17. However, it has been difficult to dissect the effect of ZJ001 on SREBP processing from mRNA transcription because the reduction of SREBP mRNA and the decrease of precursor protein levels were concomitant. The individual effect of ZJ001 on the transcriptional regulation of SREBP mRNA and SREBP processing remains to be elucidated.

In conclusion, we identified a small-molecule, ZJ001, that inhibited triglyceride and cholesterol content in vivo and in vitro. ZJ001 showed multiple beneficial effects in vivo, suggesting that modulation of the SREBP-1c pathway could be a potential therapeutic target for the treatment of lipid metabolic disorders.

Author contribution

Jia LI, Fan YANG, Jie TANG, and Jing-ya LI designed the research; Jin ZHANG, Dong-mei CHEN, Ling-ling YANG, Chun-mei XIA, Hao-wen JIANG, Chun-lan TANG, and Zhi-fu XIE performed the research; Feng ZHANG synthesized the compound; and Jin ZHANG, Li-na ZHANG, and Yan-yun FU analyzed the data and wrote the paper.