Geniposide alleviates non‐alcohol fatty liver disease via regulating Nrf2/AMPK/mTOR signalling pathways

Abstract Non‐alcohol fatty liver disease (NAFLD) is a common disease which causes serious liver damage. Geniposide (GEN), a kind of iridoid glycoside extracted from Gardenia jasminoides fruit, has many biological effects, such as resistance to cell damage and anti‐neurodegenerative disorder. Lipid accumulation was obvious in tyloxapol‐induced liver and oil acid (OA) with palmitic acid (PA)‐induced HepG2 cells compared with the control groups while GEN improved the increasing conditions. GEN significantly lessened the total cholesterol (TC), the triglyceride (TG), low‐density lipoprotein (LDL), very low‐density lipoprotein (VLDL), myeloperoxidase (MPO), reactive oxygen species (ROS) and increased high‐density lipoprotein (HDL), superoxide dismutase (SOD) to response the oxidative stress via activating nuclear factor erythroid‐2–related factor 2 (Nrf2), haeme oxygenase (HO)‐1 and peroxisome proliferator‐activated receptor (PPAR)α which may influence the phosphorylation of adenosine 5’‐monophosphate–activated protein kinase (AMPK) signalling pathway in mice and cells. Additionally, GEN evidently decreased the contents of sterol regulatory element‐binding proteins (SREBP)‐1c, phosphorylation (P)‐mechanistic target of rapamycin complex (mTORC), P‐S6K, P‐S6 and high mobility group protein (HMGB) 1 via inhibiting the expression of phosphoinositide 3‐kinase (PI3K), and these were totally abrogated in Nrf2−/− mice. Our study firstly proved the protective effect of GEN on lipid accumulation via enhancing the ability of antioxidative stress and anti‐inflammation which were mostly depend on up‐regulating the protein expression of Nrf2/HO‐1 and AMPK signalling pathways, thereby suppressed the phosphorylation of mTORC and its related protein.


| INTRODUC TI ON
Non-alcoholic fatty liver (NAFLD) is a disease that develops from hepatocyte steatosis or non-alcoholic steatohepatitis (NASH) to liver fibrosis even liver cancer. 1 During the process of NAFLD, it often occurs excessive liver and serum lipid accumulation with syndrome such as insulin resistance (IR), atherosclerosis and systemic inflammatory. 2 In liver adipocytes, IR increases the rate of lipolysis leading to the ectopic of non-esterified fatty acids (NEFA). Meanwhile, compensatory hyperinsulinaemia, which maintains normal glucose homeostasis, up-regulates the transcription factor sterol regulatory element-binding protein-1c (SREBP-1c) to drive the de novo lipogenesis. 3 NAFLD occurs through NEFA catabolism to produce intrahepatic triglycerides, which are caused by mitochondrial oxidation and are exported as very low-density lipoprotein (VLDL). 4 The increasing NEFA is related to lipid toxicity, activation of inflammatory pathways and cytokine release. 5 In the past few years, some new strategies have been proposed, such as regulating gene expression 6 and key proteins associated with oxidative stress and inflammatory responses. 7 One of the most important proteins is nuclear factor erythroid-2-related factor 2 (Nrf2) from red blood cells, 8 a nuclear factor, which was constitutively expressed in cytoplasm and inhibited by its negative regulator kelch-like ECH (enoyl-coa hydratase) related protein 1 (Keap1). 9 Once the cells accept stimulation which could cause oxidative stress, the Keap1-Nrf2 complex was disassembled and Nrf2 transferred into the nucleus to control the production of related enzymes such as haeme oxygenase 1 (HO-1) and superoxide dismutase (SOD). 10 In the process of Nrf2 working, phosphoinositide 3-kinase (PI3K) controls the translocation 11 and glycogen synthase kinase 3β (GSK3β) degrading the activity of Nrf2. 12,13 During the process of NAFLD, it often occurs a decreasing Nrf2 and applying natural antioxidants could prevent the oxidative stress by activating Nrf2 and related genes. 14 PI3K-AKT-mTOR signalling pathway is a canonical signalling pathway in cell, which participates in regulating many critical functions, such as autophagy, cell proliferation and oxidative stress. 15 Mechanistic target of rapamycin complex 1 (mTORC1) is an important growth regulator which is activated by nutrient cues, growth factors and cellular energy. 16 It could be activated through class I PI3K-mediated synthesis of phosphatidylinositol 3,4,5-trisphosphate [PI (3,4,5) P3]. PI (3,4,5) P3 stimulates AKT, which contributes to activating of mTORC1. MTORC1 stimulates the synthesis of fatty acids and sterols by regulating the expression of SREBP-1c, which is the master lipogenic transcription factor. 17-19 mTORC1 could phosphorylate S6K and S6 and participate in regulation of autophagy. mTORC2 could be activated by growth factors and phosphorylate AKT (p-ser473) in AGC kinase family to control various cellular processes. 20 Geniposide (GEN) is a natural product extracted from the dry and ripe fruit of gardenia, which has many biological effects, such as anti-inflammation and liver protection. 21,22 It has been reported that after AMPK knockout, GEN lost its protective effect against the obesity related cardiac injury. 23 However, the specific mechanism of GEN on antioxidative stress and lipid accumulation induced by NAFLD in vivo or in vitro is not clear at present. The purpose of the research was to investigate the mechanism of GEN playing a role in NAFLD-induced oxidative stress and inflammation, therefore to provide a theoretical basis for the development of new potential therapy on NAFLD.

| Cell oil red O staining
The cells were inoculated into 24-well plate with sterile cell slides and treated with 260 μmol/L GEN with or without OA (660 μmol/L) and PA (330 μmol/L) for 18 hours. The slides were incubated with the ORO fixative for 25 minutes. Then, slides were washed by distilled water twice and soaked with 60% isopropanol for 5 minutes.
ORO staining solution was added into the plates for 15 minutes. The slices were washed with distilled water till there was no ORO dye.
Haematoxylin stain was added onto the slides for 1 minute next to washing the slides with distilled water according to the manufacturer's introduction (Solarbio). The slices were observed under a microscopy (Olympus).

| Immunofluorescence
The

| Animals and ethics statement
The male wild-type (WT) and Nrf2 −/− C57BL/6 mice (6-8 weeks), weighing 18-22 g, were housed in a specific pathogen free environment fed a normal diet (24 ± 1°C, relative humidity: 40%-80%, n = 3 per cage). All experiments were carried out in an accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals and approved by the Jilin University animal administration committee.

| Western blotting
About 20 mg liver tissue of each mice was collected and made into homogenate in a 1.5 mL tube by Electric Tissue Grander (Tiangen OSE-Y30). 300 μL RIPA lysis buffer was added into the tubes for 20 minutes. The following steps were performed as described before.

| Immunohistochemistry
The liver tissue was fixed in 4% paraformaldehyde and placed in the embedding box. The blocks were located in 70% alcohol for the night, 80% alcohol for 2 hours, 90% alcohol for 1 hours, 95% alcohol for 40 minutes, anhydrous ethanol I for 40 minutes and anhydrous ethanol II for 30 minutes, respectively. The organization after dehydration was put into xylene II for 10 minutes and xylene II for 10 minutes. Tissue was paraffined for 2 hours, sliced and baked. The slices were putted into xylene for 16 minutes, anhydrous ethanol I for 20 minutes, 95% alcohol for 10 minutes, 90% alcohol for 10 minutes, 80% alcohol for 10 minutes and ddH 2 O for 10 minutes. The slices were dipped in the citrate buffer (PH

| Oil red O staining
The fresh liver tissues were embedded in optimal cutting temperature compound (OCT, SAKURA) and stored at −20°C. The tissue blocks were sliced into about 5 μm by a freezing microtome (Leica).

| Haematoxylin and eosin staining
The liver tissues were obtained, located into 4% paraformaldehyde and then embedded in paraffin. The wax blocks were cut into a thickness of 5 μm and stained with haematoxylin and eosin (H & E).
The slices were observed under a microscopy (Olympus).

| Statistical analysis
All experiments were all repeated at least 3 times. All data were expressed by Mean ± SEM and analysed by SPSS 19.0 (IBM).
Comparisons between two groups were conducted by ONE-WAY ANOVA, and comparisons between multiple groups were made by LSD method. P < .05 means significantly between two groups.

| Geniposide had no effect on cell viability and reduced lipid accumulation in HepG2 cells
From Figure 1A, it could observe that when the concentrations of GEN were 0, 65, 130, 260, 390 and 520 μmol/L, there was no significance between the GEN group with or without OA + PA group and the control group which indicated that GEN (65-520 μmol/L) had no toxicity for HepG2 cells. In Figure 1B However, the lipid droplets in cells were significantly reduced after GEN application.

| Geniposide regulated the AMPK/mTORC signalling pathways in HepG2 cells
As a key physiological energy sensor, AMPK is a major regulator of the energy balance of cells and organisms, co-ordinating multiple metabolic pathways, balancing energy supply and demand, and ultimately regulating the growth of cells and organs. Regulation of energy metabolism balance is mediated by a number of related signalling pathways, among which AMPK/mTOR signalling pathway jointly constitutes a switch for anabolic and catabolic processes in cells. In addition, AMPK/mTOR signalling pathway is an important regulatory pathway of autophagy. In Figure 2D

| Geniposide alleviated the serum dyslipidaemia of mice
In Figure 4A,B,C, after tyloxapol induced, the contents of TC, TG, LDL and VLDL were increased significantly compared with the control group, indicated the occurrence of hyperlipidemia which may lead to NAFLD. However, GEN decreased the levels of TC, TG, LDL and VLDL and increased HDL in serum compared with the tyloxapolinduced mice which suggested that GEN had the ability of reducing the serum lipid. Meanwhile, GEN also decreased the contents of MMP-9, ApoC3, VCAM-1, ICAM-1 and MCP-1 which participated in the adjusting the blood viscosity and the enrichment of inflammatory cells. While after Nrf2 knockout, the significant relief effect of GEN in TC, TG and LDL disappeared ( Figure S1).

| Geniposide up-regulated the oxidative stress ability of mice and down-regulated the inflammatory response induced by tyloxapol
As illustrated in Figure 5A-C, compared with the tyloxapol group, GEN pre-treatment elevated the content of SOD and reduced MPO and ROS, which suggested that GEN prevented the oxidative stress damage induced by tyloxapol via regulating the production of essential enzymes. GEN also decreased the over-production of proinflammatory cytokines TNF-α, IL-1β and IL-6, proved GEN could reduce the inflammation during the progress of NAFLD.

| Geniposide reduced the lipid accumulation and inflammation damages in tyloxapol-induced mice
As in Figure 8A, after tyloxapol induced, there was an amount of lipid accumulated in the hepatocyte in wild-type and Nrf2 −/− mice, while GEN treatment (75, 100 mg/kg) ameliorated the elevation gradually.
Haematoxylin and eosin staining was used to investigate whether inflammation histologic changes occur during NAFLD. In Figure 8B, tyloxapol caused the inflammatory cells infiltration and liver tissue loose compared with the control group, whereas pre-treatment with GEN obviously attenuated these changes. In Nrf2 −/− mice, the pathologic changes were more obvious than in WT mice and the improvement ability of GEN was also been impaired after Nrf2 knockout.
What's more, in Nrf2 −/− mice, after tyloxapol induced, the levels of TC, TG and LDL were increased significantly compared with the tyloxapol group of WT mice while after application of GEN, the levels were not significant reduction. It indicated that the regulating serum lipid effect of GEN may play via Nrf2 ( Figure S1).

| D ISCUSS I ON
NAFLD is the most common liver disease which acquired by metabolic stress liver injury, infected almost one-third population in some areas. 24  GEN was added into HepG2 cells prior 1 h to stimulation of OA and PA for 18 h. A-C, Protein expression of Nrf2, PPARα, PPARγ in nucleus and HO-1 in cytoplasm was detected by Western blot. (D and E) Protein expression of P-ACC, ACC, P-AKT, AKT, P-AMPKα, AMPKα, P-AMPKβ and AMPKβ was detected by Western blot. F-H, Protein expression of PI3K, P-mTORC, mTORC, P-S6K, S6K, P-S6, S6, SREBP-1c and HMGB1 was detected by Western blot. The similar results were collected from three dependent experiments. All data were expressed by mean ± SEM (n = 5 in each group). ## P < .01 vs Control Group; *P < .05 and **P < .01 vs OA and PA Group F I G U R E 3 Effect of GEN on the Nrf2 transferring into the nucleus. GEN was added into HepG2 cells for 18 h and the transferring of Nrf2 (red) was detected by immunofluorescence. The nucleus was presented as blue, and the scale bars were 50 μm Therefore, the GEN treatment also increased the activity of GSK3β.
According to the previous results, we assumed that GEN inhibited NAFLD via Nrf2 and related protein. In the experiment, we applied Nrf2 −/− mice to further confirm the role of Nrf2 in GEN regulating NAFLD. We found that after Nrf2 knockout, GEN marginally reduced the contents of TC, TG and LDL than in WT mice ( Figure S1).

F I G U R E 7
Effect of GEN on the expression of Nrf2 and PPARα in liver of WT and Nrf2 −/− mice. GEN (50, 75, 100 mg/kg) was administered to WT or Nrf2 −/− mice prior 1 h to stimulation of tyloxapol (500 mg/kg) for 18 h, and the expression of Nrf2 and PPARα in liver tissue was detected by immunohistochemistry (magnification ×100 and ×200) Furthermore, in Figure 6, GEN also could increase the phosphorylation of ACC, AKT and AMPK β and inhibit the level of P-S6K and SREBP1-c.
However, it failed to control the content of P-AMPKα, PI3K, P-mTORC and P-GSK3β. Immunohistochemistry results also confirmed the previous assumption. In WT mice, it was easily to observe an increased expression of Nrf2 and PPARα. Whereas the content of PPARα was down-regulated after Nrf2 knockout which illustrated GEN restrained oxidative stress mostly through Nrf2 (Figure 7).
According to the 'two-hit' hypothesis, during NAFLD, there is an increasing level of free fatty acid (FFA) in the adipocytes and lipid accumulation in hepatocyte resulting an inflammation damage in liver. During the NAFLD, the contents of pro-inflammatory cytokines, such as TNF-α, IL-1β and IL-6, were significantly higher than that in healthy mice ( Figure 5D) confirmed the 'second hit' hypothesis. 30 In addition, HMGB1 is a highly conserved protein which could induce inflammation once secreted outside the cell. 31 In our study, tyloxapol raised the content of HMGB1 while GEN pre-treatment improved the increasing HMGB1 which suggested that GEN may alleviate tyloxapol-induced inflammation. Tyloxapol induced the liver inflammatory reaction which created inflammatory vacuoles in the liver and large number of inflammatory cells infiltrated in the interstitium. After GEN treatment, the liver structure recovered to the normal structure and reduced the inflammatory cells gathering.
Furthermore, in liver of Nrf2 −/− mice, the structure was more porous than it in WT mice which suggested that Nrf2 knockout reduced the ability to resist oxidative stress of liver ( Figure 8B).

| CON CLUS IONS
As shown in Graphical Abstract, our study firstly demonstrated that geniposide was capable to protect mice and cells from NAFLDinduced oxidative stress and inflammation which mostly depended on up-regulating the Nrf2 and adjusting the protein expression of AMPK/PI3K/mTOR signalling pathways. This finding revealed the essential role of Nrf2 in geniposide inhibiting lipid accumulation and oxidative stress caused by NAFLD which would provide a potential therapeutic targeting at NAFLD and other liver metabolism diseases.

ACK N OWLED G EM ENTS
This work was supported by the National Natural Science Foundation of China (Grant no. 31772798, 31970507, 31672621).

CO N FLI C T O F I NTE R E S T
The authors report no conflicts of interest. The authors alone are responsible for the content of this manuscript. F I G U R E 8 Effect of GEN on liver lipid accumulation and pathological changes of WT and Nrf2 −/− mice. GEN (50, 75, 100 mg/kg) was administered to WT or Nrf2 −/− mice prior 1 h to stimulation of tyloxapol (500 mg/kg) for 18 h. A, The lipid accumulation was detected by oil red staining (magnification 100× and 200×, red: lipid; blue: nucleus). B, Haematoxylin and eosin staining method was used to observe pathological changes. (magnification ×200)

AUTH O R S' CO NTR I B UTI O N S
BYS and GWL wrote the paper and performed the experiments; HHF and JQC performed the experiments; ZL and LLZ analysed the data; and MYJ, QW and HYQ contributed to design the experiments.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support the findings of this study are available from the corresponding author upon request.