Globular adiponectin protects hepatocytes from tunicamycin-induced cell death via modulation of the inflammasome and heme oxygenase-1 induction

Accepted Manuscript Title: Globular adiponectin protects hepatocytes from tunicamycin-induced cell death via modulation of the inflammasome and heme oxygenase-1 induction Authors: Amrita Khakurel, Pil-Hoon Park PII: DOI: Reference:

S1043-6618(17)30683-7 https://doi.org/10.1016/j.phrs.2017.10.010 YPHRS 3710

To appear in:

Pharmacological Research

Received date: Revised date: Accepted date:

5-6-2017 26-9-2017 18-10-2017

Please cite this article as: Khakurel Amrita, Park Pil-Hoon.Globular adiponectin protects hepatocytes from tunicamycin-induced cell death via modulation of the inflammasome and heme oxygenase-1 induction.Pharmacological Research https://doi.org/10.1016/j.phrs.2017.10.010 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Globular adiponectin protects hepatocytes from tunicamycin-induced cell death via modulation of the inflammasome and heme oxygenase-1 induction

Amrita Khakurel1 and Pil-Hoon Park1 1:

College of Pharmacy, Yeungnam University, Gyeongsan, Republic of Korea

Address correspondence to: Pil-Hoon Park, PhD College of Pharmacy, Yeungnam University, Gyeongsan, Republic of Korea. Phone: 82-53-810-2826, Fax: 82-53-810-4654, Email: [email protected]

Graphical abstract

1

Abstract Endoplasmic reticulum (ER) stress, which is defined as the accumulation of unfolded or misfolded proteins in the ER, triggers cellular dysfunction and eventually leads to cell death. In particular, excessive and prolonged ER stress is closely related with hepatic injury. Adiponectin, a hormone predominantly produced by adipose tissue, is known to possess potent hepatoprotective properties and exhibits a cytoprotective effect in response to chronic ER stress. However, the underlying mechanisms are not clearly understood. In the present study, we examined the protective effect of globular adiponectin (gAcrp) on tunicamycin-induced cell death and further investigated its potential underlying mechanisms in rat hepatocytes. Herein, we found that treatment with gAcrp inhibited tunicamycin-induced cell death, decreased lactate dehydrogenase release (marker of pyroptotic cell death), and suppressed caspase activation; clearly indicating that gAcrp protects liver cells from ER stress. Interestingly, gAcrp prevented the tunicamycin-induced activation of the inflammasome, a key platform involved in the production of inflammatory cytokines that induces pyroptosis, determined by suppression of interleukin-1β (IL-1β) maturation, apoptosis-associated specklike protein containing a carboxy-terminal CARD (ASC) speck formation, and caspase-1 activation. Moreover, we showed that suppression of the inflammasome activation by gAcrp was mediated via modulation of reactive oxygen species (ROS) production, particularly inhibition of NADPH oxidase. In addition, inhibition of heme oxygenase-1 (HO-1) signaling by pretreatment with SnPP, a pharmacological inhibitor of HO-1, or transfection with an siRNA targeting HO-1, abrogated the protective effects of gAcrp against tunicamycin-induced cell death and abolished the suppressive effect on the 2

inflammasome activation, demonstrating that HO-1 signaling plays a crucial role in the protective effect of gAcrp against tunicamycin-induced damage in liver cells. Taken together, these results indicate that gAcrp protects liver cells from ER stress by modulating inflammasomes activation, at least in part, via HO-1 signaling-dependent mechanisms. Abbreviations ASC, Apoptosis-associated speck-like protein containing a CARD; ATF6, Activating transcription factor 6; CHOP, C/EBP homologous protein; DAMPs, Damage-associated molecular patterns; gAcrp, globular adiponectin; HBSS, Hank’s balanced salt solution; HO-1, heme oxygenase-1; IL-1β, Interleukin 1β; IRE1α, inositol-requiring protein 1α; LDH, Lactate dehydrogenase; MDMs, Monocyte-derived macrophages; NAC, Nacetylcysteine; NADPH oxidase, nicotinamide adenine dinucleotide phosphate-oxidase; NLRP3, Nod-like receptor protein containing pyrin domain; PAMPs: Pathogenassociated molecular patterns; PERK, Protein kinase RNA-like endoplasmic reticulum kinase; ROS, reactive oxygen species; SnPP, tin protoporphyrin IX; TCA, Tri-chloro acetic acid; TNF- α, Tumor necrosis factor- α; TUDCA, tauroursodeoxycholic acid; Tun, tunicamycin; UPR, Unfolded protein response.

Key words: Adiponectin, ER stress, Heme oxygenase-1, Hepatocytes, Inflammasomes 1. Introduction Hepatocytes death is the initiating factor in the development of liver diseases including the states of inflammation, fibrosis, cirrhosis, and hepatocellular carcinoma [1-3]. A 3

number of studies have proposed different mechanisms to account for hepatocyte death. Among them, endoplasmic reticulum (ER) stress has recently received attention as one of the critical causes for cellular damage and activation of hepatocellular death pathways [4]. ER stress, defined as disturbances in ER-associated functions, results in the accumulation of misfolded or unfolded proteins that triggers the unfolded protein response (UPR) [5]. The UPR is the reaction of cells exposed to stressful conditions; it conveys information about the protein folding status from the ER lumen to the nucleus and aims to restore normal function in the cells by inhibiting protein translation, degrading misfolded proteins, and increasing the production of molecular chaperons involved in protein folding. The UPR is transduced by a number of signaling molecules, including protein kinase RNA-like ER kinase (PERK), inositol-requiring protein 1α (IRE1α), and activating transcription factor 6 (ATF6). Activation of these transmembrane proteins further induces activation of specific signaling cascades, which enhances the protein folding capacity in the ER membrane, resulting in the maintenance of ER homeostasis [6]. Therefore, under moderate ER stress conditions, the UPR acts as a compensatory mechanism for the cells to adjust to the stressful condition and contributes to cell survival. In contrast, under conditions of prolonged or excessive ER stress, the UPR does not protect the cells; instead, it triggers the apoptotic cell death pathway by upregulating the transcription of pro-apoptotic genes, including CHOP (C/EBP homologous protein) [7]. In addition, ER stress induced by tunicamycin or endotoxin (lipopolysaccharide) has been shown to stimulate pyroptosis, a highly inflammatory type of programmed cell death distinct from apoptosis, through activation of caspase-1 and caspase-11, and production of interleukin-1β (IL-1) in the liver 4

[4].Therefore, excessive ER stress may lead to both apoptosis and pyroptosis in hepatocytes. Chronic inflammation is an internal milieu leading to cell death in the liver and is closely associated with the pathogenesis of various liver diseases [8]. It has long been suggested that dysregulated activation of innate immune cells, such as Kupffer cells, is directly responsible for overproduction of inflammatory mediators and hepatic inflammation [9]. However, a growing body of recent evidence has suggested that hepatocytes are also capable of producing inflammatory mediators by modulating inflammasomes [10]. Inflammasomes are multi-protein complexes composed of Nodlike receptor protein containing pyrin domain (NLRP), caspase activation and recruitment domain-NLRC4 (CARD), apoptosis-associated speck-like protein (ASC), and caspase-1. Among the various types of NLRPs, NLRP3 is the best-characterized inflammasome in hepatocytes [4, 11]. Upon exposure to cellular damage or inflammatory signals, such as microbial infection, endogenous damage-associated molecular patterns (DAMPs) or exogenous pathogen-associated molecular patterns (PAMPs), the cytosolic inflammasomes components (ASC, NLRP, and pro-caspase-1) assemble to form an oligomer that results in caspase-1 activation. This in turn leads to cleavage and activation of pro-IL-1β and pro-IL-18 to matured IL-1β and IL-18, indicating that the inflammasome acts as a signaling platform for the activation/production of pro-inflammatory cytokines [12]. During activation of the inflammasome, the adaptor protein ASC polymerizes into a large foci known as “speck”. ASC speck formation is essential for the recruitment of caspase-1 [13]. Recent studies have also demonstrated that, in addition to casapse-1, different various types of 5

caspases, particularly caspase-8 and caspase-11, contribute to the inflammasome activation [14]. There is a growing appreciation that activation of the inflammasome and subsequent maturation of the interleukin family of cytokines, including IL-1 and IL-18, are involved in the pathogenesis of a number of diseases related with inflammation. For example, inhibition of IL-1 significantly decreased the extent of inflammation in human and mouse gout models [15, 16]. Similarly, NLRP3 inflammasome activation was also found in monocyte-derived macrophages (MDMs) from type 2 diabetic patients [17]. Furthermore, activation of the inflammasome and enhanced secretion of IL-1 directly contributed to the pathogenesis of various hepatic diseases, such as steatohepatitis [18], endotoxin-induced liver injury and cholestasis, and alcoholic and non-alcoholic fatty liver diseases [19, 20]. Adiponectin, the most abundant adipokine in plasma, was initially reported to play an important role in metabolism of lipid and sensitization of insulin [21]. It has been shown that decreased levels of adiponectin is a possible risk factor for the development and/or progression of diabetes, cardiovascular diseases, and cancer [22], suggesting that adiponectin is likely to induce various beneficial physiological responses. Furthermore, in addition to its critical metabolic effects, accumulating evidence has indicated that adiponectin exerted hepatoprotective effects in various experimental conditions. For example, globular adiponectin (gAcrp) improved hepatic steatosis, increased insulin secretion, inhibited hepatomegaly, suppressed TNF-α secretion, and prevented cell death induced by various xenobiotics in the liver [23, 24]. Recent studies have also 6

indicated that adiponectin negatively regulates ER stress via multiple mechanisms. For instance, gAcrp has been shown to suppress CHOP expression and caspases activation induced by ER stress in vascular smooth muscle cells [25], to inhibit ER stress-induced apoptosis in adipocytes [5], and to decrease ER stress in a non-alcoholic steatohepatitis model [26]. Given that chronic ER stress is closely related with liver damage, these results collectively imply that hepatoprotective effects by adiponectin can be mediated via modulation of ER stress. However, the direct evidence for the involvement of ER stress modulation in the protection of liver cells by adiponectin has not been reported. Based on previous reports, adiponectin exhibits potent hepatoprotective effects from various harmful stimuli. However, cytoprotective effects of adiponectin against ER stress-induced cellular damage in hepatocytes is not well established. Moreover, its underlying mechanisms are not clearly understood, although excessive ER stress is a critical mechanism leading to liver injury and hepatocytes death. Thus, to better understand the hepatoprotective effects of adiponectin, we investigated if gAcrp protects liver cells against ER stress-induced cell death and its underlying mechanisms in rat hepatocytes. Herein, we demonstrate for the first time that gAcrp protects hepatocytes from tunicamycin-induced cell death, including both apoptosis and pyroptosis, presumably via modulation of inflammasomes activation. In addition, we found that HO-1 signaling plays a crucial role in gAcrp-modulation of the inflammasome activation and protection of liver cells via suppression of ROS production. 2. Materials and methods

7

2.1. Materials Recombinant human globular adiponectin (gAcrp) was purchased from Peprotech Inc. (Rocky Hill, NJ, USA) (450-21). Reagents for hepatocyte isolation and culture, including Hank’s balanced salt solution (HBSS), Type I collagen solution, Type IV collagenase, Williams’ media E, Hydrocortisone 21-hemisuccinate sodium salt, and insulin from bovine pancreas, were obtained from Sigma-Aldrich (St. Louis, MO, USA). Primary antibodies against total PERK (C33E10) and phospho-specific PERK (MA5-15033), CHOP (MA1-250), and β-actin (MA5-15739) were obtained from Thermo Fisher Scientific (Rockford, IL, USA), caspase-1 (AG-20B-0042) and ASC (AG-25B-0006) were purchased from Adipogen (San Diego, CA, USA), cleaved IL-1β (sc-23460) was procured from Santa Cruz (Delaware, CA), NLRP3 (MAB7578) antibody was brought from R&D (Minneapolis, MN, USA), and HO-1 (ADI-SPA-894) was brought from Enzo (Boston, MA, USA). Anti-goat (sc-2020), anti-rabbit (NCI1460KR), and anti-mouse (NCI1430KR) secondary antibodies conjugated with horseradish peroxidase (HRP) were procured from Santa Cruz (Delaware, CA). Sn (IV) Protoporphyrin IX Chloride (SnPP), a selective inhibitor of HO-1, was obtained from Frontier Scientific (Logan, UT, USA). Tunicamycin was purchased from Sigma-Aldrich (St. Louis, MO, USA).

2.2. Isolation and culture of rat primary hepatocytes All the animal experiments were conducted in accordance with the guidelines for Institutional Animal Care and Use Committee at Yeungnam University (YU-2016-009). Hepatocytes were isolated from male Sprague Dawley rats (4 to 5 weeks) via a twostep collagenase perfusion method as described previously [27]. After anaesthetization of the rats by peritoneal injection of Zoletil 50 (Vibrac S.A, France), the abdominal cavity 8

was opened, the hepatic portal vein was cannulated by an 18G catheter, and the liver was perfused with warm HBSS at a rate of 30 mL/min until the blood in the liver was completely removed. The liver was then further perfused with 0.05% Type IV collagenase and a CaCl2 anhydrous solution for 10 min. After liver digestion, the extracted cell suspension was collected, passed through a 100-M mesh, and centrifuged at 430 rpm for 30 sec. The collected hepatocytes were washed twice with HBSS buffer, seeded into the collagen-coated dishes, and incubated at 37 °C in an incubator with a humidified atmosphere of 95% O2 and 5% CO2. Hepatocytes were used for experiments when the cell viability was more than 90%, as determined by a trypan blue exclusion assay.

2.3. Measurement of cell viability Cell viability was determined using a MTS assay as described previously [28]. Briefly, hepatocytes were seeded in collagen-coated clear 96-well plates at a density of 5×104 cells/well. After treatment with gAcrp and tunicamycin, cells were incubated with 20 μL of 3-(4, 5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfopheny)-2Htetrazolium (MTS) solution for 2 h at 37 °C. Cell viability was determined by measuring the absorbance at 490 nm using SPECTROstar Nano microplate reader from BMG LABTECH (Allmendgrün, Ortenberg, Germany).

2.4. Annexin-V and Propidium iodide staining (Flow cytometric analysis)

For the evaluation of cell death, hepatocytes were seeded in collagen-coated 35 mm dishes at a density of 5 × 105 cells/well. After overnight culture, cells were incubated with Z-VAD-FMK (Promega Corporation, Madison, USA) for 2 h or gAcrp for 18 h followed 9

by further incubation with tunicamycin (1 g/ml) for 24 h. After incubation, cells were washed twice with 1X PBS and stained with Annexin V FITC and propidium iodide for 30 min in the dark. The flow cytometric measurements were conducted on BD FACS Verse flow cytometer (BD Biosciences). The annexin V/PI staining results were assessed through FSC-A vs SSC-A plots, where viable cells or non-apoptotic cells were both Annexin V and PI negative, early apoptotic cells were Annexin V positive and PI negative, late apoptotic cells were both Annexin V and PI positive whereas, necrotic cells were Annexin V negative and PI positive.

2.5. Measurement of lactate dehydrogenase (LDH) release LDH release was measured using the CytoTox 96® Non-Radioactive Cytotoxicity Assay kit (Promega Corporation, Madison, USA) according to the manufacturer’s instructions. LDH release is determined based on the conversion of a tetrazolium salt into a formazan product. Briefly, hepatocytes were seeded in a collagen-coated clear 96-well plate at a density of 5×104 cells/well. After overnight culture, cells were pretreated with gAcrp for 18 h followed by further stimulation with tunicamycin for 24 h, as indicated in the figure legends. Cell culture media were then collected and used for measuring the amount of LDH. The level of LDH was determined by measuring absorbance at 490 nm with a SPECTROstar Nano from BMG LABTECH.

2.6. Caspase-1 enzyme activity assay Caspase-1 enzyme activity was determined using a caspase-1 enzymatic assay kit (Promega Corporation, Madison, USA) according to the manufacturer’s instructions. Briefly, cells were seeded in a collagen-coated white 96-well plate at a density of 5×104 10

cells/well. After treatment with tunicamycin for 24 h in the absence or presence of gAcrp overnight, the cell culture media (50 μl) were collected and incubated with Caspase-Glo 1 reagent (1:1) for 30 min. The luminescence was measured using a BMG LABTECH plate-reading luminometer (Allmendgruen 8, Ortenberg, Germany).

2.7. Measurement of reactive oxygen species (ROS) production ROS production was determined by measuring changes in the fluorescence of 5chloromethyl-2', 7’-dichlorodihydrofluorescein diacetate (CM-H2DCFDA). Hepatocytes were seeded in a collagen-coated black 96-well plate at a density of 5×104 cells/well. After pretreatment with gAcrp overnight followed by stimulation with tunicamycin for 6 h, cells were finally incubated with 2', 7’-dichlorofluorescin diacetate (5 µM) for 30 min in the dark and washed twice with HBSS to remove excess dye. Intracellular ROS production was determined using a FLUOstar OPTIMA fluorimeter (BMG LABTECH). Excitation and emission wavelengths were set to 485 nm and 520 nm, respectively.

2.8. Preparation of cellular extracts and Western blot analysis Hepatocytes isolated from rats were seeded in collagen-coated 35 mm dishes at a density of 1×106 cells/dish. After treatment with gAcrp and tunicamycin, cells were extracted with RIPA lysis buffer containing Halt protease inhibitors cocktail. The cellular lysates were centrifuged at 13,200 rpm for 15 min at 4 °C and the supernatants were collected for further use. To measure secreted mature IL-1β after treatment with gAcrp and tunicamycin, cell culture media were collected and proteins in the media were precipitated by incubation with trichloroacetic acid (TCA) for 10 min at 4 °C. After centrifugation at 13,200 rpm for 15 min, the supernatants were 11

discarded and the pellet was washed twice with cold acetone. The pellet was dried by placing the tubes in a heating block (94 °C) and the precipitates were then dissolved with sample buffer. For immunoblot analysis, 30-40 μg of protein was loaded onto 7.5%-15% sodium dodecyl sulfate-polyacrylamide gel for electrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride (PVDF) membrane. The membrane was blocked with 5% skim milk for 1 h, incubated with designated primary antibody in 3% BSA overnight at 4 °C, and finally incubated with the secondary antibody conjugated to horseradish peroxidase (HRP) for 1 h. The images of the blots were visualized using an enhanced chemiluminiscence (ECL) detection system and images were captured using a Fujifilm LAS-4000 mini (Fujifilm, Tokyo, Japan).

2.9. Enzyme-linked immunosorbent assay (ELISA) To measure secreted total IL-1β, hepatocytes were seeded at a density of 5×104 cells/well in collagen-coated 96-well plates. After overnight incubation, cells were pretreated with gAcrp for 18 h followed by further incubation with tunicamycin for 10 h. The culture medium was then collected and used for detection of secreted IL-1 using an ELISA kit (BD Biosciences, San Diego, USA) according to the manufacturer’s instructions.

2.10. Immunocytochemistry for detection of ASC speck Hepatocytes were seeded at a density of 5×104 cells/well in 8-well chamber slides. After overnight culture, cells were pretreated with 1 g/mL of gAcrp for 18 h, followed by incubation with tunicamycin for 10 h. Cells were then washed with PBS, fixed with 4% 12

paraformaldehyde solution, permeabilized with 0.1% Triton X-100, blocked with 1% BSA in PBS and incubated with an antibody against ASC (1:300) overnight at 4 °C. After washing, cells were incubated with a FITC-conjugated goat anti-rabbit secondary antibody for 1.5 h in the dark at room temperature followed by incubation with DAPI (1:100) for 10 min. The fluorescent images were captured by fluorescent microscopy (Nikon, Tokyo, Japan).

2.11. Transient transfection with small interfering RNA (siRNA) For the transient suppression of HO-1 expression, hepatocytes were seeded in 35 mm dishes at a density of 5×105 cells/dish and cultured with Williams medium E containing Glutamax, 10% fetal bovine serum (FBS), 1% penicillin-streptomycin (P/S), and insulin (4 µg/mL). After overnight culture, hepatocytes were transfected with siRNA targeting HO-1 or scrambled control siRNA using the ICAfectin442 transfection reagent (In-Cell-Art, Nantes, France) according to the manufacturer’s instructions. Briefly, cells were incubated with siRNA in William’s Media E containing bovine insulin (4 µg/mL) and hydrocortisone hemisuccinate (50 µM) for 24 h. Gene silencing efficiency was monitored by Western blot analysis after 24 h of transfection. Oligonucleotide siRNA targeting HO-1 was obtained from Bioneer (Daejeon, South Korea). The sequences of the siRNA used were as follows: Forward; 5'ACAAGCAGAACCCAGUCUA-3', Reverse; 5'-UAGACUGGGUUCUGCUUGU-3'.

2.12. Statistical analysis Values are presented as mean ± S.E.M. from at least three independent experiments. Data were analyzed by One-Way Analysis of Variance (ANOVA) and Tukey’s multiple 13

comparison test using GraphPad Prism software version 5.01 (La Jolla, CA). Differences between groups were considered to be significant at *p<0.05. 3. Results

3.1. Globular adiponectin protects hepatocytes from tunicamycin-induced cell death by modulating both pyroptosis and apoptosis

Tunicamycin is well known to induce ER stress, which acts as an initiating factor for cell death in the liver. Contrarily, adiponectin possesses potent hepatoprotective properties against various insults. Herein, we investigated if globular adiponectin (gAcrp) protected liver cells against tunicamycin-induced hepatocyte death, and further examined its underlying mechanisms. As shown in Fig. 1A, pretreatment of hepatocytes isolated from rats with gAcrp significantly inhibited tunicamycin-induced cell death, as determined by the MTS assay. Tunicamycin is known to induce different types of cell death, including pyroptosis and apoptosis [4]. To further investigate the type of cell death modulated by tunicamycin and gAcrp, we first examined the effects of tunicamycin and gAcrp on LDH release, which is a marker of pyroptosis. As expected, tunicamycin treatment led to a significant increase in LDH release, whereas pretreatment with gAcrp suppressed the tunicamycin-induced LDH release (Fig. 1B), suggesting a protective effect of gAcrp on pyroptotic cell death induced by tunicamycin in hepatocytes. This was further confirmed by an analysis of caspase-11 activity, which is also considered a marker for pyroptosis. In initial experiments, we found that tunicamycin treatment enhanced generation of the active form of caspase-11 in a time-dependent manner without significantly affecting procaspase-11 expression (supplementary figure S1). Pretreatment with gAcrp significantly suppressed the tunicamycin-induced caspase-11 activation (Fig. 1C), whereas no significant effect on total caspase-11 expression was observed, confirming the protective effect of gAcrp on tunicamycin-induced pyroptosis. In addition, we further examined if adiponectin could modulate apoptosis by assessing caspase-3 activity, a key marker of apoptosis. As demonstrated in Fig. 1D, pretreatment with gAcrp significantly decreased the production of the cleaved active form of caspase-3. Similarly, enhanced enzymatic 14

activity of caspase-3 seen after treatment with tunicamycin was substantially suppressed by pretreatment with gAcrp (Fig. 1E). We also observed that gAcrp pretreatment significantly inhibited tunicamycin-induced caspase-8 activation (Fig. 1F); collectively indicating that gAcrp modulates tunicamycin-induced apoptosis in rat hepatocytes. To further examine the cell death modulated by tunicamycin and gAcrp, a read-out for apoptosis was evaluated by annexin-V/propidium iodide (PI) staining. As shown in Fig. 1G, tunicamycin significantly enhanced population of the cells stained with annexin V and PI, which denotes late apoptotic cells. In addition, annexin positive, but PI negative cells, which denotes early apoptotic cells, were also slightly enhanced, while annexin and PI negative viable cells were prominently decreased, collectively indicating that tunicamycin treatment induces apoptotic cell death in hepatocytes. Furthermore, tunicamycin-induced apoptosis was returned to the almost control level by pretreatment with gAcrp. Moreover, pre-treatment with Z-VAD-FMK, a pan caspase inhibitor, significantly prevented tunicamycin-induced apoptotic cell death as expected. These results further indicate that tunicamycin induces hepatocytes death via apoptosis, and gAcrp protects hepatocytes from tunicamycin-induced apoptosis.

15

Fig. 1. Protective effects of globular adiponectin on tunicamycin-induced cell death via modulation of both apoptosis and pyroptosis in rat hepatocytes. Hepatocytes isolated from rats were pretreated with 0.5 or 1 µg/mL of gAcrp for 18 h, followed by incubation with tunicamycin (1 µg/mL) for 24 h (A) & (G) or 10 h (B-F). (A) Cell viability was assessed by MTS assay as described in the materials and methods section. Values represent mean ± SEM (n=3). *P<0.05 compared with control group; #P<0.05

compared with the cells treated with tunicamycin. (B) Cell culture media were 16

collected and used for the measurement of LDH release. Values represent mean ± SEM (n=3). *P<0.05 compared with control group; #P<0.05 compared with the cells treated with tunicamycin. (C) Cellular lysates were prepared and used for the measurement of total and active form of caspase-11 expression by Western blot analysis. Representative images from three independent experiments that showed similar results are shown. Quantitative analysis of caspase-11 expression was performed by densitometric analysis and shown in the lower panel. Values presented are fold change compared to the control cells and expressed as mean  SEM (n = 3). *P<0.05 compared to the control cells. #P< 0.05 compared to the cells treated tunicamycin. (D) Cellular lysates were prepared and total/active caspase-3 levels were assessed by Western blot analysis. Representative images from three independent experiments are shown. Quantitative analysis of caspase-11 expression was performed by densitometric analysis and shown in the lower panel. Values presented are fold change compared to the control cells and expressed as mean  SEM (n = 3). *P<0.05 compared to the control cells. #P<0.05 compared to the cells treated tunicamycin. (E) Cellular lysates were prepared and incubated with Caspase-Glo 3/7 reagent. Caspase-3 enzymatic activity was measured by the cleavage of luminogenic substrate Ac-DEVD-pNA. Data are presented as mean ± SEM (n=3). *P<0.05 compared with control group; #P<0.05 compared with the cells treated with tunicamycin. (F) Total cellular lysates were prepared and used to determine the total and active form of caspase-8 by Western blot analysis. Representative images from three independent experiments are shown. Quantitative analysis of active caspase-8 expression was performed by densitometric analysis and shown in the lower panel. Values presented are fold change compared to 17

the control cells and expressed as mean  SEM (n = 3). *P<0.05 compared to the control cells. #P<0.05 compared to the cells treated tunicamycin. (G) Cells were pretreated with gAcrp for 18 h or ZVAD-FMK, a pan caspase inhibitor, for 2 h followed by incubation with tunicamycin for 24 h. The modalities of cell death were measured as described in the methods. Quantitative analysis for apoptosis from three independent experiments are shown in the right side. Values represent mean± SEM (n=3). *P<0.05 compared with control group; #P<0.05 compared with the cells treated with tunicamycin. 3.2. Globular adiponectin suppresses tunicamycin-induced cell death by modulating ER stress in hepatocytes

We next examined whether gAcrp protects liver cells from tunicamycin-induced cell death through regulation of ER stress. To this end, we first confirmed tunicamycininduced cell death was mediated via ER stress. In the preliminary experiments to ensure the pharmacological actions of tunicamycin in our experimental condition, we found that tunicamycin treatment induced phosphorylation of PERK and increased CHOP expression, which are considered markers of ER stress and integral components of the ER stress-mediated apoptosis pathway (supplementary figure S2 and S3). Moreover, treatment with TUDCA, a potent chemical chaperone that inhibits ER stress, inhibited the tunicamycin-induced PERK phosphorylation (supplementary figure S4), thus confirming the pharmacological effects of tunicamycin and TUDCA. In the following experiments, TUDCA dose-dependently restored the tunicamycin-induced decrease in cell viability (Fig. 2A), similar to the protective effect of gAcrp observed in Fig. 1A. It also markedly suppressed the tunicamycin-induced LDH release (Fig. 2B). These results ensure that ER stress plays a key role in tunicamycin-induced hepatocyte death. Then, 18

we investigated the effect of gAcrp on tunicamycin-induced ER stress. We observed that the tunicamycin-induced PERK phosphorylation and CHOP expression were returned to basal levels by pretreatment with gAcrp (Fig. 2C and 2D, respectively). Taken together, these results indicate that the hepatoprotective effects of gAcrp against tunicamycin-induced cell death could be mediated via modulation of ER stress.

Fig. 2. Protection of tunicamycin-induced cell death by globular adiponectin via modulation of ER stress in rat hepatocytes. (A) Hepatocytes were pretreated with the indicated concentrations of TUDCA for 2 h followed by treatment with tunicamycin (1 µg/ml) for 24 h. Cell viability was measured by 19

MTS assay. (B) Cells were pretreated with TUDCA (300 µM) followed by incubation with tunicamycin (1 µg/ml) for 10 h. Cell culture media were then collected for the measurement of LDH release. Results (A-B) are presented as the mean± SEM, n=4. *P<0.05 compared with the control cells; #P<0.05 compared with cells treated with tunicamycin. (C) Hepatocytes were pretreated with indicated concentration of gAcrp for 18 h, followed by stimulation with 1 µg/ml of tunicamycin for 6 h. Total and phosphoPERK levels were detected by Western blot analysis. Representative images from three independent experiments are shown. (D) Hepatocytes were pretreated with gAcrp (0.5 or 1 µg/ml) for 18 h, and further stimulated with 1 µg/ml of tunicamycin for 8 h. CHOP expression levels were determined by Western blot analysis. β-actin was used as an internal loading control. The expression levels of phospho-PERK and CHOP shown in the Western blot images were quantified by densitometric analysis and shown in the lower panel (C and D). Values are expressed as mean ± SEM, n=3. *P<0.05 compared with control; #P<0.05 compared with cells treated with tunicamycin.

3.3. Globular adiponectin suppresses tunicamycin-induced inflammasomes activation via modulation of ER stress in hepatocytes

ER stress has been shown to induce the inflammasome activation, which produces inflammatory cytokines and may lead to cell death, including pyroptosis and apoptosis [4]. To further clarify the mechanisms underlying modulation of hepatocyte death by tunicamycin and adiponectin, we examined the effects of tunicamycin and gAcrp on the inflammasome activation. First, we found that tunicamycin induced cleavage and activation of IL-1β, a final product derived from inflammasome activation, in a time20

dependent manner without significantly changing the expression of pro-IL-1β (supplementary figure S5). We further explored the effect of gAcrp on tunicamycininduced inflammasomes activation. As demonstrated in Fig. 3A, pretreatment with gAcrp significantly prevented the tunicamycin-induced maturation of IL-1β in hepatocytes. We further examined the secreted levels of cleaved IL-1β (active form) in cell culture media, because the active IL-1β is released from the cell to mediate inflammatory responses [29]. Herein, total proteins in the culture media were precipitated using TCA and cleaved IL-1β levels were assessed by Western blot analysis. As expected, tunicamycin significantly enhanced the levels of active IL-1, but it was almost completely suppressed by pretreatment with gAcrp (Fig. 3B). Similarly, secretion of total IL-1β levels were also significantly decreased by gAcrp, as determined by ELISA (Fig. 3C), collectively indicating that gAcrp suppresses tunicamycin-induced IL-1 secretion and maturation in hepatocytes. We next examined the effects of gAcrp and tunicamycin on caspase-1 activation, which is triggered during inflammasomes activation and acts as the proteolytic enzyme responsible for the conversion of pro IL-1 to active IL-1. As shown in Fig. 3D, tunicamycin treatment enhanced the levels of the cleaved (active) form of caspase-1, which was substantially suppressed by pretreatment with gAcrp without a significant effect on pro-caspase-1 levels. The suppressive effect of gAcrp on caspase-1 activation was further confirmed by a caspase-1 enzyme activity assay (Fig. 3E). Moreover, TUDCA also significantly suppressed the tunicamycininduced caspase-1 activation (Fig. 3F), an effect similar to that of gAcrp, further confirming that gAcrp suppresses the inflammasome activation by regulation of ER stress. 21

We next examined the effects of tunicamycin and gAcrp on the components of the inflammasome. Tunicamycin treatment induced an increase in the expression of NLRP3 in a time-dependent manner (supplementary figure S6), which was restored by pretreatment with gAcrp (Fig. 3G), indicating that gAcrp regulates the tunicamycininduced NLRP3 expression in rat hepatocytes. Finally, we examined the effects of tunicamycin and gAcrp on the assembly of ASC (speck formation), which is considered necessary for inflammasomes activation. As illustrated in Fig. 3H, tunicamycin treatment prominently increased ASC speck formation determined by immunocytochemical analysis (indicated as green dots), whereas pretreatment with gAcrp substantially suppressed ASC speck formation. These results are consistent with the ones observed in IL-1 maturation, NLRP3 expression, and caspase-1 activation. In this study, we also examined expression levels of ASC, but observed that ASC expression levels were not significantly affected by tunicamycin (Supplementary figure S7), implying that modulation of the inflammasome by tunicamycin and gAcrp would not be determined by the regulation of ASC expression. Collectively, all these results support that tunicamycin, an ER stress inducer, leads to inflammasome activation in hepatocytes and that gAcrp suppresses the inflammasome activation probably via modulation of ER stress.

22

Fig. 3. Suppression of tunicamycin-induced inflammasomes activation by globular adiponectin via modulation of ER stress in rat hepatocytes. (A-E) Rat hepatocytes were pretreated with the indicated concentrations of gAcrp, followed by stimulation with tunicamycin (1 µg/mL) for 10 h. (A) Total cellular lysates were prepared and used for the measurement of pro- and cleaved (mature) IL-1β 23

expression levels by Western blot analysis. β-actin was used as an internal loading control. Representative images from three independent experiments are shown. (B) After treatment with tunicamycin and gAcrp, the cell culture media were collected and used to detect the secreted levels of active IL-1β. Total proteins in the cell culture media were precipitated using trichloroacetic acid (TCA) and subjected for the Western blot analysis using cleaved IL-1β specific antibody. Representative images from three independent experiments are shown. (C) The cell culture media were collected and used for the measurement of secreted IL-1β levels by ELISA as indicated in the materials and methods section. (D) Total cellular lysates were extracted and used to measure pro- and active-caspase-1 levels by Western blot analysis. Representative images from three independent experiments are shown. (E) Cell culture media were collected and used to assess caspase-1 enzymatic activity as indicated in the materials and methods section. Values shown in C, D and E represent mean ± SEM (n=4), *P<0.05 compared with control group; #P<0.05 compared with cells treated with tunicamycin. (F) Hepatocytes were isolated and pretreated with TUDCA (300 µM) for 2 h, followed by treatment with tunicamycin for 10 h. Total cellular lysates were prepared and used for the measurement of pro- and active-caspase-1 levels by Western blot analysis. Representative images from three independent experiments are shown. (G) Hepatocytes were pretreated with the indicated concentrations of gAcrp (0.5 and 1 µg/ml) for 18 h, followed by stimulation with tunicamycin (1 µg/mL) for 10 h. NLRP3 protein expression was measured by Western blot analysis. β-actin was used as an internal loading control. Representative images from three independent experiments are shown. (H) After pretreatment with gAcrp for 18 h followed by treatment with 24

tunicamycin, cells were incubated with an antibody against ASC and a speck fluorescence images were captured (scale bar, 200 µm). Blue color represents DAPI, which labels the nucleus of the hepatocytes. Green color shown with white arrow represents ASC specks in the cytoplasm. Representative images from three independent experiments showing similar results were presented. Lower panel represents the quantification of the cells containing ASC specks from three different experiments. The expression levels of IL-1β (A and B), caspase-1(D and F) and NLRP3 (G) shown in Western blot images were quantitated by densitometric analysis and presented in the lower panel of each image. Values are expressed as mean ± SEM, n=3. *P<0.05 compared with the control cells; #P<0.05 compared with cells treated with tunicamycin.

3.4. Globular adiponectin suppresses tunicamycin-induced ER stress and inflammasomes activation via modulation of ROS in hepatocytes

ROS production is accompanied with cellular stress conditions and considered a key mechanism leading to ER stress [30]. To further elucidate the mechanisms through which adiponectin protects hepatocytes from ER stress-induced cell death, we investigated whether the regulation of ROS is implicated in the modulation of tunicamycin-induced ER stress and inflammasomes activation. In this study, we first observed that tunicamycin significantly increased ROS production in our experimental condition (supplementary figure S8). In addition, co-treatment with N-acetyl cysteine (NAC), a ROS scavenger, prominently suppressed the tunicamycin-induced PERK phosphorylation and CHOP expression (Fig. 4A and 4B). Moreover, NAC treatment 25

substantially inhibited the tunicamycin-induced IL-1β maturation (Fig. 4C), indicating that ROS production acts as an upstream signaling event in tunicamycin-induced ER stress and inflammasomes activation in hepatocytes. Finally, we found that pretreatment with gAcrp markedly prevented tunicamycin-induced ROS production (Fig. 4D). Taken together, these results imply that gAcrp inhibits tunicamycin-induced ER stress and inflammasomes activation by modulating ROS production.

Cellular ROS levels are regulated via multiple mechanisms. To further investigate the mechanisms underlying modulation of ROS production, we examined the effects of tunicamycin and gAcrp on NADPH oxidase, which is an important source for ROS production in hepatocytes [31, 32]. As shown in Fig. 4E, tunicamycin treatment significantly enhanced NADPH oxidase activity, but it was decreased to almost basal levels by pretreatment with gAcrp, raising the possibility that NADPH oxidase could be a target molecule involved in the modulation of ROS production by gAcrp in rat hepatocytes. Different types of NADPH oxidase have been identified to date. Among these, NOX-1, NOX-2, and NOX-4 are known to be expressed in hepatocytes [33]; specifically, NOX-2 is well known to be implicated in ER stress [34, 35]. Therefore, we investigated the effects of tunicamycin and gAcrp on NOX-2 expression. Herein, we found that gAcrp significantly inhibited the tunicamycin-induced increase in NOX-2 expression (Fig. 4F), similar to the regulation of ROS production, suggesting that NOX-2 could be a target molecule participating in the modulation of ROS production by tunicamycin and gAcrp.

26

Fig. 4. Inhibitory effects of globular adiponectin on tunicamycin-induced ER stress and inflammasomes activation via modulation of ROS in rat hepatocytes. (A) Rat hepatocytes were pretreated with 2 mM of NAC for 1 h followed by 1 µg/ml of tunicamycin for 6 h. Total and phospho-PERK levels were measured by Western blot analysis. (B) Hepatocytes were pretreated with 2 mM of NAC for 1 h followed by 1 µg/ml of tunicamycin for 8 h. CHOP expression levels were determined by Western blot analysis. -actin was used as an internal control. (C) Hepatocytes were pretreated with 2 mM of NAC for 1 h followed by 1 µg/ml of tunicamycin for 10 h. Pro- and cleaved IL-1 levels were measured by Western blot analysis. Representative images from three independent experiments showing similar results were presented. The expression levels 27

of PERK (A), CHOP (B) and IL-1 (C) levels shown in Western blot images were quantified by densitometric analysis and presented in the lower panel of each image. Values are expressed as mean ± SEM, n=3. *P<0.05 compared with the control cells; #P<0.05

compared with cells treated with tunicamycin. (D) Hepatocytes were pretreated

with indicated concentrations of gAcrp for 18 h followed by stimulation with tunicamycin (1 µg) for 6 h. ROS production levels were determined by measuring changes in the fluorescence of CM-H2DCFDA as described in materials and methods. (E) Cells were incubated with gAcrp (0.5 or 1 µg) followed by tunicamycin (1 µg/ml) for 6 h. NADPH oxidase activity was determined using NAPDH/Lucigenin solution as mentioned in materials and methods section. Values shown in Fig. D and E represents mean ± SEM (n=3), *P<0.05 compared with control; #P<0.05 compared with cells treated with tunicamycin. (F) Hepatocytes were pretreated with gAcrp (0.5 or 1 µg) followed by tunicamycin (1 µg/ml) for 6 h. Cellular lysates were prepared and used to determine the protein expression of NOX-2 by Western blot analysis. β-actin was used as an internal loading control. Representative images from three independent experiments were presented. The quantification of NOX-2 levels shown in the Western blot images was performed by densitometric analysis. Values represents mean ± SEM (n=3), *P<0.05 compared with control; #P<0.05 compared with cells treated with tunicamycin.

3.5. Modulation of tunicamycin-induced ROS production and hepatocytes death by globular adiponectin is mediated via HO-1 induction

Heme oxygenase-1 (HO-1), a well-known anti-oxidant signaling molecule, has been shown to mediate various physiological responses induced by adiponectin [36]. To 28

further elucidate the upstream molecular mechanisms underlying modulation of ROS production, ER stress and, hepatocytes death by gAcrp, we examined the role of HO-1 signaling. To this end, we first confirmed the effect of gAcrp on HO-1 induction. As shown in Fig. 5A, gAcrp treatment rapidly induced increase in HO-1 expression level in hepatocytes (Fig. 5A), consistent with previous reports. To corroborate the functional role of HO-1 induction, cells were pretreated with SnPP, a selective pharmacological inhibitor of HO-1. As depicted in Fig. 5B, protection of hepatocytes by gAcrp was abolished by SnPP pretreatment. In addition, genetic ablation of HO-1 by siRNA transfection also blocked the protection of liver cells by gAcrp. We confirmed that this effect was not due to the transfection procedure with the use of a scrambled control siRNA (Fig. 5C), collectively indicating the pivotal role of HO-1 signaling in the hepatoprotective effects of gAcrp against tunicamycin-induced cell death. We further investigated the role of HO-1 signaling in the modulation of ER stress by gAcrp. Interestingly, suppression of PERK phosphorylation by gAcrp was almost completely restored by either treatment with SnPP or gene silencing of HO-1 (Fig. 5D and 5E). Moreover, suppression of CHOP expression by gAcrp was also restored by SnPP treatment or gene silencing of HO-1 (Fig. 5F and 5G), clearly demonstrating the crucial role of HO-1 in the modulation of ER stress by gAcrp. Finally, we examined the role of HO-1 signaling in the regulation of ROS production in hepatocytes. As shown in Fig. 5H, suppression of ROS production by gAcrp was completely restored by transfection with HO-1 siRNA. Taken together, these results clearly indicate that HO-1 induction plays a critical role in the suppression of tunicamycin-induced cell death and ER stress presumably via modulation of ROS production. 29

Fig. 5. Protective role of HO-1 signaling in the suppression of tunicamycin-induced cell death and ER stress by globular adiponectin in rat hepatocytes. (A) Hepatocytes isolated from rat were treated with gAcrp (1 µg/ml) for 3, 8 and 24 h. Cellular lysates were prepared and used for the measurement of HO-1 protein expression by Western blot analysis. The lower panel demonstrates the quantification of HO-1 expression level in Western blot images by densitometric analysis of three 30

different experiments. Values are expressed as mean ± SEM (n=3). *P<0.05 compared with control. (B) Hepatocytes were pretreated with SnPP (20 µM) for 2 h followed by treatments with gAcrp (1 µg/ml) for 18 h and tunicamycin (1 µg/ml) for 24 h. Cell viability was measured by the MTS assay as described in materials and methods section. (C) Hepatocytes were transfected with HO-1 siRNA for 24 h. Cells were then pretreated with gAcrp for 18 h followed by tunicamycin for 24 h. Cell viability was determined by the MTS assay. The HO-1 gene silencing efficiency was evaluated via Western blot method (upper panel). Values shown in Fig. B and C represent mean ± SEM (n=3), *P<0.05 compared with control; #P<0.05 compared with the cells treated with tunicamycin; $P<0.05 compared with cells treated with gAcrp and tunicamycin. (D and E) (D) Hepatocytes were pretreated with indicated concentrations of SnPP for 2 h followed by treatments with gAcrp for 18 h and tunicamycin for 6 h. (E) After transfection with HO-1 siRNA (75 ng), cells were pretreated with gAcrp (1 µg/ml) for 18 h followed by incubation with tunicamycin (1 µg/ml) for 6 h. Total and phospho-PERK levels were measured by Western blot analysis and the expression level was quantified by densitometric analysis shown in the lower panel of each image. Values are expressed as the mean ± SEM (n=3), *P<0.05 compared with control; #P<0.05 compared with cells treated with tunicamycin; $P<0.05 compared with cells treated with gAcrp and tunicamycin. (F and G) (F) Hepatocytes were pretreated with SnPP for 2 h followed by treatments with gAcrp for 18 h and tunicamycin for 6 h. (G) After transfection with HO-1 siRNA (75 ng), cells were pretreated with gAcrp (1 µg/ml) for 18 h followed by incubation with tunicamycin (1 µg/ml) for 6 h. CHOP expression levels were measured by Western blot analysis. Expression levels shown in the blots were quantified by 31

densitometric analysis and shown in the lower panel the image. Values are presented as the mean ± SEM (n=3), *P<0.05 compared with control; #P<0.05 compared with cells treated with tunicamycin; $P<0.05 compared with cells treated with gAcrp and tunicamycin. (H) Hepatocytes were transfected with siRNA targeting HO-1 or control scrambled siRNA for 24 h. After treatments with gAcrp for 18 h followed by incubation with tunicamycin for 6 h, cells were incubated with 2', 7’-dichlorodihydrofluorescein diacetate for 30 minutes and fluorescence generated was measured. Values are presented as the mean ± SEM (n=3), *P<0.05 compared with control; #P<0.05 compared with cells treated with tunicamycin; $P<0.05 compared with cells treated with gAcrp and tunicamycin.

3.6. HO-1 induction is involved in the suppression of tunicamycin-induced inflammasomes activation by globular adiponectin in hepatocytes

We next speculated the involvement of HO-1 signaling in the modulation of inflammasomes activation by gAcrp. As shown in Fig. 6A, suppression of tunicamycininduced IL-1β maturation by gAcrp was restored by treatment with SnPP. In addition, knockdown of HO-1 by siRNA transfection also restored the tunicamycin-induced IL-1β maturation (Fig. 6B), indicating the crucial role of HO-1 signaling in the regulation of inflammasomes activation by gAcrp. Furthermore, inhibition of HO-1 signaling by SnPP treatment or genetic ablation also restored the suppression of tunicamycin-induced caspase-1 activation (Fig. 6C and 6D). Taken together, these findings demonstrate the crucial role of HO-1 induction in the suppression of the inflammasome activation by gAcrp in hepatocytes. 32

Fig. 6. Role of HO-1 signaling in the suppression of tunicamycin-induced inflammasomes activation by globular adiponectin in rat hepatocytes. (A) and (B), (A) Hepatocytes were pretreated with SnPP (10 µM) for 2 h and further incubated with gAcrp (1 µg/ml) for 18 h and tunicamycin (1 µg/ml) for 10 h. (B) After transfection with HO-1 siRNA (75 ng) for 24 h, hepatocytes were pretreated with gAcrp (1 µg/ml) for 18 h followed by incubation with tunicamycin (1 µg/ml) for 10 h. Cellular lysates were prepared and pro- and mature IL-1 levels were determined by Western 33

blot analysis. The representative image of three independent experiments are shown. Cleaved IL-1 levels were quantified by densitometric analysis and presented in the lower panel of the image. Values are expressed as mean ± SEM (n=3), *P<0.05 compared with control; #P<0.05 compared with cells treated with tunicamycin, $P<0.05 compared with cells treated with gAcrp and tunicamycin. (C) and (D), (C) Cells were pretreated with SnPP (10 µM) for 2 h followed by incubation with gAcrp for 18 h and tunicamycin for 10 h. (D) Cells were transfected with siRNA targeting HO-1. After 24 h, hepatocytes were pretreated with gAcrp (1 µg/ml) for 18 h followed by further stimulation with tunicamycin (1 µg/ml) for 10 h. Cellular lysates were prepared and proand active caspase-1 levels were determined by Western blot analysis. The representative image of three independent experiments are shown. Active caspase-1 levels were quantified by densitometric analysis. Values are expressed as mean ± SEM (n=3), *P<0.05 compared with control; # P<0.05 compared with cells treated with tunicamycin, $P<0.05 compared with cells treated with gAcrp and tunicamycin.

4. Discussion Adiponectin has exhibited many beneficial physiological actions in the liver and plays protective roles against different types of hepatic disease. In particular, adiponectin prevents hepatocyte death induced by diverse harmful stimuli [37]. Although various signaling mechanisms have been proposed for the protection of liver cells by adiponectin, including modulation of the apoptosis pathway [28], reducing TNF-α levels [38], and inhibiting up-regulation of Fas ligand [39], the underlying molecular 34

mechanisms are still largely unknown. ER stress has recently received attention as a critical cellular process that triggers death of liver cells and is implicated in the pathogenesis of liver diseases [40]. Recently, it has been shown that adiponectin regulates ER stress in a liver disease model [26]. However, its underlying molecular mechanisms are not clearly understood yet. In the present study, we examined the protective effect of globular adiponectin against tunicamycin-induced cell death in rat hepatocytes and further investigated the underlying mechanisms in an effort to expand current knowledge. Herein, we have shown that tunicamycin-induced ER stress leads to hepatocyte death via inflammasomes activation and further demonstrated, for the first time, that globular adiponectin protects hepatocytes from both tunicamycin-induced apoptosis and pyroptosis through HO-1 signaling-dependent mechanisms. In this study, we demonstrated that gAcrp protected hepatocytes from tunicamycininduced cell death (Fig. 1A), consistent with a recent study showing that adiponectin prevents tunicamycin-induced apoptosis in adipocytes. We also found that gAcrp suppressed tunicamycin-induced PERK phosphorylation (Fig. 2C) and CHOP protein expression (Fig. 2D), clearly suggesting that the protective effect of adiponectin is mediated via modulation of ER stress. The protective effects of adiponectin against ER stress-induced cellular damage were observed in various experimental conditions. For example, gAcrp protected adipocytes and endothelial cells from palmitic acid-induced ER stress via suppression of CHOP expression and Caspase-3 cleavage [5]. In addition, hypoxia-induced ER stress in rats, accompanied with enhanced PERK phosphorylation and CHOP expression, was blocked by treatment with adiponectin in cardiomyocytes [41]. Taken together, results show that modulation of ER stress by 35

adiponectin could be a critical mechanism for the prevention of aberrant cell death. Given the previous reports stating that persistent ER stress is closely associated with liver diseases and hepatocyte death, modulation of ER stress would be an important mechanism underlying the hepatoprotective effects of adiponectin. Furthermore, elucidation of the molecular mechanisms through which adiponectin regulates ER stress would provide attractive targets to develop pharmacological agents for the treatment of liver diseases. Accumulating evidence has revealed that dysregulated ER stress leads to inflammatory cell death through activation of the inflammasome [42, 43]. In the present study, we have demonstrated that inflammasomes activation is implicated in tunicamycin-induced hepatocyte death. Furthermore, we observed that gAcrp prevented the tunicamycininduced production of cleaved (active) IL-1β (Fig. 3A and 3B), a final product of the inflammasome activation. Similarly, tunicamycin-induced caspase-1 activation, an upstream signaling response that induces IL-1β maturation, was also suppressed by gAcrp (Fig. 3D), collectively suggesting that gAcrp protects hepatocytes from tunicamycin-induced cell death via modulation of inflammasome activation. The inflammasome is composed of multi-components, including ASC, NLRP3, and caspse1. Additionally, inflammasomes activity is regulated in a coordinated action of the components in the inflammasome. In this study, we observed that gAcrp pretreatment significantly inhibited the tunicamycin-induced expression of NLRP3 (Fig. 3G). Interestingly, tunicamycin itself did not affect the expression of ASC (Supplementary figure S7), however it enhanced the formation of ASC speck, an effect suppressed by pretreatment with gAcrp (Fig. 3H). ASC, an adaptor protein required for activation of 36

caspase-1, assembles into a large protein complex called speck (also known as ASC bundles), which is a distinguished characteristic and readout for inflammasomes activation [44]. The results observed in this study imply that gAcrp prevents tunicamycin-induced inflammasomes activation probably via interruption of ASC speck formation and inflammasomes assembly, rather than by modulation of ASC expression. In addition to ASC, inflammasomes activation is regulated via multiple pathways, such as NLRC4, NLRP1, AIM2, and other mechanisms. In the present study, we focused on the effects of tunicamycin and gAcrp on the modulation of NLRP3, since it is the best characterized in hepatocytes. In the present study, we did not investigate other different types of inflammasomes. Therefore, it is possible that different types of the inflammasome could be implicated in the modulation of ER stress by adiponectin. Further studies to elucidate to the effects of gAcrp and tunicamycin on the modulation of various inflammasomes would provide further insights into the mechanisms underlying protection of liver cells by adiponectin. It is well established that excessive production of ROS is a critical event leading to a number of pathophysiological states [45]. There is a growing appreciation that ROS production also plays a crucial role in mediating damage to liver cells and development of liver diseases in various conditions [46, 47]. Accumulating evidence has also revealed that dysregulation of redox homeostasis leads to ER stress and unfolded protein response (UPR). ROS are produced from various cellular organelles, including ER and mitochondria. Of the various mechanisms underlying ROS production, activation of the NOX family of NADPH oxidase plays a crucial role in ROS production in the liver upon exposure to a number of toxic molecules, which is directly responsible 37

for the induction of apoptosis in hepatocytes [30]. In the present study, we examined the involvement of NADPH oxidase in the modulation of ROS production by tunicamycin and gAcrp in hepatocytes. Herein, we found that tunicamycin treatment induced ROS production (supplementary figure S8). Furthermore, gAcrp prevented the tunicamycininduced activation of NADPH oxidase (Fig. 4E), indicating that suppression of ROS production and protection of liver cells by gAcrp are mediated by modulation of NADPH oxidase. The mammalian NOX family of NADPH oxidases comprises seven members, including Nox1 to Nox5, and Duox-1 and 2, each of which has different tissue distribution and activation mechanisms. Among them, NOX-2 has been shown to play a critical role in ROS production during ER stress in cardiovascular and renal cells [48-50]. However, in liver cells, the major isoform(s) of NOX responsible for ER stress occurrence is not clearly defined. In this study, we found that tunicamycin treatment induced an increase in NOX-2 expression, whereas pretreatment with gAcrp abolished this effect (Fig. 4F) and also the increase in ROS production (Fig. 4D). Furthermore, we also observed that N-acetyl cysteine significantly suppressed the tunicamycin-induced PERK phosphorylation (Fig. 4A) and CHOP expression (Fig. 4B). Collectively, these results indicate that modulation of ER stress and further protection of hepatocytes from ER stress-induced cell death by gAcrp could be mediated via suppression of ROS production, in particular through NOX-2-derived ROS. To the best of our knowledge, this is the first report to demonstrate the role of NADPH oxidase, in particular NOX-2, in the tunicamycin-induced ROS production in hepatocytes. In addition to NOX-2, different isoforms of NADPH oxidase may play roles in ER stress-induced ROS production 38

depending on cell type and stimulus. For example, NOX-4 has been shown to act as a principal source for ER stress-induced ROS production in PC-3 cells [51]. Herein, although we demonstrated that NOX-2 plays a critical role in tunicamycin-induced ER stress and hepatocyte death, it is possible that other isoform(s) of NOX family could be implicated in ROS production, thus leading to ER stress and hepatocyte death. Further studies identifying the role of other isoforms of NADPH oxidase in the modulation of ROS production by ER stress and gAcrp would be worth considering. Heme oxygenases (HO) are cellular enzyme systems that catalyze the conversion of heme into biliverdin, carbon monoxide, and free iron [52]. Of its three isoforms (HO-1, HO-2 and HO-3), HO-1 is well known as an inducible isoform that responds to a variety of stress conditions, such as hypoxia, inflammation, and oxidative stress. Moreover, it participates in the cellular adaptation process to oxidative stress [53]. In addition to the modulation of oxidative stress and inflammatory responses, recent studies have further demonstrated that HO-1 induction mediates cytoprotective actions by various pharmacological agents and regulates apoptosis through different signaling molecules such as p38 mitogen-activated protein kinase, autophagic proteins, and phosphatidylinositol 3-kinase/Akt [54-56]. Interestingly, HO-1 has been shown to mediate several of the beneficial actions of adiponectin. For example, HO-1 induction plays a crucial role in the protection of liver cells from chronic ethanol exposure [57, 58], suppression of LPS-induced ROS production [59], and protection against intestinal ischemia reperfusion injury [60]. Driven by these considerations, we examined whether the hepatoprotective effects of gAcrp against tunicamycin-induced cell death were mediated by HO-1 induction. We demonstrated that inhibition of HO-1 by 39

pharmacological inhibitor or genetic ablation restored the suppression of the tunicamycin-induced ER stress (Fig. 5D-5G) and prevented the protective effects of gAcrp against tunicamycin-induced cell death (Fig. 5 B and 5C). Furthermore, HO-1 inhibition abrogated the suppression of ROS production by gAcrp, collectively indicating that HO-1 signaling plays a crucial role in the hepatoprotective effects of gAcrp, presumably by modulating ROS production and ER stress. Although, during the preparation of the manuscript, Maamoun and colleagues recently reported the role of HO-1 induction in protection of endothelial cells from high glucose-induced ER stress [61], this is the first report demonstrating the role of HO-1 signaling in gAcrp-mediated protection of liver cells from ER stress-induced cell death. Previous studies have demonstrated that adiponectin modulates ER stress via Akt/PI3K-dependent mechanisms in cardiomyocytes [62] and an AMPK signaling-dependent mechanism in adipocytes [5]. Although multiple mechanisms have been proposed for the modulation of ER stress and cytoprotective action by adiponectin, the results obtained in the present study demonstrate for the first time that HO-1 could be a promising pharmacological target for the modulation ER stress and subsequent cellular damage. 5. Conclusion In conclusion, the data presented in this study have clearly demonstrated that globular adiponectin protects hepatocytes from ER stress-induced cellular damage by modulating inflammasomes activation. The hepatoprotective action of globular adiponectin was mediated via inhibition of NOX-2-derived ROS generation. Furthermore, HO-1 signaling plays a crucial role in the suppression of tunicamycininduced ROS generation and subsequent activation of inflammasomes activation. 40

These studies provide an insight into the molecular mechanisms underlying the hepatoprotective effects of adiponectin and suggest that HO-1 would be a promising target for the treatment of ER stress-induced liver cell damages.

Conflict of interest The authors have no conflicts of interest. Acknowledgement This work was supported by the Basic Science Research Program of the National Research Foundation of Korea (NRF), funded by the Ministry of Education (NRF2015R1D1A1A01058203).

References [1] T. Luedde, N. Kaplowitz, R.F. Schwabe, Cell death and cell death responses in liver disease: mechanisms and clinical relevance, Gastroenterology 147(4) (2014) 765-783. e4. [2] H. Malhi, M.E. Guicciardi, G.J. Gores, Hepatocyte Death: A Clear and Present Danger, Physiological reviews 90(3) (2010) 1165-1194. [3] L. Cao, X.-B. Quan, W.-J. Zeng, X.-O. Yang, M.-J. Wang, Mechanism of Hepatocyte Apoptosis, Journal of Cell Death 9 (2016) 19-29. [4] C. Lebeaupin, E. Proics, C.H.D. de Bieville, D. Rousseau, S. Bonnafous, S. Patouraux, G. Adam, V.J. Lavallard, C. Rovere, O. Le Thuc, M.C. Saint-Paul, R. Anty, A.S. Schneck, A. Iannelli, J. Gugenheim, A. Tran, P. Gual, B. Bailly-Maitre, ER stress induces NLRP3 inflammasome activation and hepatocyte death, Cell Death & Disease 6(9) (2015) e1879. [5] Z. Liu, L. Gan, T. Wu, F. Feng, D. Luo, H. Gu, S. Liu, C. Sun, Adiponectin reduces ER stress-induced apoptosis through PPARalpha transcriptional regulation of ATF2 in mouse adipose, Cell Death Dis 7(11) (2016) e2487. [6] H. Yoshida, T. Matsui, A. Yamamoto, T. Okada, K. Mori, XBP1 mRNA Is Induced by ATF6 and Spliced by IRE1 in Response to ER Stress to Produce a Highly Active Transcription Factor, Cell 107(7) (2001) 881891. [7] J.A. Willy, S.K. Young, J.L. Stevens, H.C. Masuoka, R.C. Wek, CHOP links endoplasmic reticulum stress to NF-kappaB activation in the pathogenesis of nonalcoholic steatohepatitis, Molecular biology of the cell 26(12) (2015) 2190-204. [8] B.-J. Park, Y.-J. Lee, H.-R. Lee, Chronic liver inflammation: Clinical implications beyond alcoholic liver disease, World Journal of Gastroenterology : WJG 20(9) (2014) 2168-2175. [9] M.W. Robinson, C. Harmon, C. O'Farrelly, Liver immunology and its role in inflammation and homeostasis, Cellular and Molecular Immunology 13(3) (2016) 267-276.

41

[10] Z. Zhou, M.-J. Xu, B. Gao, Hepatocytes: a key cell type for innate immunity, Cellular and Molecular Immunology 13(3) (2016) 301-315. [11] A. Wree, A. Eguchi, M.D. McGeough, C.A. Pena, C.D. Johnson, A. Canbay, H.M. Hoffman, A.E. Feldstein, NLRP3 inflammasome activation results in hepatocyte pyroptosis, liver inflammation and fibrosis, Hepatology (Baltimore, Md.) 59(3) (2014) 898-910. [12] Y. He, H. Hara, G. Núñez, Mechanism and Regulation of NLRP3 Inflammasome Activation, Trends in Biochemical Sciences 41(12) 1012-1021. [13] V. Compan, F. Martin-Sanchez, A. Baroja-Mazo, G. Lopez-Castejon, A.I. Gomez, A. Verkhratsky, D. Brough, P. Pelegrin, Apoptosis-associated speck-like protein containing a CARD forms specks but does not activate caspase-1 in the absence of NLRP3 during macrophage swelling, 194(3) (2015) 1261-73. [14] K. Shenderov, N. Riteau, R. Yip, K.D. Mayer-Barber, S. Oland, S. Hieny, P. Fitzgerald, A. Oberst, C.P. Dillon, D.R. Green, V. Cerundolo, A. Sher, ER stress licenses macrophages to produce mature IL-1β in response to TLR4 stimulation through a caspase-8- and TRIF-dependent pathway, Journal of immunology (Baltimore, Md. : 1950) 192(5) (2014) 2029-2033. [15] A. So, T. De Smedt, S. Revaz, J. Tschopp, A pilot study of IL-1 inhibition by anakinra in acute gout, Arthritis Research & Therapy 9(2) (2007) R28. [16] F. Martinon, V. Petrilli, A. Mayor, A. Tardivel, J. Tschopp, Gout-associated uric acid crystals activate the NALP3 inflammasome, Nature 440(7081) (2006) 237-241. [17] H.-M. Lee, J.-J. Kim, H.J. Kim, M. Shong, B.J. Ku, E.-K. Jo, Upregulated NLRP3 Inflammasome Activation in Patients With Type 2 Diabetes, Diabetes 62(1) (2013) 194-204. [18] T. Csak, M. Ganz, J. Pespisa, K. Kodys, A. Dolganiuc, G. Szabo, Fatty acid and endotoxin activate inflammasomes in mouse hepatocytes that release danger signals to stimulate immune cells, Hepatology 54(1) (2011) 133-144. [19] P. Mandrekar, G. Szabo, Signalling pathways in alcohol-induced liver inflammation, Journal of Hepatology 50(6) (2009) 1258-1266. [20] B. Guo, Z. Li, Endoplasmic reticulum stress in hepatic steatosis and inflammatory bowel diseases, Frontiers in Genetics 5(242) (2014). [21] J.M. Tishinsky, L.E. Robinson, D.J. Dyck, Insulin-sensitizing properties of adiponectin, Biochimie 94(10) (2012) 2131-2136. [22] M. Guerre-Millo, Adiponectin: An update, Diabetes & metabolism 34(1) (2008) 12-18. [23] A. Xu, Y. Wang, H. Keshaw, L.Y. Xu, K.S.L. Lam, G.J.S. Cooper, The fat-derived hormone adiponectin alleviates alcoholic and nonalcoholic fatty liver diseases in mice, Journal of Clinical Investigation 112(1) (2003) 91-100. [24] S. Nepal, P.H. Park, Activation of autophagy by globular adiponectin attenuates ethanol-induced apoptosis in HepG2 cells: involvement of AMPK/FoxO3A axis, Biochimica et biophysica acta 1833(10) (2013) 2111-25. [25] Y. Lu, Y. Bian, Y. Wang, R. Bai, J. Wang, C. Xiao, Globular adiponectin reduces vascular calcification via inhibition of ER-stress-mediated smooth muscle cell apoptosis, International Journal of Clinical and Experimental Pathology 8(3) (2015) 2545-2554. [26] T. Ueno, A. Nakamura, H. Nakayama, S. Otabe, X. Yuan, T. Fukutani, H. Iwamoto, T. Nakamura, H. Koga, T. Torimura, M. Sata, K. Yamada, Adiponectin suppresses endoplasmic reticulum stress in nonalcoholic steatohepatitis, Experimental and therapeutic medicine 2(6) (2011) 1035-1040. [27] Y. Kim, P. Rajagopalan, 3D Hepatic Cultures Simultaneously Maintain Primary Hepatocyte and Liver Sinusoidal Endothelial Cell Phenotypes, PLoS ONE 5(11) (2010) e15456. [28] S. Nepal, M.J. Kim, A. Subedi, E.-S. Lee, C.S. Yong, J.-A. Kim, W. Kang, M.-K. Kwak, D.S. Arya, P.H. Park, Globular adiponectin inhibits ethanol-induced apoptosis in HepG2 cells through heme oxygenase1 induction, Biochemical Pharmacology 84(7) (2012) 974-983. [29] G. Lopez-Castejon, D. Brough, Understanding the mechanism of IL-1β secretion, Cytokine & Growth Factor Reviews 22(4) (2011) 189-195. [30] H.M.A. Zeeshan, G.H. Lee, H.-R. Kim, H.-J. Chae, Endoplasmic Reticulum Stress and Associated ROS, International Journal of Molecular Sciences 17(3) (2016) 327. [31] S. Liang, T. Kisseleva, D.A. Brenner, The Role of NADPH Oxidases (NOXs) in Liver Fibrosis and the Activation of Myofibroblasts, Frontiers in Physiology 7 (2016) 17.

42

[32] C. Guichard, R. Moreau, D. Pessayre, Terry K. Epperson, K.-H. Krause, NOX family NADPH oxidases in liver and in pancreatic islets: a role in the metabolic syndrome and diabetes?, Biochemical Society Transactions 36(5) (2008) 920. [33] J.X. Jiang, N.J. Torok, NADPH Oxidases in Chronic Liver Diseases, Adv Hepatol 2014 (2014). [34] R.-F. Wu, Z. Ma, Z. Liu, L.S. Terada, Nox4-Derived H(2)O(2) Mediates Endoplasmic Reticulum Signaling through Local Ras Activation, Molecular and Cellular Biology 30(14) (2010) 3553-3568. [35] B. Bhandary, A. Marahatta, H.-R. Kim, H.-J. Chae, An Involvement of Oxidative Stress in Endoplasmic Reticulum Stress and Its Associated Diseases, International Journal of Molecular Sciences 14(1) (2013) 434-456. [36] M. Elfeky, R. Kaede, Y. Okamatsu-Ogura, K. Kimura, Adiponectin Inhibits LPS-Induced HMGB1 Release through an AMP Kinase and Heme Oxygenase-1-Dependent Pathway in RAW 264 Macrophage Cells, Mediators of Inflammation 2016 (2016) 5701959. [37] T.E. Silva, G. Colombo, L.L. Schiavon, Adiponectin: A multitasking player in the field of liver diseases, Diabetes & metabolism 40(2) (2014) 95-107. [38] P.-h. Park, M.R. McMullen, H. Huang, V. Thakur, L.E. Nagy, Short-term Treatment of RAW264.7 Macrophages with Adiponectin Increases Tumor Necrosis Factor-α (TNF-α) Expression via ERK1/2 Activation and Egr-1 Expression: ROLE OF TNF-α IN ADIPONECTIN-STIMULATED INTERLEUKIN-10 PRODUCTION, The Journal of biological chemistry 282(30) (2007) 21695-21703. [39] I. Wedemeyer, L.P. Bechmann, M. Odenthal, C. Jochum, G. Marquitan, U. Drebber, G. Gerken, R.K. Gieseler, H.P. Dienes, A. Canbay, Adiponectin inhibits steatotic CD95/Fas up-regulation by hepatocytes: Therapeutic implications for hepatitis C, Journal of Hepatology 50(1) (2009) 140-149. [40] L. Dara, C. Ji, N. Kaplowitz, THE CONTRIBUTION OF ER STRESS TO LIVER DISEASES, Hepatology (Baltimore, Md.) 53(5) (2011) 1752-1763. [41] W. Ding, X. Zhang, H. Huang, N. Ding, S. Zhang, S.Z. Hutchinson, X. Zhang, Adiponectin Protects Rat Myocardium against Chronic Intermittent Hypoxia-Induced Injury via Inhibition of Endoplasmic Reticulum Stress, PLoS ONE 9(4) (2014) e94545. [42] S. Kim, Y. Joe, S.O. Jeong, M. Zheng, S.H. Back, S.W. Park, S.W. Ryter, H.T. Chung, Endoplasmic reticulum stress is sufficient for the induction of IL-1beta production via activation of the NF-kappaB and inflammasome pathways, Innate immunity 20(8) (2014) 799-815. [43] J. Zhang, Z. Li, B. Guo, ER stress-induced inflammasome activation promotes liver steatosis (IRM9P.723), The Journal of Immunology 192(1 Supplement) (2014) 128.6. [44] D.P. Sester, S.J. Thygesen, V. Sagulenko, P.R. Vajjhala, J.A. Cridland, N. Vitak, K.W. Chen, G.W. Osborne, K. Schroder, K.J. Stacey, A novel flow cytometric method to assess inflammasome formation, Journal of immunology (Baltimore, Md. : 1950) 194(1) (2015) 455-62. [45] A.A. Alfadda, R.M. Sallam, Reactive Oxygen Species in Health and Disease, Journal of Biomedicine and Biotechnology 2012 (2012) 14. [46] J.X. Jiang, X. Chen, N. Serizawa, C. Szyndralewiez, P. Page, K. Schroder, R.P. Brandes, S. Devaraj, N.J. Torok, Liver fibrosis and hepatocyte apoptosis are attenuated by GKT137831, a novel NOX4/NOX1 inhibitor in vivo, Free Radic Biol Med 53(2) (2012) 289-96. [47] K. Wang, Molecular mechanisms of hepatic apoptosis, Cell Death Dis 5 (2014) e996. [48] G. Li, C. Scull, L. Ozcan, I. Tabas, NADPH oxidase links endoplasmic reticulum stress, oxidative stress, and PKR activation to induce apoptosis, The Journal of cell biology 191(6) (2010) 1113-25. [49] E. Pedruzzi, C. Guichard, V. Ollivier, F. Driss, M. Fay, C. Prunet, J.C. Marie, C. Pouzet, M. Samadi, C. Elbim, Y. O'Dowd, M. Bens, A. Vandewalle, M.A. Gougerot-Pocidalo, G. Lizard, E. Ogier-Denis, NAD(P)H oxidase Nox-4 mediates 7-ketocholesterol-induced endoplasmic reticulum stress and apoptosis in human aortic smooth muscle cells, Mol Cell Biol 24(24) (2004) 10703-17. [50] R.F. Wu, Z. Ma, Z. Liu, L.S. Terada, Nox4-derived H2O2 mediates endoplasmic reticulum signaling through local Ras activation, Mol Cell Biol 30(14) (2010) 3553-68. [51] S.-H. Kim, K.-Y. Kim, S.-N. Yu, Y.-K. Seo, S.-S. Chun, H.-S. Yu, S.-C. Ahn, Silibinin induces mitochondrial NOX4-mediated endoplasmic reticulum stress response and its subsequent apoptosis, BMC Cancer 16(1) (2016) 452. [52] S.W. Ryter, J. Alam, A.M.K. Choi, Heme Oxygenase-1/Carbon Monoxide: From Basic Science to Therapeutic Applications, Physiological Reviews 86(2) (2006) 583. [53] J. Araujo, M. Zhang, F. Yin, Heme Oxygenase-1, Oxidation, Inflammation, and Atherosclerosis, Frontiers in Pharmacology 3(119) (2012).

43

[54] D. Morse, L. Lin, A.M. Choi, S.W. Ryter, Heme oxygenase-1, a critical arbitrator of cell death pathways in lung injury and disease, Free Radic Biol Med 47(1) (2009) 1-12. [55] R. Fürst, S.B. Blumenthal, A.K. Kiemer, S. Zahler, A.M. Vollmar, Nuclear Factor-κB-Independent AntiInflammatory Action of Salicylate in Human Endothelial Cells: Induction of Heme Oxygenase-1 by the cJun N-Terminal Kinase/Activator Protein-1 Pathway, Journal of Pharmacology and Experimental Therapeutics 318(1) (2006) 389. [56] H.J. Jang, E.M. Hong, M. Kim, J.H. Kim, J. Jang, S.W. Park, H.W. Byun, D.H. Koh, M.H. Choi, S.H. Kae, J. Lee, Simvastatin induces heme oxygenase-1 via NF-E2-related factor 2 (Nrf2) activation through ERK and PI3K/Akt pathway in colon cancer, Oncotarget 7(29) (2016) 46219-46229. [57] B. Bakhautdin, D. Das, P. Mandal, S. Roychowdhury, J. Danner, K. Bush, K. Pollard, J.W. Kaspar, W. Li, R.G. Salomon, M.R. McMullen, L.E. Nagy, Protective role of HO-1 and carbon monoxide in ethanolinduced cell death in hepatocytes and liver injury in mice, Journal of hepatology 61(5) (2014) 1029-1037. [58] Z. Lijie, F. Ranran, L. Xiuying, H. Yutang, W. Bo, M. Tao, Soyasaponin Bb Protects Rat Hepatocytes from Alcohol-Induced Oxidative Stress by Inducing Heme Oxygenase-1, Pharmacognosy Magazine 12(48) (2016) 302-306. [59] A. Shrestha, P.H. Park, Globular adiponectin attenuates LPS-induced reactive oxygen species production in HepG2 cells via FoxO3A and HO-1 signaling, Life sciences 148 (2016) 71-9. [60] X.-H. Liu, Y.-W. Yang, H.-T. Dai, S.-W. Cai, R.-H. Chen, Z.-Q. Ye, Protective role of adiponectin in a rat model of intestinal ischemia reperfusion injury, World Journal of Gastroenterology 21(47) (2015) 1325013258. [61] H. Maamoun, M. Zachariah, J.H. McVey, F.R. Green, A. Agouni, Heme oxygenase (HO)-1 induction prevents Endoplasmic Reticulum stress-mediated endothelial cell death and impaired angiogenic capacity, Biochem Pharmacol 127 (2017) 46-59. [62] J. Guo, Y. Bian, R. Bai, H. Li, M. Fu, C. Xiao, Globular adiponectin attenuates myocardial ischemia/reperfusion injury by upregulating endoplasmic reticulum Ca(2)(+)-ATPase activity and inhibiting endoplasmic reticulum stress, Journal of cardiovascular pharmacology 62(2) (2013) 143-53.

44