Nrf2 activation is required for curcumin to induce lipocyte phenotype in hepatic stellate cells

Biomedicine & Pharmacotherapy 95 (2017) 1–10

Available online at

ScienceDirect www.sciencedirect.com

Original article

Nrf2 activation is required for curcumin to induce lipocyte phenotype in hepatic stellate cells Chunfeng Lua,b,c,1, Wenxuan Xua,b,c,1 , Shizhong Zhenga,b,c,* a

Department of Pharmacology, School of Pharmacy, Nanjing University of Chinese Medicine, Nanjing, Jiangsu, China Jiangsu Key Laboratory for Pharmacology and Safety Evaluation of Chinese Materia Medica, Nanjing University of Chinese Medicine, Nanjing, Jiangsu, China c Jiangsu Key Laboratory of Therapeutic Material of Chinese Medicine, Nanjing University of Chinese Medicine, Nanjing, Jiangsu, China b

A R T I C L E I N F O

Article history: Received 4 May 2017 Received in revised form 20 July 2017 Accepted 7 August 2017

Keywords: Hepatic fibrosis Hepatic stellate cell Lipocyte phenotype Curcumin Nuclear factor (erythroid - derived 2) - like 2

A B S T R A C T

Hepatic fibrosis is a reversible scarring response that commonly occurs with chronic liver injury. During hepatic fibrogenesis, the major effector hepatic stellate cells (HSCs) become activated, featured by disappeared intracellular lipid droplets, decreased retinoid storage, and dysregulated expression of genes associated with lipid and retinoid metabolism. Compelling evidence suggested that recovery of retinoid droplets could inhibit HSC activation, while the precise molecular basis underlying the phenotypical switch still remained unclear. In this study, curcumin increased the abundance of lipid droplets and content of triglyceride in activated HSCs. In addition, curcumin could concentration-dependently regulate genes associated with lipid and retinoid metabolism. Further, consistent results were obtained from in vivo experiments. Curcumin increased Nrf2 expression and nuclear translocation, and its binding activity to DNA, which might be associated with suppression of Kelch-like ECH-associated protein 1 in HSCs. Of interest was that Nrf2 overexpression plasmids, in contract to Nrf2 siRNA, strengthened the effect of curcumin on induction of lipocyte phenotype. In in vivo system, Nrf2 knockdown mediated by Nrf2 shRNA lentivirus not only accelerated the lipid degradation in HSCs but also promoted the progression of CCl4-induced hepatic fibrosis in mice. Noteworthily, Nrf2 knockdown abolished the protective effect of curcumin. In conclusion, curcumin could induce lipocyte phenotype of activated HSCs via activating Nrf2. Nrf2 could be a target molecule for antifibrotic strategy. © 2017 Elsevier Masson SAS. All rights reserved.

Abbreviations: ALT, alanine aminotransferase; ANOVA, one way analysis of variance; a-SMA, alpha-smooth muscle actin; AST, aspartate aminotransferase; C/ EBPa, CCAAT/enhancer-binding protein alpha; CCl4, carbon tetrachloride; DMEM, Dulbecco’s modified Eagle’s medium; DMSO, dimethylsulfoxide; ECM, extracellular matrix; FBS, fetal bovine serum; GAPDH, glyceraldehyde phosphate dehydrogenase; HA, hyaluronic acid; H&E, haematoxylin-eosin; HRP, horseradish peroxidase; HSCs, hepatic stellate cells; i.p., intraperitoneally; IV-C, type IV collagen; LN, laminin; NC, negative control; Nrf2, nuclear factor (erythroid-derived 2) - like 2; NS, normal saline; PBS, phosphate buffered saline; PCR, polymerase chain reaction; PCIII, procollagen type III; PPARa, peroxisome proliferator-activated receptor alpha; PPARg, peroxisome proliferator-activated receptor gamma; RARb, retinoic acid receptor beta; RXRa, retinoid X receptor alpha; SD, standard deviation; shRNA, short hairpin RNA; siRNA, small interfering RNA; TG, triglyceride; TNF-a, tumor necrosis factor-alpha. * Corresponding author at: Department of Pharmacology, School of Pharmacy, Nanjing University of Chinese Medicine, 138 Xianlin Avenue, Nanjing, Jiangsu 210023, China. E-mail address: [email protected] (S. Zheng). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.biopha.2017.08.037 0753-3322/© 2017 Elsevier Masson SAS. All rights reserved.

1. Introduction Hepatic fibrosis is a reversible wound-healing response featured by excessive deposition of extracellular matrix (ECM) to liver injury [1]. Hepatic stellate cells (HSCs), also known as lipocytes or fat-storing cells earlier, are the primary source of ECM in normal and fibrotic liver [2]. Under physiological conditions, quiescent HSCs function as retinol-storing cells and control the absorption, storage, and metabolism of vitamin A (retinoid). Anatomically, HSCs simultaneously secret moderate ECM and matrix metalloproteinase to dynamically maintain the threedimensional structure of normal liver. When stimulated by liver injury irrespective of etiologies, HSCs undergo activation, changing from a quiescent fat-rich phenotype characterized by storage of multiple lipid droplets in cytoplasm into a fibrotic myofibroblastic phenotype [3]. Thus targeted regulation of HSC phenotypic

2

C. Lu et al. / Biomedicine & Pharmacotherapy 95 (2017) 1–10

transformation is a key approach to reverse hepatic fibrogenesis. Currently, apoptosis induction is identified as a mainstream therapeutic strategy against hepatic fibrosis [4], which, however, may destroy liver structure. Also, HSCs can gain the ability to escape from apoptosis during the regression of hepatic fibrosis [5]. Notably, increasing studies provided another perspective of the cell fate of HSCs during fibrogenesis that lipid droplets in cells determined the functions of HSCs. It was reported that HSCs containing abundant lipid droplets could not produce collagen fibers in spite of the direct contact between HSCs and the injured parenchymal cells [6]. This finding highlighted a potential correlation between restoration of lipocyte phenotype in HSCs and regression of hepatic fibrosis. Accumulative evidence also consolidated the concept that regaining intracellular lipid droplets could significantly inhibit HSC activation [7,8]. Thus, reversion of activated myofibroblasts into inactivated lipocytes is anticipated to be a classical and straightforward antifibrotic strategy [9]. However, precise elucidation of the potential regulatory mechanisms underlying the phenotypical switch of HSCs remains difficult. A complexity in the understanding of lipocyte phenotype in HSCs is underscored by transcriptional regulation. There are no simple rules that govern mechanisms of transcriptional manipulation in HSCs though, continued advances have yielded novel insights into their roles in lipid homeostasis in HSCs. Although peroxisome proliferator-activated receptor alpha (PPARa) is lowly expressed in HSCs, it has been defined as a dominant contributor for initiating mitochondrial and peroxisomal fatty acid b-oxidation enzymes [10,11]. Peroxisome proliferator-activated receptor gamma (PPARg), CCAAT/enhancer-binding protein alpha (C/EBPa), retinoic acid receptor beta (RARb), and retinoid X receptor alpha (RXRa) are highly expressed in quiescent HSCs but progressively decreased in culture-activated stellate cells [12,13]. Overexpression of PPARg or agonism of PPARg, RARb, and RXRa by specific agonists could effectively inhibit HSC activation accompanied by induction of lipocyte phenotype [8,12,14–16]. All these findings suggested a critical role of the regulators in the regulation of retinoid homeostasis and HSC states. However, the mechanisms by which they are affected and regulated are unclear. Nuclear factor (erythroid
derived 2)
like 2 (Nrf2), a critical factor involved in HSC activation and hepatic fibrosis, belongs to the leucine zipper transcription factor family [17]. Nrf2 expression is gradually decreased in the wake of HSC activation [18]. Systematic Nrf2 deficiency severely delayed the spontaneous cure from acute and chronic toxin-mediated liver damage and fibrosis [19]. Genetic or pharmacological modulation by chemicals implied that Nrf2 enhancement could facilitate the formation of lipid droplets and inactivation of HSCs in our previously published study [20]. Curcumin, an active component of turmeric derived from the rhizome of the tropical plant Curcuma longa, has been used in Chinese medicine. Abundant studies, including ours, suggested that curcumin had potent inhibitory effect on hepatic fibrosis [21– 23]. However, few studies focused on the lipocyte phenotype of HSCs. Therefore, we aimed to investigate the role of curcumin in lipocyte phenotype of HSCs and further uncover the potential mechanisms.

saline (NS). Carbon tetrachloride (CCl4, C822981) and olive oil (O815211) were purchased from Shanghai Macklin Biochemical Co., Ltd., (Shanghai, China). The pCMV-XLS-Nrf2 and pcDNA were kind gifts from Dr. Peng Cao (Nanjing University of Chinese Medicine, Jiangsu, China). Nrf2 small interfering RNA (siRNA, sc156128) and control siRNA (sc-37007) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Lentivirus vectors encoding negative control (NC) shRNA and Nrf2 shRNA (forward: gatccAAGCCTTACTCTCCCAGTGAATCGAAATTCACTGGGAGAGTAA GGCTTtttt, reverse: aattAAAAAAAAGCCTTACTCTCCCAGTGAA TTTCGATTCACTGGGAGAGTAAGGCTTg) were constructed by Nanjing Dirui Biological Technology Co., Ltd. (Nanjing, Jiangsu, China). The primary antibody against Nrf2 (KG22587-1) was purchased from Nanjing KeyGEN Biotech (Nanjing, Jiangsu, China). The primary antibody against alpha-smooth muscle actin (a-SMA, GB13044) was purchased from Wuhan Servicebio Technology Co., Ltd. (Wuhan, Hebei, China). The primary antibodies against PPARa (ab8934), C/EBPa (ab15047), PPARg (ab209350), RARb (ab53161), Kelch-like ECH-associated protein 1 (Keap1, ab150654), and b-actin (ab8226) were purchased from Abcam (Cambridge, MA, USA). The primary antibody against RXRa (21218-1-AP) was purchased from Proteintech Group, Inc. (Rosemont, IL, USA). The primary antibody against lamin B (sc-6217) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). The horseradish peroxidase (HRP)-conjugated anti-mouse and anti-rabbit IgG antibodies (7076 and 7074) were purchased from Cell Signaling Technology (Danvers, MA, USA). The fluorescein-conjugated anti-mouse and anti-rabbit IgG antibodies (SA00003-1 and SA00003-2) were purchased from Proteintech Group, Inc. 2.2. Animals Male ICR mice (20–25 g) were purchased from Nantong University (Nantong, Jiangsu, China). All mice were kept in the specific pathogen free clean room under a controlled condition of 21–25  C and a 12-h light/dark cycle, and had free access to standard chow diet (Xietong Organism Co., Ltd., Nanjing, Jiangsu, China) and water. All experimental procedures were approved by the institutional and local committee on the care and use of animals of Nanjing University of Chinese Medicine (Nanjing, Jiangsu, China), and all animals received human care in accordance with the National Institutes of Health guidelines. 2.3. Experimental procedures 2.3.1. Study I In this four-week experiment, sixty mice were randomly divided into five groups (12 mice/group). Group 1 was the vehicle control in which mice were intraperitoneally (i.p.) injected with olive oil and orally administrated with NS once daily. Group 2 was the model group in which mice were i.p. injected with CCl4 oily solution and administrated with NS by gavage once every day. Groups 3, 4, and 5 were trial groups in which mice were i.p. injected with CCl4 and orally administrated with curcumin at 100, 200, and 400 mg/kg body weight once a day. A mixture of CCl4 (1 mL/kg body weight) and olive oil [1:1 (v/v)] was injected every other day for four weeks to establish fibrotic model in mice.

2. Materials and methods 2.1. Reagents and antibodies Curcumin (C1386) was purchased from Sigma-Aldrich and dissolved in dimethylsulfoxide (DMSO; Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) for in vitro experiments. For in vivo experiments, curcumin was suspended in sterile normal

2.3.2. Study II In this four-week experimental protocol, sixty mice were randomly divided into five groups (12 mice/group). Mice in groups 1–5 were orderly administrated as follows: group 1, NC shRNA lentivirus, olive oil, and NS; group 2, NC shRNA lentivirus, CCl4, and NS; group 3, Nrf2 shRNA lentivirus, CCl4, and NS; group 4, NC shRNA lentivirus, CCl4, and curcumin; group 5, Nrf2 shRNA

C. Lu et al. / Biomedicine & Pharmacotherapy 95 (2017) 1–10

lentivirus, CCl4, and curcumin. Nrf2 shRNA lentivirus was biweekly injected into murine caudal vein at a titer of 1  108 TU/mice. CCl4 was i.p. injected every other day and curcumin (200 mg/kg) was orally given once daily. At the end of experiments, all mice were anesthetized by intraperitoneal injection with pentobarbital sodium (50 mg/kg; Shanghai Haling Biological Technology Co., Ltd., Shanghai, China). Blood was collected from orbital vein and livers were harvested. Half of each liver was fixed in 4% paraformaldehyde solution (I0042; Nanjing Jiancheng Bioengineering Institute, Nanjing, China) and embedded in paraffin for histopathological studies. The remaining livers were rapidly stored in liquid nitrogen for protein extraction and hydroxyproline analysis. 2.4. Cell culture Human LX-2 cells are immortalized HSCs used in this study and were purchased from the Cell Bank of Chinese Academy of Sciences (Shanghai, China). Primary HSCs were isolated from ICR mice as we previously described [24]. Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; 12800082, Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS; 10099141, Gibco), 100 U/mL of penicillin (BS215B, Biosharp, Hefei, Anhui, China), and 100 mg/mL of streptomycin (BS077B, Biosharp). Cells were incubated at 37  C in a humidified atmosphere of 95% air and 5% CO2.

3

Biosharp), phenylmethylsulfonyl fluoride (BL507A, Biosharp), and phosphatase inhibitor (KGP602, Nanjing KeyGEN Biotech). Protein concentration was detected using a bicinchoninic acid assay kit (23227, Pierce Biotechnology, Rockford, IL, USA) in accordance with the manufacturer’s recommendation. Proteins (50 mg per well) were separated by sodium dodecyl sulfate-polyacrylamide gel, transferred to a polyvinylidene fluoride membrane (IPVH00010, Millipore, Burlington, MA, USA). The membrane was blocked with 5% skim milk (BS075B, Biosharp) in Tris-buffered saline (BS002B, Biosharp) containing 0.1% Tween 20 (BS361A, Biosharp) for 2 h. The membrane was incubated with primary antibodies and then HRPconjugated secondary antibodies to detect the abundance of target proteins. In this study, b-actin and lamin B were used to ensure equivalent loading. The abundance of target proteins was densitometrically determined using Quantity One 4.4.1 (Bio-Rad Laboratories, Berkeley, CA, USA). Variations in the densities of bands were expressed as fold changes compared with the control in the blot after being normalized to b-actin or lamin B. 2.8. Quantitative real-time PCR analysis RNA isolation and quantitative real-time PCR were performed according to our previous description [26]. Glyceraldehyde phosphate dehydrogenase (GAPDH) was used as an invariant control, and the mRNA expression of target genes was expressed as fold changes after normalization to GAPDH. The primers were listed in Table 1.

2.5. Plasmids and transient transfection 2.9. Nile red staining Cell transfection was performed according to the protocol as we previously described [25]. In brief, overexpression plasmid or siRNA was well mixed with 150-mL medium without FBS and antibiotics. 5 mL of Lipofectamine 2000 Reagent (Life Technologies, New York, NY, USA) was added to another 150 mL of DMEM. The two solutions above were respectively incubated at room temperature for 5 min and then gently mixed to prepare 300 mL of transfection complex, which was incubated for 10 min. Then, 700 mL of medium was introduced to obtain a 1000-mL transfection mixture. Finally, each well of 6-well plates was filled with 1000-mL transfection mixture for 24 h in the incubator. The empty vectors and control siRNA were used as negative controls. 2.6. Histopathological studies Haematoxylin-eosin (H&E) and Masson’s trichrome staining were performed using standard methods. Sirius Red staining was performed to determine collagen abundance. In brief, liver sections of 5 mm were deparaffinized and stained with picro-sirius red for 1 h at room temperature. After washes, sections on the slides were dehydrated in 100% ethanol and in xylene. Then they were mounted in PermountTM mounting medium (00-4960-56, eBioscience, Inc., San Diego, CA, USA). Immunofluorescence double staining was performed to visualize the abundance of target genes particularly in HSCs. In brief, after deparaffinization, liver sections of 5 mm were blocked with 1% bovine serum albumin and then incubated with primary antibodies overnight at 4  C. After three washes with phosphate buffered saline (PBS), liver sections were then incubated with fluorescein-conjugated secondary antibodies for 1 h at room temperature. DAPI (KGA215-50, Nanjing KeyGEN Biotech) was used to stained cell nuclei. A fluorescence microscope (Nikon, Tokyo, Japan) was used to observe immunofluorescent staining. 2.7. Western blot analysis Total proteins were extracted from liver tissues and HSCs using the mixture of radioimmuniprecipitation assay buffer (BL504A,

Nile Red (S19279, Shanghai Yuanye Biotechnology Co., Ltd., Shanghai, China) staining was used to visualize lipid droplets. Cells were washed twice with PBS and fixed with 4% paraformaldehyde for 20 min at room temperature. After immobilization, cells were stained with Nile Red (3.3 mg/mL) for 15 min at room temperature. DAPI was introduced to stain cell nuclei. Images were taken using a fluorescence microscope (Nikon, Tokyo, Japan). 2.10. Electrophoretic mobility shift assay (EMSA) Nuclear extracts from primary murine HSCs were prepared using the NE-PERTM nuclear and cytoplasmic extraction kit according to the protocol from the manufacturer (78833, Thermo Fisher Scientific, Waltham, MA, USA). Biotin-labeled Nrf2 probe was purchased from Beyotime Biotech Inc. (Nantong, Jiangsu, China). The extracted nuclear proteins (10 mg) were incubated in a binding reaction mixture containing 1.5-mL 10  binding buffer, 1.5-mL poly (dI-dC) (1 mg/mL), and ddH2O to a final volume of 14.4 mL for 20 min at room temperature. The probe of 0.6 mL (300 fmol) was added and incubated for 20 min at room temperature. Where indicated, 2 mL of specific, cold-competitor oligonucleotides in 100  competing buffer was added before the labeled probe, and the reaction was incubated for 20 min. Protein-DNA complexes were subjected to electrophoresis in a 6.5% acrylamide gel at 4  C for 1 h. The gels were transferred to the bonding membrane for 40 min at room temperature. After crosslinking for 10 min with an ultraviolet crosslinking apparatus, the membrane Table 1 Primers used for analysis of mRNA expression in primary murine HSCs. Gene

Orientation

Primers sequences (50 –30 )

Nrf2

forward reverse forward reverse

CGAGATATACGCAGGAGAGGTAAGA GCTCGACAATGTTCTCCAGCTT CTATGACCACAGTCCATGC CACATTGGGGGTAGGAACAC

GAPDH

4

C. Lu et al. / Biomedicine & Pharmacotherapy 95 (2017) 1–10

was blocked, streptavidin-HRP labeled, washed again, and equilibrated. Images were obtained using Gel Doc2000 system (Bio-Rad).

Institute) according to the protocol from manufacturer. The absorbance values were detected using a SpectraMax microplate spectrophotometer.

2.11. Biochemical analysis 2.13. Statistical analysis After centrifugation, sera were collected from blood and stored at
80  C for further analyses. Serum activities of aspartate aminotransferase (AST, C010-2) and alanine aminotransferase (ALT, C009-2), and content of triglyceride (TG, A110-1) in HSCs were detected using commercial assay kits according to the protocols from the manufacturer (Nanjing Jiancheng Bioengineering Institute). Serum levels of tumor necrosis factor-alpha (TNF-a, SBJM0030), laminin (LN, SBJ-M0668), hyaluronic acid (HA, SBJM0567), procollagen type III (PC-III, SBJ-M0654), and type IV collagen (IV-C, SBJ-M0653) were analyzed using commercial kits according to the protocols from the manufacturer (Nanjing SenBeiJia Biological Technology Co., Ltd., Nanjing, Jiangsu, China). The serum level of interleukin-6 (IL-6, ml002293) was detected using a commercial kit from the manufacturer (Shanghai MLBIO Biotechnology Co. Ltd., Shanghai, China). 2.12. Hydroxyproline analysis The content of hydroxyproline in liver tissues was measured by a commercial kit (A030-2, Nanjing Jiancheng Bioengineering

Data were presented as mean  standard deviation (SD), and results were analyzed using GraphPad Prism Software Version 5.0 (Graphpad Software, La Jolla, CA, USA). The significance of difference was determined by one-way analysis of variance (ANOVA) with the post hoc Dunnett’s test. Values of P < 0.05 were considered statistically significant. 3. Results 3.1. Curcumin induces lipocyte phenotype in HSCs Firstly, we investigated the effect of curcumin on inducing lipocyte phenotype in HSCs. Nile Red staining visually represented that curcumin caused a concentration-dependent increase in the numbers and sizes of lipid droplets in HSCs (Fig. 1A). Given that TG is the main component of lipid droplets, TG content was then detected. We found that curcumin also increased TG content in activated HSCs in a concentration-dependent manner (Fig. 1B). In addition, curcumin concentration-dependently reduced PPARa

Fig. 1. Curcumin induces lipocyte phenotype in HSCs. (A–C) LX-2 cells were treated with or without curcumin (10 mM, 20 mM, or 40 mM) for 24 h. DMSO (0.02%, w/v) was used as a negative control. (A) Nile Red staining. Scale bar = 25 mm. (B) TG content in HSCs. (C) Western blot analyses of PPARa, C/EBPa, PPARg, RXRa, and RARb in HSCs. Data were expressed as mean  SD. P values were determined by one-way ANOVA with the post hoc Dunnett’s test, *P < 0.05, **P < 0.01, and ***P < 0.001 versus DMSO. (D–F) Hepatic fibrosis was induced in mice by intraperitoneal injection of CCl4 once every other day for four weeks. Curcumin was administrated by gavage once a day for four weeks. Mice were randomly divided into five groups (n = 12 in every group): Vehicle, CCl4, CCl4 + Curcumin (100 mg/kg), CCl4 + Curcumin (200 mg/kg), and CCl4 + Curcumin (40021, 4 mg/ kg). (D) Experimental design. (E) Nile Red staining for primary murine HSCs freshly isolated from control, fibrotic, and curcumin-treated fibrotic mice. Scale bar = 25 mm. (F) Immunofluorescence double staining for PPARa, PPARg, and RXRa with a-SMA of murine liver sections. Scale bar = 100 mm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

C. Lu et al. / Biomedicine & Pharmacotherapy 95 (2017) 1–10

abundance but increased the protein abundance of C/EBPa, PPARg, RXRa, and RARb in HSCs (Fig. 1C). In vivo system was accordingly established (Fig. 1D). Primary murine stellate cells (mHSCs) were freshly isolated to integrally investigate the changes in lipid droplets in HSCs. Consistent with the in vitro data, few lipid droplets were observed in activated HSCs isolated from CCl4caused fibrotic mice. However, a dose-dependent increase in lipid droplets was found in HSCs isolated from livers of curcumintreated mice (Fig. 1E). Immunofluorescence double staining consistently showed that curcumin felicitously regulated the expression of genes associated with lipid and retinoid metabolism in HSCs in a dose-dependent manner (Fig. 1F). Taken together, these findings strongly suggested that curcumin could induce lipocyte phenotype in HSCs. 3.2. Curcumin stimulates nrf2 expression and nuclear translocation, and enhances its binding activity to DNA in HSCs Curcumin has been reported to activate Nrf2 and contribute to the regression of hepatic fibrosis [23,27]. We next detected the protein abundance of Nrf2 in HSCs. Results showed that curcumin increased the protein abundance of Nrf2 in HSCs in CCl4-treated mice (Fig. 2A). Western blot results showed a clear curcumininduced increase in the level of nuclear Nrf2 but a slight decrease in the level of cytoplasmic Nrf2 in primary HSCs, indicating that curcumin could promote the nuclear translocation of Nrf2 (Fig. 2B). Further, curcumin also strengthened the interaction of Nrf2 with DNA sequence as demonstrated by the EMSA data (Fig. 2C). Of great interest was that curcumin only at high dose induced Nrf2 transcription (Fig. 2D). Yet the total protein expression of Keap1, an endogenous inhibitory molecular that could specifically destabilize Nrf2 and facilitate its degradation via ubiquitin-proteasome system, was evidently disrupted by curcumin in a dose dependent manner (Fig. 2E). Collectively, these data illustrated that curcumin could impressively increase the protein abundance and nuclear translocation, and enhance the activity of Nrf2 in HSCs, which might be associated with its inductive effect on HSC lipocyte phenotype. 3.3. Curcumin induces lipocyte phenotype in HSCs via activating nrf2 To further explore the role of Nrf2 in inducing HSC lipocyte phenotype, we used gain- or loss-of function mutations of Nrf2 in HSCs. Western blot analyses showed that transfection of Nrf2 overexpression plasmids and Nrf2 siRNA was successful (Fig. 3A and B). We noted that overexpression of Nrf2 in HSCs using pCMVXLS-Nrf2 strengthened the regulatory effects of curcumin on the expression of PPARa, C/EBPa, PPARg, RXRa, and RARb. Nrf2 silencing using siRNA almost completely abolished the effects of curcumin (Fig. 3C and D). Further, Nrf2 shRNA lentivirus was introduced to establish Nrf2-knockdown mice to validate the function of Nrf2 in mediating the curcumin effect in vivo (Fig. 3E). Results from Nile Red staining for primary mHSCs and immunofluorescence double staining revealed that Nrf2 knockdown nearly abolished the induction of curcumin on lipid accumulation in HSCs and the regulatory effects on PPARg and C/EBPa (Fig. 3F and G). In summary, these findings implied that Nrf2 activation was required for curcumin to induce the lipocyte phenotype in HSCs. 3.4. Nrf2 activation by curcumin alleviates murine hepatic fibrosis Given that Nrf2 activation and curcumin could reverse the lipocyte phenotype in activated HSCs, the specific effect of curcumin on Nrf2 was further linked to the amelioration of hepatic fibrosis in mice. Elevated serum AST and ALT activities caused by CCl4 were markedly reduced in curcumin-treated mice

5

but not shRNA-mediated Nrf2-knockdown mice (Fig. 4A and B). Curcumin obviously decreased the elevated serum levels of proinflammatory cytokines TNF-a and IL-6 aroused by CCl4, however, which were abrogated by Nrf2 shRNA lentivirus (Fig. 4C and D). Results also indicated that Nrf2 knockdown abolished the reduction of curcumin on LN, HA, PC-III, and IV-C in serum (Fig. 4E
H). From H&E and collagen staining for murine liver sections, we observed severe hepatocyte adipose degeneration, inflammatory cell infiltration, fibrous hyperplasia, and collagen deposition, especially in areas near lobular inflammation and fibrotic septa, in fibrotic livers exposed to CCl4. Curcumin dramatically alleviated CCl4-caused a sequence of pathological changes. Notably, Nrf2 knockdown partially aggravated the toxic actions of CCl4 and abrogated the ameliorative effect of curcumin, evidenced by sustained hepatic cord disorder, vacuolation, and collagen accumulation (Fig. 4I). The determination of hepatic hydroxyproline content quantificationally confirmed the results above (Fig. 4J). Altogether, these results suggested that Nrf2 activation contributed to curcumin-caused reversion of hepatic fibrosis. 4. Discussion The incidence of hepatic fibrosis is rising worldwide, thus there is an urgent need to identify specific targets and design novel innocuous therapeutic strategies to prevent disease onset and progression. HSC activation is defined as an essential step for trigging fibrosis and advanced cirrhosis. Thus previously, most of the researches in this field were focused on preventing HSC activation and inducing HSC apoptosis; yet, the interventions did not achieve the desired results. Noticeably, activation of HSCs is accompanied by loss of its characteristic retinoid droplets. It is unclear, however, whether retinoid loss is a requirement for HSC activation, and if preventing retinoid loss could alter the activation cascade. A recent study highlighted a concept that accumulative lipid droplets were strong accelerants for reversion of activated HSCs to quiescent lipocytes [28], while the precise underlying signaling events have not been illuminated. Therefore, our goal was to dissect if Nrf2 was a critical molecule responsible for the metabolic dysregulation of retinoid droplets and HSC perpetual activation, and whether curcumin targeted Nrf2 to recover the lipocyte phenotype in activated HSCs and attenuate hepatic fibrosis. Nrf2, a distinguished regulator of antioxidant enzymes, is a ubiquitous “messenger” protein in various types of cells [29]. Except for its typical role in redox homeostasis, Nrf2 has been newly identified as a critical controller of lipid metabolism. Our prior studies have revealed that Nrf2 activation in hepatocytes inhibited alcohol-caused lipid deposition in whole livers, which suggested that Nrf2 could be recognized as a new therapeutic target for alcoholic liver diseases [26,30]. Regrettably, to date, few studies have uncovered the specific role of Nrf2 in lipid homeostasis in HSCs. Previously, we found that Nrf2 expression were markedly reduced in HSCs in CCl4-induced fibrotic mice [20]. Gain- or loss-of-function approaches were used to elucidate whether Nrf2 was closely involved in the metabolism of retinoid droplets. Since Nile Red is a kind of fluorescent oxazine dye with potent lipophilic properties, it was utilized as a selective dye to qualitatively detect lipid droplets in HSCs. Lipid droplets are comprised of sorts of lipids, among which TG is the most ample ingredient [3]. Results of the determination on intracellular TG content quantificationally confirmed the results of Nile Red staining. Here, it was interesting to find that in HSCs, Nrf2 expression positively correlated with lipid abundance, which was in sharp contrast with hepatocytes in our previous studies [30,31]. Retinoid metabolism is a complex and tightly programmed process

6

C. Lu et al. / Biomedicine & Pharmacotherapy 95 (2017) 1–10

which involves diverse genes correlated with lipometabolism. PPARa is a nuclear hormone receptor that facilitates the transportation and oxidation of fatty acids. C/EBPa and PPARg mainly initiate lipogenesis [32]. We found that curcumin could inhibit fatty acid oxidation and induce lipid synthesis. RXRa and

RARb are acknowledged as dominant transcriptional factors participated in the regulation of retinoid metabolism [33]. From our results, we observed that curcumin further stimulated the expression of RXRa and RARb, suggesting that curcumin may regulate retinoic acid responsiveness and signaling.

Fig. 2. Curcumin stimulates Nrf2 expression and nuclear translocation, and enhances its binding activity to DNA in HSCs. Hepatic fibrosis was induced in mice by intraperitoneal injection of CCl4 once every other day for four weeks. Curcumin was administrated by gavage once a day for four weeks. Mice were randomly divided into five groups (n = 12 in every group): Vehicle, CCl4, CCl4 + Curcumin (100 mg/kg), CCl4 + Curcumin (200 mg/kg), and CCl4 + Curcumin (400 mg/kg). (A) Immunofluorescence double staining for Nrf2 with a-SMA of murine liver sections. Scale bar = 100 mm. (B–D) Primary murine HSCs were freshly isolated from control, fibrotic, and curcumin-treated fibrotic mice. (B) Western blot analysis of Nrf2 in primary murine HSCs. (C) EMSA analysis for investigating the binding capacity of Nrf2 to DNA sequences in primary murine HSCs. (D) Quantitative real-time PCR analysis of Nrf2 in primary murine HSCs. (E) Western blot analysis of Keap1 in primary murine HSCs. Data were expressed as mean  SD. P values were determined by one-way ANOVA with the post hoc Dunnett’s test, ***P < 0.001 versus vehicle control, #P < 0.05, ##P < 0.01, and ###P < 0.001 versus CCl4.

C. Lu et al. / Biomedicine & Pharmacotherapy 95 (2017) 1–10

Curcumin has a long history of use as a dietary agent, which is different from most other natural active compounds. Several studies have shown that curcumin has no toxicity in vivo and in vitro. In a study of high dose oral curcumin in Taiwan, researchers administered up to 8 g daily of curcumin for 3 months to patients with pre-invasive malignant or high risk pre-conditions, stating that no toxicity was observed [34]. In patients with advanced colorectal cancer treated in the UK, curcumin was well tolerated at all dose levels up to 3.6 g daily for up to 4 months [35]. In our study, mice were treated with curcumin at 400 mg/kg body weight for attenuating hepatic fibrosis. The equivalent human dosage using body surface area conversion would be 0.04 g/kg body weight which was lower than that commonly used. Previous results showed that the viability of hepatocytes treated with 10 to 70 mM curcumin for 24 h was not significantly different from that of the

7

control group, suggesting that curcumin concentrations up to 70 mM have no significant cytotoxic effect on hepatocytes [36]. Given that curcumin had potent protective effect on hepatic fibrosis and was used in clinical trials for some diseases, such as advanced pancreatic cancer [37], we believe that curcumin will be used in clinical therapy for liver fibrosis in the near future. Effective scavenging of toxic substances is necessary to protect organisms from liver injury. Hepatotoxic chemicals are usually divided into two kinds, hepatotoxins that require metabolic activation in livers and others that do not, among which CCl4 belongs to the first class. CCl4 is metabolized by hepatic microsomal cytochromes P450 to form the hepatotoxic metabolites trichloromethyl (CCl3) and trichloromethylperoxyl radicals (CCl3OO). These free radicals trigger intense oxidative stress in hepatocytes and aggravate cell injury, such as apoptosis. Injured

Fig. 3. Curcumin induces lipocyte phenotype in HSCs via activating Nrf2. (A and B) LX-2 cells were transfected with or without pCMV-XLS-Nrf2 plasmids (3 mg) or Nrf2 siRNA (100 pmol) for 24 h. Control pcDNA and siRNA were used as negative controls. Western blot analysis of Nrf2 in HSCs. Data were expressed as mean  SD. P values were determined by one-way ANOVA with the post hoc Dunnett’s test, ***P < 0.001 versus blank control. (C and D) LX-2 were transfected with or without pCMV-XLS-Nrf2 (3 mg) or Nrf2 siRNA (100 pmol), and treated with or without curcumin (20 mM) for 24 h. Control pcDNA and siRNA were used as negative controls. Western blot analyses of PPARa, C/ EBPa, PPARg, RXRa, and RARb in HSCs. Data were expressed as mean  SD. P values were determined by one-way ANOVA with the post hoc Dunnett’s test, *P < 0.05 and ** P < 0.01 versus pcDNA or control siRNA, #P < 0.05 and ##P < 0.01 versus Curcumin + pcDNA or curcumin + control siRNA. (E–G) Hepatic fibrosis was induced in mice by intraperitoneal injection of CCl4 once every other day for four weeks. Curcumin was administrated by gavage once a day for four weeks. Nrf2 shRNA lentivirus was biweekly injected into murine caudal vein at a titer of 1 108 TU/mice. Mice were randomly divided into five groups (n = 12 in every group): Vehicle, CCl4, CCl4 + Nrf2 shRNA, CCl4 + Curcumin (200 mg/kg), and CCl4 + Curcumin (200 mg/kg) + Nrf2 shRNA. (E) Experimental design. (F) Nile Red staining for primary murine HSCs freshly isolated from mice treated with indicated agents. Scale bar = 25 mm. (G) Immunofluorescence double staining for PPARg and C/EBPa with a-SMA of murine liver sections. Scale bar = 100 mm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

8

C. Lu et al. / Biomedicine & Pharmacotherapy 95 (2017) 1–10

hepatocytes release various cytokines to stimulate activation of HSCs. Several approaches to induction of fibrosis have been described. Of these, CCl4 intoxication in mice is probably the most widely studied. Chronic CCl4-caused experimental hepatic fibrosis is an accepted common model used in preclinical researches because of its relevance to clinical human hepatic fibrosis [38]. The CCl4 model is the best characterized with respect to histological, biochemical, cell, and molecular changes associated with the development of fibrosis [39]. Development of efficient therapies for hepatic fibrosis has long plagued the medical profession because breakthrough targets from

the multifactorial pathogenesis of hepatic fibrosis are far from clear and definite. We found that curcumin not only increase the total protein abundance in HSCs but also promoted Nrf2 translocation. The potent induction of Nrf2 by curcumin on lipocyte phenotype in activated HSCs suggested that Nrf2 could be a potential target for curcumin to reverse HSC phenotype in the recovery from experimental hepatic fibrosis. Thus, we believed that Nrf2 molecular targeted therapy was promising. Plentiful studies in the past decades, including ours, have provided compelling supportive evidence showing that the natural product curcumin inhibited HSC activation in vitro system and

Fig. 4. Nrf2 activation by curcumin alleviates murine hepatic fibrosis. Hepatic fibrosis was induced in mice by intraperitoneal injection of CCl4 once every other day for four weeks. Curcumin was administrated by gavage once a day for four weeks. Nrf2 shRNA lentivirus was biweekly injected into murine caudal vein at a titer of 1 108 TU/mice. Mice were randomly divided into five groups (n = 12 in every group): Vehicle, CCl4, CCl4 + Nrf2 shRNA, CCl4 + Curcumin (200 mg/kg), and CCl4 + Curcumin (200 mg/kg) + Nrf2 shRNA. (A and B) Activities of serum AST and ALT. (C and D) Levels of TNF-a and IL-6 in serum. (E–H) Levels of LN, HA, PC-III, and IV-C in serum. (I) Microphotographs of H&Estained, Masson-stained, and Sirius Red-stained murine liver sections. Scale bar = 100 mm. (J) Hepatic hydroxyproline content. Data were expressed as mean  SD. P values were determined by one-way ANOVA with the post hoc Dunnett’s test, **P < 0.01 and ***P < 0.001 versus vehicle control, #P < 0.05, ##P < 0.01, and ###P < 0.001 versus CCl4, $ P < 0.05, $$P < 0.01, and $$$P < 0.001 versus CCl4 + Curcumin.

C. Lu et al. / Biomedicine & Pharmacotherapy 95 (2017) 1–10

exhibited excellent antifibrotic potential in vivo system [22]. In addition, curcumin induced Nrf2 expression in murine hepatocytes and immortalized HSCs [25,40]. In this integral study, we found that curcumin also induced the expression, nuclear translocation, and transcriptional activity of Nrf2 in primary mHSCs in vivo. Curcumin promoted the preservation of lipid droplets in HSCs and regulated the genes relevant to lipid metabolism and retinoic acid responsiveness and signaling, which was completely abrogated in Nrf2-silenced HSCs and enhanced in Nrf2 over-expressed HSCs. These findings implied that curcumin induced the lipocyte phenotype in activated HSCs possibly by enhancing Nrf2 expression. Lipocyte phenotype is a characteristic of quiescent HSCs, we next explored whether induction of lipocyte phenotype in activated HSCs by curcumin ultimately contributed to the recovery of hepatic fibrosis and whether Nrf2 induction was a prerequisite. Chronic inflammation aggravates the liver damage and facilitates the development of hepatic fibrosis. Various proinflammatory factors participate in inflammatory responses, among which TNFa and IL-6 are central factors. TNF-a plays an important role in immunoregulation and cellular homeostasis. In addition, it has been recognized as a controller of cell survival, necroptosis, and apoptosis [41]. Serum levels of HA, LN, PC-III, and C-IV are the most sensitive indicators for hepatic fibrosis, which have an important diagnostic value for judgment of the disease progression and curative outcomes of therapies [42]. Hydroxyproline exclusively exists in collagens, which has been identified as a biomarker to quantify hepatic fibrosis [43]. The effects of curcumin were linked to the significant amelioration of CCl4-caused hepatic fibrosis in mice, marked by decreases in multiple liver damage biomarkers, such as serum AST, ALT, hepatocyte steatosis, fibrous hyperplasia, collagen accumulation, and liver hydroxyproline content. Of great note was that Nrf2 knockdown aggravated CCl4-triggered hepatic fibrosis and seriously damaged the protective effects of curcumin. Therefore, induction of Nrf2 to promote the restoration of lipid droplets in activated HSCs could be a novel strategy to inactivate HSCs and deal with hepatic fibrosis. In conclusion, our present work demonstrated that curcumin induced lipocyte phenotype of activated HSCs, where Nrf2 activation was a prerequisite. Targeting at regulating Nrf2 to force HSCs to dedifferentiate into lipocytes may serve as a novel strategy for the reversion of hepatic fibrosis (Fig. 5). These novel discoveries

Fig. 5. Schema of underlying mechanism of curcumin induction of lipocyte phenotype of HSCs. Curcumin activates Nrf2, which is possibly associated with disruption of keap1. As a consequence, curcumin induces lipocyte phenotype leading to improved hepatic fibrosis. The identified mechanism possibly accounts for curcumin attenuation of hepatic fibrosis.

9

not only provided a new strategy for the treatment of hepatic fibrosis but also uncovered the potential mechanisms involved. Conflict of interest The authors declare that there are no conflicts of interest. Acknowledgments This work was supported by 2015 Program for Graduate Scientific Innovation of Jiangsu Higher Education Institutions (KYLX15_0999), the National Natural Science Foundation of China (81270514, 31571455, 31401210, 31600653, and 81600483), the Natural Science Foundation of Jiangsu Province (BK20140955), the Open Project Program of Jiangsu Key Laboratory for Pharmacology and Safety Evaluation of Chinese Materia Medica (JKLPSE201502), and the Project of the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD). References [1] V. Hernandez-GeaS.L. Friedman, Pathogenesis of liver fibrosis, Annu. Rev. Pathol. 6 (2011) 425–456. [2] S.L. Friedman, Hepatic fibrosis
overview, Toxicology 254 (3) (2008) 120–129. [3] W.S. Blaner, S.M. O'Byrne, N. Wongsiriroj, J. Kluwe, D.M. D'Ambrosio, H. Jiang, R.F. Schwabe, E.M. Hillman, R. Piantedosi, J. Libien, Hepatic stellate cell lipid droplets: a specialized lipid droplet for retinoid storage, Biochim. Biophys. Acta 1791 (6) (2009) 467–473. [4] A.M. Elsharkawy, F. Oakley, D.A. Mann, The role and regulation of hepatic stellate cell apoptosis in reversal of liver fibrosis, Apoptosis 10 (5) (2005) 927– 939. [5] N. Kawada, Human hepatic stellate cells are resistant to apoptosis: implications for human fibrogenic liver disease, Gut 55 (8) (2006) 1073–1074. [6] K. Imai, M. Sato, N. Kojima, M. Miura, T. Sato, T. Sugiyama, K. Enomoto, H. Senoo, Storage of lipid droplets in and production of extracellular matrix by hepatic stellate cells (vitamin A-storing cells) in Long-Evans cinnamon-like colored (LEC) rats, Anat. Rec. 258 (4) (2000) 338–348. [7] T. Miyahara, L. Schrum, R. Rippe, S. Xiong, H.F. Yee Jr., K. Motomura, F.A. Anania, T.M. Willson, H. Tsukamoto, Peroxisome proliferator-activated receptors and hepatic stellate cell activation, J. Biol. Chem. 275 (46) (2000) 35715–35722. [8] H. Tsukamoto, H. She, S. Hazra, J. Cheng, T. Miyahara, Anti-adipogenic regulation underlies hepatic stellate cell transdifferentiation, J. Gastroenterol. Hepatol. 21 (Suppl. 3) (2006) S102–S105. [9] T. Kisseleva, M. Cong, Y. Paik, D. Scholten, C. Jiang, C. Benner, K. Iwaisako, T. Moore-Morris, B. Scott, H. Tsukamoto, S.M. Evans, W. Dillmann, C.K. Glass, D.A. Brenner, Myofibroblasts revert to an inactive phenotype during regression of liver fibrosis, Proc. Natl. Acad. Sci. U. S. A. 109 (24) (2012) 9448–9453. [10] E.L. Guimaraes, M.F. Franceschi, C.M. Andrade, R.M. Guaragna, R. Borojevic, R. Margis, E.A. Bernard, F.C. Guma, Hepatic stellate cell line modulates lipogenic transcription factors, Liver Int. 27 (9) (2007) 1255–1264. [11] W.S. Cook, A.V. Yeldandi, M.S. Rao, T. Hashimoto, J.K. Reddy, Less extrahepatic induction of fatty acid beta-oxidation enzymes by PPAR alpha, Biochem. Biophys. Res. Commun. 278 (1) (2000) 250–257. [12] H. She, S. Xiong, S. Hazra, H. Tsukamoto, Adipogenic transcriptional regulation of hepatic stellate cells, J. Biol. Chem. 280 (6) (2005) 4959–4967. [13] M.I. LefterovaM.A. Lazar, New developments in adipogenesis, Trends Endocrinol. Metab. 20 (3) (2009) 107–114. [14] E. Sharvit, S. Abramovitch, S. Reif, R. Bruck, Amplified inhibition of stellate cell activation pathways by PPAR-gamma, RAR and RXR agonists, PLoS One 8 (10) (2013) e76541. [15] R. Bruck, S. Weiss, H. Aeed, M. Pines, Z. Halpern, I. Zvibel, Additive inhibitory effect of experimentally induced hepatic cirrhosis by agonists of peroxisome proliferator activator receptor gamma and retinoic acid receptor, Dig. Dis. Sci. 54 (2) (2009) 292–299. [16] L. Wang, J.J. Potter, L. Rennie-Tankersley, G. Novitskiy, J. Sipes, E. Mezey, Effects of retinoic acid on the development of liver fibrosis produced by carbon tetrachloride in mice, Biochim. Biophys. Acta 1772 (1) (2007) 66–71. [17] D.J. Reuland, S. Khademi, C.J. Castle, D.C. Irwin, J.M. McCord, B.F. Miller, K.L. Hamilton, Upregulation of phase II enzymes through phytochemical activation of Nrf2 protects cardiomyocytes against oxidant stress, Free Radic. Biol. Med. 56 (2013) 102–111. [18] J.J. Yang, H. Tao, W. Hu, L.P. Liu, K.H. Shi, Z.Y. Deng, J. Li, MicroRNA-200a controls Nrf2 activation by target Keap1 in hepatic stellate cell proliferation and fibrosis, Cell Signal. 26 (11) (2014) 2381–2389. [19] W. Xu, C. Hellerbrand, U.A. Kohler, P. Bugnon, Y.W. Kan, S. Werner, T.A. Beyer, The Nrf2 transcription factor protects from toxin-induced liver injury and fibrosis, Lab. Invest. 88 (10) (2008) 1068–1078. [20] C. Lu, W. Xu, F. Zhang, J. Shao, S. Zheng, Nrf2 knockdown attenuates the ameliorative effects of ligustrazine on hepatic fibrosis by targeting hepatic stellate cell transdifferentiation, Toxicology 365 (2016) 35–47.

10

C. Lu et al. / Biomedicine & Pharmacotherapy 95 (2017) 1–10

[21] H. Jin, Y. Jia, Z. Yao, J. Huang, M. Hao, S. Yao, N. Lian, F. Zhang, C. Zhang, X. Chen, M. Bian, J. Shao, L. Wu, A. Chen, S. Zheng, Hepatic stellate cell interferes with NK cell regulation of fibrogenesis via curcumin induced senescence of hepatic stellate cell, Cell Signal. 33 (2017) 79–85. [22] F. Zhang, Z. Zhang, L. Chen, D. Kong, X. Zhang, C. Lu, Y. Lu, S. Zheng, Curcumin attenuates angiogenesis in liver fibrosis and inhibits angiogenic properties of hepatic stellate cells, J. Cell. Mol. Med. 18 (7) (2014) 1392–1406. [23] Y. Fu, S. Zheng, J. Lin, J. Ryerse, A. Chen, Curcumin protects the rat liver from CCl4-caused injury and fibrogenesis by attenuating oxidative stress and suppressing inflammation, Mol. Pharmacol. 73 (2) (2008) 399–409. [24] Z. Zhang, Y. Guo, S. Zhang, Y. Zhang, Y. Wang, W. Ni, D. Kong, W. Chen, S. Zheng, Curcumin modulates cannabinoid receptors in liver fibrosis in vivo and inhibits extracellular matrix expression in hepatic stellate cells by suppressing cannabinoid receptor type-1 in vitro, Eur. J. Pharmacol. 721 (1-3) (2013) 133–140. [25] C. Lu, W. Xu, F. Zhang, J. Shao, S. Zheng, Nrf2 knockdown disrupts the protective effect of curcumin on alcohol-Induced hepatocyte necroptosis, Mol. Pharm. 13 (12) (2016) 4043–4053. [26] C. Lu, F. Zhang, W. Xu, X. Wu, N. Lian, H. Jin, Q. Chen, L. Chen, J. Shao, L. Wu, Y. Lu, S. Zheng, Curcumin attenuates ethanol-induced hepatic steatosis through modulating Nrf2/FXR signaling in hepatocytes, IUBMB Life 67 (8) (2015) 645–658. [27] S. Shishodia, Molecular mechanisms of curcumin action: gene expression, Biofactors 39 (1) (2013) 37–55. [28] S.L. Friedman, Hepatic stellate cells: protean, multifunctional, and enigmatic cells of the liver, Physiol. Rev. 88 (1) (2008) 125–172. [29] H. Motohashi, T. O'Connor, F. Katsuoka, J.D. Engel, M. Yamamoto, Integration and diversity of the regulatory network composed of Maf and CNC families of transcription factors, Gene 294 (1-2) (2002) 1–12. [30] C. Lu, W. Xu, J. Shao, F. Zhang, A. Chen, S. Zheng, Nrf2 activation is required for ligustrazine to inhibit hepatic steatosis in alcohol-preferring mice and hepatocytes, Toxicol. Sci. 155 (2) (2016) 432–443. [31] C. Lu, Y. Jiang, F. Zhang, J. Shao, L. Wu, X. Wu, N. Lian, L. Chen, H. Jin, Q. Chen, Y. Lu, S. Zheng, Tetramethylpyrazine prevents ethanol-induced hepatocyte injury via activation of nuclear factor erythroid 2-related factor 2, Life Sci. 141 (2015) 119–127. [32] Y.C. Liang, M.T. Yang, C.J. Lin, C.L. Chang, W.C. Yang, Bidens pilosa and its active compound inhibit adipogenesis and lipid accumulation via down-modulation of the C/EBP and PPARgamma pathways, Sci. Rep. 6 (2016) 24285.

[33] A. Chawla, J.J. Repa, R.M. Evans, D.J. Mangelsdorf, Nuclear receptors and lipid physiology: opening the X-files, Science 294 (5548) (2001) 1866–1870. [34] A.L. Cheng, C.H. Hsu, J.K. Lin, M.M. Hsu, Y.F. Ho, T.S. Shen, J.Y. Ko, J.T. Lin, B.R. Lin, W. Ming-Shiang, H.S. Yu, S.H. Jee, G.S. Chen, T.M. Chen, C.A. Chen, M.K. Lai, Y.S. Pu, M.H. Pan, Y.J. Wang, C.C. Tsai, C.Y. Hsieh, Phase I clinical trial of curcumin, a chemopreventive agent, in patients with high-risk or pre-malignant lesions, Anticancer Res. 21 (4B) (2001) 2895–2900. [35] R.A. Sharma, S.A. Euden, S.L. Platton, D.N. Cooke, A. Shafayat, H.R. Hewitt, T.H. Marczylo, B. Morgan, D. Hemingway, S.M. Plummer, M. Pirmohamed, A.J. Gescher, W.P. Steward, Phase I clinical trial of oral curcumin: biomarkers of systemic activity and compliance, Clin. Cancer Res. 10 (20) (2004) 6847– 6854. [36] M. Manju, A.S. Vijayasree, M.A. Akbarsha, O.V. Oommen, Protective effect of dietary curcumin in Anabas testudineus (Bloch) with a special note on DNA fragmentation assay on hepatocytes and micronucleus assay on erythrocytes in vivo, Fish Physiol. Biochem. 39 (5) (2013) 1323–1330. [37] N. Dhillon, B.B. Aggarwal, R.A. Newman, R.A. Wolff, A.B. Kunnumakkara, J.L. Abbruzzese, C.S. Ng, V. Badmaev, R. Kurzrock, Phase II trial of curcumin in patients with advanced pancreatic cancer, Clin. Cancer Res. 14 (14) (2008) 4491–4499. [38] P. StarkelI.A. Leclercq, Animal models for the study of hepatic fibrosis, Best Pract. Res. Clin. Gastroenterol. 25 (2) (2011) 319–333. [39] C. Constandinou, N. Henderson, J.P. Iredale, Modeling liver fibrosis in rodents, Methods Mol. Med. 117 (2005) 237–250. [40] Z. Liu, W. Dou, Y. Zheng, Q. Wen, M. Qin, X. Wang, H. Tang, R. Zhang, D. Lv, J. Wang, S. Zhao, Curcumin upregulates Nrf2 nuclear translocation and protects rat hepatic stellate cells against oxidative stress, Mol. Med. Rep. 13 (2) (2016) 1717–1724. [41] H. Blaser, C. Dostert, T.W. Mak, D. Brenner, T.N.F ROS crosstalk in inflammation, Trends Cell Biol. 26 (4) (2016) 249–261. [42] Z. Wang, Z. Zhang, N. Du, K. Wang, L. Li, Hepatoprotective effects of grape seed procyanidin B2 in rats with carbon tetrachloride-induced hepatic fibrosis, Altern. Ther. Health Med. 21 (Suppl. 2) (2015) 12–21. [43] M. Kusunose, B. Qiu, T. Cui, A. Hamada, S. Yoshioka, M. Ono, M. Miyamura, S. Kyotani, Y. Nishioka, Effect of Sho-saiko-to extract on hepatic inflammation and fibrosis in dimethylnitrosamine induced liver injury rats, Biol. Pharm. Bull. 25 (11) (2002) 1417–1421.